Retreatment of Pediatric Brain Tumors With
Radiation and Misonidazole
Results of a CCSG/RTOG Phase 1/11 Study
WILLIAM M. WARA, MD,* KENT E. WALLNER, MD,* VICTOR A. LEVIN, MD,t
HSIU-CHIH LIU, MD,t AND MICHAEL S. B. EDWARDS, MDt
Twenty-nine patients with recurrent pediatric brain tumors after full-dose treatment with radiation were
retreated with misonidazole (9 gm/m2) and whole-brain irradiation (300 rad X 10). Seventeen of 29
patients (59%) experienced nausea and vomiting with misonidazole administration and 6 patients (21%)
developed peripheral neuropathy. Serum levels of misonidazole were similar to those reported for adults.
One patient died of trauma shortly after retreatment. For the remaining 28 patients, the median time-to-
progression was 5.5 months. Median survival was 13 months. Six patients (21%) developed radiation
toxicity, two of whom died from the toxicity. Whereas the results on this study are similar to reported
series of patients receiving salvage chemotherapy for recurrent pediatric brain tumors, more effective
treatment clearly is needed.
Cancer 58: 1636-1 640,1986.
D nostic, surgical and radiotherapeutic techniques
over the last 20 years, patients with malignant brain tu-
mors have a high chance of local tumor recurrence after
definitive treatment.'-3 Several authors have reported
some success with combination chemotherapy for patients
who have failed definitive However, most,
if not all patients with recurrent brain tumors treated with
chemotherapy ultimately will succumb to their disease.
Data concerning retreatment of brain tumors with ra-
diation is sparse.' Two report^^^'^ have shown some long-
term palliation in a sizeable percentage of patients who
are retreated. Most of the retreated patients, however, had
received suboptimal doses of radiation for their initial
therapy. Radiation therapists are generally reluctant to
offer patients a second course of radiation for fear of ex-
ceeding normal tissue tolerance.
There has been considerable interest over the last decade
in the use of radiosensitizer drugs that increase the effect
of radiation on tumors, presumably by mimicking the
effect of oxygen in poorly vascularized tumors. Misoni-
dazole, the most extensively studied radiosensitizer, has
shown considerable promise in experimental tumors.''
ESPITE THE INTRODUCTION of more effective diag-
From the *Department of Radiation Oncology, and ?Brain Tumor
Research Center, Department of Neurological Surgery, University of
California, San Francisco, California.
Address for reprint: William Wara, MD, Department of Radiation
Oncology, University of California, San Francisco, CA 94143.
Accepted for publication February 14, 1986.
We report here the results of a clinical Phase 1/11 trial
designed to test the efficacy and safety of combining mis-
onidazole with re-irradiation for recurrent pediatric brain
tumors. Re-irradiation was done with large fractions (300
rad X 10) over a relatively short period of time (2 weeks)
to allow maximum interaction with a limited number of
doses of misonidazole.
Patients with recurrence of intracranial brain tumors
after definitive irradiation were eligible for entry into the
protocol. In all cases, prior irradiation was done with con-
ventional fractionation of 150 to 200 rad per day to a
total dose of 4500 to 6000 rad. Tumor recurrence was
documented in all patients by computerized tomography
(CT) scan. Five patients underwent a repeat biopsy or
subtotal resection of their tumor recurrence before entry
into the study. Patients were not excluded if they had
received prior chemotherapy, either at the time of first
diagnosis or as part of an earlier attempt at salvage (more
than 1 month before entry). The study was limited to
patients 18 years of age or less at original diagnosis, or to
older patients with medulloblastoma. Patients had to be
in functional classes I or 11, and have a Karnofsky per-
formance status of 40 or more. Written, informed consent
was required from the patient or a responsible family
member. Hematologic, hepatic, and renal function pa-
rameters had to be within normal limits at the start of
OF BRAIN TUMORS -
WUrU et Ul.
