Cell Transplantation, Vol. 7, No. 3, pp. 227–238, 1998
© 1998 Elsevier Science Inc.
Printed in the USA. All rights reserved
0963-6897/98 $19.00 1 .00
PII S0963-6897(97)00161-9
Original Contribution
ONE-YEAR CHROMAFFIN CELL ALLOGRAFT SURVIVAL IN CANCER PATIENTS
WITH CHRONIC PAIN: MORPHOLOGICAL AND FUNCTIONAL EVIDENCE
JEAN C.
BÉS,* JEAN
TKACZUK,* KIMBERLY A. CZECH,† MATHIEU TAFANI,* RAYMOND BASTIDE,* CLAUDE CARATERO,*
GEORGE D. PAPPAS,† AND YVES LAZORTHES*1
*Laboratory of Pain and Cell Therapy, Faculty of Medicine, University Paul Sabatier, 133 route de Narbonne, Toulouse 31062, France, and
†Department of Anatomy & Cell Biology, University of Illinois at Chicago, Chicago, IL 60612
M Abstract — The control of chronic pain through transplantation of chromaffin cells has been reported over the past few
years. Analgesic effects are principally due to the production
of opioid peptides and catecholamines by chromaffin cells.
Clinical trials have been reported with allografts consisting of
whole-tissue fragments implanted into the subarachnoid space
of the lumbar spinal cord (14,19,36). In the present study,
allogeneic grafts were successfully used to control chronic pain
in two patients over a period of 1 yr based on patient reported
pain scores, morphine intake, and CSF levels of Met-enkephalin. Macroscopic examination at autopsy located the transplanted tissue fragments in the form of multilobulated nodules
at the level of the spinal axis and cauda equina. Immunocytochemical microscopy showed neuroendocrine cells are positive for chromagranin A (CGA), and enzymes tyrosine hydroxylase (TH) and dopamine-b-hydroxylase (DbH). The
results suggest that there is a relationship between analgesic
effect, Met-enkephalin levels in CSF, and the presence of
chromaffin cells surviving in spinal subarachnoid space.
© 1998 Elsevier Science Inc.
ble pain in terminal cancer patients. However, these
routes of morphine administration introduce problems
related to mechanical malfunctions of tunneled catheters,
injection ports, and implantable pumps as well as an
increased risk of infection (12,13). These drawbacks in
using morphine pumps have triggered research in finding
alternative therapies for treating severe pain conditions
(26,36).
The implantation of adrenal medullary chromaffin cell
tissue into the subarachnoid space of the spinal cord has
been shown to provide stable relief of chronic pain that
surpasses the conventional use of morphine in both
animal and clinical trials [for review, see (4). The
theoretical basis for this pain relief is thought to be the
production and release of opioid peptides and catecholamines by chromaffin cells (20,22,24). When implanted, adrenal chromaffin cells release a “cocktail” of
peptides that diffuse locally into the dorsal horn of the
spinal cord where they modulate, and possibly block the
transmission of somatosensory information sent to
higher brain centers. Although the mechanism of pain
relief with adrenal medullary implants has not been
thoroughly characterized, several studies point to attenuation of NMDA receptor-mediated hyperalgesia as the
primary action of chromaffin cell products (10,28).
Because chromaffin cells release a combination of peptides including growth factors (epidermal growth factor,
fibroblast growth factor, nerve growth factor), catecholamines (epinephrine, norepinephrine, and dopamine), and other proteins (enkephalins and endorphins)
(30,31), it is probable that all of these products contribute
M Key Words — Neural transplants; Adrenal medulla;
Chromaffin cells; Cancer pain; Opioid peptides.
INTRODUCTION
Pharmacologic approaches that are currently available
for the management of chronic cancer pain have certain
limitations. Some patients rapidly develop tolerance to
parenteral opioid peptides, necessitating progressively
increasing dosages that often produce undesirable side
effects. In most pain clinics intrathecal administration
and/or intracerebroventricular administration of morphine is the therapy of choice to combat severe intractaACCEPTED11/26/97.
1
Correspondence should be addressed to Yves Lazorthes,
M.D., Laboratory of Pain and Cell Therapy, Faculty of Medi-
cine, University Paul Sabatier, 133 route de Narbonne, Toulouse 31062, France.
227
228
Cell Transplantation ● Volume 7, Number 3, 1998
either directly or indirectly to the analgesic effect of
chromaffin cell implants.
A series of studies by our laboratory at The University
of Illinois at Chicago showed that chromaffin cells from
both allogeneic and xenogeneic sources can release
catecholamines and opioid peptides into the CSF, and
they reduce pain sensitivity in well-known chronic pain
paradigms (8,9,23,35). Interestingly, no immunosuppressant or antiinflammatory agents were required in any
behavioral studies performed with rodents using allografts. Supporting the validity of this methodology,
healthy TH-positive implants of adrenal medullary tissue
have been shown in the spinal cord several weeks
following implantation (18). In addition, the survival
period was reported as long as 6 mo for intrathecal
allografts in rats based on increased CSF levels of opioid
peptides and catecholamines compared to pretransplant
levels (24). Our initial clinical experience at Toulouse,
France, has indicated that immunosuppression with cyclosporin A (10 mg/kg) for not more than 2 wk after
transplantation is apparently sufficient for long-term
allograft survival and pain reduction (7,14,15).
Immunosuppression is required, at least initially, for
animals implanted with the chromaffin cells obtained
from xenogeneic sources (18). More recently, experiments with xenogeneic chromaffin cells that are encapsulated using a semipermeable polymer membrane
(without immunosuppression) have defined a new methodology that has proven functional in animal, preclinical,
and clinical trials (1,3,11,25). These studies have also
shown that the released products from chromaffin cell
implants can significantly reduce pain sensitivity for
limited periods of time and have corroborated the findings using unencapsulated allogeneic tissue.
