Adult medulloblastoma

SUMMARY Twelve cases of adult onset medulloblastoma are presented. Clinical features, treatment and outcome are discussed. It was found that the survival rates for adults are no better than for children. There were no clinical or histological features that distinguished these tumours from those occurring in childhood, except for a higher incidence of hemisphere lesions. Since its initial recognition as a separate entity by relationship of survival to the age of onset. In an Bailey and Cushing in 1925,' medulloblastoma has effort to define our own experience with these become well recognised as a malignant cerebellar tumours we have reviewed the records of the 12 tumour occurring mainly in childhood. The name adult patients with medulloblastoma admitted to medulloblastoma was chosen by Bailey and Cushing Addenbrooke's Hospital after June, 1961. The to illustrate their view that the tumour arose from a records were examined with a view to finding any hypothetical primitive pluri-potential stem cell cal-clinical or histological features which might dis-led a medulloblast. This theory has now become tinguish cases arising after the 18th birthday from traditional. Weight was added to it in 1968 when those arising in childhood. In addition, treatment Zulch and Wechsler2 suggested that the tumour and mortality were reviewed. arose from the external granular cell layer of Obers-teiner. This neuro-ectodermal remnant can be read-Material and methods ily demonstrated as an external granular layer in the vermis in normal brains up to the age of 16 months. After this age the cells disappear, presumably mi-occurring after the 18th birthday have been identified. grating into deeper granular layers. tumours

Since its initial recognition as a separate entity by relationship of survival to the age of onset. In an Bailey and Cushing in 1925,' medulloblastoma has effort to define our own experience with these become well recognised as a malignant cerebellar tumours we have reviewed the records of the 12 tumour occurring mainly in childhood. The name adult patients with medulloblastoma admitted to medulloblastoma was chosen by Bailey and Cushing Addenbrooke's Hospital after June, 1961. The to illustrate their view that the tumour arose from a records were examined with a view to finding any hypothetical primitive pluri-potential stem cell cal-clinical or histological features which might disled a medulloblast. This theory has now become tinguish cases arising after the 18th birthday from traditional. Weight was added to it in 1968 when those arising in childhood. In addition, treatment Zulch and Wechsler2 suggested that the tumour and mortality were reviewed. arose from the external granular cell layer of Obersteiner. This neuro-ectodermal remnant can be read-Material and methods ily demonstrated as an external granular layer in the vermis in normal brains up to the age of 16 months. Between 1961 and 1982 12 cases of medulloblastoma After this age the cells disappear, presumably mioccurring after the 18th birthday have been identified.
After thsaeteclsdsperrsDuring this period a total of 37 patients, both adults and grating into deeper granular layers. These tumours children, with medulloblastoma were seen, making adults have therefore always been considered congenital 32% of the cases. There were two females and ten males. and most likely to arise within the first 16 months of The age of the patients at the time of diagnosis ranged life. This supposition has even led to the prediction from 18 to 49 years, with an average age of 39 years. One of a period of risk for recurrence which may be calpatient was in the second decade of life, three in the third, culated by adding nine months to the patient's age at three in the fourth and five in the fifth. the time of diagnosis.3 The appearance in adults of medulloblastomas histologically indistinguishable Clinical features from those in children has therefore remained a Most patients presented with very short histories (table 1). great puzze ad a cIn seven (58%) the history was 3 months or shorter. The great puzzle and a challenge to tne traditional other five patients had histories of between 3 and 6 theory. Cushing and Bailey recognised this In their months. The longest history was 6 months, the shortest 2 original paper. Nine of their 61 patients were 18 months. The most common symptom was headache, folyears old or over at the time their tumours arose. lowed by ataxia, nausea and vomiting (table 2). These They could find no differences in histology, natural symptoms usually occurred in combination. Headache history or response to treatment in patients of differ-occurred in ten (85%) of the patients at some time before ing ages. Since their original paper the literature has their diagnosis. Seven (58%) suffered from ataxia and six become replete with conflicting evidence as to the (50%) from nausea and vomiting. Two patients (16%) were confined strictly to the cerebellar hemisphere, with five on the right and four on the left. One left-sided tumour presented as a cerebello-pontine angle mass in a patient with unilateral cranial nerve signs. Three patients had midline tumours.

