Chemotherapy and Radiation Therapy

for Anal Carcinoma

Survival and Late Morbidity
G. Tanum, MD, K. Tveit, MD, K. 0. Karlsen, MD,
and M. Hauer-Jensen, MD

This study was performed to evaluate the survival and late morbidity rates of a
widely used combined chemotherapy and radiation therapy regimen given to
patients with carcinoma of the anal canal. One hundred six patients received
radiation therapy (5000 cGy given by two anteroposterior-posteroanterior[AP-PA]
opposed fields) and chemotherapy (mitomycin C plus 5-fluorouracil [5-FU])from
1983 to 1989. Patients with primary tumors (n = 86) had a complete response rate of
84% and a 5-year survival rate of 72%. There was no significant difference in
survival rate according to tumor stage. Patients with local recurrence (n = 20) after
primary surgery had a complete response rate of 50% and a 5-year survival rate of
40%. Fifteen percent of the patients experienced late treatment-related symptoms
including anal incontinence, intestinal obstruction, and chronic pelvic pain. The
current treatment regimen is effective but carries a considerable risk of
complications. As survival rate was independent of tumor stage, the locoregional
treatment should probably be less extensive for small tumors than for advanced
tumors. This strategy may reduce the late side effects for patients with small
tumors without reducing the survival rate. Cancer 679462-2466,1991.

S TANDARD TREATMENT of squamous cell carcinoma

of the anal canal has changed radically during the
last decade. Combined chemotherapy and radiation ther-
lished. Moreover, randomized trials evaluating various
chemotherapy and radiation therapy regimens have not
been reported. The main reason for this is that the disease
apy or radiation therapy alone has gained general accep- is uncommon and thus most hospitals have only small
tance as primary treatment, whereas abdominoperineal series of patients. Also, due to the lack of a uniform staging
resection (APR) is now being used only as salvage treat- system, comparing treatment results by stage is difficult.
ment in (chemo) radio-resistant The reason for Consequently, only a few reports on treatment results in
the change from surgery to conservative treatment is partly large series have been p ~ b l i s h e d . ~The
- ~ current study
improved survival and partly preservation of anal func- consists of a comparatively large series of patients with
tion. Anal carcinoma often grows in a diffuse manner and anal canal carcinoma undergoing uniform treatment in
infiltrates widely beyond the palpable mass. The main a single institution during a relatively short period. All
problem associated with surgical treatment has been the patients were diagnosed, staged, treated, and observed after
inability to achieve local tumor control. Thus, a substan- treatment in accordance with a standardized protocol.
tial number of patients treated with surgery alone die of Patients with primary anal canal carcinoma and patients
local r e ~ur r e nc e .~Chemotherapy
-~ and radiation therapy with locally recurrent tumors after previous APR were
has improved locoregional control; consequently, the rel- included in the series. This study was performed to eval-
ative impact of distant metastases on survival has in- uate a commonly used chemotherapy and radiation ther-
creased. However, no randomized study comparing sur- apy regimen with respect to long-term survival and late
gery and chemotherapy/radiation therapy has been pub- morbidity rates.

From the The Norwegian Radium Hospital, Oslo, Norway. Patients and Methods
Address for reprints: G. Tanum, MD, Department of Oncology, The
Norwegian Radium Hospital, N-03 LO Oslo 3, Norway. One hundred seventeen patients were admitted to the
Accepted for publication February 15, 1991. Norwegian Radium Hospital, Oslo, Norway, for squa-

