Dose response and factors related to interstitial pneumonitis after bone marrow transplant

Total body irradiation (TBI) and chemotherapy are common components of conditioning regimens for bone marrow transplantation. Interstitial pneumonitis (IP) is a known regimen-related complication. Using published data of IP in a multivariate logistic regression, this study sought to identify the parameters in the bone marrow transplantation conditioning regimen that were significantly associated with IP and to establish a radiation dose-response function.

A retrospective review was conducted of articles that reported IP incidence along with lung dose, fractionation, dose rate, and chemotherapy regimen. In the final analysis, 20 articles (n = 1090 patients), consisting of 26 distinct TBI/chemotherapy regimens, were included in the analysis. Multivariate logistic regression was performed to determine dosimetric and chemotherapeutic factors that influenced the incidence of IP.

A logistic model was generated from patients receiving daily fractions of radiation. In this model, lung dose, cyclophosphamide dose, and the addition of busulfan were significantly associated with IP. An incidence of 3%-4% with chemotherapy-only conditioning regimens is estimated from the models. The alpha/beta value of the linear-quadratic model was estimated to be 2.8 Gy. The dose eliciting a 50% incidence, D50, for IP after 120 mg/kg of cyclophosphamide was 8.8 Gy; in the absence of chemotherapy, the estimated D50 is 10.6 Gy. No dose rate effect was observed. The use of busulfan as a substitute for radiation is equivalent to treating with 14.8 Gy in 4 fractions with 50% transmission blocks shielding the lung. The logistic regression failed to find a model that adequately fit the multiple-fraction-per-day data.

Dose responses for both lung radiation dose and cyclophosphamide dose were identified. A conditioning regimen of 12 Gy TBI in 6 daily fractions induces an IP incidence of about 11% in the absence of lung shielding. Shielding the lung to receive 50% of this dose lowers the estimated incidence to about 2.3%. Because the lungs can be adequately shielded, we recommend against using busulfan as a substitute for fractionated TBI with cyclophosphamide.