The efficacy of adrenal medullary implants for the
treatment of chronic pain has been demonstrated in
rodents by several groups, using different pain paradigms, and have provided a sound foundation for initiating clinical trials in humans. The first application of
these cellular properties of chromaffin cells as a therapeutic strategy for controlling cancer pain in humans was
reported by the Vaquero group (32,33). The results from
this study were considered unsatisfactory. Histological
findings indicated necrotic reduction in 50% of the
grafted tissue. A second series of five patients underwent
trials of the same nature in our preliminary study at the
University of Illinois at Chicago. However, tissue was
maintained in culture for several days before transplantation to insure viability with more positive results
showing the reduction of both morphine intake and pain
scores in four of the patients (26,36). Unfortunately,
autopsy permission was not obtainable so the viability
and integration of grafted tissue fragments were not
confirmed by histology. However, CSF samples taken
periodically throughout the study showed Met-enkephalin and catecholamine levels at several magnitudes
higher than pretransplant levels.
A further study using the same procedure with eight
patients at Toulouse University Hospital (France) was
recently reviewed (14), and this data was combined with
the clinical studies at The University of Illinois Medical
Center (36). Similar findings were reported by both
groups that showed that 10 out of 13 patients benefited
from adrenal medullary implants and, in some patients,
morphine intake ceased whereas in other cases it was
reduced and, more importantly, it had stabilized.
If chromaffin cell therapy is proposed as an alternative
to such traditional approaches to pain relief, as chronic
intrathecal morphine administration after failure of oral
therapy, it is important to establish its long-term efficacy
in humans. In this study we report on the clinical and
pharmacotherapeutic findings in two terminal cancer
patients who were followed for 12 mo after receiving
adrenal chromaffin cell tissue transplants. Graft function
was assessed using subjective measures, such as pain
scores, and objective measures such as morphine intake,
CSF levels of Met-enkephalin, and histology of grafted
tissue obtained from autopsy. Our results show that both
patients benefited from transplants of adrenal medullary
implants with a significant reduction in pain scores. In
addition, biochemical assays and histological studies
indicate that chromaffin cells were present in grafted
tissue and producing catecholamines and enkephalins. In
summary, our results indicate that adrenal medullary
transplantation has the potential to provide significant
analgesia for as least 1 yr in cancer patients suffering
from chronic pain.
MATERIALS AND METHODS
Patient Selection
This study was carried out at Toulouse University
Hospital (France) after examination and approval of the
protocols by the Committee of Ethics and People Protection (CCPPRB). Patient selection included informed
consent and both participants in the study presented with
identified progressing cancerous lesions associated with
chronic pain that initially responded well to oral morphine but, because of irreversible side effects, oral intake
was replaced by intrathecal administration. Ambulatory
follow-up was possible in both reported cases.
Tissue Preparation
Adrenal medullary glands were obtained from male
organ donors who tested negative for viral serologies
HIV-l and 2, HTLV-l and 2, and hepatitis B/C. Following dissection from the kidneys, glands were placed in
Chromaffin cell allograft and cancer pain ● J.C. BÉS
cold Belzer liquid and transferred to the laboratory. The
duration of the protocol for isolation of adrenal medullary tissue from cortical tissue, modified from Winnie et
al. (36), did not exceed 2-1/2 h. During the isolation procedure, each gland was maintained in a media
solution (Dulbecco’s modified Eagle’s—DMEM—and
Ham’s F12—1:1 DMEM: F12) supplemented with antibiotics (penicillin/streptomycin, 100 U/mL; gentamicin.
50 mg/mL; kanamycin, 25, mg/mL), antifungicide (Fungizone) (amphotericin [B, 0.125 mg/mL), and 50% FBS
(Source of supplies: GibcoBRL, Life Technologies,
France). Adrenal medullary tissue was dissected from
cortical tissue with microscissors under binocular loop
(Olympus), cut into small pieces of 0.5 to 2.0 mm3 in
size and cultured at standard conditions (37°C with 5%
CO2) in media containing 50% fetal bovine serum (FBS).
During the course of a week, the percentage of FBS in
media was gradually reduced and adrenal medullary
pieces were maintained in culture at standard conditions
with media containing 10% FBS until needed for transplantation.
Grafts of adrenal chromaffin cells were performed on
the eighth day following isolation of adrenal medullary
tissue for the first patient (P1), on the 10th day for the
second (P2). In both cases, a sample of tissue was taken
for morphological analysis at the light microscopic level
prior to transplantation to confirm viability of the graft
tissue. Light microscopic analysis was assessed by immunoreactivity of chromaffin cells in tissue pieces to
antityrosine hydroxylase (TH) and/or antidopamine-bhydroxylase (DbH) markers on frozen sections as described previously (18). Neurochemical analysis of the
culture media for Met-enkephalin was also performed to
provide chemical support of chromaffin cell secretion
from the graft tissue and to parallel the morphological
findings. Met-enkephalin levels were determined by
radioimmunoassay using a commercially available kit
(Methionine Enkephalin 125 RIA kit, Incstar, Sorin,
France). Met-enkephalin levels were determined using
standard radioimmunoassay procedures (5).
Transplantation Procedure and Patient Follow-Up
The graft procedure and the follow-up of patients were
performed according to previously described methods
(14,19). In brief, an injection access Miniport (Cordis)
was situated in both cases on the level L4 –L5, of which
the distal end of catheter was respectively at T10 for P1
and T9 for P2. Then, a small CSF sample was obtained
for measuring normal CSF levels of Met-enkephalin to
establish a baseline control for future comparisons. Then,
approximately 1 mL of adrenal medullary tissue suspended in a balanced salt solution (without FBS) was
carefully infused into the CSF of the lumbar cistern.
ET AL.