2462
No. 10 TREATMENT
OF ANALCA - Tunurn d ul. 2463
TABLE1. Stage of Primary Anal Canal Carcinoma at Admittance were examined as described above and multiple biopsy
to Hospital (TNM system) (n = 94)
specimens were taken from the anal canal. Salvage surgery
No. YO of all (APR) was performed on patients with primary tumors
if malignant tissue was present after treatment. Patients
Tumor stage
TO 2 2. I who had distant metastases received chemotherapy (either
T1 11 11.7 cisplatin plus 5-FU or mitomycin C plus 5-FU). Two pa-
T2 23 24.5 tients with solitary liver metastasis were treated with rad-
T3 31 33.0
T4 27 28.7 ical surgery, and one patient with multiple lung metastases
Node stage received total lung irradiation in addition to chemother-
NO 70 74.5 apy. After evaluation the patients were examined every 3
NI 8 8.5
N2 10 10.6 months for 2 years and every 6 months for an additional
N3 6 6.4 3 years as described above (except MRI or CT scans).
Metastasis stage Forty-nine patients were observed for more than 3 years
MO 88 93.6
MI 6 6.4 (at the end of 1989).
Statistical calculations were performed with the BMDP-
PC program package (BMDP Statistical Software Inc., Los
Angeles, CA). Actuarial survival was obtained using Kap-
mous cell carcinoma ofthe anal canal from 1983 to 1989. lan-Meier estimates. Survival curves were compared with
All tumors were proven by histopathologic examination, the Mantel-Haensel test and the Tarone-Ware test for
but no biologic tumor markers were studied. The series trend as appropriate, using a level of significance of 0.05.
included 25 (2 I %) men and 92 (79%)women. The mean
age of the patients was 64.3 years (range, 35 to 91 years) Results
for women and 67.8 years (range, 37 to 85 years) for men.
Ninety-four patients had primary tumor and 23 had local Seventy-nine patients (84%) with primary tumors had
recurrence after previous surgery. Seven patients initially a complete response to chemotherapy and radiation ther-
had distant metastases and were in poor condition at the apy, as judged by multiple negative biopsy specimens,
first admittance to the hospital, whereas four patients with whereas 15 patients (16%) later underwent an APR due
TO and early T1 tumors were treated with a limited tumor to persistent malignant tumor (positive biopsy specimen)
excision. These 1 I patients did not receive chemotherapy after treatment. Approximately 66% of the latter cases
and radiation therapy and were therefore excluded from had T4 tumors and 33% had T3 tumors before treatment.
the study. Altogether 106 patients received the combined N o additional chemotherapy and radiation therapy was
chemotherapy and radiation therapy regimen described given after APR. One patient had a local recurrence after
below. salvage APR. Eleven patients (48%)with local recurrent
The primary tumors (previously untreated) were staged tumor after primary APR responded completely. Patients
according to the TNM system (International Union with primary tumors had a significantly better 5-year dis-
Against Cancer [UICC] 1987) (Table 1). All patients were ease-free survival rate (72%) than those with local recur-
examined clinically (digital anal and rectal examination, rence after primary APR (40%) (Fig. 1). No significant
proctoscopy, computed tomography [CT] [ 1983 to 19881 correlation was found between tumor stage and survival
or magnetic resonance imaging [MRI] [ 19891scans of the rate in patients with primary tumors (Fig. 2).
pelvis, ultrasonography of the liver, chest radiograph, and Eighteen patients had metastases in 24 sites after che-
blood tests). motherapy and radiation therapy. The liver was the most
Patients with primary tumors and patients with recur- frequent site (ten patients), followed by the pelvic lymph
rent tumors after APR received radiation therapy, 200 nodes (six patients), the skin (three patients), the lungs
cGy daily 5 days per week, up to 5000 cGy given by two (two patients), and the distant lymph nodes (one patient).
anteroposterior-posteroanterior (AP-PA) opposed fields Patients with liver metastases appeared to survive longer
that covered the pelvis from the perineum to the upper (median survival time of 12 months after chemotherapy
part of the sacrum (Sl to S3). Chemotherapy consisted and radiation therapy) compared with those with pelvic
of mitomycin C (10 to 15 mg/m2) given as a bolus on day lymph node and skin metastases (median survival time
1 of radiation therapy and 5-fluorouracil (5-FU) (1000 of 4 months) (Table 2). Two patients are without evident
mg/m'/d) given as continuous infusions on days 1 to 4. disease 6 and 42 months after resection of liver metastases,
Forty-eight patients required a split course of 2 to 3 weeks and one patient 37 months after chemotherapy and total
that was compensated for by two additional fractions at lung irradiation of multiple lung metastases, respectively.
the end of the treatment period. The response was eval- Fourteen of 89 patients who were observed for more
uated 1 month and 3 months after treatment. All patients than 3 months experienced treatment-related morbidity
2464 CANCER
May 15 1991 Vol. 61

loo
80
hill I, I 1 1 1 1 , qr;mary tumors

L
Site

Liver
TABLE2. Metastatic Sites of Anal Carcinoma and Survival*

No. (%) of eases

10 (42)
Median survivalt

12
4
Pelvic lymph nodes 6 (25)
Skin 3 (13) 4
Lung 2 (8) >I4
Distant lymph nodes l(4) 9

LT 40 - * Twenty-four sites were recorded in 18 patients.

