TH Neal

03-Clinical Oncology-ch03-ccp:Layout 1 12/3/09 16:25 Page 22
DECISION-MAKING
3 AND COMMUNICATION
■ Treatment options 23 ■ Clinical evidence and clinical trials 28

■ Quality of life 25 ■ Conclusion 35
■ Communication 26 ■ Further reading 35
■ Conclusion 28
The following chapters describe clear treatment Diagnosis of malignant condition

policies for different types of malignant disease. Staging and histology reviewed
These policies are based upon a knowledge of
the natural history of the disease and, as DECISION 1 No treatment
detailed in the previous chapter, details of its
extent – the clinical or pathological stage of the
tumour. Active treatment
However, the practice of clinical oncology
demands more than simple application of these
instructions in an uncritical fashion. The indi- DECISION 2
vidualization of treatment for any given patient

will be influenced by sociological, economic RADICAL PALLIATIVE
Radiotherapy
and psychological factors as well as oncological Systemic therapy
principles. The three levels of decision-making DECISION 3 Surgery
used when formulating a treatment policy for
an individual are:
Primary Adjuvant
■ the decision to treat or not to treat Surgery Radiotherapy
Radiotherapy Systemic therapy
■ treatment intent, whether radical or pallia- Systemic therapy
tive, and
■ specific aspects of treatment policy regard- Figure 3.1 Treatment options.
ing local, systemic and supportive therapy.
Figure 3.1 illustrates the various options for
treatment.
Treatment options 23
TREATMENT OPTIONS they are almost certainly cured and require no

further treatment other than close follow-up
To treat or not to treat for a finite period.
Treatment should always have a positive ben-
Not every patient in whom a diagnosis of efit for the patient but treatment outcome for
cancer is made will benefit from active treat- any individual is not predictable. The decision
ment of their disease. There is, for example, to treat or not then becomes a matter of balanc-
good evidence to show that active local treating the probability of improving a patient’s con-
ment in the form of radiotherapy for inoperable dition, whether by symptom control with
carcinoma of the bronchus in poor performance palliative treatment or by cure with radical
status patients will have no impact whatsoever treatment, against the toxicity and disturbance
on the survival of a patient. It follows therefore, to lifestyle that treatment will entail. Thus it is
that in asymptomatic patients diagnosed with possible to justify an intensive course of
this condition, treatment will only be meddle- chemotherapy with major side-effects when
some and indeed may detract from that there is a high probability of cure but more dif-
patient’s quality of life by invoking side-effects ficult to do so in a patient with limited life
for no positive outcome. In contrast, treatment expectancy in whom only minor symptom
with radiotherapy or chemotherapy in good improvement can be anticipated. Wherever
performance status patients may result in a 2- possible, the patient should be allowed to deter-
month gain in survival; it is then a judgement as mine the level of input they wish to undergo;
to whether for an individual asymptomatic there is good evidence that, as a group, patients
patient this is appropriate treatment, and it is will often wish to undertake treatments with
important that those patients who develop significant toxicity and very limited chances of
symptoms from a bronchial carcinoma are not possible benefit in contrast to the views of
denied treatment by operating a blanket policy health professionals.
of no treatment for these patients. The actual
decision to treat a patient will be based not only Radical or palliative treatment
on their clinical state but also on the availability
of treatment facilities and the emotional Radical treatment is that which is given with
response of the patient and their relatives. It is the intent of long-term control or cure for the
often very difficult for a patient to accept that, patient. Palliative treatment is that which is
having been told they have cancer, no treatment given to improve the quality of life for a patient
is proposed other than symptomatic measures, with no implied impact upon their survival.
even though specific cancer treatment may While this may seem to be a clear distinction,
have no proven benefit. in practice it may be difficult to define treat-
A more comfortable scenario in which a no- ment aims in these terms. For example, a
treatment decision may be taken is when the patient presenting with small cell lung cancer
prognosis is so good and the risk of relapse so may be offered a course of radical treatment
small that treatment for all patients will result comprising intensive chemotherapy and per-
in overtreatment with consequent side-effects haps also chest radiotherapy. However, it is rec-
for the majority. An example of this situation ognized that, while this treatment may prolong
is the management of stage 1 testicular ter- survival, the likelihood of cure is less than 15
atoma following orchidectomy where the per cent. Thus the treatment intent, although
probability of relapse is around 20 per cent radical, will result for most patients in only very
and the use of tumour markers and scans limited benefit. Conversely a patient with
enables early diagnosis of relapse in a tumour breast cancer relapsing with a single site of
readily cured on exposure to appropriate painful bone metastases may receive a low pal-
chemotherapy. This is a much easier scenario liative dose of radiation for pain relief only, yet
for the patient, who will easily accept that live, pain free, for several years.
24 DECISION-MAKING AND COMMUNICATION
In most cases, however, the decision of treat- spread in a predictable fashion. In practice most
ment intent if not outcome will be clear. Those common tumours are thought to spread at an
patients who present with localized tumours early stage in their evolution through blood and
accessible to local therapy and those with lymphatic dissemination of tumour cells. For
metastatic disease from chemosensitive this reason the use of radical regional surgery is
tumours, such as germ cell tumour or lym- usually inappropriate. This is well illustrated in
phoma, will be offered radical treatment. Those the case of breast cancer where it is now clear
who present with metastatic disease from other that survival is largely independent of the type
tumours and those who relapse after primary of local treatment for a given stage of disease.
treatment will, with few exceptions, fall into Thus radical mastectomy is no better at curing
the palliative group. breast cancer than simple removal of the lump
Palliative treatment has the aim of improving from the breast.
a patient’s wellbeing usually through treatment Although local control of a tumour is an
for specific local symptoms such as pain, important goal, most patients die from cancer
obstruction or haemorrhage. It follows there- because of metastatic disease. Where a cancer
fore that patients in whom radical treatment is has been detected prior to the establishment of
not appropriate and who have no symptoms do metastases, then radical local treatment can
not require palliative treatment. The concept of result in cure. However, the natural history of
prophylactic palliative treatment, in other many cancers is such that even relatively early
words treatment to prevent symptoms emerg- tumours will already be associated with distant
ing, is for most patients inappropriate and, in micrometastases. In these cases improvements
the asymptomatic patient, introduces treatment in survival are likely to come only from the use
toxicity with little likelihood of benefit. There of adjuvant systemic treatment.
may be occasional exceptions to this concept,
for example the prophylactic fixation of a bone Adjuvant treatment
damaged by extensive osteolytic metastases
with impending pathological fracture. As always Adjuvant treatment is the prophylactic use of
the probability of treatment benefit must be local or systemic treatment following primary
weighed against the natural history of the con- treatment of a malignant tumour to prevent
dition and probability of disease-related and recurrence.
treatment-related symptoms. One of the most common adjuvant treat-
ments given today is postoperative radiotherapy
Local, regional or systemic treatment following excision of a malignant tumour, for
example irradiation of the breast following local
As a general principle, a primary malignant excision of an early carcinoma. Such adjuvant
tumour will require ablation with local treat- treatment may add significantly to patient mor-
ment, which may be surgical excision and/or bidity and it is important therefore to consider
radiation treatment. Similarly, metastatic dis- the relative merits of treatment in these situa-
ease requires systemic treatment, which may be tions. For example, the risk of local relapse in
chemotherapy, hormone therapy or a biological the breast following simple excision with no
agent. radiotherapy is around 30–50 per cent depend-
Local treatment may involve simple excision ing upon tumour size. On this basis, if all
of a tumour, removal of the entire organ or patients are treated following lumpectomy, half
removal of the involved organ and regional tis- may never have required treatment; the diffi-
sues at risk of tumour involvement. In the past culty lies in predicting accurately those who
there have been advocates of extensive regional will relapse. A further consideration is the fact
surgery around a tumour site in the hope of that local relapse following lumpectomy may
improving cure; however, such approaches are be treated successfully in many women and will
only rational where a tumour is known to still occur in 5–10 per cent even with radiother-
Quality of life 25
apy. A small survival advantage with postopera- It can be difficult to translate clinical trial and
tive radiotherapy has also been shown. The population-based data to the individual patient.
decision to offer a woman breast radiotherapy Breast cancer is a common cancer and therefore
following excision of a malignant breast lump even a very small treatment effect of, say, an
therefore has to balance the potential benefits improvement in survival of 5 per cent over 10
with the likely side-effects for each individual years will result in many thousands of women
patient. The substantial reduction in local worldwide living for longer after breast cancer.
relapse in the breast from 50 to 5 per cent will However, for the individual woman the odds
be seen in most cases to justify a relatively that she will benefit from adjuvant treatment in
simply, low morbidity treatment. that setting are small and it is possible that she
The use of adjuvant systemic therapy is also will undergo a toxic treatment with no effect
one in which the decision to offer a particular on her ultimate survival.
treatment must be balanced against the possible It is also difficult for those around the
acute toxicity which, with chemotherapy, may patient, be they family, friends or physicians, to
be significant. There are, in fact, few sites of evaluate and balance the risks of having a toxic
cancer where the use of adjuvant systemic ther- treatment against the risk of relapse, particu-
apy is of proven value. The area where there has larly when there can never be the certainty of
been the greatest endeavour and the most reli- cure. There is evidence to suggest that patients
able information is once again breast cancer. have a much lower threshold for accepting
There is now little doubt that an overall survival treatment, even where there is significant asso-
benefit is achieved by offering women with ciated toxicity, than would the nurses or doc-
early operable breast cancer some form of adju- tors who look after them.
vant systemic therapy. In postmenopausal
women tamoxifen or anastrazole is recom-
mended. This reduces the likelihood of dying QUALITY OF LIFE
from breast cancer by one-third and is a simple
treatment involving the administration of a Whilst survival will always be the most impor-
single tablet daily for a number of years with tant end-point of cancer treatment, the quality
few, if any, associated side-effects. Therefore the of life experienced and the impact of different
decision in this case is reasonably straightfor- treatment regimens upon this is an important
ward. In contrast there is a less reliable effect of further consideration. Measurement of quality
anti-oestrogen treatments in premenopausal of life has become increasingly sophisticated.
women with an oestrogen receptor-negative Early attempts relied heavily upon assessments
tumour, but proven advantage with chemother- by a physician on a broad scale measuring phys-
apy for those with high-risk tumours defined by ical performance status as illustrated in Chapter
high tumour grade or positive regional lymph 6 (see Table 6.2, page 55). A true measure of
nodes. However, this treatment involves 6 the quality of life for an individual can, how-
months of intravenous chemotherapy with ever, be acquired only from that individual, and
attendant side-effects. The greatest efficacy is a number of formal patient questionnaire-based
seen in those patients with positive axillary scales have now been developed. These include
lymph nodes, but even so the difference in sur- the Rotterdam symptom checklist, the Hospital
vival at 10 years between women who receive Anxiety Depression (HAD) score and the
chemotherapy and those who do not in European Organization for Research and Treat-
prospective randomized trials is only 6 per cent. ment of Cancer (EORTC) standardized quality
Therefore the treatment decision for adjuvant of life questionnaire (EORTC QLQ-C30), con-
chemotherapy in a premenopausal patient with sisting of a core questionnaire to which can be
negative lymph nodes in whom there may be added disease site-specific modules. The place
only a small probability of survival gain for the for quality of life assessments in clinical trials
individual becomes more difficult. will be discussed later in this chapter.
COMMUNICATION process of physical examination or while still

undressed.
Whilst the principles of communication in If possible and provided this is desired by the
oncology should be no different to any other patient they should have a relative or friend
branch of medicine there are specific features accompanying them or alternatively a health
that require consideration. In Western society professional with whom they have empathy, for
the term ‘cancer’ remains one that is frequently example their ward nurse. Despite inevitable
surrounded by fear and dread for most of the pressures on the time available for such inter-
population. In other cultures it may not even be views the patient should not be aware of time
mentioned in open discussion. Many patients pressures that may hinder their willingness to
passing through an oncology unit will have an interact and ask questions.
incurable and ultimately fatal disease. This
requires special skills in communication and Culture
support, which will span the entire range of
oncological practice. Radical treatment might There will be cultural differences in the way in
require lengthy discussion to consider the which a diagnosis of cancer, a potentially fatal ill-
choice between different treatment options; at ness, and death are approached. Increasingly in
the other extreme, specific skills are required in Western civilization a direct approach is pre-
imparting the news that a cancer may be incur- ferred. The word ‘cancer’ should not be avoided
able and death imminent. or hidden behind euphemisms. A clear and
Patients are increasingly well informed and simple explanation of the disease relevant to that
wish to discuss the details of their disease and particular patient and the treatment options
options for treatment. The diagnosis of cancer is should be given. Patients may be unfamiliar with
often associated with many emotional surgical procedures, and the processes of radio-
responses, which will colour the consultation therapy and chemotherapy; these will require
including anger, frustration, guilt and despair. additional explanation. It is unwise to assume
These may present considerable barriers to that a patient has any background information
open discussion particularly where there is a unless confirmed. It is often helpful early in a
perception that earlier medical care has been consultation to ask the patient to give their
inadequate or delayed the diagnosis of cancer. understanding of their illness or the treatments
Communication with a patient is affected by being offered, against which background further
many factors relating to both the patient and discussion can proceed. A particular pitfall to
the healthcare professional including environ- avoid relates to fellow health professionals who
ment, culture and content. may be disadvantaged by the assumption that
they already are familiar with their condition
and will have decided on the treatment they
Environment
wish to receive; usually this is far from the case
Wherever possible important discussions and they will welcome putting aside any back-
relaying news of a diagnosis of cancer, discus- ground knowledge they may have acquired for a
sions relating to treatment options and the full and simple explanation of their position.
‘bad news interview’ relaying news that the
condition may be incurable and terminal Content
should be carried out in an environment com-
fortable for the patient. Privacy is important Three common consultations can be identified
but often difficult to achieve in a busy ward in oncology that present specific challenges in
and wherever possible a dedicated side room communication; these are the choice between
or communication room should be available. different radical or palliative treatments, the
The patient should be adequately clothed and discussion of likely prognosis and discussion of
not given important information during the matters related to the process of dying.
Communication 27
Treatment choices viving, say, 5 years can be given. This should,