Whole-brain irradiation was given with lateral, opposed
portals, delivering 300 rad midplane dose in 10 fractions
given on a daily basis. Treatment was given with mega-
voltage radiation, using 6oCo or a linear accelerator (4 or
6 MeV). Electron, particle, or implant boosts were not
Misonidazole, obtained from the National Cancer In-
stitute, Division of Cancer Treatment, was given orally
on alternate days (Monday, Wednesday, Friday) during
radiation. The planned daily dose was 1.5 gm/m2 for a
total dose of 9.0 gm/m2. The drug was administered 4 to
6 hours before irradiation. The use of antiemetics was
allowed, as needed. Complete blood count (CBC), plate-
lets, and serum chemistries were monitored weekly during
therapy and monthly thereafter.
When possible, misonidazole serum levels were mon-
itored at 4 to 6 hours and 24 to 30 hours after each oral
dose. Serum was collected in a heparinized syringe and
the drug concentration was measured by ultraviolet spec-
troscopy or high-pressure liquid chromatography. The re-
ported serum concentrations include both misonidazole
and its primary metabolite, desmethylmisonidazole.
Toxicity associated with misonidazole was monitored
closely. Nausea and vomiting was rated on a scale of 0 to
4 (Table 1). Peripheral neuropathy was rated on a scale
of 0 to 4 (Table 2). Any other toxicity thought to be related
to misonidazole was separately noted.
Glucocorticoids were used as needed in the smallest
doses possible, while still affording patients satisfactory
neurologic function. A significant increase in steroid dos-
age was interpreted as evidence of clinical deterioration.
Upon completion of protocol therapy, no further treat-
ment was to be given until evidence of tumor progression,
at which time individual investigators were free to employ
further salvage treatment, if such was deemed appropriate.
Time-to-progression was calculated from the first day
of treatment until definite clinical or CT evidence of in-
tracranial progression. Time-to-death was calculated from
the first day of treatment. Postmortem examination was
obtained whenever possible.
Thirty-three patients were enrolled in the protocol from
December 1978 through November 1980. One patient
deteriorated suddenly and died after her fourth fraction
of radiation and is omitted from the results. Three patients
were treated on protocol but received combination che-
motherapy before evidence of disease progression; they
are omitted from the results. The remaining 29 patients
form the basis of this report.
There were 9 females and 20 males. Patient age at the
time of retreatment ranged from 4 to 27 years (median,
14 years; mean, 13 years). Patient Karnofsky status ranged
from 40 to 90 (mean, 7 1 : median, 70). The types of tumors
TABLE I. Misonidazole Toxicity
Level of toxicity
Nausea and vomiting
0 No nausea or vomiting
1 Slight nausea, no vomiting
2 Nausea, occasional vomiting
3 Nausea, vomiting 1-3 times daily
4 Severe and continuous nausea, continuous vomiting
requiring hospitalization 0
0 None measurable
I Decreased deep tendon reflexes or absent ankle jerk;
mild constipation; mild parasthesias
2 Absent deep tendon reflexes excluding ankle jerk;
moderate parasthesias; mild weakness
3 Disabling sensory loss or pain; peripheral nerve
weakness or foot drop; severe parasthesias; severe
4 Unable to walk or sleep due to paresthesias or pain;
confined to bed or chair due to peripheral neurologic
treated, patient age at treatment, median survival, and
time-to-progression are summarized in Table 2.
All patients had completed definitive irradiation, from
11 months to 6.6 years before enrollment in the study
(mean, 34 months; median, 27 months). Previous treat-
ment generally consisted of whole-brain irradiation plus
a boost to the area of the primary lesion. Patients with
medulloblastoma had received spinal irradiation as part
of their initial therapy. The spinal cord was included in
the retreatment volume at the discretion of the individual
investigators (1 5 patients). Twenty-two of the 29 patients
(76%) had been treated with chemotherapy as part of their
previous therapy, either as part of their initial treatment,
or as earlier salvage treatment.
A single patient died from trauma, 1 month after com-
pleting retreatment, and is omitted from the survival and
time-to-progression data, but was included in the toxicity
analysis. Time-to-progression ranged from 0 to 48 months,
with a median of 5.5 months (Fig. 1). In three cases, the
first evidence of progression was increased tumor size on
routine follow-up CT scan. In all other patients, clinical
deterioration was the first evidence of disease progression.