229
For P1, two grafts were performed at different times
with adrenal glands taken from two male donors, 40 and
35 years old, respectively. The dissection product was
limited because of the dysmorphia of the first donor’s
glands, and the availability of only one gland for the
second donor (the second gland was inadvertently destroyed by the surgeon). Following transplantation of the
first adrenal medullary graft, P1 was immunosuppressed
for a period of 15 days with cyclosporin A (l0 mg/kg
daily, Sandoz): P1 was not immunosuppressed after the
second graft. For P2, two grafts were also performed:
donors were male 23 and 42 years old, respectively. In
the first donor, only the right gland was used because the
left gland was hypotrophied. Medullary tissue was very
sparse in the second donor’s glands and difficult to
separate from cortical tissue. No immunosuppressive
treatment was used following transplantation of the
second adrenal medullary allograft.
The implanted access port was kept in place for the
duration of the study, and this allowed for periodic
sampling of CSF for monitoring Met-enkephalin levels
without subjecting patients to the pain and discomfort
from repeated lumbar punctures. CSF examination
started on the day of the graft (day 0) with a second
sample taken 8 days later (day 8) and then periodically
every month subsequent to obtaining these controls.
Cytological examinations of CSF were performed the
routine way, on Nageotte’s cell after cytocentrifugation.
Two histological slides were stained with Maygrunwald
Giemsa, four slides were fixed with acetone at 220°C
and saved for eventual immunocytochemical tests. Concentrations of Met-enkephalin were measured by radioimmunoassay as described above. At day 0, pain intensity was determined in both using a visual analog scale
(VAS, where 0 5 no pain and 10 5 the most severe pain)
and behavioral evaluation was carried out according to
the Karnofsky scale. After transplantation, patient-reported pain scores were noted on a monthly basis.
Autopsy
Graft tissue was removed 8 h after P1’s death and 12 h
after P2’s. Autopsy was limited to the dorsal spinal axis
and cauda equina. After laminectomy, a large exeresis of
the dural sheath and its contents was performed between
T7 and sacrum, after ligature and section of the roots at
the internal orifice of the intervertebral foramen.
Autopsy specimens were ligated at their upper portion.
After fluid was aspirated for cytological analysis, specimens were then put into IX HlSTOChoicey Tissue
Fixative (Amresco, Solon, OH) with 20% ethanol. After
opening the dura mater and examination with a surgical
microscope (OPMI-l, Zeiss), the implant access ports
230
Cell Transplantation ● Volume 7, Number 3, 1998
were noted and photographed before the removal of the
grafts.
Histological Study
Dopamine-b-Hydroxylase (DbH) Immunostaining.
Following fixation of tissue, 1 mm3 cell blocks were cut
from the transplant, embedded in paraffin, and 10 mm
sections were cut and placed on slides. Tissue sections
were deparaffinized and rehydrated using standard methods. Slides with sections were placed in xylene for 5–10
min, dipped in a series of alcohol solutions for 3 min
each, then followed by phosphate-buffered saline (PBS)
for 15 min.
Endogenous peroxide activity was removed by a 40%
methanol, 3% hydrogen peroxide solution in PBS for 20
min. Nonspecific binding sites on tissue sections were
then blocked with 5% donkey serum (Amersham) in
PBS-BSA (PBS with 0.8% bovine serum albumin, BSA)
for 30 min. After washing three times (5 min) with
PBS-BSA, sections were incubated overnight (4°C) in
DbH (lncstar, diluted 1:1000). The following day, sections were rinsed in PBS-BSA three times (5-min intervals), followed by incubation in secondary antirabbit
biotinylated antibody (Amersham, 1:200 dilution) for 60
min at 4°C. Sections were washed again in PBS-BSA
three times (5 min), streptavidin biotinylated horseradish
peroxidase complex (Amersham, 1:100 dilution) was
added, and incubated for 30 min at 4°C. Sections were
rinsed in PBS three times (10 min), reacted with DAB
(3,39-diaminobenzidine, Sigma) for 1–10 min, dehydrated with a series of alcohol solutions, followed by
xylene, and mounted with permount.
Tyrosine Hydroxylase (TH) Staining. Following fixation with HISTOChoicey, areas of the spinal cord
containing transplanted adrenal chromaffin cell pieces
were dissected and placed in 20% sucrose solution in
0.1% phosphate buffer overnight at 4°C for cryoprotection. Cryostat sections of isolated tissue were cut at
10 –12 mm, mounted on slides, and incubated overnight
at 4°C with TH monoclonal antibody (1:500 dilution,
lncstar). Preimmune serum instead of primary antibody
was used on control sections. Next day, sections were
rinsed three times in PBS followed by incubation for l h
(at room temperature) with rhodamine-linked secondary
antibody (1:100 dilution, Cappel). After mounting on
slides and coverslipped with Fluoromount, sections were
observed through an Olympus BX6O epifluorescence
microscope.
Chromagranin A Immunostaining. A technique similar to the one described above (avidin– biotin–peroxidase) was performed on deparaffinized paraffin sections
for chromogranin A (lncstar, diluted 1:300) immunoreactivity.
RESULTS
Clinical and Biological Data
Patient 1. The first patient (P1), who had been operated on 6 yr previously, presented with lung cancer
localized in the lower right lobe. Lateral-thoracic, homolateral pain occurred subsequent to neoplastic pleural
invasion. This pain, which failed to respond to different
therapies, was partially controlled (VAS, 3–5) for a year
with buprenorphine (Temgesic three times/24 h). Slowrelease morphine (Moscontin) was then prescribed but
had to be discontinued, despite much significant analgesic action, because of severe and irreversible digestive
troubles. After 5 yr of chronic evolution, pain gradually
became more disabling (VAS between 9 –10), limiting
this patient’s autonomy and he subsequently required
permanent assistance (Karnofsky index: 50). Because of
the side effects with morphine taken orally, intrathecal
administration of morphine (2.5 mg/24 h, single dose)
was advised and this provided good, but incomplete
analgesia (VAS, 3–5) without the associated digestive
complaints common to oral morphine.