3
v)
t Months after completed chemotherapy and radiation therapy.
20 -

0
Discussion
1 ’ 1 ’ 1 ’ 1 ’ 1 ’ 1

Chemotherapy and radiation therapy is accepted as the

primary treatment of anal carcinoma. However, a wide
range of radiation therapy and chemotherapy regimens
are being used. 133.8-’2 Our patients were treated with the
widely used Nigro-regimen’’ except that no second se-
quence of 5-FU was administered to reduce acute toxicity.
The current study represents a comparatively large series
of patients with anal carcinoma from a single institution
that prevented a normal social life or required surgical undergoing uniform conservative treatment and system-
intervention (Table 3). All of these problems turned up atic follow-up at regular intervals.
within 2 years after treatment. Six patients experienced
The overall long-term survival rate for patients with
anal problems with permanent incontinence, and two had
primary carcinomas was found to be approximately 70%
bladder dysfunction. Two patients with anal incontinence
(Fig. I), which is similar to previous report^.^,^,',^,^ Th’1s
later underwent APR with colostomy to improve their
survival rate is very satisfactory when considering the high
social life. Drug-resistant chronic diarrhea was observed
proportion of advanced tumors in the current study com-
in three patients, one of whom had to have an intestinal
pared with most previous Patients with
resection. Three patients died without evident carcinoma,
local recurrence after primary surgery have a less favorable
two after laparotomy for intestinal obstruction due to fi-
prognosis (Fig. 1). The main reason for this is that recur-
brosis and one of fistulae after salvage APR.
rent tumors respond less favorably than primary tumors
to chemotherapy and radiation therapy, as 52% of these
patients have persistent malignant tumor after treat-
100 T1
ment.I6 Consequently, these patients often die of local
recurrence.

80
T3
T4
TABLE3. Serious Late Morbidity* Observed More Than 3 Months
60 T2 After Complction of Chemotherapy-Radiation Therapy
for Anal Carcinoma?

2- Problem No. % of all

3
v) Anal function 6 6.7
a Intestinal obstruction$ 4 4.5
Diarrhea 3 3.4
Bladder function 2 2.2
Pelvic pain 2 2.2
Fistulae4 1 1.1
0 12 24 36 48 60 72
* “Serious late morbidity” means that normal social life was not pos-
OBSERVATION TIME (months) sible or that surgery had to be performed.
t Eighteen symptoms observed in 14 patients (15.7% of all).
FIG. 2. Survival time of patients receiving combined chemotherapy $ Of which two died after laparatomy, without evidence of tumor.
and radiation therapy for anal carcinoma, relative to tumor stage (TI- 5 Died 18 months after salvage surgery (APR) without evidence of
T4) (TNM 1987). tumor.
No. 10 TREATMENT
OF ANALCA - Tanum et af. 2465