The time-to-progression was not significantly affected by
the interval from prior therapy or by initial Karnofsky
2. Survival and Time-to-Progression Versus Tumor Type
(6, 18 mo)
(2, 7 mo)
1 0 0 r
CANCER October 15 1986
FIG. 1. Survival and freedom-
from-progression for 28 patients.
Follow-up was complete for all pa-
MONTHS AFTER RETREATMENT
status (P > 0.1). Nine patients received salvage chemo-
therapy after evidence of disease progression, none of
Survival from the time of retreatment is shown in Figure
1. Survival ranged from 3 to 52 months (median, 13
months). Patients with recurrent medulloblastomas sur-
vived somewhat longer than those with other tumors (Ta-
ble 2). The interval from previous therapy or Karnofsky
status did not affect survival (P > 0.1).
Six patients (2 1 %) were reported to experience long-
term impairment due to radiation toxicity. One patient
died with autopsy-confirmed bulbar radionecrosis (de-
scribed below). A second patient died 3 years after re-
treatment of a medulloblastoma from what was clinically
believed to be severe degenerative encephalopathy. A re-
peat CT scan was not done to rule out recurrent tumor
and no autopsy was performed. Three additional patients
had neurologic impairment (seizures, leukoencephalop-
athy, cerebellar dysfunction) attributed to radiation tox-
icity at 0,6 and 12 months after retreatment, respectively,
but all eventually died of recurrent tumor. A sixth patient
developed deafness, attributed to radiation. Five of six
patients (83%) who developed radiation toxicity had re-
ceived combination chemotherapy as part of previous
therapy, versus 16/22 (73%) of those who did not develop
Twenty-six of 28 patients (93%) died of what was clin-
ically judged to be recurrent tumor. Tumor regrowth was
confirmed by CT scan in 13 patients, but not in the re-
mainder. Five patients underwent autopsy. Four of the
five were found to have extensive intracranial tumor,
causing their death. Of those four patients, two were noted
to have changes consistant with radiation, but not con-
tributing to their death. The fifth patient was found to
have diffuse, severe radiation changes, including radione-
crosis of the brainstem, which caused his death from ap-
nea. He had been treated in 1975 with 5000 rad to the
posterior fossa for a high-grade astrocytoma, followed by
2 years of treatment with vincristine, prednisone, and
CCNU. He recurred locally in 1979 and was treated with
misonidazole and 3000 rad whole-brain irradiation. He
did well until 1980, when he developed bulbar palsies
which progressed and caused his death in 1983.
Seventeen of 29 patients (59%) experienced nausea/
vomiting during misonidazole administration. The
symptoms were generally easily controlled with anti-
emetics. In no case did the nausea and vomiting require
termination of misonidazole and in no case did it persist
Six patients (2 1 %) developed peripheral neuropathy at-
tributed to misonidazole (Table 1). One patient had hy-
poreflexia only. Five of the six developed parasthesias of
the extremities. Four were grade I1 and one was grade 111.
Peripheral neuropathies developed from 1 to 44 days after
the first dose of misonidazole and lasted from 14 to 350
days. The neuropathies resolved with time in all cases.
Three patients (among the six with peripheral neuropa-
thies) were reported to develop ototoxicity, consisting of
bilateral sensorineural hearing loss, from 10 to 18 days
after their first dose of misonidazole. In no case did the
hearing loss per se impair daily function. In two patients,
the hearing loss was noted to be improving at 30 and 90
days. The third patient was still noted to have diminished
hearing two years after retreatment. One patient had an
isolated seizure after administration of the fourth dose of
misonidazole. No definite etiology for the seizure was de-
OF BRAIN TUMORS -
Wara et al.
termined, and she had no further seizures after medical
therapy. It is unclear whether the single seizure was due
Serum misonidazole levels were drawn from 4 to 6
hours after administration in 13 patients. The levels
ranged from 40 to 95 pg/ml, with a mean value of 68 pg/
ml (+15 pg/ml). Serum levels were drawn at 24 to 30
hours in 10 patients, ranging from 3 to 35 pg/ml, with a
mean value of 18 pg/ml (k9.7 pg/ml).