Where there is a choice of treatment, this however, be carefully interpreted since it is
should be laid out for the patient and in each often difficult for patients to relate life-table
case the possible advantages and disadvantages curves giving a probability of survival at any
discussed. It may be difficult to give a balanced point to their own individual survival.
view of treatment options particularly for spe- The point at which patients wish to discuss
cialists outside their own area of interest; this is such matters in detail will vary, some wishing to
the advantage of multidisciplinary discussions. consider this in detail from the outset but many
Realistic estimates of success should be given preferring to leave the matter unspoken for
with a clear explanation of the expected end- some time or on occasions throughout their ill-
point, whether ‘cure’, by which the patient will ness. This should be respected and the issue
interpret a complete return to normal life and dealt with sensitively, giving patients the oppor-
normal life expectancy; ‘control’ of the disease, tunity to ‘set the pace’, by asking specific details
which may be only temporary; or simply con- of their outlook, rather than giving bald and
trol of symptoms without any impact perhaps alarming statements of a short prognosis. Open
on survival. There is a recognized need to questions such as ‘Is there anything else you
clearly highlight possible side-effects of treat- would like to ask?’ may give them the lead to
ment, in particular those that may result in a pursue specific areas of concern when they wish
long-term alteration in lifestyle for the patient. to do so.
Again realistic estimates of their likelihood and Dying
the balance between side-effects and positive
benefit should be discussed. The mention of death and dying is often
It is well recognized that most patients do avoided by both carers and patients perhaps in
not retain information from their initial inter- an attempt to avoid facing the inevitability of
view for long and that a significant amount of this event. Meaningful discussions with a
the information will be forgotten or misinter- patient who has advanced incurable cancer can,
preted. It is therefore important that written however, only progress once the notion of
information is given to the patient to support dying has been accepted, if only to subse-
the facts imparted in the interview and that this quently agree not to mention it further.
is followed up by an opportunity for the patient Patients and their relatives will often hide
to return for further discussion. A readily avail- behind euphemisms such as ‘the end’ or ‘when
able point of contact should be given to the it’s all over’, whilst healthcare professionals
patient and often it is helpful for another may use terms such as ‘very serious’ or ‘incur-
healthcare professional such as a specialist able’ to introduce the idea that an illness is
nurse to be available for them to discuss matters fatal. Realistic expectations are important for
further. practical issues to be faced such as where and
by whom the patient wishes to be cared for in
his or her last days, clarification of their wishes
Prognosis in the form of a will and, where dependent
Prediction of life expectancy is notoriously dif- children are involved, arrangements for their
ficult and there is considerable evidence that future care. Patients and their relatives are
estimates by healthcare professionals are often often concerned over the process of dying
widely inaccurate. It is often difficult for rather than the event itself, and it is important
patients and their relatives to accept that their to explore these fears with appropriate reassur-
medical advisers cannot give an accurate pic- ance over the availability and efficacy of symp-
ture of the future. It is, however, usually possi- tom control and nursing care. Above all, it is
ble to estimate survival in terms of days, weeks, important not to avoid the issue when faced
months or years and, where lengthy survival is with direct questions from the patient but to
to be expected, a percentage likelihood of sur- reply honestly and with sensitivity.
Difficulties in communication bers. There is often guilt that the patient has
not been given sufficient attention, symptoms
Expectations versus reality have been missed and that more should have
For many patients the reality of a serious illness been done. This may be transferred to those
and the possibility of death is dealt with by looking after the patient as complaints relating
unrealistic expectations from their treatment to their diagnosis and care. Other more com-
and their carers. This can be projected as anger plex tensions may emerge with unmasking of
and dissatisfaction when treatment failure is relationship difficulties between partners or
encountered. It is important that unrealistic parents and children. There may be attempts to
expectations are not fuelled by unrealistic pre- ‘protect’ the patient by asking professionals to
dictions of outcome from treatment and that withhold information, which relatives may see
such responses are met sympathetically but as distressing. Throughout the important princi-
with clear and accurate information of realistic ple is that the patient is pre-eminent and, while
outcomes. although you must deal with relatives sympa-
thetically, the patient’s wishes should at all
Lack of knowledge times be respected.
Many patients may find it difficult to accept
that their advisers do not understand all the
details of their disease process and rarely can CONCLUSION
they attribute a cause to it. Patients often focus
on events in their life and seek confirmation Decision-making in oncology is not as straight-
that this was the possible cause for their cancer, forward as a simple knowledge of the disease
which it is not possible to confirm or refute. process may imply. Relating the large body of
Similarly events may occur in their disease knowledge regarding disease outcome, treat-
course that are unexpected and cannot neces- ment outcome and treatment-related morbidity
sarily be explained by previous experience. to an individual patient can be difficult. The
Healthcare professionals should not be reluc- probability of benefit from any treatment can
tant to admit that they do not know or under- be interpreted differently by each patient so
stand events and this is far better than that, while one might accept a 90 per cent
fabrication of possible mechanisms which have probability of cure as very favourable, another
no basis. will find a 10 per cent probability of relapse
unacceptable. Treatment decisions rely on an
Identification accurate and realistic presentation of the facts
but individual interpretations might come to
Patients who may be contemporaries of their very different conclusions from the same facts.
carers and children are particularly likely to
provoke strong emotional responses. This is a
normal reaction to events and should be recog- CLINICAL EVIDENCE AND
nized as such but not be allowed to cloud
objectivity. If this is found to be an overwhelm-
CLINICAL TRIALS
ing difficulty, healthcare professionals should Ultimately treatment decisions rely upon inter-
have access to an another colleague who can pretation of the available data for a given clini-
take their place or support them through diffi- cal situation. Different sources of data,
cult events. however, have different levels of reliability
when applying them to a patient population.
Conflict within families
Perhaps the least reliable, but often most mem-
The diagnosis of a life-threatening illness or orable, is that of the anecdote recalling the
impending death may unveil underlying con- course of a similar patient treated in the past.
flicts between a patient and other family mem- The most reliable data will come from the
Clinical evidence and clinical trials 29
results of a large randomized trial addressing a a population of patients. The characteristics of

specific question. Clinical evidence is graded the population are defined and patients fulfill-
according to its level of reliability as follows: ing these criteria are allocated to one of the
treatment options by a random process akin to
■ Level la – Meta-analysis of randomized tossing a coin, but more usually derived from a
controlled trials computer-generated list of random numbers. In
■ Level lb – Evidence from one or more ran- this way fluctuations in the population that
domized trial may affect the outcome of treatment are dis-
■ Level lla – Evidence from a non-random- tributed evenly across the treatment groups
ized trial allowing a true comparison of the treatment to
■ Level lll – Evidence from descriptive studies be made.
■ Level lV – Evidence from expert committee
■ Reports or clinical experience. End-points
A particular clinical question, whether in It is important at the outset of a trial to define