The most frequent late problem in the current series

was anal insufficiency. It is difficult to determine whether
a0 Node negative
this was caused by previous tumor invasion or by che-
I I motherapy and radiation therapy. Pelvic pain (two pa-
a tients) was probably caused by compression of peripheral
Node positive nerves due to tissue swelling and fibrosis after irradiation.
Two of four patients with intestinal obstruction died after
laparotomy, indicating the difficultiesinvolved in surgery
for intestinal radiation injury. Although the long-term
survival rate reported here is excellent, the frequency of
2o 1 treatment-related morbidity seems unacceptable. Radia-
tion with high doses and large fields may be necessary to
achieve cure in patients with advanced tumors, but prob-
0
0 12 24 36 4a 60 12
ably not in those with smaller tumors. This is supported
by the observation of a similar survival rate irrespective
OBSERVATION TIME (months) of tumor stage. Other investigations have reported excel-
FIG.3. Survival time of patients receiving combined chemotherapy lent results using far more lenient treatment regimens for
and radiation therapy for anal carcinoma, relative to pelvic lymph node patients with small t ~ m o r s . ' . ~Consequently,
.'~ treatment
status (node negative and positive) (TNM 1987). regimens should be differentiated according to tumor
stage.
In the current study the long-term survival rate for pa- Few, if any, previous studies have correlated survival
tients with primary tumors was independent of tumor rate for patients with distant metastases with metastatic
size and lymph node status before treatment (Figs. 2 and site. The current series indicates that the liver is the most
3). This is in contrast to most previous studies in which frequent initial site of metastasis, followed by the pelvic
a negative association has been found between tumor stage lymph nodes (Table 2). In this context, only metastatic
and survival rate.i,9,i4,15One reason for this discrepancy nodes present after chemotherapy and radiation therapy
may be that a majority of the patients in previous are included. Survival time appeared to be shorter for
reports799914did not receive the treatment regimen used in patients with lymph node metastases and skin metastases
the current study. In these reports irradiation was given than for those with liver metastases (Table 2). A possible
only to the tumor site itself by a sacral and perineal field explanation for the poor prognosis of patients with per-
followed by an interstitial implant without chemother- sistent pelvic node metastases may be that these tumor
Survival results in these series were similar to cells have survived chemotherapy and radiation therapy
the current results for small tumors, whereas in our study and hence comprise a selection of treatment-resistant cells.
patients with T4 tumors had improved survival rates. The One patient with multiple bilateral lung metastases was
reason for this may be that the large tumors have salelites treated with four courses of cisplatin and 5-FU followed
that are included in the current irradiation field but are by total lung irradiation. He is alive without evident dis-
partly located outside the smaller fields described above. ease 42 months after treatment. Two patients with liver
This indicates that only patients with advanced tumors
(T3 to T4) benefit from the extensive therapy regimen
TABLE4. Proposal for a New Treatment Protocol
used in the current study, whereas in patients with smaller for Anal Canal Carcinoma
tumors (T 1 to T2) a less extensive treatment may be used
without compromising survival. Adjuvant
T stage N stage Radiation therapy* Chemotherapy?
A relatively high number of patients exhibited chronic
treatment-related symptoms. There was no correlation TO NO None (local excision) None
between tumor stage and late complications. The fre- TI Il c m NO None (local excision) None
quency of late morbidity in the current series is higher TI > I c m NO Local None
T2 NO Local Yes
Less treat-
than that in most previous s t u d i e ~ . ~i714,ii-i9
,~,~,~ T3 NO Regional Yes
ment-related morbidity would be expected in patients re- T4 NO Regional Yes
All N 1-3 Regional Yes
ceiving less extensive treatment,7,9,i4 but the reason for
this discrepancy is unclear for studies using treatment * Local: 5400 to 5800 cGy given to tumor site itself by external ir-
regimens5,i1,17-19 similar to the current regimen. There radiation using a three-field technique (two lateral and one sacral field).
Regional: 5400 to 5800 cGy given to tumor site and regional lymph
are, however, studies that report serious late morbidity in nodes by external irradiation using a two-field technique (2 AP-PA op-
as much as 12% of the patients, even after less extensive posed fields).
treatment regimens.'* 7 Combination of cisplatin and 5-FU.
2466 CANCER
May 15 199 1 Vol. 67

