CONTROVERSY Avoiding chemotherapy related late effects in children with curable tumours

A number of articles have highlighted the dilemma facing those involved with the design of treatment programmes for patients with curable tumours, in whom long term survival necessitates the avoidance, if possible, of late sequelae.' 2 Nowhere is this problem more apparent than in the management of paediatric cancer. With almost 60% of children who have been treated for cancer in the past decade achieving long term survival and probable cure, it is becoming increasingly clear that a variety of problems are being created as a direct consequence of curative treatment. Many paediatric subspecialists other than oncologists, such as endocrinologists, cardiologists, respirologists, and neurologists are involved in the joint long term follow up of these children into adulthood. The problem of avoiding mutilating surgery has largely been achieved by the use of preoperative chemotherapy. Similarly, extended field high dose radiotherapy is now rarely used in paediatric practice. These improvements that avoid late sequelae are, however, dependent on the effective use of chemotherapy to achieve maximum response and reduce or obviate the need for further treatment directed towards local tumour control. In an attempt to reduce chemotherapy related late sequelae there is at present considerable debate concerning which drugs could be dropped or replaced by less toxic agents. Virtually all chemotherapeutic agents are non-specific cell poisons; their effectiveness often depending on a narrow therapeutic ratio (that is, tumour cell kill to normal tissue toxicity). Up until recently the classic 'bogymen' of chemotherapy were the alkylating agents. The use of regimens such as mustine, vincristine, procarbazine, prednisolone (MOPP) or chlorambucil, vincristine, procarbazine, prednisolone (ChlVPP) in childhood Hodgkin's disease is undesirable if effective alternatives exist because of the inevitable sterilising effect in males and the possible risk of second tumours and leukaemia. There is little doubt about the sterilising effect in boys after standard chemotherapy for Hodgkin's disease, although girls are generally spared this adverse effect. Data for cyclophosphamide alone are less clear, but it seems likely that a third to a half of boys will be sterilised after conventional 12-24 months regimens for various childhood solid tumours.3 It has been suggested that a total cumulative dose of up to around 4 g/m2 of cyclophosphamide can be given without sterilisation.46 With regimens containing a small number of courses using high doses of cyclophosphamide, for example in non-Hodgkin's lymphoma, where up to 3 g/m2 is given per course,7 8 the sterilising effect may be less than when the same total dose is given over several months. This remains to be clarified, however, in prospective studies. Ifosfamide, an analogue of cyclophosphamide, which appears to be less myelosuppressive with comparable antitumour alkylating activity, is now in widespread use for Ewing's sarcoma, rhabdomyosarcoma, and some protocols for leukaemia/lymphoma. Unfortunately the lack of myelosuppression has been offset by a disturbing incidence of renal tubulopathy. The precise incidence of significant dysfunction remains unclear pending adequate studies of unselected patients, but there is little doubt that a Fanconi-like syndrome may occur particularly in the smaller child who has received high doses of ifosfamide. The sterilising effect of four to six courses of 6-9 g/m2 ifosfamide is unclear but seems likely. The incidence of second cancers in children ranges from 1-4% and is dependent on the nature of prior chemotherapy and the use of radiotherapy.9 In adults an incidence of up to 18% at 15 years has been reported in those with Hodgkin's disease who have received alkylating agents and radiotherapy. 10 In one study of childhood Hodgkin's disease using an alternating MOPP/doxorubicin bleomycin, vinblastine, dacarbazine (ABVD) regimen where the median follow up was only 35 months three out of 62 patients developed second malignancies. The latter attempt to reduce the late effects of a MOPP based regimen could in fact result in 'the worst of both worlds', namely sterilisation, second tumours, anthracycline related cardiotoxicity, and bleomycin related lung toxicity." With a conservative radiotherapy policy risks are lower and only three second malignancies have occurred in 232 children mostly receiving ChlVPP alone on the current United Kingdom Children's Cancer Study Group (UKCCSG) regimen. High incidences have been reported in children with Ewing's sarcoma treated with cyclophosphamide based regimens and high dose radiotherapy'2 and in children with nonHodgkin's lymphoma receiving cyclophosphamide and chloroethylcyclohexyl nitrosourea (CCNU) regimens. 13 In solid tumours such as rhabdomyosarcoma and Wilms' tumour a series of prospective randomised studies have evaluated the need to include cyclophosphamide. The American Intergroup Rhabdomyosarcoma Study group has demonstrated clearly that this drug is not necessary to cure patients with completely resected rhabdomyosarcoma or in those with microscopic residual disease provided they are Royal Marsden Hospital, Downs Road, Sutton, Surrey SM2 5PT