cancer treatment or any other speciality, can be the end-points by which the results will be
addressed in a number of ways. Types of trial judged. This will also guide the assessments that
include: will be required during the trial to provide the
necessary information to answer the main ques-
■ randomized controlled trials tions posed. End-points are often defined as pri-
■ case–control studies mary, reflecting the main question addressed by
■ cohort studies. the trial, and secondary, which will include
other important outcomes such as side-effects.
A clearly defined path for new treatment Common end-points will include the following:
development in the clinical setting is defined as
follows: ■ survival
■ disease-free survival
■ Phase 1 ■ response
– First exposure in man ■ toxicity
– Low doses initially based on animal data ■ quality of life.
– Dose escalation to maximum tolerated
dose (MTD) Survival is the commonest end-point for a
– Measurement of pharmacokinetics in man large trial comparing two or more treatments
– Definition of significant and limiting for cancer. Whilst apparently straightforward in
toxicities its definition time, cause of death may be diffi-
■ Phase 2 cult to trace, particularly in trials continuing for
– Testing for activity in a range of tumours many years and where the condition has a long
– Define response rates natural history, for example patients in trials of
■ Phase 3 prostate and breast cancer. It is important to
– Randomized controlled trial to compare define the cause of death. This will allow a com-
with best current treatment or placebo parison of not only overall survival but also dis-
■ Phase 4 ease-specific survival, i.e. counting only those
– Post marketing surveillance following patients dying from the disease under investiga-
successful phase 3 trial as use becomes tion. It is always important, however, to analyse
widespread outside trial setting. all causes of death, since this may on occasions
reveal an excess of deaths from complications
of the treatment. A typical example of this is
Randomized controlled trials the long-term analysis of the results of radio-
therapy for breast cancer, where a reduction in
These are statistically the most reliable way of breast cancer death rate is seen in patients
comparing two or more different treatments in receiving radiotherapy, but overall survival
between those receiving radiotherapy and those Response may be an important end-point
who did not is no different. The explanation for when testing new treatments to see whether
this apparent anomaly was explained by an they are effective. There are recognized formal
excess of non-cancer deaths, predominantly car- definitions of response. The most commonly
diovascular disease in the radiotherapy group used are the RECIST response criteria as fol-
which negated the reduction in breast cancer lows:
deaths.
■ Complete response (CR): complete disap-
Disease-free survival is defined by the period
pearance of all detectable disease for a
during which a patient remains in remission,
period of at least 1 month
without detectable disease, following treat-
■ Partial response (PR): a 30 per cent reduc-
ment. It may be further subdefined as ‘local dis-
tion in measurable disease for at least 1
ease-free survival’ taking into account only
month
relapse within a specified site, typically the pri-
■ Stable disease (SD): a reducing of <30 per
mary site of the cancer, after local treatment
cent in measurable disease or no increase in
such as surgery or radiotherapy, which would
size for at least 1 month
not be expected to directly influence other dis-
■ Progressive disease (PD): any increase in size
tant sites. ‘Distant disease-free survival’ may
of measurable disease during the assess-
also be used to assess a treatment for metastatic
ment period.
disease. Where this is used as an end-point in a
trial, it is important to account for the fact that Response may be quoted as overall reponse
relapse may only be defined by specific tests, rates (OR or RR), which will usually include
which will be performed at specific time points. CR + PR, but it is important to check on precise
For example if a patient is most likely to relapse definitions as some investigators will include
with lung metastases detected on a chest X-ray, stable disease in this category.
where the trial design defines a chest X-ray to Toxicity is a vital component in assessing the
be performed once a year, relapse can only be outcome of any clinical trial since ultimately
defined once a year, assuming the patient did any benefit will be balanced with side-effects.
not become symptomatic and obtain a chest X- Toxicity will include the immediate effects
ray for their clinical management in the mean- observed during the trial and may take the form
time. This would give only a very crude picture of laboratory measures, such as full blood count
of the pattern of relapse, and a delay in appear- or renal function, and symptom scores for
ance of metastases by less than 12 months expected side-effects of treatment. It is also
would not be detected; a design, which called important to consider later effects after treat-
for a chest X-ray every 2 to 3 months during ment has been completed, and specific issues
the period of risk, might be more suitable in that might arise from cancer trials include the
such a scenario. It is also important to realize effects of treatment on fertility and the induc-
that patients can only be compared by disease- tion of second malignancies.
free interval if they become disease free; for Quality of life measures will give a global
example, it is a very common end-point for view of the impact of the intervention to be
trials after local treatment of early breast cancer tested taking into account the response of the
where virtually all patients are ‘disease free’ disease, side-effects of treatment and also the
after treatment, but of little value in trials of more complex issues of the psychological and
treatment for metastatic disease where only sociological consequences of the disease and its
modest response rates might be anticipated. treatment.
In these patients a similar concept may, how- Measurement of end-points is a very impor-
ever, be applied using progression or symptom tant component of the design of any trial and
response rather than tumour response and requires a knowledge of the natural history of
measuring ‘progression-free survival’ or ‘symp- the disease being studied together with a
tom-free survival’. knowledge of the effects of the treatment inter-
ventions that will be available from preceding to use a single reference laboratory distant
use of the agents in phase 1 and 2 trials (see from the treating centre with mechanisms
below). There are two important considerations in place to notify investigators of clinically
in deciding how measurements should be relevant abnormal results.
included in a trial. ■ Measurement of symptoms is more com-
Timing must be defined to give useful infor- plex. There are a number of accepted and
mation but without burdening the patient or validated measuring tools for common
investigators with unnecessary measurements. problems such as pain or vomiting and
In the early part of a trial more measures are wherever possible these should be
likely than later on, the most intensive being employed rather than individualized scales
during treatment. However, as illustrated by the from a single trial or centre. These will typ-
example of annual chest X-rays in a disease that ically take the form of categorical scales
may relapse within a few months, appropriate based on none, mild, moderate, severe
time-points for assessment after treatment format or visual analogue scales, using a 10
must also be incorporated. The nature of most cm scale, on which patients mark the sever-
cancer is for these to be most intensive over the ity of a symptom. Critical to achieving good
first 2–3 years after treatment and then to data is careful explanation and continued
become less intrusive as the high-risk period for support during the data collection. Wher-
relapse passes. ever possible it is important to have data
Measurement may require a range of meth- completed by the patient themselves rather
ods to be applied: than a surrogate.
■ Quality of life instruments are detailed
■ Death is a clear end-point but tracing questionnaires relating to the patients emo-
patients after treatment to different centres tional and physical wellbeing. They can be
and defining cause of death can be more general or have additional ‘disease-specific
troublesome. In some countries cancer reg- modules’ addressing issues that may be per-
istries might be able to supply the informa- tinent to a particular tumour site, for exam-
tion independently. ple urinary and sexual function in
■ Clinical measurement of a tumour mass carcinoma of the prostate. Common exam-
may appear straightforward but clear ples used in cancer trials are the FACT and
parameters will need to be defined, for EORTC QLQ-C30 quality of life question-
example whether maximum diameter, min- naires.
imum diameter or a calculated area or ■ Health economic assessments are increas-
volume is to be used. Ideally measurements ingly an integral component of randomized
are made by independent observers or, if trials assessing new treatments as it
made by the investigators, then they should becomes apparent that the resources allo-
be blind to the treatment received. Trials cated to health care in the developed world
that do not include these safeguards could can no longer match the availability and
be open to bias. demand for increasingly complex and
■ Radiological measurements should similarly expensive treatments. Demonstration of
be carefully defined and ideally judged by cost-effectiveness is therefore an important
an independent radiologist or preferably a component of any clinical trial today evalu-
central panel convened for the trial. This ating new cancer treatments. The tools for
allows any variation between observers to evaluating this are also becoming more
be ironed out and greater uniformity in the complex as it becomes clear that a simple
observations will be achieved. balance sheet approach cannot address the
■ Laboratory measures are in general more issues that may arise. Attempts are made to
objective but validation of the laboratory relate the treatment effect with the length
methods used is important. It is preferable of time over which benefit may occur,
resulting in concepts such as the Quality trials there will be balanced randomization, i.e.
Adjusted Life Year (QALY) to measure the trial will comprise equal numbers in each of
treatment outcome. the arms of the trial. Occasionally, however, a
design can be used in which there is a weighted
Placebo-controlled trials randomization, for example a 2:1 distribution
Placebo-controlled trials will be appropriate of subjects, usually in favour of the experimen-
when a new treatment is being tested in a situa- tal arm, i.e. twice as many subjects in the exper-
tion where there is no recognized standard imental arm as the control. Randomization
treatment with which to compare it. Although should take place through an independent trials
it may be possible to compare the new treat- office totally separate from the facilities in
ment with no treatment there is always the pos- which the trial will be undertaken; usually this
sibility that simply the process of delivering a will involve computer-generated blocks of
treatment, whether a drug or other interven- random numbers, which will then allocate sub-
tion, will have an unexpected beneficial or jects to their arm of the trial. This may be
harmful effect. The use of a placebo, which will accessed by researchers through a telephone
be made to resemble the active treatment as call, fax or, increasingly, a voice-activated com-
closely as possible, reduces this possibility. puter program.
Placebos are most effective in trials that are
blinded. This refers to a situation in which Stratification
either the investigator or subject do not know Stratification can be built into the design of a
the difference between the placebo and active trial. This is to ensure that subgroups within the
drug (single blind design). The optimum design trial population, which may have different
is a double blind trial in which neither the prognoses, are equally distributed. In oncology
investigator nor the subject are aware whether a this might mean that trials are stratified by
placebo or active drug is being used. tumour stage or performance status. At ran-
Where there is already a recognized treat- domization each tumour stage or performance
ment against which the new treatment is to be status group will be independently randomized
tested then placebo-controlled trials are inap- ensuring an equal distribution of stages or per-
propriate and the control arm will be the stan- formance status across the arms of the trial.
dard best available treatment. Alternatively, This precaution also aids later analysis using the
particularly in testing new treatment in stage or performance status subgroups without
advanced cancer, the control arm could be best prejudice. Multicentre trials often stratify by
supportive care. This is optimized symptomatic treatment centre also.
supportive care distinct from the use of ‘anti-
cancer’ treatment and permits the test of a new Trial statistics
treatment in the setting of symptomatic, The major drawback of a randomized trial is
advanced malignancy to evaluate the addition that for a statistically robust result large num-
of an anticancer therapy where standard treat- bers of patients are required. The total number
ment is symptomatic care only. required will be defined beforehand by the trial
statistician who will be able to predict the
Randomization number needed to give a defined level of statis-
Randomization describes the method by which tical certainty for a given change in response
subjects in a trial are distributed into different rate from a known baseline. For example if a
treatment groups. Most trials have two ‘arms’: standard treatment has a known response rate
the standard arm, which should be the recog- of 70 per cent and 5-year survival of 45 per
nized best available treatment or, where there is cent, the trial investigators must decide what
none, then this will be a ‘no treatment’ or improvement upon this they would expect, or
placebo arm and the experimental arm in wish to detect, in their trial. They will then
which the new treatment will be tested. In most need to define the power of the trial, i.e. how
certain they wish to be of their result at the unit that will co-ordinate the trial and provide a
end. As an example, if a particular treatment central point for randomization, data collection
has a survival rate of 50 per cent, in order to and analysis. It should be independent from the
reliably detect an absolute improvement of 10– investigators entering and treating patients in
60 per cent from a new treatment, approxi- the trial who are usually based in many differ-
mately 400 patients will be required to enter a ent centres all accruing relatively small num-
randomized trial comparing the two. The pre- bers of patients. A randomized clinical trial may
cise number of patients will be affected by the take several years before it is completed and
degree of error in the trial results that will be analysed to give reliable results that will be
accepted by those designing the trial. There are translated into clinical practice. It is important
two types of error recognized: during the running of such a trial that any new
information from other trials that may affect
■ type I(α) error in which a difference may
the relevance of the outcome are considered
be observed which does not really exist;
and that any unexpected toxicity or difference
and
in response is evaluated within the trial. For this
■ type II(β) error in which a true difference
reason most large trials will have associated
may be missed.
with them a data monitoring committee
These will commonly be set at an α error of (DMC) usually comprising a statistician and
0.05 and a β error of 0.2 (often referred to as 80 two or three clinicians or scientists who are not
per cent power). This means that it will be involved in that particular trial. They will
accepted that there is a 5 per cent chance of the review the results of the trial at defined time-
result observed being false. As a general princi- points perhaps annually or after accrual of a
ple the more strict the statistical constraints, i.e. predetermined number of subjects, e.g. after
the smaller the α and β errors accepted and the every 100 patients entered. They may recom-
smaller the difference to be detected between mend early closure of a trial if a large number of
the arms of the trials, the larger the numbers unexpected toxicities are found in one arm, or
required. As an example, for a trial in a condi- if a sufficiently large difference emerges
tion where, with standard treatment, 5-year sur- between the arms of the trial early on that was
vival is 50 per cent, to detect a 10 per cent unexpected and unlikely to be reversed by
increase in survival with a new treatment in a larger numbers. The data seen by the DMC
two-arm randomized trial accepting an α error must at all times remain confidential from the
of 0.05 and β error of 0.2, then a total of 760 investigators involved in running the trial. In
patients will be required. Trials designed to show practice early closure is rare in a well-designed
equivalence, i.e. no difference between the clinical trial but the DMC is an important com-
treatment arms, require the highest numbers. ponent to ensure that patients are not given
The ‘p’ value is often quoted alongside results inappropriate treatment.
of a clinical trial. This essentially relates to the α
error accepted in the result. The conventional Ethics of clinical trials
value required before a result is considered sta- It is a fundamental principle in designing and
tistically significant is p = 0.05, which relates to partaking in a clinical trial comparing two treat-
an α error of 0.05 and means that there is a 5 ments that the patient should not be disadvan-
per cent chance of the results not being a true taged compared with other patients not taking
reflection of the comparison in the trial, or in part in the trial. For this to be satisfied the
other words there can be a 95 per cent certainty investigators must be confident that there is no
that the result is a true result. proven advantage for the new treatment over
the existing management and equally that there
Trial infrastructure is no evidence of any greater toxicity that
Clinical trials demand an extensive infrastruc- would disadvantage the patient. This will be
ture within a dedicated central clinical trials based on information gained from earlier phase
I and phase II trials. In fact it has been shown women that adjuvant tamoxifen conveys a 6.2
that patients taking part in clinical trials have a per cent survival advantage after 10 years and in
better outcome than those who do not, possibly 10 000 women that polychemotherapy carries a
because of adherence to a rigid protocol and the 6.3 per cent survival advantage after 10 years.
greater clinical input that inevitably accompa- These figures illustrate the very small overall
nies trial participation. survival effect seen and explain why series of
small trials gave conflicting results. With the
Meta-analysis large numbers in the meta-analysis, small differ-
A meta-analysis is a means of overcoming the ences can be found with a high level of statisti-
problem of patient numbers. This arises because cal reliability as demonstrated by the very low
most innovations in treatment have only a ‘p’ value associated with these data; the tamox-
modest impact upon outcome; an improvement ifen result above has a p value of <0.00001, i.e.
in survival of more than 10 per cent from any only a 1 in 10 000 chance that the observation
new intervention is very unusual. Other new is not real.
treatments might not even be expected to
improve survival but aimed at reducing toxicity Case–control studies
or improving quality of life. Any clinical obser-
vation is inherently unreliable in statistical In many situations, particularly in rare tumours
terms and will carry a range around which a where it is not possible to study large numbers
repeat observation may be expected to fall – of patients, randomized trials are not practical.
this is often quoted as the 95 per cent confi- The next best type of trial is the case–control
dence interval within which 95 per cent of trial in which a new treatment is given to a
repeated observations might be expected. The series of patients and then compared with pre-
smaller the number of observations the wider vious patients treated for that condition. It is
will be the confidence interval. To be certain important in such a comparison to exclude any
that the observations in two groups of patients factors that might affect the result other than
are truly different, there must be no overlap of the treatment being tested. This is achieved by
the confidence intervals. Where there is a big selecting matched controls, i.e. patients with
difference, even relatively wide confidence otherwise identical characteristics to one of the
limits will not overlap, but for a small true dif- new treatment group with whom they are
ference the confidence interval needs to be as matched. Parameters important to match for
small as possible to demonstrate the difference, include age, sex, tumour stage and histological
and this will only be achieved with a large type, together with any other known prognostic
number of patients. factors. The comparison may be refined by
A meta-analysis therefore attempts to selecting two matched controls for each patient
include as many patients as possible who have receiving the new treatment.
entered trials addressing a particular question. The difference between the two groups is
This approach has been used in several tumour often quoted as an odds ratio comparing the
sites of which breast cancer provides an excel- probability of an event (e.g. death or tumour
lent example. For a number of years a series of recurrence) occurring in the control group with
individual clinical trials across the world has that in the new treatment group.
addressed the question as to whether adding
hormone therapy or chemotherapy to the pri- Cohort studies
mary treatment can prolong survival. The
results of individual trials failed to give a clear Cohort studies are not usually of value in com-
answer, some showing benefit and others no paring two different treatments but are used in
benefit. A meta-analysis has therefore been per- particular to investigate aetiological factors. As
formed combining the results of all known trials their name suggests they are observational stud-
in this area and demonstrating in 30 000 ies in which a group of patients (a cohort) is
Further reading 35
monitored over a period of time for a particular that there are many pitfalls in obtaining and
outcome, e.g. development of or death from interpreting the results of clinical data sets. Any
cancer. This type of study is usually reported in advances in cancer management depend criti-
terms of relative risk, comparing the incidence cally upon rigorous evaluation with close atten-
in the cohort with that of a control cohort. The tion to the design and interpretation of clinical
value of this type of study is in defining the nat- trials to ensure that reliable outcome data are
ural history of a particular tumour type and acquired to further advance knowledge. It also
evaluating the impact of potential aetiological demands that clinicians and patients of the
agents. future are sufficiently informed to understand
and partake in clinical trials.
CONCLUSION
FURTHER READING
Subsequent chapters will describe for different
tumour types their aetiology, natural history Girling DJ, Parmar MKB, Stenning SP, Stephens RJ
and various treatment options. The accuracy of and Stewart LA. Clinical Trials In Cancer: princi-
this depends critically upon the data from ples and practice. Oxford University Press, Oxford,
which this information is derived. It can be seen 2003
PRINCIPLES OF
SURGICAL
4 ONCOLOGY
■ Management of the primary tumour 36 ■ Palliative surgery 38

■ Combined surgery and radiotherapy 37 ■ Further reading 40
■ Management of regional lymph nodes 37 ■ Self-assessment questions 41
The local treatment of malignant disease The type of surgery required will be deter-
involves the use of surgery, radiotherapy or a mined by the type of tumour and the anatomi-
combination of the two. Surgery is essentially a cal site, and may include:
local treatment while radiotherapy offers
■ wide local excision of the tumour mass, e.g.
locoregional treatment covering a wider area
local excision of a breast lump
less constrained by anatomical boundaries and
■ removal of part of an organ and surround-
surgical technique. Optimum initial manage-
ing tissue, e.g, partial glossectomy and neck
ment of the primary tumour and regional
dissection for a carcinoma of the tongue
metastases is vital if later relapse is to be
■ removal of an entire organ, e.g. laryngec-
avoided, and close liaison between surgeon and
tomy, cystectomy or hysterectomy.
oncologist is required to enable the best use of
each modality. En bloc removal of the immediate lymphatic
drainage areas is usually an integral part of any
MANAGEMENT OF THE cancer surgery, e.g. hysterectomy for cervical
cancer includes pelvic lymphadenectomy.
PRIMARY TUMOUR Radiotherapy will give equivalent local con-
trol rates to surgery for small tumours (<5 cm)
Surgery for a malignant tumour may have sev-
in many anatomical sites and has the potential
eral components:
advantage in certain sites of being able to pre-
■ tissue biopsy to establish the diagnosis serve anatomical structure and function, e.g. in
■ removal of malignant disease with a clear the treatment of cancer of the larynx where
margin of normal tissue radical radiotherapy can result in tumour eradi-
■ repair, reconstruction and restoration of cation with voice preservation in contrast to the
function. This may vary according to the surgical alternative, which is total laryngectomy.
extent of resection and anatomical site, However, against this must be balanced the
from simple primary wound closure to need for close postradiotherapy surveillance,
major reconstruction of bone and soft tissue which can be difficult if there are florid late
with vascularized grafts and prostheses. radiation tissue changes, and that a significant
Management of regional lymph nodes 37
TABLE 4.1 Relative merits of pre- and postoperative radiotherapy
Preoperative RT Postoperative RT
Early radiotherapy No delay to surgery
Enables preoperative preparation True pathological staging may be masked by preoperative RT
for planned surgery
Surgery may be easier if tumour Pathology of surgical specimen may guide later RT
shrinks preoperatively Lower dose of RT needed for microscopic disease
proportion (around 20 per cent) of patients will MANAGEMENT OF

still come to surgery. Furthermore, surgery fol-
lowing radiotherapy can be technically more REGIONAL LYMPH NODES
difficult and have a higher complication rate.
The common epithelial tumours, such as those
As with surgical management, irradiation of a
of the lung, breast, head and neck region and
local tumour must include areas of potential
pelvis, metastasize through two routes, the lym-
spread. A common approach to enable high
phatic system and the blood circulation. Clini-
radiation doses to be concentrated at the site of
cally involved lymph nodes and those at high
the tumour mass is to use a shrinking field tech-
risk of microscopic involvement require active
nique. This involves:
treatment.
■ initial treatment of a wide area to cover all
potential routes of microscopic spread Surgery
together with the primary tumour and
regional lymphatics Surgery is indicated for the immediate draining
■ a boost to the primary tumour site alone nodes around a primary tumour. Radical dissec-
with a margin of 1–2 cm. tion of the involved node chain should be con-
sidered, and may involve:
COMBINED SURGERY ■ axillary dissection for breast cancer

