Impaired diffusion capacity predicts for decreased treatment tolerance and survival in limited stage small cell lung cancer patients treated with concurrent chemoradiation

Purpose: To determine if stratification of limited stage small cell lung cancer (LSCLC) patients by pre-treatment pulmonary function test (PFT) prognostic indicators predicts for treatment-related toxicity risks and survival following concurrent chemoradiation. Materials and methods: From 1989 to 1999, 215 LSCLC patients received six cycles of alternating cyctophosphamide/doxorubicin/vincristine and etoposide/cisplatin (EP). Thoracic radiation (RT) was initiated only with EP and at cycle 2 or 3. RT dose was: 40 Gy/15 fractions/3 weeks or 50 Gy/25 fractions/5 weeks. RT fields encompassed gross and suspected microscopic disease with a 2 cm margin. Pre-treatment PFT values analyzed included forced expiratory volume in 1 s (FEV1) (in titer and as % predicted) and diffusion capacity for carbon monoxide (DLCO) (as % predicted). The "marker" for toxicity during concurrent chemoradiation was the duration of any RT breaks initiated for severe hematologic or tocoregionat symptomatology. Patient outcomes were analyzed for associations between recognized PFT cut-offs (FEV1 < 2 l, greater than or equal to 2 l; FEV1 < 60%, greater than or equal to 60% predicted; DLCO < 60%, greater than or equal to 60% predicted), toxicity rates, and survival. Results: For the whole study cohort, median, 2- and 5-year overall survivals were: 14.7 months, 22.7 and 7.2%, respectively. Fifty-six patients (26%) required treatment breaks due to toxicity. FEW and DLCO results were available for 96 (45%) and 86 (40%) patients, respectively. Two thirds of FEV1s measured were <2 l. On statistical analysis, the incidence of toxicity-related interruptions was significant for DLCO < 60% (P = 0.043), suggestive for FEV1 < 2 l; (P = 0.1) and non-significant for FEV1 < 60%. Patients with simultaneous DLCO < 60% and FEV1 < 2 l showed a trend toward increase toxicity risk (P = 0. 1). For selected PFT measures, median overall survivals were: 12.7 months versus 14.8 months for DLCO < 60% versus greater than or equal to60%; 13.4 months versus 17.7 months for FEV1 < 2 l versus greater than or equal to2 l; 15.4 months versus 19.9 months for DLCO < 60% + FEV1 < 2 l versus DLCO greater than or equal to 60% + FEV1 greater than or equal to 2 l. Although absolute differences favored all patients with PFT values above the prognostic cut-offs, differences were not statistically significant on this analysis. Patients with both a treatment break and a DLCO < 60% had the poorest median survival of all patient subsets, at 11.4 months (P = 0.09). Conclusions: Impaired DLCO (i.e. < 60%) is a novel predictor of increased treatment-related toxicity leading to interruptions. The present study suggests a probable rote for DLCO and FEV1 (in 1) as prognostic factors for predicting survival but larger patient samples are required for confirmation. Patients with impaired DLCOs experiencing treatment interruptions have the poorest survival. Assessment of pre-treatment PFTs contributes to determining optimal management strategies for LSCLC patients receiving definitive chemoradiation. (C) 2003 Elsevier Ireland Ltd. All rights reserved.