Histology
The picture was that of a cellular tumour composed of small, dark cells, with scanty cytoplasm. The cell nuclei were small, round or ovoid, and frequently hyperchromatic. Cellular and nuclear pleomorphism were not conspicuous. Mitoses were present in either moderate numbers (five tumours) or large numbers (four tumours). Necrosis was not an outstanding feature. Desmoplastic change was seen in two tumours. Both of these tumours were characteristically lateral in position in the cerebellar hemisphere. As conventional wisdom would predict,4 we only found desmoplastic change in those tumours which were laterally placed and superficial enough to pick up a mesenchymal element from the dura. Possible spongioblastic differentiation was present in one tumour. Rosettes were not seen in any of the tumours. Though rosettes are not present in all childhood medulloblastomas, the total absence of rosettes in the present series is the most significant difference from the appearances of the childhood tumour.
Pobereskin, Treip Treatment All patients underwent some form of direct operative intervention on the tumour. In two cases this was limited to biopsy only, and in a further two cases to what was considered to be a sub-total removal. In eight cases there was considered to be a macroscopic total removal of the tumour. None of the patients was staged by CSF cytology or myelography at the time of original diagnosis. All patients received radical radiotherapy to the brain and spine following surgery. This was given on a 6meV linear accelerator for the brain and a Cobalt gamma 60 machine for the spine. Fifteen to twenty fractions were given over three weeks. Eight patients received a total dose of 4000-4500 cGy to the posterior fossa, with 2000 to the spinal canal. Three patients received 5000-5500 cGy to the posterior fossa, with 2000 cGy to the spinal canal. Eleven patients completed the full course, but one died during radiotherapy. Four underwent a second course of radiotherapy at the time of recurrence, and three patients received chemotherapy with 1-(2-chloroethyl)-3cyclohexyl-l-nitrosourea (CCNU). One patient received both. Two patients underwent a second posterior fossa exploration at the time of recurrence. For the group as a whole there was no operative mortality been reported varying from zero6 to 58%7 and no (death within 6 weeks). The follow up period ranged from clear relationship between the various histological, 5 to 10 years. The overall one year survival rate was 75%. clinical and surgical factors and survival can be The five year survival rate in 12 patients is 50%, and at ten established.

years 11% (nine patients). Three patients remain alive and
Recent studies in-children,89 have reported five well without recurrence at five, five and nine years from Recent survIn res 9 appro rted fiv diagnosis. One patient is alive 10 years following diagnosis, and ten year survival rates of approximately 5 but now has widespread cerebral and spinal metastases 55% and 40% respectively. Comparisons are (table 3). Of the four patients alive now, two underwent difficult because, as in our series, radiation therapy biopsy only and radiotherapy and are alive five and ten modalities have changed in recent years. In addition, years later. Two patients are still alive following total sur-reports from radiotherapeutic centres such as those gical removal and radiotherapy five and nine years ago quoted above deal only with patients who survive respectively. Of the patients still alive, three had hemioperation and a full course of radiotherapy. It seems spheric lesions and one a midline lesion.
The length of the initial history had no bearing on survi-gen ly a t tat "sua l rates yaproxival. The average survival was 32 months fo'r those with a y y y history of less than three months, and 27 months in those the best that can be expected in children with medulwhose history was between three and six months long. The loblastoma.0... '°W ithin these limitations our surclinical presentations were broadly similar in all patients vival figures for adults are comparable to those in except for the two who presented with cranial nerve signs. children, especially considering one of our survivors Of those one died at 11 months and one is alive without has only a few more months to go to reach the ten recurrence after 5 years.
year milestone. The type of surgical intervention would appear to have We could not confirm the suggestion of Quest et made little difference to survival. There was a 50% five a13 that the time interval between original diagnosis year survival rate in those whose tumours were not com--pletely removed and a 42% five year survival rate in the and recurrence would vary directly with the patient' s group whose tumours were macroscopically completely age at diagnosis. For those with tumour diagnosis removed. before the thirtieth birthday the average survival Unlike surgery, the total dose of radiation to the poswas 53 months, whereas after the thirtieth birthday terior fossa did appear to affect survival. The three patients it was 47 months. Chatty and Earle" found that who received 5000-5500 cGy to the posterior fossa are patients with hemispheric lesions had increased suralive and well at five, five and nine years respectively, but vival. Our numbers are too small to draw statistically of the eight patients who received only 4000-4500 cGy to valid conclusions, but tend to support this with three the posterior fossa, only one is still alive at ten years andof eight patients with hemispheric lesions surviving, has widespread spinal metastases. One patient died while and one of four patients with midline lesions still undergoing radiotherapy. and In of serients c ld fidln elation be-In ten patients the histological appearances were classi-alive. In this series we could find no correlation becal of medulloblastoma, and in two cases the medulloblas-tween the thoroughness of surgical tumour removal tomas were desmoplastic in appearance. One of the and survival. Our study confirms the data of Kopelpatients originally harbouring a desmoplastic tumour is son'2 that higher radiation dosage does increase suralive and without recurrence at five years. However, the vival. other died of recurrence two years after original diagnosis.
In conclusion, our series demonstrates that the Three patients suffered from spinal metastases at one, survival rates for adults with medulloblastoma are two and nine years following their original diagnosis. One no better than for children. Our series showed a patient died without evidence of recurrence. Seven patients had posterior fossa recurrence at the time of death higher percentage of hemispheric tumours, which is and these recurrences occurred at one to six years from the not found in children, but we could not distinguish original diagnosis. As other studies have shown,5 once any histological factors that would suggest that these there was recurrence the prognosis was very grave, all -but tumours are fundamentally different from those to one patient dying within six months. The one exception is a be found in children, nor do there seem to be any patient who is still alive thirteen months after the diagnosis features that would allow one to predict which of spinal metastases. Metastases outside the nervous sys-patients will enjoy long survival. The efficacy of tem were found in only one patient who interestingly did higher dosage radiotherapy to the posterior fossa is not have a shunt. confirmed.