metastasis received two courses of cisplatin and 5-Fu fol- 6. Brown DK, Oglesby AB, Scott DH, Dayton AT. Squamous cell
lowed by resection and have now been under observation carcinoma of the anus: A twenty-live year retrospective.Am Surg 1988;
54:337-342.
for 6 and 42 months, respectively,without evident disease. 7. Papillon J, Mayer M, Montbarbon JF, Gerard JP, Chassard JL,
Thus, the current study indicates the possibility of cure Bailly C. A new approach to the management of epidermoid carcinoma
of the anal canal. Cancer 1983; 51:1830-1837.
in some patients with distant metastases. 8. Bomanm BM, Moertel CG, O’Connell MG et a!. Carcinoma of
Approximately 25% of the patients with primary anal the anal canal: A clinical and pathological study of 188 cases. Cancer
carcinoma died of metastases. This raises the question of 1984; 54114-125.
indications for adjuvant chemotherapy. Although adju- 9. Papillon J, Montbarbon JF. Epidermoid carcinoma of the anal
canal: A series of 276 cases. Dis Colon Rectum 1987; 30:324-333.
vant chemotherapy has not been shown to be effective in 10. Nigro ND, Vaitkevicius VK, Considine B. Combined therapy for
anal carcinoma, it is probable that potent cytostatic agents cancer of the anal canal: A preliminary report. Dis Colon Rectum 1974;
may eliminate subclinical metastases.’ Several courses of 17:354-356.
1 I . Flam MS, John MJ, Mowry PA, Lovalvo LJ, Ramalho LD, Wade
cisplatin plus 5-FU are probably more suitable for this J. Definitive combined modality therapy of carcinoma of the anus. Dis
purpose than the current combination of mitomycin C Colon Rectum 1987; 30:495-502.
plus 5-FU that is mainly thought to act as radiation sen- 12. Ng Ying Kin NYK, Pigneux J, Auvray H, Brunet R, Thomas L,
Denepoux R. Our experience of conservative treatment of anal canal
sitization. Adjuvant chemotherapy should thus replace carcinoma combining external irradiation and interstitial implant: 32
the current chemotherapy. The radiation therapy dose cases treated between 1973 and 1982. Int JRadiat Oncol Biol Phys 1988;
should be somewhat increased to compensate for the 141253-259.
13. Goldman S, lhre T, Seligson U. Squamous-cell carcinoma of the
missing radiation sensitization. Based on the results of the anus. Dis Colon Rectum 1985; 28:143-146.
current study, we advocate a treatment protocol for anal 14. Salmon RJ, Fenton J, Asselain B el al. Treatment of epidermoid
cancer that differentiates chemotherapy and radiation anal canal cancer. Am JSurg 1984; 147:43-48.
15. Hughes LL, Rich TA, Delclos L, Ajani JA, Martin RG. Radio-
therapy according to tumor stage (Table 4). therapy for anal cancer: Experience from 1979-1 987. Int JRadiat Oncol
REFERENCES Biol Phys 1989; 17:1153-1160.
16. Tanum G, Tveit KM, Karlsen KO. Chemoradiotherapy of anal
1. Cummings BJ. The place of radiation therapy in the treatment of carcinoma: Tumour response and acute toxicity. Eur J Cancer 1991.
carcinoma of the anal canal. Cancer Treat Rev 1982; 9: 125- 147.
17. Sischy B, Remington JH, Hinson EJ. Definitive treatment ofanal
2. Cummings BJ. Treatment of primary epidermoid carcinoma of
canal carcinoma by means of radiation therapy and chemotherapy. Dis
the anus. Colorecful Disease 1987; 2:107-1 12.
3. Cummings BJ. Current management of epidermoid carcinoma of
Colon Rectum 1982; 25:685-688.
the anal canal. Gastroenterol Clin Norlh A m 1987; 16: 125-1 42. 18. Sischy B. The use of radiation therapy combined with chemo-
4. Dougherty BG, Evans HL. Carcinoma of the anal canal: A study therapy in the management of squamous cell carcinoma ofthe anus and
of79 cases. JCIin Pathol 1985; 83:159-164. marginally resectable adenocarcinoma of the rectum. Int J Radiat Oncol
5. John MJ, Flam M, Lovalvo L, Mowry PA. Feasibility of non-sur- BiolPhys 1985; 11:1587-1593.
gical definitive management of anal canal carcinoma. Int J Radiat Oncol 19. Glimelius B, Pihlman L. Radiation therapy of anal epidermoid
Biol Phys 1987; 13:299-303. carcinoma. Int J Radiat Oncol Biol Phys 1987; 13:305-312.