Correspondence IX19

thoracic or thoraco-lumbar. Figure 3’ attempted to present a

representative case of the series.

The four patients who received partial brain irradiation with fields that ranged from 64 cm* to 100 cm* were treated prior to 1972. None

of these fields encompassed the cervical spinal cord. However. these fields should not be considered “limited” since three of these patients

were younger than IO years of age: the remaining patient was I I years at the time of therapy. Whole brain fields employed in two patients were set by the author

himself. As the author has explained,‘.* whole brain fields for

posterior fossa lesions encompass the first four cervical vertebrae to

avoid contiguous spread into this area. Cranio-spinal irradiation in these patients began in 1974 after

observing previous spinal cord failures. Techniques for crania-spinal

irradiation have also been published by the author on repeated

occasions.” The technique resembles that employed for medullo-

blastomas and includes in the whole brain field the first few cerebral

vertebrae and this is treated at I50r/day (if the child is c IO years of

age) to 4500 rad followed by a boost of 1000 rad to the posterior fossa

together with spinal irradiation.

It must be clearly understood that the average age in the Rochester series was IO years with 9 of I3 children 510 years of age at

diagnosis. The author has never recommended using 200 rad/

fraction in these very young patients.

Prior to 1974. a dose of 4500~5000 rad for a malignant glioma in

young adults was considered adequate therapy. This is why 2 of the 4 patients received 4500 rad and the other 2 received 5000 rad to

limited brain portals. The author considers this adequate therapy for

the time it was given; the tumor volume was certainly adequately

covered.

In essence, Dr. Kopelson‘s letter has been, in fact, answered. However, the Rochester experience was aimed to alert investigators to a potential spread factor which should be taken into consideration when faced with this problem. If, in fact. the 500 rad difference in dose given in Rochester to the target volume prior to 1974 would make any difference as to seeding or not, is an exercise of academic futility since no randomized study could ever be attempted in such a rare tumor.

Finally, I do not believe that a local recurrence preceeded and was the causative agent for spinal cord metastases since the iatter preceded the other in every instance. It must also be stressed that every spinal cord metastases in the Rochester experience was documented (2 radiologi- cally and 3 radiologically and pathologically). Therefore. the reader should be alerted that a malignant cerebellar astrocytoma in a child carries a potential danger of spread through cerebra-spinal fluid pathwaya. and this must be weighed in the treatment approach which will be offered to the child.

O.M. SAI.AZAR. M.D

I. Salazar, O.M.: Moments of Decision in Primary Brain Tumors. P. Rubin (Ed.). Chicago, Illinois. American College of Radiology. 1977. pp. 43-154.

2. Salazar, O.M.: Primary brain tumors of the posterior fossa. In Modern Radiation Oncology, H.A. Gilbert and A.R. Kagan (Eds.). New York, Harper & Rowe. 1978. pp. 105%143.

3. Salazar, O.M.: Primary malignant cerebellar astrocytomas m children: A signal for postoperative crania-spinal irradiation. Inr. J. Radial. Oncol. Biol. Phys. 7:1661-1665, 1981.

4. Zulch. K.J.: Brain Tumors: Their Biology and Pathology. New York, Springer & Verlag Publishing Co. 1957. pp. 62-198.

CRITIQUE OF “RADIOBIOLOGICAL BASIS OF TOTAL BODY IRRADIATION WITH DIFFERENT DOSE RATE AND

FRACTIONATION: REPAIR CAPACITY OF HEMOPOIETIC CELLS”

To rhe Editor: We are at a loss to understand the motivation for the recent paper by Song et al.’ The authors go to great lengths to dispute the generally accepted view that hemopoietic cells are characterized by

limited radiation repair capacity, and have selected experimental data which show that some hemopmetic and culrured non-hemopoietlc ccII\ have similar survival curve parameters and split-dose recovery ratio5 From this, they apparently seek to Justify the practice in their Institution of using single dose, high dose-rate TBI as a conditioning procedure for bone marrow transplantation. However. they specitically state that it 1s not their contention “that normal or malignant hemopoietic cells are able to repair radiation damage as efficiently as are other normal (ISUIC’S such as the lung.“Surely. the verq crux of the question in deciding which radiation dose schedule will provide the best therapeutic ratio is the relative repair capacities of the normal and malignant hemopoictic clonogens and the cells of the dose-limiting normal tissues.

We maintain quite simply that the cxperimcntal data in rodcnta shoH that the sparing effect of dose fracttonation is greater for lung death than for bone marrow death. Hence. in order to ablate normal hemopoietlc tissue with least risk of lung toxIcitS. one should ubc fractionated (or low-dose rate) TBI. The same argument holds true for other potentially dose-limiting normal organs such as CNS. kldncq\, and heart. The calculations In our paper (Radiology I31 :243 747, 1979) of the anticipiated log kill with different TBI radiation techniques referred to normal hemopoietic stem cells. We acknowledge that there ilr more heterogeneity of response to radiation of leukemia/lqmphoma cells than of normal hemopoietic cells. Obviously. it would be desirable to have detailed information regarding the dobe survival characlerlstics 01 leukemia clonogens in individual patients. but. in the absence of such information, it is reasonable to presume that their response would approximate more closely that of their normal cells of origin than that of organized pulmonary tissues.

Therefore, on the basis of the most relevant data available we inslsl that the most logical approach is to replace single dose, high dose-rale treatments with fractionated regimens that maximiTe the diferential between hemopoietic cells and organized tissues such as lung. In the final analysis. the proofof the pudding is in the eating. and the reduction in toxicity of TBI without an increase in leukemia relapse rate5 using fractionated schedules at both Seattle and Houston would appear to have vindicated our recommendation to convert to such regimens

LESTER J. P[.TI-~5. M.D Professor and Head Division of Radiotherapy U.T.M.D. Anderson Hospital and Tumor lnatitute Houston. TX 77030 H. Ronht:y WITHt,RS. M D.. Pti.D. Professor of Radiotherap) Head, Section of ExperImental Radiation Oncology UCLA Los Angeles. CA 90024

I. S0ny.C.W.. Kim.T.H.. Khan, F.M.. Kersey. J.H.. Lcvitt.S.H.: Int. J. Radial Oncol. Biol. Phv.( 7. I695 1701. 19x1

REBUTTAL

To the Ediror: We appreciate the critique of our article by Peters and Withers. but wedo not agree with the radiobiological justification in their criticism of our technique for TBI. Furthermore, we could not lind the data supporting their claim of better clinical results with their technique in literature. Simply stated. since our clinical results for high dose rate TBI were and are equal to any other results reported in the literature, and in view of the discomfort and morbidity of patients stemming from prolonged treatment sessions associated with TBI xt a low dose rate. we feel it is justified to continue using high dose rate for single dose TBI.’

Contrary to the statement of Peters and Withers. we did not “select” but actually discussed all the available experimental data on the effect of split-dose irradiation on normal and malignant hemopoietic stem cells and the most frequently quoted reports on the effect of single dose irradiation on these cells. Based on these data, we concluded that the repair capacity of hemopoietic stem cells, in particular malignant hemopoietic cells. may have been underestimated. We agree with Peters and Withers that the differential response of hemopoietic cells and dose-limiting normal tissues to irradiation should be exploited maxi-