Most brain tumors are treated initially with maximally
tolerated doses of radiation to achieve the highest likeli-
hood of local control. Doses of 5000 to 6000 rad given
over 5 to 6 weeks are generally used, whereas slightly lower
doses are commonly given to children. Radiation thera-
pists are generally reluctant to offer a patient a second
course of irradiation at the time of recurrence because of
the risk of exceeding normal tissue tolerance. Several ret-
rospective studies have, however, demonstrated significant
palliation in a relatively large percentage of patients who
are retreated. Horns and Webber' reported significant
palliation in 6 of 12 patients retreated with cranial irra-
diation for intracranial tumors. Clinical improvement in
the six palliated patients lasted from 8 to 3 1 months. Drit-
schillo et al." reported a median survival of 36 months
in 32 patients retreated for recurrent brain tumors. The
median survival after retreatment for our patients was 13
months. The higher survival reported by Dritschillo
et al." is most likely due to the fact that most of their
patients received suboptimal doses at initial therapy,
whereas all of our patients had received what is currently
believed to be a maximum safe dose. Other authors also
have reported occasional long-term disease remission after
retreatment of medulloblastomas'2-15 or gliomas16 in
children. In most cases, the primary treatment was with
a suboptimal radiation dose.
No patient in this trial was cured of their recurrent
brain tumor. Nonetheless, the delay in tumor progression
after retreatment with irradiation plus misonidazole was
similar to that reported by other authors after salvage che-
motherapy. Levin et aL6 and Crafts et ~ 1 . ~
times-to-progression of 36 and 45 weeks, respectively, after
treatment with procarbazine, vincristine, and CCNU for
recurrent medulloblastomas, compared to our median
time-to-progression of 22 weeks. Most of our patients had
already failed chemotherapy by the time they were en-
rolled in this study; it is likely that our results would have
been more favorable if all patients received cranial irra-
diation with misonidazole before receiving (and failing)
Misonidazole was relatively well tolerated by the pa-
tients. Nausea and vomiting occurred in 59% of our pa-
tients; it has been reported to occur in 65% of adults re-
was a common occurrence, it was generally easily con-
trolled with antiemetics and did not necessitate discon-
tinuation of the drug.
Data pooled from several Phase I1 Radiation Oncology
Study Group (RTOG) studies with adults showed an
overall incidence of peripheral neuropathy of 2O%.I7
Twenty-one percent of our patients developed peripheral
neuropathies, demonstrating that pediatric patients are
not at higher risk of developing neuropathies than adults.
As reported previou~ly,'~ the peripheral neuropathies were
generally of a mild nature, with resolution over 1 to 3
months. Ototoxicity has been reported to occur in 6% of
patients receiving misonidazole. l7 Three of our patients
developed neurosensory hearing loss, an incidence of 10%.
It is unclear, however, if the ototoxicities were due to
misonidazole, radiation, and/or to tumor progression.
Serum levels after drug ingestion were similar to values
reported in adults after similar dosages. Wasserman
reported mean serum levels of 57 pg/ml (+ 23 pg/
ml) at 4 to 6 hours after ingestion of 1.5 g/m2 of miso-
nidazole, compared to a mean level of 68 pg/ml (+ 15 pg/
ml) in our patients. Extrapolating from murine and hu-
man experimental data, a serum concentration of 50 to
70 pg/ml would yield an enhancement ratio of roughly
1.2 to 1.5 compared to the effect of radiation alone.''
Mean serum levels in our pediatric patients at 24 to 30
hours were similar to reported pharmokinetic values in
Because this was not a randomized trial, it is not pos-
sible to discern what, if any, effect misonidazole contrib-
uted to radiation. Although in vivo data show that mi-
sonidazole enhances the effect of radiation by a factor of
1.2 to 1.8 at clinically achievable serum levels," a number
of clinical trials have failed to show a benefit by adding
misonidazole to radiation in the treatment of intracranial
turn or^'^-^^ or to tumors in other site^.^^,^^ It is unclear
at this time why misonidazole has not shown some benefit
when administered with radiation. The drug has been
shown to penetrate well into tumor tissues,24 including
intracranial tumors.25 Either tumor hypoxia is not a lim-
iting factor in the curability of human tumors, or mison-
idazole is not sensitizing hypoxic cells as well as predicted.
