Introduction: Chest radiation-related lung cancer has been previously reported in cancer survivors (Lorigan et al, Lancet Oncology, 2005). The risk increases with dose (20-40Gy) and volume of radiation; cigarette smoking multiplies the risk associated with higher doses of radiation. Total body irradiation (TBI) delivers a smaller dose but larger volume of radiation to the lung compared to conventional radiation therapy (e.g., mediastinal radiation). Whether exposure to TBI (in the absence of chest radiation) increases the risk of lung subsequent neoplasms (SNs) and whether cigarette smoking interacts with TBI in increasing the risk of lung SNs remains unknown. We leveraged BMTSS to address these gaps in knowledge.

Methods: BMTSS is a retrospective cohort study of individuals who were treated with BMT from 1974 to 2014 at three transplant centers and survived ≥2y. This report includes BMT survivors who completed a BMTSS survey at age ≥18y, providing information on demographics, health habits (including smoking history), and occurrence of chronic health conditions and SNs (including lung SNs). Disease and treatment characteristics were obtained from medical records and included exposure to pre-BMT radiation and chemotherapy (alkylating agents, anthracyclines), BMT conditioning (TBI, chemotherapies), donor type (autologous, allogeneic). The present analysis excluded patients who received chest radiation. Study participants were classified as never smokers, former smokers, and current smokers. TBI exposure was classified as: 0–799cGy and ≥800cGy. Fine-Gray proportional sub-distribution hazard models were used determine the hazard of lung SNs, treating other causes of death as competing risks. Time at risk was calculated from the date of BMT to the earliest of lung SN diagnosis, death, or last known follow-up (May 2024). Standardized incidence ratios (SIRs) quantified the excess risk of lung SNs among BMT recipients when compared with the general population.

Results: A total of 3,767 BMT survivors were included. Median age at BMT was 48y (range, 0-78y), and median follow-up was 15.2y. Overall, 57% were male; 43% had received TBI ≥800 cGy; 34% were former smokers; and 4% were current smokers. Fifty-four individuals developed lung SNs at a median of 7.5y post-BMT. A significant multiplicative interaction between TBI and current smoking suggested effect modification; thus, a six-category combined exposure variable was created to examine joint effects. Standardized incidence ratios: Compared with the general population, the current smoker + TBI ≥800cGy group had a 4.6-fold increased risk of lung SNs (p=0.002). The risk of lung SNs was not statistically-significantly increased in the current smoker + TBI 0-799cGy group (SIR=2.6, p=0.2), the former smoker + TBI ≥800cGy group (SIR=1.4, p=0.3), or the never smoker + TBI ≥800cGy group (SIR=0.4, p=0.1). Multivariable models: Using the never smoker + TBI 0–799cGy as the reference group, the highest hazard for lung SNs was observed in the current smoker + TBI ≥800cGy group (aHR=5.9, p=0.003), followed by the current smoker + TBI 0–799cGy group (aHR=3.7, p=0.05) and the former smoker + TBI ≥800cGy group (aHR=2.6, p=0.03). The former smoker + TBI 0–799cGy group was not at increased risk of lung SNs (aHR=1.7, p=0.3). Notably, exposure to TBI ≥800cGy alone (in the absence of smoking) was not associated with an increased risk of lung SNs (aHR=0.9, p=0.8).

Conclusions: This study shows no association between TBI and lung SNs in the absence of current cigarette smoking. The multiplicative interaction between TBI and current smoking informs the need for targeted surveillance and interventions in the highest risk group.

This content is only available as a PDF.
2025
Sign in via your Institution