Morphine therapy through intrathecal administration
provided adequate pain relief initially, but with time P1
developed tolerance and the chronic pain associated with
his condition returned (VAS 7– 8). Because no other
alternative pain therapies were available to provide
relief, P1 consented to participate in the study to see
whether a transplant of adrenal medullary chromaffin
cells could alleviate his chronic pain.
The first intrathecal graft of about 1 mL of adrenal
medullary tissue was performed 4 mo later following
patient consent and committee approval for admission
into the study. Preparation conditions were fulfilled
(measurement of Met-enkephalin in culture media at
2500 pg/mL) as described, and the patient was given
cyclosporin A (according to the protocol, 10 mg/kg 24 h,
Sandoz) as a precautionary measure to preclude any risk
of graft rejection. On day 8, analgesia was excellent
(VAS 2) and it was noted that this patient’s general
condition had improved (Karnofsky index 70); however,
he still required intrathecal morphine, which had stabilized at the initial dose. After 15 days, cyclosporin A
treatment was stopped because the patient complained of
nausea.
Despite a significant increase in CSF Met-enkephalin
levels and a reduction in pain scores, a second graft was
performed on day 135 in an attempt to decrease morphine intake. Again, about 1 mL of adrenal medullary
tissue pieces suspended in a balanced salt solution was
Chromaffin cell allograft and cancer pain ● J.C. BÉS
injected into the lumbar cistern through the implanted
access port. The measurement of Met-enkephalin, from
supernatant media obtained from cultured adrenal medullary tissue to be used as the second graft, revealed basal
secretion at 2800 pg/mL. The patient was not given any
immunosuppressive treatment following the second graft
procedure.
Results over a period of 360 days are shown in Fig.
1. Pain scores (Fig. 1A, P1) go from 8 to 2 at the end
of the first week and remained stable at this level.
After the second graft, pain levels dropped to between
0 and 1 and functional activity remained excellent
until the eighth month (Karnofsky index .70) when
P1’s general condition quickly worsened with pain
occurring again.
The patient was maintained at 2.5 mg/24 h of morphine for 190 days and then weaned to 1.0 mg/24 h
by 1 120 days (Fig. 1B, P1). During the terminal period,
however, the dosage was increased to 3 mg/24 h over 30
days. After a period of increased Met-enkephalin in CSF
(Fig. 1C, P1) between day 0 (80 pg/mL) and day 190
(320 pg/mL), variations occurred with lower values until
day 1210 (between 120 and 150 pg/mL), with gradual
increase during the last 3 mo (around 200 pg/mL). Death
occurred on day 1385.
Patient 2. The second patient (P2) was operated on
for infiltrating cancer of the right urethra (enlarged
nephrectomy followed by radiotherapy). He presented
with subsequent progressive pain in the bilateral lumbar
area and in the right perineal and crural areas. At first the
pain responded well to slow-release morphine (100
mg/24 h), but after 6 mo treatment, morphine tolerance
developed and efficacy was under 50% with undesirable
digestive side effects when doses were increased. As this
patient’s general condition worsened (Karnofsky index
5 50) and pain became more severe, he was placed on
intrathecal morphine at 10 mg/24 h . This allowed for
greater pain relief without the side-effects associated
with higher oral doses of morphine. This pain therapy
provided relief for the patient for only 2 mo before his
condition worsened and the pain returned (VAS 8 –9).
Because intrathecal morphine administration is the
option that provides the most potent pain relief and this
therapy was inadequate, P2 consented to be a participant
in this study to see if an adrenal medullary transplant
could alleviate the chronic cancer pain that had been
unyielding to intrathecal morphine therapy. A graft of
about 1 mL adrenal medullary tissue pieces was performed into the lumbar cistern through an access port.
Prior to transplantation, a sample of culture medium was
obtained from adrenal medullary tissue pieces maintained in culture, and this revealed a rate of Metenkephalin at 2600 pg/mL.
ET AL.
231
Following transplantation, P2’s general condition had
improved (Karnofsky index 5 70), as demonstrated by
excellent pain control from day 18 to day 130 (VAS
0.5–1), and this finding was complemented by reduced
morphine doses that had stabilized to 2.5 mg/24 h. But
because of renewed pain, doses of daily morphine had to
be progressively increased to 12 mg/24 h between day
160 and day 190. Therefore, a second graft was
performed on day 1136. The graph in Fig. 1 shows the
results obtained for both adrenal medullary grafts in P2
over a period of 360 days where pain scores (Fig. 1A, P2)
are reduced from 9 to 1 after the first week and are
between 0 and 1 on day 130; afterwards, they increase
until day 1120 (VAS 5) and appear to stabilize after the
second graft (VAS, 0 –1) until day 1330.
Biochemical data on P2 shows that Met-enkephalin
levels in CSF (Fig. 1C, P2) showed two peaks on day
130 and day 1150 (respectively at 270 and 300 pg/mL)
from initial values of about 100 pg/mL; Table 1 and Fig.
1C contains all CSF neurochemical data. Although
Met-enkephalin levels fell, after an initial peak, within 3
mo of receiving the first transplant, following a second
transplant at day 1136, high levels were maintained and
appear to stabilize for the subsequent 8 mo until P2’s
death.
During the terminal period, perineal lumbar pain and
the radix pain of the inferior right member were completely alleviated. There was only distal dysesthesia of
inferior members related to lymphedema subsequent to
pelvic extension of the cancer, which did not respond
well to opioids. Analgesic control remained excellent
(Fig. 1A, P2) until the patient’s death (day 1362),
despite no increases in morphine doses throughout the
follow-up period (Fig. 1B, P2).