■ radical neck dissection for head and neck
AND RADIOTHERAPY sites
■ inguinal node dissection for vulval, anal
The combination of surgery with radiotherapy
or penile cancers treated with primary
has two potential advantages and applications:
surgery.
■ It enables the extent of surgery to be lim-
A more conservative approach, sentinel
ited by treating sites of microscopic disease
lymph node mapping, is now the standard of
immediately adjacent to the primary site
care for breast cancer and melanoma. This
with radiotherapy, e.g. the use of local exci-
entails identification of the ‘sentinel’ lymph
sion and radiotherapy in place of mastec-
node. This is the node that is first to receive
tomy for breast cancer.
lymph from the tumour-bearing tissue. A vital
■ For large tumours (>5 cm) local control
blue dye and radiolabelled colloid micro-
rates are in general better for combined
spheres are injected preoperatively to allow
therapy than either modality alone.
perioperative identification of the lymph
Combined treatment may involve the use of node(s) to be removed. If they are found to be
preoperative or postoperative radiotherapy. The involved by frozen section analysis or on defin-
relative merits of each approach are shown in itive histological examination, a full lymph
Table 4.1. There is no evidence that pre- or node dissection is mandatory. If this is negative,
postoperative radiotherapy is more effective in it is highly unlikely that other lymph nodes in
ultimate tumour control. that group will harbour cancer cells, and the
38 PRINCIPLES OF SURGICAL ONCOLOGY
(a) (b)
Figure 4.1 Pathological fracture of the right neck of femur (a) before and (b) after internal fixation.
patient may be spared the additional morbidity beyond the capsule of the gland. Prophylactic
of the larger operation. The technique can radiotherapy to sites of potential microscopic
allow identification of unusual patterns of lym- disease is often given when a primary site is
phatic spread, e.g. to the internal mammary being treated, e.g. pelvic nodes for gynaecologi-
lymph nodes in breast cancer, or the axilla in cal malignancy, neck nodes for head and neck
truncal melanoma. tumours.
Radical excision of nodes is not indicated for:
■ lymphomas (Hodgkin’s disease or non- Chemotherapy
Hodgkin’s lymphoma) or leukaemias
Chemotherapy might be indicated for enlarged
■ those with metastatic involvement at vis-
lymph nodes owing to germ cell tumour, lym-
ceral sites where there may be little or no
phoma, metastatic breast cancer or small cell
quality of life gain by removing enlarged
lung cancer.
lymph nodes.
PALLIATIVE SURGERY
Radiotherapy
Even when cure is no longer a realistic aim, sur-
Radiotherapy is indicated for surgically inopera-
gery can have an important role in the palliation
ble nodes and might succeed in rendering inop-
of local symptoms. Specific examples where
erable lymph nodes operable. Postoperative
surgery should be considered include:
radiotherapy following surgical dissection is
recommended for those patients with multiple ■ palliation of obstructive symptoms
node involvement and extension of tumour ■ control of haemorrhage
Palliative surgery 39
■ Intubation of the oesophagus with a rigid

tube or flexible stent provides rapid and
effective relief of dysphagia.
■ Nephrostomy or passage of ureteric catheters
will relieve obstructive hydronephrosis.
■ Biliary stents or choledochojejunostomy
will relieve obstructive jaundice owing to
extrahepatic bile duct obstruction as in car-
cinoma of the pancreas.
■ Gastroenterostomy will relieve gastric out-
flow obstruction.
■ Rectal stents can be used for rectal obstruc-
tion.
■ Ventriculoperitoneal shunting of hydro-
cephalus may result in dramatic improve-
ment of headache and neurological deficits
even though the underlying tumour may be
incurable.
Control of haemorrhage
■ Diathermy at bronchoscopy or cystoscopy
will control haemoptysis or haematuria
respectively.
■ Bleeding tumours in the oesophagus, stom-
ach or in the large bowel can be controlled
by using diathermy or laser coagulation at
endoscopy.
Figure 4.2 Uncontrolled bone metastases. This woman Palliation of tumour fungation
had internal fixation for a pathological fracture of the
mid-shaft of the humerus. This was not followed by Local resection, even if not complete, may be of
radiotherapy. She has now developed significant bone value for a locally advanced tumour mass that is
lysis compromising the mechanical integrity of the bone. necrotic and breaking down. Toilet mastectomy
for a progressive breast cancer is perhaps the
most common example of this.
■ palliation of tumour fungation Fracture reduction and fixation
■ fracture reduction and fixation.
Pathological fracture of a weight-bearing bone
Palliation of obstructive symptoms is best dealt with by internal fixation, as shown
in Figure 4.1, followed by postoperative radio-
■ Palliative resection of a bowel tumour or a therapy. Radiotherapy consolidates the surgical
simple bypass procedure such as a gastro- fixation, maximizing the probability of its secu-
jejunostomy will avoid symptoms of intes- rity in the longer term (Fig. 4.2). In certain cir-
tinal obstruction. cumstances prophylactic fixation may also be
■ Laser or cryotherapy resection of an indicated; specific indications for this include
obstructing tumour mass will restore the diffuse lytic bone disease in a weight-bearing
lumen of an obstructed bronchus or area and destruction of more than 50 per cent
oesophagus. of the cortex by a lytic deposit.
40 PRINCIPLES OF SURGICAL ONCOLOGY
FURTHER READING Sabel MS, Sondak VK, Sussman JJ. Essentials of Surgi-
cal Oncology. Mosby, London, 2007.
Pollock RE, Curley S, Ross M, Perrier N. Advanced Schauer A, Becker W, Reiser M, Possinger K. The Sen-
Therapy. In: Surgical Oncology. BC Decker, tinel Lymph Node Concept. Springer, Berlin, 2005.
Hamilton, Ontario, 2008.
Self-assessment questions 41
SELF-ASSESSMENT QUESTIONS
1. Which three of the following are indica- 2. Which one of the following is the most
tions for the surgical treatment of cancer? important feature of preoperative radio-
a. Long waiting time for radiotherapy therapy?
b. Fracture of a weight-bearing bone a. Allows treatment of regional lymph
c. Acute intestinal obstruction nodes
d. Hypercalcaemia b. Downstages disease to facilitate surgical
e. Anaemia resection
f. Hypoproteinaemia c. Removes need for radiotherapy after
g. Intestinal perforation surgery
d. Lessens morbidity from radiotherapy
e. Cost-effective
PRINCIPLES OF
5 RADIOTHERAPY
■ Types of radiotherapy 42 ■ Side-effects of radiotherapy 48

■ Other forms of radiation 44 ■ Radiation protection 49
■ Biological actions of ionizing radiation 44 ■ Further reading 50
■ Clinical use of radiotherapy 45 ■ Self-assessment questions 51
Radiotherapy is the use of ionizing radiation to structures such as ribs, scapula or sacrum but
treat disease. Ionizing radiation can be delivered are not sufficiently powerful to reach deeper
by X-ray beams, beams of ionizing particles, e.g. internal organs.
electrons, or by beta or gamma irradiation pro-
duced in the decay of radioactive isotopes. Megavoltage machines
These are either linear accelerators producing
TYPES OF high energy X-ray beams of 4–20 MV or cobalt
machines containing a source of cobalt-60,
RADIOTHERAPY which decays spontaneously to nickel-60
External beam radiotherapy releasing gamma rays of 1.2 and 1.3 MV. Linear
accelerators form the mainstay of modern clini-
External beam radiotherapy is the most cal radiotherapy and have largely replaced older
common form of treatment in clinical use. A cobalt machines in modern radiotherapy
range of X-ray beams is available, varying departments. A modern linear accelerator is
according to their energy. shown in Figure 5.1.
Electron beams are also produced by linear
Superficial voltage accelerators and have the advantage that, as par-
These X-ray beams are of energy 50–150 kV ticulate radiation beams, they have a defined
and are, as their name implies, suitable for the range in tissue with a sharp cut-off at the point
treatment of superficial lesions in the skin; their where they deliver their energy. This is of value
useful treatment energy penetrates no more where a tumour is superficial and particularly
than 1 cm beneath the surface. when it is overlying a radiosensitive structure
such as the spinal cord. In many centres elec-
Orthovoltage machines tron beam treatment has replaced older superfi-
These produce X-rays of energy 200–300 kV cial and orthovoltage X-ray machines, providing
and penetrate to a depth of approximately 3 a wide range of beams with varying effective
cm. These are therefore useful for treating treatment depths from the linear accelerator.
Types of radiotherapy 43
Live source implants

Direct handling of live sources is most com-
monly used today when iodine or palladium
seeds are used for prostate implants. While the
radiation from iodine-125 is short range and
poses no major radiation hazard, that from irid-
ium is more penetrating and its use has major
disadvantages in exposure of staff and patients
within the hospital to radiation from the time of
insertion to the time of removal, which may be
several days. This means that careful monitoring
is required and that there are limits to the time
Figure 5.1 A modern linear accelerator.
that individuals can spend in caring for the
patient. It also means that friends and family are
not permitted to be with the patient.
Brachytherapy Manual afterloading

In this method, inactive source carriers are used
This is the use of radioactive sources placed
initially to enable accurate siting of the treat-
either on or within a site involved with tumour.
ment so that the patient can be moved within
The great advantage of this form of treatment is
the hospital from theatre or the X-ray depart-
the rapid fall-off in dose at only a short distance
ment without radioactive sources in place. The
from the source (obeying the inverse square
active isotope is introduced manually once the
law). There are three types of brachytherapy in
inactive carrier is correctly placed. This enables
use:
more accurate placement not constrained by
■ Mould treatment This uses sources placed the need for rapid placement and long handled
directly over a superficial tumour of the instruments required when a live source is used.
skin fixed in a plastic mounting (the Loading can be done in the operating theatre or
mould). once the patient has returned to a protected
■ Intracavitary treatment This uses radioactive room and postoperative recovery is complete.
sources placed within a body cavity. An example is the use of iridium wire hairpins
Intrauterine tubes and vaginal sources in the for tongue and floor of mouth implants.
treatment of gynaecological tumours are a
common application of this technique. Remote afterloading
■ Interstitial treatment This uses radioactive Following placement of the source carriers, the
sources, which may be in the form of nee- radiation sources are introduced by remote con-
dles or wire inserted directly within the trol via pneumatic pipes or a cable connected to
area of interest. This technique is used for the source carriers within the patient. This
cancer of the tongue, floor of mouth and system minimizes exposure to staff and in prac-
breast. tice means that only the patient is exposed to
The first isotope to be employed was radium. radiation. It is also possible to interrupt the
This is no longer used because it constitutes a treatment so that the patient can receive atten-
major radiation hazard as it has a long half-life tion without further exposing hospital person-
(1620 years) and decays to a radioactive gas nel and there are no time limits on their stay. A
(radon). It has been superseded initially by modern afterloading machine is shown in
cobalt sources and currently by caesium and Figure 5.2.
iridium as the principal isotopes used for All such treatments require the patient to be
brachytherapy. isolated in a protected room with thickened or
44 PRINCIPLES OF RADIOTHERAPY
phosphorus (32P) for polycythaemia rubra vera

and strontium (89Sr) for bone metastases.
OTHER FORMS OF
RADIATION
There are types of radiation other than X-rays
or gamma rays that can be used therapeutically.
These are either radiation, which causes greater
biological damage along each path track it trav-
erses – high linear energy transfer (LET) radia-
tion – or beams that have a highly focused
deposition of radiation (Bragg peak).
Neutrons
Neutrons have been evaluated extensively in a
number of centres across Europe and the US.
Their advantages are that they have less
dependence on oxygen for cell killing, a recog-
nized limitation for X-rays, and they cause
more direct damage to the cellular DNA. While
Figure 5.2 A modern brachytherapy afterloading there is little doubt that they are more effective
machine, which uses a small iridium source located in at achieving cell kill, they are not selective for
the head of the machine that passes out by remote cancer cells and their use has been accompa-
control through a series of channels to deliver radiation nied by unacceptably high levels of normal
into applicators within a tumour area connected to these tissue damage and severe late side-effects. They
outlets. (Courtesy of Varian Medical Systems) are therefore not in routine clinical use.
shielded walls, floors and ceilings while the Protons

source is in position in the patient.
Protons are available in an increasing number of
Remote afterloading can deliver radiation at
centres around the world. Their advantage is
different dose rates but increasingly older
the production of a highly localized, high-
‘medium dose rate’ systems are being replaced
energy peak of energy deposition, which, by
by high dose rate machines, which deliver the
manipulation of the beam energy and the use of
dose in only a few minutes. This has substantial
absorbing materials, can be focused to a defined
advantages in radiation protection and means
position in a patient. Biologically they are simi-
that the patient no longer needs prolonged peri-
lar in action to X-rays. They have been used
ods of isolation to receive brachytherapy.
particularly for tumours in inaccessible sites
such as the back of the eye, pituitary and base
Internal isotope therapy of skull, and are being applied to many other
sites as experience and accessibility increases.
Internal isotope treatment involves the adminis-
tration of a radioactive isotope systemically,
which is then concentrated within the body at BIOLOGICAL ACTIONS OF
certain sites. Specific examples of this form of
treatment are the use of radioiodine (131I) for thy-
IONIZING RADIATION
roid cancer and also neuroblastoma when conju- Ionizing radiation causes damage to cellular
gated in meta-iodobenzyl guanidine (mIBG), DNA by two mechanisms:
Clinical use of radiotherapy 45
■ direct damage caused by the passage of the Particular care is required with certain normal
photon through the nucleic acid structure tissues when irradiation is delivered:
■ indirect damage owing to the production of ■ CNS tissue has a relatively low threshold for
toxic free hydroxyl radicals from the inter- damage and has little or no capacity for
action of radiation with water within the repair. Radiation damage to the CNS is
cell. This results in single and double strand therefore often irreversible and can result in
breaks in the DNA, which, unless repair catastrophic morbidity if, for example, necro-
occurs, will accumulate, resulting in repro- sis of the spinal cord or brainstem results.
ductive death of the cell. ■ The small bowel is also relatively sensitive
Other factors important in the response to and doses well below those required to ster-
radiation include: ilize an epithelial tumour will cause serious
damage.
■ Oxygenation Hypoxic tissues are considered ■ The lens of the eye will develop cataract
relatively radioresistant. Delivery of radio- after exposure to only small doses and must
therapy in multiple fractions allows reoxy- therefore be carefully shielded whenever
genation to occur between each treatment possible.
as the tumour shrinks and blood flow ■ Bladder and rectum are often dose-limiting
improves. tissues when pelvic tumours are treated.
■ Repopulation Both tumour tissue and ■ Lung damage occurs at around half the
normal tissues continue to divide during a dose required to treat a lung cancer;
course of radiotherapy and there is even damage to a large volume of lung will result
some evidence to suggest that repopulation in respiratory distress and even death from
may increase during this time as cells are respiratory failure.
lost. Gaps within a radiotherapy schedule The balance between causing serious normal
should therefore be avoided wherever pos- tissue damage and curing a malignant tumour is
sible to avoid significant repopulation termed the ‘therapeutic ratio’. Compromise of
undoing any effects from irradiation in the radiation dose to avoid excessive toxicity is the
previous days. major limitation in successfully achieving local
■ Repair Much of the damage produced by cure of malignant tumours with radiotherapy.
X-rays and gamma rays is not sufficient to
kill the cell immediately (sublethal
damage). Repair of such damage will mean CLINICAL USE OF
that the cell retains its viability. This capac-
ity is better developed in normal tissues
RADIOTHERAPY
than in many malignant tumours, which The patient attending for radiotherapy treat-
accounts for much of the differential cell ment will pass through a series of steps to
kill between tumour and normal cells. ensure that treatment is given as accurately and
■ Redistribution Radiosensitivity varies as safely as possible. These include:
cells progress through the cell cycle, being
maximum during the periods of active ■ patient positioning and immobilization
DNA synthesis in late G1 and S phases and ■ tumour localization
least during G2 and early M phases. ■ treatment planning
■ verification
There is a spectrum of radiosensitivity for ■ treatment.
both tumour and normal cell types. Few, if any,
tumours are truly radioresistant, although some Positioning of a patient
may require a larger dose of radiation than
others to achieve the same effect. There is also a This is very important in order to enable an X-
wide spectrum of sensitivity in normal tissues. ray beam to reach a certain site. It is particularly
for tumours in the thorax, abdomen or pelvis.