Other drugs with more favorable pharmacologic param-
eters and more effective sensitization properties are cur-
rently being sought.
Surprisingly, there have been remarkably few instances
reported of frank radionecrosis after retreatment of brain
tumors. Dritschillo et al." reported only two cases of ra-
dionecrosis in 32 patients retreated with radiation. Their
low incidence of radionecrosis probably was due to careful
patient selection, low doses given at the original treatment,
and the short survival of the majority of patients. Nev-
ertheless, there were a number of long-term survivors with
good neurologic function after retreatment. Marks et ~
Although nausea and vomiting
CANCER October 15 1986
reported a 5% risk of radionecrosis at doses of 4500 rad
or greater, with most cases occurring at greater than 6000
rad. Although there is no data concerning the degree of
recovery of normal brain tissue after irradiation, retreat-
ment with 3000 rad as done in this study almost certainly
exceeds normal tissue tolerance. We proceeded only be-
cause the prognosis without retreatment is uniformly poor.
Whereas only 2 of 28 patients died from radionecroses,
that is likely a conservative estimate of the incidence of
serious toxicity, as more patients surely would have de-
veloped severe radiation sequelae had they lived long
enough; radionecrosis generally does not develop until 6
months to 3 years after irradiation.26 Additionally, some
patients on whom autopsies were not performed may have
developed symptomatic radionecrosis that was erro-
neously attributed to tumor progression.
Although the prognosis for children with recurrent brain
tumors is dismal with any currently available therapy,
clinicians are occasionally faced with such patients who
might benefit from palliative treatment. If re-irradiation
is contemplated, it is doubtful, in light of randomized
clinical trials, that misonidazole offers a significant benefit
over radiation alone. A daily fraction size of 300 rad was
chosen for this study to maximize the potential interaction
of radiation with a limited number of doses of misoni-
dazole. In the absence of a compelling reason to use large
radiation fractions, conventional fractionation of I 50 to
200 rad per day would be prudent. The central nervous
system may be more sensitive to large radiation fractions
than are other normal tissues,27 and retreatment with
conventional fractionation may decrease the risk for ra-
diation toxicity. Additionally, partial-brain irradiation, as
opposed to whole-brain irradiation as used in this study,
would likely decrease the risk of causing normal tissue
damage in patients receiving a second course of irradia-
Although this was not a comparative trial, it appears
that retreatment of pediatric brain tumors with radiation
offers some palliation by delaying disease progression.
That the vast majority of patients died of recurrent intra-
cranial disease despite a total dose of radiation in excess
of 8000 rad is a clear indication of the need for more
effective chemotherapy to be used in conjunction with
radiation in the treatment of these tumors.
I. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect
relationship in the radiotherapy of malignant gliomas. Int J Radiat Oncol
BiolPhyS 1979; 5:1725-1731.
2. Silverman CL, Simpson JR. Cerebellar medulloblastoma: The im-
portance of posterior fossa dose to survival and patterns of failure. Int J
Radiat Oncol Biol Phys 1982; 8:1869-1876.
3. Pierre-Kahn A, Hirsch JF, Roux D et al. Intracranid ependymomas
in childhood: Survival and functional results of 47 cases. Child Brain
1983; 10: 145-156.
4. Crafts DC, Levin VA, Edwards MS et al. Chemotherapy of recurrent
medulloblastoma with combined procarbazine, CCNU, and vincristine.
J Neurosurg 1978; 49589-592.
5. Duffner PK, Cohen ME, Thomas PRM et al. Combination che-
motherapy in recurrent medulloblastoma. Cancer 1979; 43:41-45.
6. Levin VA, Edwards MS, Wright DC et al. Modified procarbazine,
CCNU, and vincristine (PCV 3) combination chemotherapy in the treat-
ment of malignant brain tumors. Cancer Treat Rep 1980; 64:237-241.
7. Levin VA, Vestnys PS, Edwards MS et al. Improvement in survival
produced by sequential therapies in the treatment of recurrent medul-
loblastoma. Cancer 1983; 5 1: 1364-1370.