Morphological Analysis
Autopsies. Access ports are generally situated between T12 and the sacrum, but one was located at the
level of T9 in P1 (Fig. 2A). In both cases, access ports
were located at points where isolated nodular formations
(0.2– 0.3 mm in diameter) and, as was typically the case,
multilobulated nodules of irregular size (up to 10 mm
length) could be seen (Fig. 2B). These formations were
located along thoracic and lumbosacral roots of the
cauda equina (Fig. 2C) and seemed to adhere to the
radicular sheath and were without visible macroscopic
extension into the roots.
Only one nodular location was found on the internal
face of the dura mater on the level of the external
foramen of the posterior root (Fig. 2D), whereas filum
terminale and conus medullaris were spared in both
cases. On the whole, injected material spread proximally
from the injection area at the level of lumbar enlarge-
232
Cell Transplantation ● Volume 7, Number 3, 1998
Fig. 1. (A) Patient-reported pain score (from 0 5 no pain to 10 5 the most intense pain). P1: patient 1; P2: patient 2; Single arrow
5 initial chromaffin cell graft; double arrows 5 second transplant. (B) The complementary daily intrathecal morphine intake (mg/24
h). (C) The Met-enkephalin release into the CSF (pg/mL). The mean basal level of CSF Met-enkephalin was done for 10 nongrafted
patients (hatched areas). This was the reference value for the CSF Met-enkephalin rate of grafted patients.
Chromaffin cell allograft and cancer pain ● J.C. BÉS
ET AL.
233
Table 1. The complementary intrathecal morphine intake (mg/24 h) and the CSF Met-enkephalin release (pg/mL) data
Following (days)
J0
18
130
160
190
2,5
2,5
2,5
1
1
1
1
1
3
3
3
3
170
280
320
150
160
130
120
—
210
230
200
190
5
2,5
12
12
12
10
12
12
12
12
12
12
270
210
105
110
300
280
280
260
250
280
190
210
P1
Intrathecal morphine 2,5 2,5
(mg/24 h)
Met-enképhaline/CSF 80 130
(pg/mL)
P2
Intrathecal morphine 10
5
(mg/24 h)
Met-enképhaline/CSF 100 230
(pg/mL)
1120
ment and conus medullaris and distally into the area of
dural pouch without precise systematization between the
two areas. The graft pieces furthest from the injection
sites were found at 18 cm, but because the anatomic
elements studied did not contain higher cervicodorsal
spine and brain, we could not exclude further locations.
Histology
Nerve roots did not present demyelination or segmentary degeneration at the level of the grafts. They did not
adhere to the radicular sheath and were irregularly
limited by a thin encapsulating connective tissue lamina.
A fibroconnective tissue structure that sometimes developed between the fragments of adrenal medullary tissue
pieces constitutes an important polymorphic conglomerate in both patients. Rare mononuclear infiltrates were
observed in P2’s biopsies; however, no inflammatory
cells could be seen in P1’s graft. In some sections,
microscopic examination revealed some necrosis involving a part of the implanted tissue. The chromaffin
granules appeared dense and dispersed throughout the
cytoplasm, (Fig. 5), while large areas of degranulation
can be observed with characteristic vacuolization with
the electron microscope [cf. (16)]. With the light microscope, the structural heterogeneity of antichromagranin
A tissue immunoreactive areas (Fig. 3) confirms the
neuroendocrine specificity of these nodular formations
(i.e., the grafted tissue). In the same way, positive
reaction to enzymatic markers (TH and DbH) reveals a
functional identification of chromaffin cells (Fig. 4).
DISCUSSION
Several animal studies have provided substantial evidence supporting the use of adrenal chromaffin cell
implants as a potentially useful therapy for chronic pain
(8,23,36). These studies show that implants of chromaffin cells to the subarachnoid space can consistently
reduce pain sensitivity in such well-established chronic
pain paradigms as the adjuvant-induced arthritic rat
1150
1180
1210
1240
1270
1300
1330
1360
model (23) and the Bennett-Xie model (8). From previous biochemical studies, the mechanism of pain reduction is postulated to occur by the release of natural
pain-relieving substances such as catecholamines and,
especially, by opioid peptides that are coreleased (20,22).
These neuroactive substances secreted from chromaffin
cells travel locally through the subarachnoid space and
diffuse into the dorsal horn of the spinal cord where
primary sensory afferents synapse. It is hypothesized that
the secreted products of chromaffin cells either modulate
the synaptic function of somatosensory afferents by
inhibiting synaptic activity or by activating local sensory
afferent inhibitory circuits. Most recent studies suggest
that the mechanism of pain relief in chromaffin cell
implants may be through inhibition of the NMDA
receptor-mediated events in secondary afferent neurons
(10,28).
Based on these animal studies that have characterized
the mechanism and efficacy of pain relief with chromaffin cell implants in rats, our group received permission
by ethics committees to initiate preliminary clinical trials
on human patients with chronic intractable pain resulting
from terminal cancer. The collaborative study between
the Toulouse group at the University of Paul Sabatier and
the Chicago group at The University of Illinois at
Chicago has shown that 11 out of 15 cancer patients
benefited from chromaffin cell allografts (19). If chromaffin cell transplantation is proposed as an alternative
therapy to the conventional use of intrathecal morphine,
it is important to establish the long-term efficacy of
chromaffin cell grafts in humans. The purpose of this
article was to report on two of the patients with longterm chromaffin cell allografts that survived and functioned for 1 yr. Transplant function in patients was
assessed using several subjective and objective analyses
that included patient-reported pain scores, morphine
intake, CSF levels of Met-enkephalin, and histological
staining of transplant tissue obtained at autopsy.