Alongside this, it is important to identify vital
structures to be avoided such as the spinal cord
or kidney.
Treatment planning
This defines the optimal arrangement of X-ray
beams to cover the treatment area as evenly as
possible while avoiding structures around it. At
its simplest level this may be a single direct
electron beam for a skin tumour; for internal
tumours a more complex arrangement of two,
three or four X-ray beams approaching from
Figure 5.3 Patient in position for treatment to the orbit
demonstrating immobilization mask. different directions may be required. Treatment
plans of multiple beams are drawn up using
computerized data describing the characteris-
tics of the beam and enabling rapid addition of
the case in the head and neck region where their effects at a point.
small changes in the position of the neck or Conformal radiotherapy using external X-ray
chin can greatly affect the tissues included in an beams shapes the high-dose treatment area as
X-ray beam. closely as possible to that of the tumour
In order to reproduce a particular position volume, even where this may be a highly irreg-
accurately, some form of immobilization may be ular shape, while avoiding adjacent sensitive
designed, such as a plastic head shell to hold the normal structures. This requires accurate three-
head in a fixed position as shown in Figure 5.3. dimensional reconstruction of the volume to be
treated and irregular beam shaping using lead
Tumour localization blocks, or more commonly a multileaf collima-
tor shaping the beam typically by 80 or 120
Localization can be achieved by simple clinical interleaved projections into the beam, which
examination as in the case of a skin tumour; can be adjusted as required. An example is
plain X-rays as in bone metastases; or a CT scan shown in Figure 5.4.
Figure 5.4 On left, beam’s eye view

(BEV) of planning scan pictures from a
radiotherapy plan to treat the pelvis,
demonstrating position of the beam
and shaping using the multileaf
collimator. On the right, an electronic
portal image (EPI) has been taken
using the high energy linear
accelerator beam prior to a treatment
exposure to verify the beam size,
shape and position. Note the loss of
definition between bone and soft
tissue characteristic of high energy
(megavoltage) X-rays.
Clinical use of radiotherapy 47
Intensity-modulated radiotherapy (IMRT) is a tube on the gantry, which enables cross-

further development, which varies the dose sectional imaging to be acquired at the time of
delivered within a tumour volume. This means treatment. This has the advantage of providing
that a central high-dose region can be defined information on the soft tissues within the beam
surrounded by a lower dose region, minimizing to be exposed, which cannot be identified from
inclusion of normal tissues, all within the same conventional beam verification images.
treatment volume. Treatment only proceeds once it is certain
Stereotactic radiotherapy is a means of treating that the defined area to be treated will be
a small spherical volume to a high dose and is encompassed using the planned X-ray beams.
typically used for localized lesions in the brain. During treatment the patient will be continu-
This can be achieved using a standard linear ally monitored by the radiographers and seen
accelerator with a modified beam producing a regularly by the medical staff. In a radical treat-
small focused field, which builds up the high- ment, verification images will typically be
dose volume using a series of rotating arc move- repeated daily for the first 3 days and then once
ments. Modern developments of this include a week to ensure the beam positioning is consis-
tomotherapy and the ‘cyberknife’, which incor- tent throughout. Specific additional measure-
porate a megavoltage X-ray beam in a gantry ments of radiation exposure can be taken using
capable of multiple arcs controlled by software dosimeters placed on vital areas, such as the eye
that enables variable beam profiles during the or testes, and further verification X-rays taken
arc to build up a complex localized dose distri- during the course of treatment.
bution. The alternative approach is to use the Image-guided radiotherapy is the latest devel-
Gammaknife®. This is a machine containing opment to ensure that the radiation beam fol-
multiple cobalt sources each with its independ- lows as closely as possible the planned volume
ent shield which can be opened to focus on the and avoids normal structures. This has evolved
treatment volume in a spherical pattern. Such with the increasing recognition that, during a
machines are only available in a few specialized few minutes of radiation exposure and between
units. daily exposures, there may be critical changes,
for example bladder filling, rectal and bowel dis-
Verification tension, respiratory and cardiac movements,
which will alter the anatomical relations of the
Verification entails ensuring that the treatment treated radiation volume. The use of imaging
plan can be translated back to the patient in the immediately prior to exposure with markers,
defined treatment position. This is usually done such as small metallic clips or gold grains fixed
on a machine called a treatment simulator, in the tissue, will allow adjustments to the beam
which reproduces precisely the movements of to be made immediately prior to radiotherapy
the treatment machine but produces only a being delivered. Complex techniques for expo-
diagnostic X-ray beam so that an X-ray picture sure to be matched to phases of the respiratory
of the proposed treatment beam can be taken. cycle, termed gating, are also available to opti-
Further verification when the actual treat- mize delivery to tumours in the lung.
ment starts is essential, particularly if complex
beam shapes or lead shielding are to be Treatment duration
employed. Megavoltage images can be taken
using the linear accelerator beam. Their disad- There is considerable variation in the dose and
vantage is that, because they interact in a differ- duration of radiotherapy treatment given to
ent manner to low-energy X-rays, there is no patients, which can appear confusing. Neither
differential absorption between bone and soft the total number of treatments nor the total
tissue, resulting in poor anatomical definition; dose given are necessarily a guide to the biolog-
this is shown in Figure 5.4. Modern linear accel- ical dose of radiation delivered. There are three
erators also incorporate a kilovoltage X-ray components to biological radiation dose:
■ total dose this routinely, while others treat over 1–2

■ number of treatments (fractions) of a given weeks. Examples include:
dose
■ 8–10 Gy in a single dose
■ overall time of treatment.
■ 20 Gy in 5 daily fractions over 1 week or
The following general principles can be 30 Gy in 10 daily fractions over 2 weeks.
applied:
■ The same total dose given in a short time or SIDE-EFFECTS OF
fewer fractions has greater effect than the
same dose given over a longer time in many
RADIOTHERAPY
fractions. The toxicity of radiotherapy is divided into two
■ Fraction size is an important determinant of distinct groups: the early effects, which occur
normal tissue damage: small fractions mini- during treatment, and the late effects, which
mize normal tissue effects. However, it fol- come on months or years following treatment.
lows from the previous point that, if used,
small fractions demand a longer course of
treatment with a higher total dose to have Early effects
the same effect. These develop during treatment usually in the
■ Overall treatment time is important in second to third week of a course of radical irra-
achieving tumour control. Delays and inter- diation. They may be divided as follows:
ruptions in a course of treatment should be
avoided as should overall treatment times ■ Non-specific effects Many patients feel tired
of greater than 6–7 weeks. and lack energy during treatment. There
may be many factors in addition to radia-
The unit of radiation dose used is the Gray tion exposure to account for this, including
which is a measure of absorbed energy (1 Gy = depression, anxiety, travelling daily to treat-
1 joule per kilogram). This has replaced the rad ment and concomitant medication.
but the conversion is simple: 1 Gy = 100 rads. ■ Specific local effects related to the area being
In some centres in order to keep the actual treated It is important to note that areas
numbers the same, doses are described in outside the irradiation field do not exhibit
centigray since 1 cGy = 1 rad. acute toxicity. Examples are shown in Table
Precise radiation schedules vary from centre 5.1 together with suggested treatment. As a
to centre and between individual clinicians, general principle, these are all self-limiting
however, some generalizations can be made. A effects, which resolve spontaneously after
fundamental difference lies between palliative treatment, their pathogenesis being related
and radical doses. to temporary loss of cell division at an
Radical treatments require high doses of radi- epithelial surface.
ation to eliminate tumour but are divided into
smaller fractions to minimize side-effects and
remain within the tolerance of normal tissues. Late effects
Examples of radical schedules include: These are potentially the most serious effects of
■ 60–70 Gy in 30–35 fractions in 6–7 treatment since, unlike the acute effects, they
weeks are not self-limiting and indeed tend to be pro-
■ 55 Gy in 20 fractions in 4 weeks or 50 Gy gressive and irreversible. They arise owing to
in 15 fractions in 3 weeks. loss of stem cell recovery potential and progres-
sive damage to small blood vessels resulting in
Palliative treatments require short schedules their occlusion (endarteritis obliterans).
with few acute side-effects. The ideal palliative Fortunately in practice they are rare but any
treatment is a single dose and many centres use radical treatment dose will carry a risk of late
Radiation protection 49
TABLE 5.1 Acute effects after radical radiotherapy
Site Effect Treatment

Skin Erythema leading, if severe, Minimal; avoid irritants, trauma
to desquamation Aqueous or weak hydrocortisone cream
Bowel Diarrhoea/colic Low residue diet
Codeine or loperamide
Bladder Frequency/dysuria Exclude infection
Scalp Hair loss Order wig in advance; hair will regrow after
palliative but not radical doses
Mouth/pharynx Mucositis Avoid irritants and alcohol; treat candidiasis;
topical chlorhexidine or benzydamine
TABLE 5.2 Late effects after radical radiotherapy
Site Effect Treatment

Skin Fibrosis; telangiectasia; rarely necrosis Usually none
Bowel Stricture; perforation; bleeding fistulae If severe, resection of affected segment
Bladder Fibrosis causing frequency; haematuria; fistulae If severe, surgical resection
CNS Myelitis causing paraplegia; cerebral necrosis None
Lung Fibrosis None
damage becoming manifest in the months and ments in whom the risk of a second tumour
years ahead. They are not usually seen before 6 after 20 years is around 15 per cent.
months after treatment but there is an ongoing
risk, which is never entirely lost. Examples of RADIATION PROTECTION
late effects are given in Table 5.2.
Finally, the risk of inducing second malignancy Indiscriminate use of radiation is dangerous.
should be considered. In practice this is exceed- There are strict regulations surrounding its use
ingly rare and invariably outweighed by the risk for medical purposes to ensure that exposure to
from the established malignancy for which radia- staff, visitors and patients is kept to an absolute
tion is given. A clear pattern is recognized with minimum. This is achieved by physical separa-
leukaemia or lymphoma seen in the first few tion and protection from the sources of radia-
years after exposure, reaching a peak incidence tion and subsequent monitoring of exposure in
around 3 years and not seen after 10 years. In those at particular risk.
contrast solid tumours have an increasing risk
with time, are rarely seen before 10 years and Separation and protection
may occur 30 years or more after exposure.
The greatest risk may be associated with low- Radiotherapy machines are localized in one
dose exposure and there are many historical area or even in a separate hospital. Their design
examples after treatment of benign disease such incorporates shielding of scattered radiation to
as tinea capitis, goitre and ankylosing spondyli- produce a defined beam, the penetration and
tis. After therapeutic radiation the risk is small qualities of which are carefully measured and
but there is some evidence that this may be maintained. They are housed within a room
increased in patients who receive combined designed to contain the radiation, using lead
modality treatment with the addition of barriers within the walls for low-energy beams
chemotherapy. This has recently become appar- and thick high-density concrete walls for high-
ent in long-term survivors of lymphoma treat- energy beams.
Similarly, radioisotopes are stored under tions in context, while not relaxing the rules
carefully controlled conditions in a radiation governing its use. There is no evidence that hos-
safe and administered to the patient in a desig- pital workers are at greater risk than the general
nated area, again designed to contain any radia- population from malignant disease. The entire
tion exposure. Patients having implants or high population is exposed to levels of background
doses of radioiodine will be isolated in single radiation and many other common daily activi-
rooms that have additional shielding to prevent ties carry a greater risk than low-level exposure.
radiation reaching other areas of the ward. Staff For example, it has been estimated that all of
are given strict time limits to be spent with the the following activities carry a one in a million
patient. Visitors are usually not permitted while risk of death:
there are high levels of radioactivity present.
■ driving for 65 miles
■ flying in civil aircraft for 400 miles
Radiation monitoring ■ smoking less than one cigarette
■ drinking half a bottle of wine.
All staff involved in the direct care of patients
receiving radiation treatment will be designated
as workers who require regular monitoring for FURTHER READING
radiation exposure. The mainstay of monitoring
is the film badge worn by these staff, which, Bomford CK, Kunkler IH. Walter and Miller’s Text-
when processed, will detect those who may book of Radiotherapy, 6th edn. Churchill Living-
have been inadvertently exposed to radiation. stone/Elsevier, London, 2002
There are clearly defined exposure limits for Hoskin P (ed.). Radiotherapy in Practice: external beam
adults within which all personnel must remain. therapy. Oxford University Press, Oxford, 2007
All departments must have local rules regarding Hoskin P, Coyle C (eds). Radiotherapy in Practice:
the handling and use of radiation sources and brachytherapy. Oxford University Press, Oxford,
staff must be fully acquainted with these. 2005
It is important to keep the dangers of radia- Joiner M, van der Kogel A. Basic Clinical Radio-
tion as used under controlled medical condi- biology. 4th edn. Arnold, London, 2009
Self-assessment questions 51
1. Which of the following best describes 5. When the clinical use of radiotherapy is
radiotherapy? considered, which of the following is
a. It uses non-ionizing radiation correct?
b. It is a useful systemic treatment for cancer a. Palliative treatments will require high
c. It can use beams of ionizing particles total doses
d. Toxicity is usually limited to acute b. Radical treatments are best given in
effects during treatment large daily fractions
e. It is suitable for tumours at any site c. A typical radical dose would be 30 Gy
in 10 daily fractions
2. Which of the following applies to d. Single doses are the best palliative
megavoltage X-ray machines? treatment if effective
a. They are useful for diagnostic X-ray e. Radical doses work best when delivered
production slowly over a long time
b. They are usually portable
c. They can be used to produce neutrons 6. Which three of the following are
for superficial radiotherapy recognized acute effects of radiotherapy?
d. They can have a radioisotope for their a. Constipation
radiation source b. Peripheral neuropathy
e. They are commonly used for c. Skin erythema
brachytherapy d. Cataract
e. Dysphagia
3. Which three of the following are important f. Dysuria
in modifying the response of radiotherapy g. Lymphoedema
in the cell? 7. Which three of the following are
a. White cell count during irradiation recognized late effects of radiotherapy?
b. Oxygen levels during irradiation a. Peripheral neuropathy
c. RNA repair after irradiation b. Skin fibrosis
d. Cell repopulation after irradiation c. Cataract
e. Phase of cell cycle during irradiation d. Constipation
f. Water content of the cell during e. Dysuria
irradiation f. Lymphoedema
g. Interstitial pressure during irradiation g. Fatigue
4. Which of the following statements is 8. With regard to radiation protection, which
correct regarding radiation dose? of the following is correct?
a. Total dose alone will indicate the a. There is no significant risk from doses
biological effect used in clinical practice
b. The same dose given over a longer time b. High energy beams are shielded behind
is more effective lead screens
c. A dose is most effective given in c. Patients receiving radioisotope therapy
multiple small doses (fractions) can be visited regularly
d. Large doses per fraction are more d. Hospital workers have higher rates of
effective than smaller fractions exposure than the public
e. The overall time is more important than e. The entire population is exposed to
the dose per fraction radiation
PRINCIPLES
OF SYSTEMIC
6 TREATMENT
■ Chemotherapy agents 52 ■ Hormone therapy 61