8. Sheline GE. Radiation therapy of recurrent gliomas. Semin Oncol
9. Horns J, Webber MM. Retreatment of brain tumors. Radiology
10. Dritschillo A, Bruckman JE, Cassady JR et al. Tolerance of brain
to multiple courses of radiation therapy: Clinical experiences. Br J Radiol
11. Denekamp J, Fowler JF. Radiosensitization of solid tumors by
nitroimidazoles. Int J Radiat Oncol Biol Phys 1978; 4:143-151.
12. Harisiadis L, Chang CH. Medulloblastoma in children: A corre-
lation between staging and results of treatment. Int J Radiat Oncol Biol
PhyS 1977; 21833-841.
13. Smith CE, Long DM, Jones TK et al. Medulloblastoma: An anal-
ysis of time-dose relationships and recurrence patterns. Cancer 1973;
14. Tice GM. Treatment and prognosis of medulloblastoma. JAMA
15. Bloom HJG, Wallace ENK, Henk JM. The treatment and prog-
nosis of medulloblastoma in children: A study of 82 verified cases. Am
J Roentgen01 Radiat Ther Nucl Med 1969; 105:43-62.
16. Abramson N, Raben M, Cavanaugh PJ. Brain tumors in children:
Analysis of 136 cases. Radiology 1974; 112:669-672.
17. Wasserman TH, Stetz J, Phillips TL. Radiation therapy oncology
group clinical trials with misonidazole. Cancer 198 I; 47:2382-2390.
18. Nelson DF, Gillespie BW, Diener MD et al. Is misonidazole neu-
rotoxicity altered by the use of phentytoin and/or dexamethasone in
RTOG 79-18 and RTOG 79-16? Int J Radiat Oncol Biol Phys 1984;
1 0 : 173 1-1734.
19. Nelson DF, Schoenfeld D, Weinstein AS et al. A randomized
comparison of misonidazole sensitized radiotherapy plus BCNU and
radiotherapy plus BCNU for treatment of malignant glioma after surgery:
Preliminary results of an RTOG study. Int J Radiat Oncol Biol Phys
20. Hatlevoll R, Lindegaard K, Hagen S et al. Combined modality
treatment of operated astrocytomas grade 3 and 4: A prospective and
randomized study of misonidazole and radiotherapy with two different
radiation schedules and subsequent CCNU chemotherapy. Stage I1 of a
prospective multicenter trial of the Scandinavian glioblastoma study
group. Cancer 1985; 56:41-47.
2 1. Levin VA, Edwards MS, Wara WM et al. 5-Fluorourad and 1-
I-nitrosourea (CCNU) followed by hy-
droxyurea, misonidazole, and irradiation for brain stem gliomas: A pilot
study of the brain tumor research center and the Childrens Cancer Group.
Neurosurgery 1984; 14:679-68 1.
22. Panduro J, Kjaer M, Wolff-Jensen J et al. Misonidazole combined
with radiotherapy in the treatment ofinoperable squamous cell carcinoma
of the lung: A double-blind randomized trial. Cancer 1983; 52:20-24.
23. Panis X, Nguyen T, Froissart D et al. Hyperfractionated radio-
therapy with or without misonidazole: Results of a prospective random-
ized study in Stage 111-IV squamous cell carcinoma of the head and neck.
Int J Radiat OncolBiol Phys 1984; 10:1845-1849.
24. Urtasun RC, Band P, Chapman JD et al. Clinical Phase I Study
of the hypoxic cell radiosensitizer RO-07-0582: A 2-nitroimidazole de-
rivative. Radiology 1977; 122:SOl-804.
25. Bleehen NM, Grant RM, Jones DH et al. The penetration of
misonidazole and desmethylmisonidazole into brain tumours and other
central nervous system tissues in man. Br J Radiol 1983; 56:859-863.
26. Marks JE, Baglan RJ, Prassad SC et al. Cerebral radionecrosis:
Incidence and risk in relation to dose, time, fractionation and volume.
Int J Radiat Oncol Biol Phys 198 I; 7:243-252.
27. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and
brain injury. Int JRadial Oncol Biol Phys 1980; 6:1215-1228.