For subjective analysis of transplant function, P1 and
P2 rated the severity of their pain on the Visual Analog
234
Cell Transplantation ● Volume 7, Number 3, 1998
Fig. 2. Autopsy specimens from P1 show the dispersion of grafted tissue fragments (A, whole spinal cord, reduced, ruler length 16
cm) from the injections site (L4/L5). Note (arrows), multilobulated nodules of irregular size (B), clusters of adrenal medullary tissue
between the roots of the cauda equina (C), and one nodular location in the inner aspect of the dura mater (D) (mag: B, 33; C, 43;
D, 43).
Scale (VAS) where 1 represented the least and 10 the
most intense. Pain scores reported by both patients
showed that within the first week following implantation
of adrenal medullary tissue there was a rapid decrease in
pain scores from 7 to 2 in P1, and 8 to 1 in P2.
Essentially, in the course of 1 wk, P1’s and P2’s
perception of pain changed from severe pain to having
mild or no pain.
Besides pain scores, additional evidence that adrenal chromaffin cell allografts are functional for pain
relief is based on biochemical studies showing that
CSF levels of Met-enkephalin increased following
Chromaffin cell allograft and cancer pain ● J.C. BÉS
Fig. 3. Low-power photomicrograph (A) of a section through
two fragments of adrenal medullary tissue isolated from the
lumbar-sacral roots and immunostained for chromagranin A
(CGA) reactivity. At higher magnification (B), CGA-positive
remnants revealed the rounded morphology of chromaffin cells
and cellular and collagenous elements (arrowheads) between
these nodular formations (mag: A, 3203; B, 22003).
transplantation. Synthetic opioids have well-known
pharmacological applications in the treatment of severe acute and chronic pain symptoms (37,38). The
natural analog to morphine are enkephalins, which are
produced by chromaffin cells and coreleased with
catecholamines (17). In both patients (P1 and P2),
normal CSF levels of Met-enkephalin prior to transplantation were 80 and 100 picograms, respectively.
Following the first transplant, Met-enkephalin levels
increased to 170 picograms (130 days) in P1 and 270
picograms (130 days) in P2. This demonstrated a high
correlation to pain scores (which fell from 10 to 1 and
10 to 2, respectively, in patients P1 and P2). After the
second transplant, Met-enkephalin levels reached as
high as 320 picograms in P1 (at 190 days) and 300
picograms in P2 (at 1 150), which correlated again to
reduced pain scores. Although Met-enkephalin is only
one of many opioid peptides believed to be secreted by
chromaffin cells, this biochemical assay has provided
evidence that chromaffin cells are probably releasing
many other neuroactive substances (30,31). Further-
ET AL.
235
Fig. 4. Low-power photomicrograph (A) of dopamine-b-hydroxylase (DbH)-stained section through the graft site. (B)
Chromaffin cells immunohistochemically stained with tyrosine
hydroxylase (TH) antibody with a rhodamine-linked secondary
antibody marker. Note the intensity of TH labeling. (mag: A,
1603: B, 12003).
more, the high correlation of CSF Met-enkephalin
levels to patient pain scores indicates that opioid
production by chromaffin cells is a major factor
Fig. 5. The electron microscopy study (330,000) show an
important tissue alteration with degenerative chromaffin cells
in necrotic zone but also the evidence of some secretory
granuli.
236
Cell Transplantation ● Volume 7, Number 3, 1998
responsible for pain relief. Among the other released
products from chromaffin cells that may contribute to
pain relief are several neurotrophic factors including
fibroblast growth factor (FGF), transforming growth
factor, and ciliary neurotrophic factor (CNTF) (31).
Recent reports suggest that a diminution of neurotrophic factors may contribute to the pathophysiology of
certain neuropathies, and artificial replacement strategies of growth factors have been shown to reverse
peripheral neuropathy [for review, see (2)].
In both cases, the decision to give a second transplant
was based either on a modest increase in Met-enkephalin
levels or when pain scores/morphine intake increased.
Therefore, a second transplant was performed on P1 at
day 140 and on P2 at day 1136 after the first transplant.
Following a second transplant, patient P1 was relatively
pain free for the next 330 days, which corresponded to
only a slight increase in morphine intake. In patient P2,
a second transplant resulted in pain relief for an additional 230 days but with no reduction in morphine intake.
It should be noted in both cases that morphine intake was
stabilized and patient Met-enkephalin levels in CSF were
double and close to triple their baseline value prior to
transplantation. It is believed that the apparent weaker
analgesic effect with the first transplant in both patients
was due to an inadequate amount of adrenal chromaffin
tissue transplanted. We have found that about 2 mL of
small, viable adrenal medullary tissue fragment (0.5–2.0
mm3 in size) should be transplanted initially to obtain
optimal results so that additional transplants may not be
necessary (36).
The problems associated with obtaining enough
healthy human donor tissue is not uncommon, and has
sparked interest in alternative donor sources such as
xenografts (1,6). The difficulty faced with obtaining
human adrenal medullary tissue presents a unique problem for several reasons. Although adrenals are routinely
removed from cadaver donors together with the kidneys,
it can be difficult obtaining glands with a large amount of
intact adrenal medullary tissue. Also, a thick adrenal
medulla is needed because some chromaffin cells/tissue
are lost during microdissection and culture. In addition,
because of the awkward position of the adrenals when
cadavers are supine and their close proximity to the
kidney, many surgeons fear that donor kidneys may be
injured while trying to dissect the adrenal gland from the
perirenal fascia. Therefore, often only a fraction of the
adrenal tissue may be obtained instead of the entire
gland.
Both patients, P1 and P2, died as a result of their
terminal cancer condition at 1385 days and 1362 days,
respectively. Prior to death, consent was obtained permitting autopsy to assess graft survival and morphological changes, if any, of the grafted and host tissue.