■ Efficacy and toxicity of ■ Biological therapy 63
chemotherapy 54 ■ Growth factors 64
■ Clinical use of chemotherapy 56 ■ Experimental chemotherapy 65
■ Drug resistance 56 ■ Further reading 65
■ Administration of chemotherapy 57 ■ Self-assessment questions 66
Cancer chemotherapy is the treatment of ■ Signal transduction inhibitors

malignant disease with drugs rather than radia- ■ Drugs inducing apoptosis
tion or surgical removal. The drugs used are ■ Drugs targeting the tumour vasculature
often highly toxic since they are rarely totally – antiangiogenesis
selective for cancer cells. They should therefore – vascular disrupting agents.
be given within an oncology unit by those expe-
rienced in their use. Drugs acting on the structure of DNA
Systemic therapy can be used alone or, more
commonly, in combination with surgery or radio- Antimetabolites
therapy as an adjuvant to enhance local control These function at the level of DNA synthesis,
and to attack potential sites of metastases. interfering with the incorporation of nucleic
acid bases (cytosine, thymine, adenine and gua-
nine). These can be classified into two groups:
CHEMOTHERAPY
■ Drugs based on chemical modification of a
AGENTS nucleic acid so that the drug rather than
the true nucleic acid is incorporated into
Cancer chemotherapy agents act upon cell divi-
the DNA, thereby preventing accurate
sion, interfering with normal cell replication.
replication of the complementary base
They can be broadly classified as follows:
sequence. 5-Fluorouracil (5FU) is the clas-
■ Drugs acting on the structure of DNA sical example now available as an oral pro-
– antimetabolites drug capecitabine. Others include
– alkylating agents gemcitabine, cytosine arabinoside and flu-
– intercalating agents darabine. These drugs may also have addi-
– topoisomerase inhibitors tional modes of inhibiting DNA synthesis,
■ Drugs acting on mitosis e.g. the inhibition of thymidine synthetase
Chemotherapy agents 53
by 5FU and of ribonucleotide reductase by Drugs acting on mitosis

gemcitabine.
■ Drugs which inhibit reduction of folic acid Mitosis requires spindle formation essential in
(essential for the transfer of methyl groups the sorting and moving of chromosomes fol-
in DNA synthesis) from its inactive dihy- lowing replication at the end of mitosis. Spin-
drofolate form to the active tetrahydrofo- dle formation is affected through two
late, e.g. methotrexate. mechanisms:
■ spindle poisons: the vinca alkaloids (vin-
Alkylating agents
cristine, vinblastine and vindesine)
These directly interfere with the DNA double ■ tubulin polymerization resulting in abnor-
strand of base pairs by chemically reacting with mal spindle formation and thereby cell
the structure, forming methyl cross-bridges. death. The taxane group of drugs (pacli-
These then prevent the two DNA strands taxel and docetaxel) are cytotoxic through
coming apart in mitosis to form daughter DNA this mechanism.
fragments, and division therefore fails. Drugs in
this group include cyclophosphamide, chloram- Microtubule formation is also affected by
bucil, melphalan, mitomycin C and the podophyllin and so etoposide and teniposide
nitrosoureas BCNU and CCNU. both inhibit spindle formation in addition to
their effects through topoisomerase.
Intercalating agents
These act in a similar way to the alkylating
agents but, rather than directly forming cross- Signal transduction inhibitors
strands in the DNA molecule, they bind Communication within and between cells is
between the base pair molecules, i.e. binding mediated by extensive biochemical interactions
adenine to thymine and cytosine to guanine. triggered by cell surface receptors. Tyrosine
This again prevents the DNA double strand kinases are a frequent component of these cas-
from dividing in order to replicate, preventing cades, particularly those related to the epider-
cell division. Platinum compounds (cisplatin, mal growth control receptor (EGFR) and
oxaliplatin and carboplatin) act by binding angiogenesis (VEGF and PDGF). They are
specifically to guanosine, forming DNA adducts therefore a potential target for new drug devel-
that cross-link either within one DNA strand or opment. Drugs which act primarily through
across strands. Other compounds, active inhibition of tyrosine kinases include imatinib,
through DNA intercalation, are the anthracy- gefitinib, sorafenib and sunitinib.
cline group of drugs (Adriamycin, epirubicin
and idarubicin).
Bleomycin is not a true intercalating agent Drugs inducing apoptosis
but does partially intercalate with DNA and
produces direct DNA damage through forming A new class of drugs, proteosome inhibitors,
a complex with iron, resulting in the produc- has emerged, which enhance apoptosis in a
tion of reactive toxic products. population of tumour cells. Proteosomes are
large protein complexes within the cell, which
Topoisomerases inhibitors regulate the degradation and removal of pro-
Topoisomerases are enzymes that control the teins. Inhibitors of proteosomes have been
tertiary coiling of DNA molecules. Two main shown to enhance apoptosis by disrupting the
classes are recognized: topoisomerase I and II. ordered degradation of proteins controlling the
The camptothecin group of drugs including cell cycle. The most successful agent in this
topotecan and irinotecan are topoisomerase I group is bortezomib which is now the recog-
inhibitors, and the podophyllotoxins, etoposide nized treatment for recurrent multiple
and teniposide, are topoisomerase II inhibitors. myeloma.
54 PRINCIPLES OF SYSTEMIC TREATMENT
Drugs targeting the tumour lites during S phase and vinca alkaloids in
vasculature M phase
■ cycle-specific drugs, which require that cells
New blood vessel formation is an essential pre- are only dividing and passing through the
requisite for a tumour to become established cell cycle rather than being in the resting
and remain viable. VEGF is an important com- G0 phase. These include the alkylating
ponent in this process. Tumour blood vessels agents and intercalating agents.
differ from normal vasculature forming a
random chaotic network with little structure
and variable flow controlled primarily by fluctu- EFFICACY AND TOXICITY
ations in interstitial pressure. Both the formation
of blood vessels (angiogenesis) through VEGF
OF CHEMOTHERAPY
and the new blood vessels themselves are there- Although in principle all cycling cells should be
fore potential targets to inhibit tumour growth. sensitive to drugs that act on the cell cycle, in
Antiangiogenic drugs include bevacizumab, practice chemotherapy for most cancers is only
which is a monoclonal antibody to the VEGF modestly effective. This is due to a variety of
receptor, and sorafenib which inhibits signalling reasons related to drug delivery to the cell and
from the VEGF receptor through tyrosine activation or deactivation within the target
kinase inhibition. Thalidomide and lenalido- cells. Precise indications for the use of
mide, both used in multiple myeloma, also have chemotherapy are detailed in the following
antiangiogenic activity, although their main chapters but in general the common malignant
mode of action may relate to inhibition of IL-6 diseases can be broadly classified into three
and immunomodulatory effects. groups:
Vascular disrupting agents (VDAs) target the
stucture of the blood vessels and prevent their ■ those extremely sensitive to chemotherapy
proliferation. Examples include combretastatin- where this is the treatment of choice
4 phosphate and its analogues. ■ those with modest sensitivity where
Because of their mode of action, certain chemotherapy may play a part in their
drugs require cells to be in specific phases of the management but usually in combination
cell cycle to have any effect. For example, those with other treatment as an adjuvant or for
acting on synthesis of DNA will act only on recurrent disease
cells actively synthesizing at the time they are ■ those where chemotherapy is only of lim-
exposed to the drug. Cytotoxic drugs are there- ited value in the palliation of advanced dis-
fore further classified into: ease.
■ phase-specific drugs, which act only in a spe- Examples of this classification are given in
cific phase of the cell cycle, e.g. antimetabo- Table 6.1.
TABLE 6.1 Classification of common malignant diseases according to their sensitivity to chemotherapy
Group 1: High sensitivity Group 2: Modest sensitivity Group 3: Low sensitivity

Leukaemias Breast Prostate
Lymphomas Colorectal Kidney
Germ cell tumours Bladder Primary brain tumours
Small cell lung cancer Ovary Adult sarcomas
Myeloma Cervix Melanoma
Neuroblastoma
Wilms’ tumour
Embryonal rhabdomyosarcoma
Efficacy and toxicity of chemotherapy 55
Response and survival plete response in a significant proportion of

patients, it is unlikely to have any impact on
In describing the efficacy of chemotherapeutic overall survival when used to treat that disease
agents, there are internationally accepted in the general population. Partial responses may
response criteria, the RECIST (Response Evalu- be of value in delaying progression of a disease
ation Criteria in Solid Tumours) criteria (see but, since they represent only a very small
also page 30), defined as follows: decrement in total cancer cell burden, signifi-
■ Complete response (CR) – complete reso- cant effects on overall survival are unlikely.
lution of all clinically detectable disease With this in mind, performance status and
■ Partial response (PR) – a reduction in the quality of life measures are becoming an
longest diameter of a tumour of at least 30 increasingly important measure of the efficacy
per cent of cancer treatment. Numerous scales of per-
■ Stable disease (SD) – small changes with a formance status have been devised, usually
response less than a partial response or pro- based on a 4 or 5 point scale ranging from
gression less than a 20 per cent increase in normal to moribund. A common example is the
measurable disease during the period of WHO performance scale shown in Table 6.2.
observation Quality of life measures are more complex;
■ Progressive disease (PD) – an increase in they should be completed by the patient and
measurable disease of at least 20 per cent will involve a structured questionnaire often
during the period of observation, or the divided into domains evaluating different
development of any new lesions. aspects of the patient’s activity, e.g. physical
activity, pain and specific symptoms, emotional
It is important to interpret with care response, social interaction and financial
response data derived from clinical trials. Unless impact. The most widely used quality of life
an agent or drug combination achieves a com- questionnaires are the EORTC QLQ-C30 and
TABLE 6.2 Scales for performance status
Score Status
Karnofsky
100 Normal; no complaints; no evidence of disease
90 Able to carry on normal activities; minor signs or symptoms
80 Normal activity with effort, some signs or symptoms of disease
70 Cares for self; unable to carry on normal activity or do active work
60 Requires occasional assistance but able to care for most needs
50 Requires considerable assistance and frequent medical care
40 Disabled; requires special care and assistance
30 Severely disabled; hospitalization indicated although death not imminent
20 Very sick; hospitalization necessary; active supportive treatment necessary
10 Moribund; fatal processes progressing rapidly
0 Dead
WHO
0 All normal activity without restriction
1 Restricted in physically strenuous activity but ambulatory and able to do light work
2 Ambulatory and capable of all self-care but unable to carry out any work. Up and about
>50% of waking hours
3 Capable of only limited self-care, confined to bed or chair >50% of waking hours
4 Completely disabled. Cannot carry on any self-care. Totally confined to bed or chair
the FACT (Functional Assessment of Cancer geting as many cells in the population as possi-
Therapy). ble. An example of spreading limiting toxicities
is shown in Table 6.4 with R-CHOP used for
non-Hodgkin’s lymphoma.
CLINICAL USE OF Combination therapy can also involve the
CHEMOTHERAPY use of a non-chemotherapy agent with a cyto-
toxic drug to enhance its activity. The most
Cytotoxic drugs can be given as single agents common example of this in clinical use is the
but are more usually given as multiple drug addition of folinic acid to 5-fluorouracil (5FU),
combinations. The major limitation in using which approximately doubles its response rates
these agents is toxicity to normal tissues. in the treatment of colorectal cancer. This
Because the majority are non-specific in their works because administered folinic acid
action, both malignant and normal cells are increases intracellular concentrations of folate,
damaged when exposed to a cytotoxic drug. which is an essential cofactor in the incorpora-
When the drugs are used within their defined tion of 5FU into RNA and also its action in
dose limits, the normal cells will recover and it is inhibiting thymidylate synthetase, an important
the need for this window of recovery that results enzyme in DNA synthesis.
in the typical intermittent scheduling of these
agents, most being given at 3–4 weekly intervals.
Common limiting toxicities are bone marrow
DRUG RESISTANCE
suppression, bowel toxicity, and renal and neu- Resistance to chemotherapy drugs can be an
rological damage. The major limiting factors for intrinsic property of a malignant cell but can
specific drugs are shown in Table 6.3. also be acquired after exposure to individual
drugs. Mechanisms of resistance include:
Combination chemotherapy
■ altered biochemical pathways to avoid spe-
For most malignancies, combinations of drugs cific pathway blocks, e.g. modified folate
are more effective than single agents. These are use to avoid dihydrofolate reductase block
designed by adding together modestly active with methotrexate
agents to give greater overall activity and choos- ■ altered cell transport mechanisms to pre-
ing agents with different dose-limiting toxicities vent drug concentration in cancer cell by
so that the antitumour effect but not the toxic either reduced uptake or enhanced efflux
effect is additive or synergistic. Combinations ■ altered drug metabolism increasing clear-
also aim to incorporate drugs from different ance or reducing drug activation; and
classes thereby attacking cells with both phase- ■ impaired mechanisms of apoptosis (pro-
and cycle-specific agents with the intent of tar- grammed cell death).
TABLE 6.3 Major toxicities for chemotherapeutic agents
Tissue affected Toxic agents

