Background and purpose: Over a 20 year period, four out of 2000 paediatric radiotherapy patients, treated at St. Bartholomew's Hospital (three with lymphoma, one with angiosarcoma), have revealed extreme/fatal clinical hypersensitivity in normal tissues.
Patients and methods: Cellular hypersensitivity was confirmed in vitro and attributed to the ataxia-telangiectasia (A-T) gene in cases I and II, a newly described defect in the DNA ligase 4 gene in case III, and a novel and as yet incompletely defined, molecular defect in case IV who presented with xeroderma pigmentosum (XP).
Results: The severe clinical hypersensitivity preceded the cellular and molecular analysis, but did not manifest as a clinically exaggerated normal tissue reaction until 3+ weeks after the start of a conventionally fractionated course of radiotherapy, by which time the latent damage had been inflicted. There were no clinical stigmata to alert the clinician to a predisposing syndrome in two patients (cases I and II). We point out that approximately 20% of A-T patients are classified as variants with delayed expression of clinical symptoms, and case II falls into this category.
Conclusions: As lymphoma (incidence, one in 100000 children) constituted the majority of the diagnoses, questions arise as to: (1), the probability of other centres having experienced and being presented in the future with similar problems (particularly bearing in mind that other oncologically predisposing radiosensitivity syndromes have not been not represented in our experience); and (2), the appropriateness, efficiency and applicability of predictive assays. Unambiguous cellular radiosensitivity would have been apparent from clonal assays on fibroblast cultures from all four cases prior to treatment, but such assays take 4-6 weeks to produce results. While estimates of chromosome damage or clonal assays on pre-treatment blood derived cells would be faster, there is a health economics issue as to the general applicability of such 'screening' assays.