However, the tissue was not preserved with fixative until
8 h (in P1) and 12 h (P2) following death, so it was
recognized that cellular morphology would not be in
pristine condition. Gross morphology of the spinal cord
and nerves in the lumbar cistern and in rostral spinal
segments indicated that there was postimplantation migration. In P1, adrenal tissue fragments were identified
as high as T6 and as low as S1. This migration of tissue
may reflect the response of tissue fragments to gravity as
a result of the sedentary behavior of P1 and P2. However,
the histology of implants from our rodent studies have
not indicated that there is migration of graft because
adrenal medullary tissue appears to be firmly attached to
the pia mater of the dorsal spinal cord and in close
proximity to the implantation site (23). This is not
surprising if one considers that, because human and
rodent donor adrenal medullary tissue pieces are of
similar size (1 mm3), the lumbar cistern in our clinical
trials is a much larger space and can accommodate more
movement of tissue than in the rat subarachnoid space.
Transplant survival in patients P1 and P2 was assessed
by immunostaining for specific chromaffin cell markers
in grafted tissue fragments from autopsy specimens. This
reveals that cells in the graft were tyrosine hydroxylase,
dopamine-b-hydroxylase, and chromagranin A positive.
The presence of these enzymes confirms that chromaffin
cells were a component of the tissue fragments, and they
were producing catecholamines providing morphological
evidence of graft function in P1 and P2.
Biochemical and morphological analyses of graft
function are evidence that adrenal medullary implants
into the subarachnoid space can control pain in cancer
patients for at least 1 yr. Patient pain scores decreased
drastically in both cases. A reduction in pain scores
corresponded to an increase in CSF Met-enkephalin
levels that were double and triple their preimplantation
level. Furthermore, the morphology of implanted adrenal
medullary tissue fragments indicated that chromaffin
cells were implanted and they were actively producing
and secreting catecholamines (19). Despite the fact that
no effort was made to crossmatch donor and recipient,
there appeared to be no rejection of implanted chromaffin cells. Infiltrating leukocytes were absent in graft
fragments and glial tissue around it. The adrenal medullary tissue graft appeared intact.
In summary, the use of implanted adrenal medullary
tissue represents a new approach to treating cancer pain
that rivals current pharmacological approaches such as
implantable morphine pumps, which are expensive and
have many side effects. Furthermore, recent studies
indicate that adrenal chromaffin cell implants inhibit the
development of morphine tolerance and opens the possibility for combined therapy of oral morphine and
adrenal medullary transplantation (34). Because our
Chromaffin cell allograft and cancer pain ● J.C. BÉS
results have been positive with cancer patients, in the
future we hope to extend our studies to include other
chronic pain syndromes.
14.
Acknowledgments — This work was supported, in part, by the Pain
Institute UPS(IUD) grants, and by NIH Grant NS-28931.
REFERENCES
1. Aebischer, P.; Buchser, E.; Joseph, J.M.; Favre, J.; DeTrebolet, N.; Lysaght, M.; Rudnick, S.; Goddard, M.
Transplantation in humans of encapsulated xenogeneic
cells without immunosuppression: A preliminary report.
Transplantation 58:1275–1277; 1994.
2. Apfel, S.C.; Kessler, J.A. Neurotrophic factors in the
therapy of peripheral neuropathy. Bail. Clin. Neurol.
4:593– 629; 1995.
3. Buchser, E.; Goddard, M.; Heyd, B.; Joseph, J.M.; Favre,
J.; DeTrebolet, N.; Lysaght, M.; Aebischer, P. Immunoisolated xenogeneic chromaffin cell therapy for chronic pain:
Initial clinical experience. Anesthesiology 85:1005–1012;
1996.
4. Czech, K.; Sagen, J. Update on cellular transplantation into
the CNS as a novel therapy for chronic pain. Prog.
Neurobiol. 46:507–529; 1995.
5. Duka, T.; Holt, V.; Przewlocki, R.; Wesche, O. Distribution of methionine and leucine enkephalin within the rat
pituitary gland measured by highly specific radioimmunoassays. Biochem. Biophys. Res. Commun. 85:119; 1978.
6. Dunning, J.J.; White, D.J.; Wallwork, J. The rationale for
xenotransplantation as a solution for donor organ shortage.
Pathol. Biol. 42:231–235; 1994.
7. Fazakerley, J.K.; Webb, H.E. Cyclosporin enhances virally
induced T-cell-mediated demyelination. The effect of cyclosporine on a demyelinating virus infectione. J. Neurol.
Sci. 78:35–50; 1987.
8. Hama, A.T.; Sagen, J. Reduced pain-related behavior by
adrenal medullary transplants in rats with experimental
pain neuropathy. Pain 52:223–231; 1993.
9. Hama, A.T.; Sagen, J. Alleviation of neuropathic pain
symptoms by xenogeneic chromaffin cell grafts in the
spinal subarachnoid space. Brain Res. 651:183–193; 1994.
10. Hama, A.T.; Unnerstall, J.R.; Siegan, J.; Sagen, J. Modulation of the NMDA receptor expression in the rat spinal
cord by peripheral nerve injury and adrenal medullary
grafting. Brain Res. 687:103–113; 1995.
11. Joseph, J.M.; Goddard, M.B.; Mills, J.; Padrum, V.; Zurn,
A.; Zielinski, B.; Favre, J.; Gardez, J.P.; Mosimann, F.;
Sagen, J.; Christenson, L.; Aebischer, P. Transplantation of
encapsulated bovine chromaffin cells in the sheep subarachnoid space. A preclinical study for the treatment of
chronic paine. Cell Transplant. 5:355–364; 1994.
12. Lazorthes, Y.; Verdié, J.C.; Bastide, R.; Lavados, A.;
Descouens, D. Spinal vs. intraventricular chronic opiate
administration with implantable drug delivery devices for
cancer pain. Appl. Neurophysiol. 48:234 –241; 1985.
13. Lazorthes, Y.; Sallerin-Caute, B.; Verdié, J.C.; Bastide, R.
Advances in drug delivery systems and applications in
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
ET AL.