Bone marrow* All drugs, in particular vinblastine, etoposide, carboplatin, cyclophosphamide,
melphalan, ifosfamide, paclitaxel
Bowel Melphalan, cisplatin, 5FU, irinotecan, methotrexate
Renal Methotrexate, cisplatin, ifosfamide
Neurological Cisplatin, vincristine, ifosfamide, paclitaxel
Cardiac Adriamycin, 5FU
Bladder Cyclophosphamide
* In practice, bone marrow toxicity is becoming less important as a limiting toxicity as techniques for bone marrow
support, such as autologous marrow or peripheral stem cell transplantation, become available.
Administration of chemotherapy 57
TABLE 6.4 Drugs in R-CHOP
Drug Action Main toxicity

Rituximab Complement activation Allergy, anaphylaxis
Cyclophosphamide Alkylating agent Bone marrow, bladder
Adriamycin Intercalating Bone marrow, cardiac
Vincristine Spindle poison Neurological (peripheral neuropathy)
Prednisolone Unknown Fluid retention and weight gain, gastric irritation,
hyperglycaemia
It is often found that resistance to a number ■ haemoglobin >10 g/dL – although this is
of drugs develops together, a phenomenon rarely a limiting factor and it is acceptable
called ‘multidrug resistance’ (MDR). This is to proceed at lower levels and transfuse if
thought to result from common molecular necessary
mechanisms related to specific DNA sequences. ■ total white count >3.0 × 109/L and total
Identification of one particular genetic change neutrophil count >1.0–1.5
in drug resistance has led to isolation of the ■ platelets >50–80 × 109/L.
MDR1 gene and the protein for which it codes,
For certain drugs renal function must also be
a transmembrane glycoprotein important in cell
carefully checked prior to administration of
transport. As mechanisms of resistance are elu-
drugs. It is not usually sufficient to rely on
cidated, strategies to overcome them are
serum urea or creatinine, and a measure of cre-
emerging. For example, transmembrane trans-
atinine clearance should be performed. This
port can be modified by drugs such as vera-
applies in particular to the following:
pamil and cyclosporin inhibiting the efflux of
drugs away from the cell. ■ cisplatin
■ carboplatin
■ ifosfamide
ADMINISTRATION OF ■ methotrexate (high dose).
CHEMOTHERAPY
Because of the close relation between serum
Chemotherapy should be administered only in levels and creatinine clearance, the dose of car-
specialized units with the experience and sup- boplatin is usually defined by a formula (the
port to do so safely. Most of the agents used can Calvert formula), which relates the predicted
be harmful if there is continuous contact with area under the serum concentration versus time
the skin and even more so if accidentally curve (AUC) to the clearance. This enables the
ingested through contamination of hands and drug to be given relatively safely, even where
work surfaces. Agents should therefore be pre- renal function is impaired, with a predictable
pared by a specialized chemotherapy pharma- AUC after administration.
cist under strictly controlled conditions, and Chemotherapy drugs can be given orally, by
gloves should be worn by the clinical staff when intravenous or intramuscular bolus injection or
they are administering the drugs. by intravenous infusion. Examples of these are
Prior to administration of drugs, all patients shown in Table 6.5.
should have a full blood count measured. In Many chemotherapy drugs are severe irri-
varying circumstances and with different drugs tants when injected outside a vein and extreme
the criteria for safe administration will vary but care must be taken during intravenous injec-
a general guide is that chemotherapy should tions. These should be performed ideally by
only be given if the following parameters are experienced staff working in a dedicated
met: chemotherapy administration unit. Large
TABLE 6.5 Routes of administration
Oral IV injection Infusion Intrathecal

Chlorambucil Cyclophosphamide Cisplatin Methotrexate
Busulphan Methotrexate Carboplatin Cytosine arabinoside
Melphalan (ld) 5FU Mitozantrone
Procarbazine Adriamycin Ifosfamide
Etoposide Vincristine Methotrexate (hd)
Capecitabine Vinblastine Melphalan (hd)
Temozolamide Vinorelbine Paclitaxel
Gemcitabine
Oxaliplatin
5FU
Key: ld, low dose; hd, high dose
visible veins should be chosen as far as possible saline at least weekly is required, and some doc-
and it is safest to administer the drugs into a tors also recommend continous low-dose war-
running intravenous drip. farin 1 mg daily while the line remains to
Where venous access is difficult: prevent thrombotic complications. More
sophisticated devices place a reservoir under
■ for schedules requiring lengthy infusions
the skin of the chest wall rather than an exter-
■ when continuous infusion pumps are to be
nal line; this enables multiple access with less
used
potential risk of infection, and it is more con-
■ in procedures likely to require the adminis-
venient for the active patient.
tration of many intravenous drugs, and
An alternative type of central line, which
■ in transfusions such as bone marrow trans-
may be less robust but is simpler to insert, is a
plantation or high-dose chemotherapy
line such as a PICC line, which enters through
an indwelling central venous catheter has many the antecubital vein and is then directed
advantages. A common type is the Hickman through the brachial and subclavian veins into
catheter as shown in Figure 6.1. These can be the vena cava; an example is shown in Figure
inserted under radiological control with local 6.2.
anaesthetic and remain in situ for many The major complication from central lines in
months. Regular flushing with heparinized chemotherapy patients is that of infection, and
Subclavian vein Superior vena

cava
Incision
Dacron
cuff LA
Axillary RV LV
vein
Incision
(a) (b)
Figure 6.1 Central Hickman line shown (a) diagrammatically and (b) in situ in a patient.
Administration of chemotherapy 59
■ The chemotherapy injection or infusion

must be stopped immediately.
■ The cannula is left in situ, residual drug is
aspirated and the area flushed with saline.
■ Ice packs may be applied and some recom-
mend administration of local steroids or
hyaluronidase, except for vincristine and
oxaliplatin when gentle warming of the
area is recommended.
In severe cases with irritant drugs such as
Adriamycin or epirubicin, there may be exten-
Figure 6.2 A PICC line in situ entering an antecubital sive soft tissue damage as shown in Figure 6.3,
vein. particularly if the problem is not identified
immediately and action taken. Liaison with a
infected lines or subcutaneous tracts require plastic surgery unit for such eventualities is of
immediate removal of the catheter together great value. Occasionally damage may be such
with high-dose antibiotic cover. Subclavian vein as to require surgical repair.
thrombosis may also occur, particularly around
an infected line. Avoiding side-effects
If extravasation of chemotherapy does occur
then there should be clear guidelines for its As a general rule side-effects are best antici-
management. Table 6.6 details the drugs for pated and prevented. Chemotherapy agents
which this is an important issue. vary in their emetic potential but all may cause
The following principles apply: nausea and some severe vomiting. Anti-emetics
TABLE 6.6 Vesicant properties of common cytotoxic drugs
Vesicants Exfoliants Irritants Inflammatory agents Neutral

Causing pain, Causing Causing Causing mild to Causing no
inflammation and inflammation inflammation moderate inflammation or
blistering, leading and shedding and irritation inflammation and flare damage
to tissue death of skin in local tissues
and necrosis
Carmustine Cisplatin Carboplatin Fluorouracil Aspariginase

Dacarbazine Docetaxel Etoposide Methotrexate Bleomycin
Dactinomycin Mitozantrone Irinotecan Cyclophosphamide
Daunorubicin Oxaliplatin Cytarabine
Doxorubicin Topotecan Fludarabine
Epirubicin Gemcitabine
Idarubicin Ifosfamide
Mitomycin Melphalan
Mustine
Paclitaxel
Vinblastine
Vincristine
Vindesine
Vinorelbine
pramide 10 mg intravenously preceding

chemotherapy followed by dexametha-
sone 2 mg tds orally for 2–3 days with
metoclopramide 10 mg tds orally
■ Highly emetogenic schedules
– dexamethasone 8 mg and ondansetron 8
mg orally or intravenously preceding
chemotherapy followed by dexametha-
sone 4 mg bd and ondansetron 8 mg bd
(a)
orally daily for 2–3 days.
Anticipatory nausea and vomiting may be a
particular problem for some patients. This
occurs when they experience symptoms with
any visit to the hospital, often on arrival, with-
out exposure to the drugs. Lorazepam may be
helpful in these instances. It is also often easier
for inpatients to receive chemotherapy in the
evening when they can receive added sedation
and sleep through the administration.
(b) Other side-effects must also be anticipated
and steps taken to prevent them. For example:
Figure 6.3 (a) Erythema following minor extravasation of
epirubicin and (b) more severe soft tissue damage ■ diarrhoea (e.g. from cisplatin, 5FU or
following extravasation of chemotherapy given irinotecan) – prescribe loperamide or
intravenously into the back of the hand. codeine phosphate
■ mucositis (e.g. from 5FU, methotrexate) –
use regular chlorhexidine and benzydamine
should therefore be considered for all but the
mouthwashes
most gentle of agents. A three-stage anti-emetic
■ alopecia (e.g. from Adriamycin) – use of
policy based on emetic potential will work for
scalp cooling and provision of wig.
most patients.
Emetic potential of common chemotherapy Scheduling

agents
Chemotherapy schedules vary from simple
Chlorambucil and vincristine/vinblastine, used
single oral drugs to complex multiple drug regi-
as single agents, rarely cause significant nausea
mens. The design of drug schedules is based on
and either anti-emetics are not required or
a consideration of the effects on both normal
simple oral agents are sufficient. Other drugs or
tissues and the tumour cells as discussed above.
combinations fall into the groups shown in
Normal tissues are inevitably damaged by
Table 6.7
chemotherapy agents. Particularly sensitive are
The use of specific anti-emetic drugs will
the dividing cells of the bone marrow and
vary but a simple anti-emetic protocol is as
mucosal epithelial cells lining the oropharynx,
follows.
gut and bladder. The general principle of sched-
■ Low potential schedules uling chemotherapy is that each successive
– metoclopramide 10 mg orally or course should be given only when damage from
prochlorperazine 25 mg rectally preced- the previous drug exposure has been repaired.
ing chemotherapy Fortunately both bone marrow and epithelial
■ Moderately emetogenic schedules lining cells have a large capacity for tolerating
– dexamethasone 8 mg and metoclo- and recovering from damage and usually do so
Hormone therapy 61
TABLE 6.7 Classification of chemotherapy drugs according to their potential to provoke emesis
Low (<10%) Moderate (10–30%) High (31–90%) Very high (>90%)

Vincristine Cyclophosphamide Adriamycin Cisplatin
Vinblastine Methotrexate Epirubicin Dacarbazine
Bleomycin Mitomycin Ifosfamide Carmustine
Fludarabine Mitozantrone Carboplatin Mustine
Rituximab Etoposide Oxaliplatin Cyclophosphamide (>1.5 g/m2)
Bevacizumab 5FU Irinotecan
Paclitaxel
Gemcitabine
Cetuximab
Transtuzumab
within 2–3 weeks. On this basis most HORMONE THERAPY

chemotherapy is given at 3–4 weekly intervals.
Damage to tumour cells may depend on both A small number of tumours are influenced by
the absolute levels of drug that can be achieved therapeutic changes in their hormone environ-
within them and the duration of exposure. In ment. In practice, hormone therapy is of value
order to achieve the maximum levels possible, a for breast and prostate cancer and to a lesser
fine line may be drawn between serious side- extent endometrial cancer. The response of
effects and maximizing tumour cell kill. In some breast cancer is based primarily upon the influ-
circumstances where there is evidence that very ence of oestrogen, and that of prostate cancer
high doses of drugs beyond normal tolerance upon the influence of androgens.
will achieve more, e.g. in acute leukaemia, then Occasional responses with other tumour types
bone marrow, or more usually bone marrow have been reported, particularly to the drug
stem cells, can be removed prior to chemother- tamoxifen, but their basis remains uncertain.
apy and stored to be used as an autograft once
high-dose therapy has been given. Breast cancer
In order to achieve maximum levels for a
suitable duration a knowledge of the pharmaco- Oestrogen exerts its effect by binding to a
kinetics of the drug is important. Drugs with receptor within the cell nucleus called the
short half-lives, such as 5FU, may have a greater ‘oestrogen receptor’. This process is thought to
effect when given by infusion or in divided be fundamental to the way in which oestrogen
doses. can influence the development of breast cancer.
The design of chemotherapy drug schedules Drugs that alter the balance of activity at the
is often a combination of elegant hypothesis, oestrogen receptor are therefore often of value
serendipity and pragmatism. However, when in the treatment of breast cancer. Such drugs
the results of chemotherapy trials are assessed may act directly by binding to the receptor and
and translated into general oncological practice, thereby blocking its activity as typified by
careful attention to the drug dosing and sched- tamoxifen, or indirectly on the production and
uling is important. There is some evidence that peripheral activation of oestriol and oestrone to
patients who fail to receive full doses of drugs oestradiol by inhibiting the enzyme aromatase;
at the designated minimum intervals have a commonly used aromatase inhibitor drugs
reduced benefit from the treatment and this is a include anastrozole and letrozole.
strong argument for chemotherapy being man- The presence of oestrogen receptors can be
aged only in experienced units familiar with the demonstrated histologically using immunohisto-
complications and tolerance for each schedule. chemistry and the demonstration of oestrogen
receptors (ER) in an individual tumour is impor- Progestogens and androgens can also be
tant in predicting the likelihood of response to effective as second- and third-line treatments in
hormonal treatment. Overall around 50 per breast cancer. Drugs such as megestrol and
cent of patients with breast cancer will respond medroxyprogesterone do indirectly affect activ-
to hormone therapy, but this figure reflects a ity at the oestrogen receptor causing downregu-
response rate of up to 80 per cent when oestro- lation (i.e. making them less sensitive to
gen receptors are positive compared with stimulation by oestrogen), but in addition spe-
around 10 per cent where they are negative. cific progestogen receptors have been identified
The observation that some patients who have in breast cancers and found to correlate with
no demonstrable receptors will respond may be response to treatment.
explained by the presence of receptors in these An overview of hormone therapy in breast
cancer cells that have a slightly different structure cancer is given in Table 6.8.
not detected by the routine methods of assay.
Characteristically, hormone responses have a Prostate cancer
finite duration and it would seem that all breast
cancers eventually become resistant to the first The basis of hormone therapy in prostate
hormone treatment to which they are exposed. cancer is the dependence on androgen for its
This may, however, be followed by second, third growth. The drugs used to treat prostate cancer
and even fourth successive responses to further therefore are anti-androgens working either
hormone manipulations. directly by antagonism at the androgen receptor
The basis for hormone resistance developing or indirectly on the hypothalamic–pituitary axis
is thought to be changes in the characteristics of regulation of androgen release. Similar effects
the oestrogen receptor, varying from complete are also achieved by surgical orchidectomy
loss of receptor to alterations in its binding sites removing the main site of androgen production.
or in its transcriptional properties within the An overview of androgen blockade is shown in
nucleus. In other cases access to the receptor Figure 6.4. The relative clinical merits of the
may be denied because of alterations of trans- available antiandrogen therapies are shown in
port mechanisms and metabolism within the Table 10.1 (see Chapter 10, page 171).
cell. Exposure to an alternative hormone ther- None of the individual anti-androgen treat-
apy may in many cases achieve a second ments offer complete blockade of both testicu-
response, which will be seen in up to 45 per lar and adrenal androgens. This may be achieved
cent of patients after an initial exposure. by combining a centrally acting gonadotrophin-
TABLE 6.8 Hormone therapy in breast cancer
Drug Mode of action Other actions and common side-effects

Tamoxifen Anti-oestrogen Hot flushes, fluid retention, vaginal dryness/discharge,
Faslodex uterine bleeding, deep vein thrombosis
Exemestane Aromatase inhibitor Hot flushes, nausea, rashes, joint stiffness, vaginal
Anastrozole dryness, raised cholesterol, osteoporosis
Letrozole
Medroxyprogesterone Progestogen Nausea, fluid retention, weight gain
Megestrol
Biological therapy 63
HYPOTHALAMUS treatment with progestogens or gonadotrophin-

releasing hormone agonist/antagonists such as
+ve goserelin or leuprorelin can be effective in
Oestrogen metastatic disease. Response is predicted by the
presence of oestrogen or progestogen receptors
Goserelin with around 65 per cent of patients having posi-
Leuprorelin
PITUITARY tive receptors responding to treatment. A reduc-
Cyproterone tion in oestrogen receptor concentration has been
seen after progestogen treatment, suggesting that
–ve –ve the efficacy of hormone treatment in endometrial
cancer reflects inhibition of oestrogenic stimulus
+ve
at the tumour cell.
+ve
BIOLOGICAL THERAPY
ADRENAL
A number of the chemicals produced by the
body in response to injury or infection have
TESTES been explored as potential new treatments to
Orchidectomy
eradicate cancer cells. Unfortunately only lim-
+ve +ve ited success has so far been achieved. Invariably
Cyproterone such preparations are immunogenic and are
TARGET TISSUE therefore associated with side-effects such as
Flutamide
Bicalutamide malaise and low-grade fever. These are often
debilitating but major toxicities are rare with
Figure 6.4 Mechanisms of androgen release and commercially available compounds. In clinical
blockade. practice only α-interferon is in occasional use.
releasing hormone agonist/antagonist, e.g. α-Interferon

goserelin, with a peripherally acting drug such as
cyproterone or bicalutamide, and is known as This has been the treatment of choice in the
‘maximal androgen blockade – MAB’. This may rare haematological malignancy hairy cell
be slightly more more effective than single drug leukaemia (see Chapter 17). In renal cancer a
therapy in advanced disease, particularly in modest impact in high-risk and advanced dis-
younger patients but will be associated with ease is seen and some responses are also seen in
more severe toxicity from androgen withdrawal, melanoma. In each of these instances interferon
including lethargy, hot flushes, gynaecomastia is believed to function by enhancing the natural
and loss of sexual interest and function. defences available within the body, facilitating
An important distinction between the hor- removal of malignant cells.
mone response of prostate cancer and that of Alpha-interferon is given by subcutaneous
breast cancer is seen on relapse to first-line anti- injection three times per week. In doses above 3
androgen therapy; second responses with megaunits side-effects of fever and malaise can
prostate cancer are seen only occasionally. be dose limiting. Treatment can be continuous
for many months or even years.
Endometrial cancer
Monoclonal antibody treatments
Endometrial cancer is recognized as a tumour
related to high levels of circulating oestrogen. Monoclonal antibodies (MABs) are the equiva-
Both oestrogen and progesterone receptors can be lent of the magic bullet, their principle being to
demonstrated in cells of endometrial cancer and carry a toxic agent to a cell defined by specific
surface antigens against which the antibody is Clinical trials are underway to define the pre-
targeted. In recent years this approach has cise role of these new agents in the management
resulted in a number of new clinical agents of these diseases; they will probably be incorpo-
becoming available. rated into existing chemotherapy regimens both
Rituximab is targeted against the CD20 anti- in the adjuvant setting and in primary treat-
gen on B-cell lymphomas; in addition to stan- ment.
dard chemotherapy it improves survival.
Ibritumomab is an anti-CD20 antibody tagged GROWTH FACTORS
with a radioisotope yttrium-90, which is highly
effective in follicular lymphoma. One of the major dose-limiting toxicities encoun-
Herceptin is targeted against the HER2 anti- tered in the use of chemotherapy is that of bone
body on breast cancer cells. and is now a valu- marrow toxicity. In recent years the availability of
able additional treatment option for patients colony-stimulating factors (CSF), in particular
with breast cancer which is HER2 positive; this granulocyte colony-stimulating factor (G-CSF)
can be demonstrated immunohistochemically has enabled chemotherapy dose to be intensified
and the extent of staining can be graded to pre- within the limits of tolerance of bone marrow.
dict the likelihood of response, as shown in G-CSF is an analogue of naturally occurring
Figure 6.5. growth factors given by subcutaneous injection
Alentuzumab is an anti-CD52 antibody, following exposure to chemotherapy. It stimu-
which is highly active in chronic lymphocytic lates the granulocyte production lines in the bone
leukaemia. marrow, reducing the period of neutropenia after
Bevacizumab, discussed earlier, is a mono- intensive chemotherapy. The growth factors cur-
clonal antibody, targeting tumour vasculature rently available have no significant effect upon
through VEGF. platelet and red cell lines. Their use has also facil-
Figure 6.5 Histological section showing breast cancer cells with a range of positive staining for the HER-2 receptor,
which may be seen to predict varying degrees of response to the monoclonal antibody transtuzumab (Herceptin®).
Further reading 65
itated the process of peripheral blood progenitor results and most of these studies are there-
cell (PBPC) harvesting, thereby simplifying the fore multicentre studies.
use of ultra-high-dose chemotherapy which is
Only following satisfactory passage through
being increasingly applied to the treatment of
the above steps is a drug incorporated in the
solid as well as haemopoietic tumours.
routine treatment of cancer; this may be many
years after the first identification of the com-
EXPERIMENTAL pound and will be followed by further evalua-
CHEMOTHERAPY tion of the drug to establish its full potential
and application in adjuvant and primary treat-
The chemotherapy drugs and combination regi- ment and its role in palliation both alone and in
mens widely used in routine oncological prac- combination. Its toxicity in wider use must also
tice today have arisen out of carefully designed be continuously monitored and notified (post-
and regulated drug development programmes. marketing surveillance).
From discovery of a promising new drug in the Current areas of research in drug develop-
laboratory a lengthy and expensive period of ment alongside the more conventional classes
evaluation starting with animal pharmacology of chemotherapy drugs include the following:
and toxicology studies and ending in successful ■ drugs targeted against specific biochemical
drug marketing has to be embarked on. The pathways involved in cellular growth con-
clinical studies within which new drugs may be trol, for example the epithelial growth
given are outlined below and discussed in factor receptor (EGFR), tyrosine kinases
greater detail in Chapter 3. and proteosomes
■ Phase 1 studies These are studies in which ■ anti-angiogenic agents and vascular target-
the new agent is first tried in patients. Such ing agents exploiting the different charac-
agents are offered only to patients for whom teristics of tumour vasculature from normal
there is no other recognized effective treat- blood vessels to selectively damage tumour
ment and who may wish to try other drugs blood supply
in the slim hope of benefit. In phase 1 stud- ■ immunotherapy through the development
ies activity against the tumour is not an end- of vaccines against tumour antigens.
point but these studies are designed to
assess the maximum tolerated doses and FURTHER READING
define the toxicity profile in humans.
■ Phase 2 studies These are the earliest studies Brighton D, Wood M. Royal Marsden Hospital Hand-
in which antitumour activity is sought, book of Chemotherapy. Churchill Livingston,
giving the drug in doses and schedules London, 2005
defined from the phase 1 studies to patients Buzdar AU (ed.). Endocrine Therapies in Breast
who have failed previous conventional Cancer. Oxford University Press, Oxford, 2007
therapy or for whom there may be no Hesketh PJ. Chemotherapy-induced nausea and
effective recognized treatment. vomiting. New England Journal of Medicine 2008;
■ Phase 3 studies With activity in phase 2 358:2482–92
established, the new agent is compared Summerhayes M and Daniels S. Practical Chemother-
with the standard best treatment (or, if apy: a multidisciplinary guide. Arnold, London,
there is none, with placebo) in a large 2003
prospective randomized trial. Large num- Young A, Rowett L, Kerr D (eds). Cancer Biotherapy:
bers of patients are entered into such trials an introductory guide. Oxford University Press,
in order to achieve statistically robust Oxford, 2006
1. Which three of the following drugs are 6. Which three of the following are highly
antimetabolite chemotherapy agents? sensitive to chemotherapy?
a. Gemcitabine a. Breast cancer
b. Topotecan b. Non-Hodgkin’s lymphoma
c. Cisplatin c. Small cell lung cancer
d. Fludarabine d. Neuroblastoma
e. Adriamycin e. Renal cell carcinoma
f. Cyclophosphamide f. Phaeochromocytoma
g. Cytosine arabinoside g. Glioblastoma
2. Which of the following applies to 7. For which three of the following drugs is
intercalating agent chemotherapy drugs? neurological toxicity a major concern?
a. They act by alkylation of DNA a. Cyclophosphamide
b. They inhibit RNA synthesis b. Carboplatin
c. They prevent separation of the DNA c. Paclitaxel
strands d. Irinotecan
d. They are topoisomerase inhibitors e. Cisplatin
e. They include irinotecan f. Ifosphamide
3. Which of the following statements is true 8. For which of the following drugs must
regarding those chemotherapy agents renal function be measured before
acting on mitosis? administration?
a. They interfere with synthesis of DNA a. Vincristine
b. They affect spindle formation b. Adriamycin
c. They are independent of topoisomerase c. Carboplatin
d. They include cisplatin d. Etoposide
e. They are independent of the cell cycle e. Epirubicin
4. The following is true of signal transduction 9. Which of the following drugs are used for
inhibitors: anti-androgen activity?
a. They act through tyrosine synthase a. Fluoxetine
b. They can inhibit angiogenesis b. Tamoxifen
c. They include bortezomib c. Exemestane
d. They are cell cycle dependent d. Bicalutamide
e. They act on cell division e. Cyclophosphamide
5. Which three of the following are true of 10. Which three of the following are
drugs targeting the tumour vasculature? monoclonal antibodies used in cancer
a. They will reduce blood flow to normal treatment?
tissues a. Rituximab
b. They can act through VEGF b. Bortezomib
c. They include thalidomide c. Herceptin
d. They cause extensive thrombosis d. Lapatinib
e. Their effect is reversed by heparin e. Erythropoietin
f. They can alter blood vessel structure f. Combretastatin
g. Bevacizumab

TH Neal

Uploaded by

Document Informationclick to expand document informationTerapija citostaticima

Document Informationclick to expand document information

Copyright:

Available Formats

TH Neal

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TH Neal

Uploaded by

Copyright:

Available Formats

03-Clinical Oncology-ch03-ccp:Layout 1 12/3/09 16:25 Page 22

■ Treatment options 23 ■ Clinical evidence and clinical trials 28

The following chapters describe clear treatment Diagnosis of malignant condition

vidualization of treatment for any given patient

TREATMENT OPTIONS they are almost certainly cured and require no

24 DECISION-MAKING AND COMMUNICATION

26 DECISION-MAKING AND COMMUNICATION

COMMUNICATION process of physical examination or while still

Treatment choices viving, say, 5 years can be given. This should,

28 DECISION-MAKING AND COMMUNICATION

Clinical evidence and clinical trials 29

results of a large randomized trial addressing a a population of patients. The characteristics of

A particular clinical question, whether in It is important at the outset of a trial to define

30 DECISION-MAKING AND COMMUNICATION

Clinical evidence and clinical trials 31

32 DECISION-MAKING AND COMMUNICATION

Clinical evidence and clinical trials 33

34 DECISION-MAKING AND COMMUNICATION

■ Management of the primary tumour 36 ■ Palliative surgery 38

Management of regional lymph nodes 37

TABLE 4.1 Relative merits of pre- and postoperative radiotherapy

proportion (around 20 per cent) of patients will MANAGEMENT OF

COMBINED SURGERY ■ axillary dissection for breast cancer

38 PRINCIPLES OF SURGICAL ONCOLOGY

■ Intubation of the oesophagus with a rigid

40 PRINCIPLES OF SURGICAL ONCOLOGY

■ Types of radiotherapy 42 ■ Side-effects of radiotherapy 48

Live source implants

Brachytherapy Manual afterloading

phosphorus (32P) for polycythaemia rubra vera

shielded walls, floors and ceilings while the Protons

Clinical use of radiotherapy 45

for tumours in the thorax, abdomen or pelvis.

Figure 5.4 On left, beam’s eye view

Clinical use of radiotherapy 47

Intensity-modulated radiotherapy (IMRT) is a tube on the gantry, which enables cross-

■ total dose this routinely, while others treat over 1–2

TABLE 5.1 Acute effects after radical radiotherapy

Site Effect Treatment

TABLE 5.2 Late effects after radical radiotherapy

Site Effect Treatment

■ Chemotherapy agents 52 ■ Hormone therapy 61

Cancer chemotherapy is the treatment of ■ Signal transduction inhibitors

by 5FU and of ribonucleotide reductase by Drugs acting on mitosis

54 PRINCIPLES OF SYSTEMIC TREATMENT

Group 1: High sensitivity Group 2: Modest sensitivity Group 3: Low sensitivity

Efficacy and toxicity of chemotherapy 55

Response and survival plete response in a significant proportion of

TABLE 6.2 Scales for performance status

56 PRINCIPLES OF SYSTEMIC TREATMENT

TABLE 6.3 Major toxicities for chemotherapeutic agents

Tissue affected Toxic agents

TABLE 6.4 Drugs in R-CHOP

Drug Action Main toxicity

58 PRINCIPLES OF SYSTEMIC TREATMENT

TABLE 6.5 Routes of administration

Oral IV injection Infusion Intrathecal