237
neurosurgery. In: Advances and technical standards in
neurosurgery, vol. 18. Berlin: Springer Verlag; 1991:143–
192.
Lazorthes, Y.; Bés, J.C.; Sagen, J.; Tafani, M.; Tkaczuk, J.;
Sallerin, B.; Nahri, I.; Verdié, J.C.; Ohayon, E.; Caratero,
C.; Pappas, G.D. Transplantation of human chromaffin
cells for control of intractable cancer pain. Acta Neurochir.
64:97–100; 1995.
Lemaire, M.; Maurer, G.; Wood, A.J. Pharmacokinetic and
metabolism in cyclosporin. In: Broel, J.F., ed. Cyclosporin
prog. allergy, vol. 38. Basel: Karger; 1986:93–107.
Nahas, G.D.; Zagury, D.; Milhaud, A.; Manger, W.M.;
Pappas, G.D. Acidemia and catecholamine output of the
isolated canine adrenal gland. Am. J. Physiol. 213:1186 –
1192; 1967.
Nguyen, T.T.; De Leane, A. Noradrenergic modulation of
clonidine of the cosecretion of catecholamines and enkephalins in adrenal chromaffin cells. Can. J. Physiol.
Pharmacol. 65:823– 827; 1987.
Ortega, J.; Sagen, J.; Pappas, G.D. Short-term immunosuppression enhances long-term survival of bovine chromaffin
cell xenografts in the rat CNS. Cell Transplant. 1:33– 41;
1992.
Pappas, G.D.; Lazorthes, Y.; Bés, J.C.; Tafani, M.; Winnie,
A. Relief of intractable cancer pain by human chromaffin
cell transplants: Experience at two medical centers. Neurol.
Res. 19:71–77; 1997.
Sagen, J.; Pappas, G.D. Morphological and functional
correlates of chromaffin cell transplants in CNS pain
modulatory regions. Ann. NY Acad. Sci. 495:306 –333;
1987.
Sagen, J.; Kemmler, J. Increased levels of met-enkephalin-like immunoreactivity in the spinal cord CSF of rats
with adrenal medullary transplants. Brain Res. 502:1–
10; 1989.
Sagen, J.; Pappas, G.D. Increased opioid peptide and
catecholamine production in human adrenal medullary
explants. Soc. Neurosci Abstr. 16:835; 1990.
Sagen, J.; Wang, H.; Pappas, G.D. Adrenal medullary
tissue implants in the rat spinal cord reduce nociception in
a chronic pain model. Pain 42:69 –79; 1990.
Sagen, J.; Kemmler, J.; Wang, H. Adrenal medullary
transplants increase spinal cord cerebrospinal fluid catecholamine levels and reduce pain sensitivity. J. Neurochem. 56:623– 627; 1991.
Sagen, J.; Wang, H.; Tresco, P.; Aebischer, P. Transplants
of immunologically isolated xenogeneic chromaffin cells
provide a long-term source of pain-reducing substances.
J. Neurosci. 13:2415–2423; 1993.
Sagen, J.; Pappas, G.D.; Winnie, A. Alleviation of pain in
cancer patients by adrenal medullary transplants in the
subarachnoid space. Cell Transplant. 2:259 –266; 1993.
Schug, S.A.; Zech, D.; Grond, S.; Jung, H.; Meuser, T.;
Stobbe, B. A long-term survey of morphine in cancer
patients. J. Pain Manage. 7:250 –266; 1992.
Siegan, J.; Sagen, J. Attenuation of NMDA-induced spinal
hypersensitivity by adrenal medullary transplants. Brain
Res. 680:88 –98; 1995.
238
Cell Transplantation ● Volume 7, Number 3, 1998
29. Tkaczuk, J.; Bes, J.C.; Duplan, H.; Sallerin, B.; Tafani, M.;
Charlet, J.P.; Abbal, M.; Lazorthes, Y.; Ohayon, E. Characterization of CSF lymphocytes before and after intrathecal chromaffin cells allografts for intractable pain treatment (submitted).
30. Unsicker, K. The trophic cocktail made by adrenal chromaffin cells. Exp. Neurol. 123:167–173; 1993.
31. Unsicker, K.; Krieglstein, K. Growth factors in chromaffin
cells. Prog. Neurol. 48:307–324; 1996.
32. Vaquero, J.; Martinez, R.; Oya, S.; Coca, S.; Salazar, F.G.;
Colado, M.I. Transplantation of adrenal medulla into
spinal cord for pain relief. Disappointing outcome. Lancet
12:1315; 1988.
33. Vaquero, J.; Martinez, R.; Coca, S.; Oya, S.; Salazar, F.G.;
Colado, M. Pain relief in humans by chromaffin tissue
graft into medullary arachnoid. Transplantation 1:3942;
1989.
34. Wang, H.; Sagen, J. Absence of appreciable tolerance and
35.
36.
37.
38.
morphine cross-tolerance in rats with adrenal medullary
transplants in the spinal cord. Neuropharmacology 33:
681– 692; 1994.
Wang, H.; Sagen, J. Attenuation of pain-related hyperventilation in adjuvant arthritis rats with adrenal medullary
transplants in the spinal subarachnoid space. Pain 63:313–
320; 1995.
Winnie, A.; Pappas, G.D.; Das Gupta, T.K.; Wang, H.;
Ortega, J.; Sagen, J. Subarachnoid adrenal medullary
transplants for terminal cancer pain. Anesthesiology 79:
644 – 653; 1993.
Yaksh, T.L.; Reddy, S.V.R. Studies in the primate on the
analgetic effects associated with intrathecal actions of
opiates, a-adrenergic agonists and baclofen. Anesthesiology 54:451– 467; 1981.
Zenz, M.; Strumpf, M.; Tryba, M. Long-term opioid
therapy in patients with chronic nonmalignant. J. Pain
Manage. 69 –77; 1992.