Association of pretreatment chest CT abnormalities with trastuzumab deruxtecan–associated pneumonitis.

Abstract

1106 Background: Trastuzumab deruxtecan (T-DXd) is an antibody-drug conjugate approved for HER2+ metastatic breast cancer (MBC) and HER2-low (IHC 1+ or 2+ with FISH-) MBC . Pneumonitis with T-DXd has been observed in 13.6% of patients (pts) on DESTINY-Breast 01 (including 2.2% fatal events) and 12.1% in pts on DESTINY-BREAST 04 (including 0.8% fatal events). There is an unmet need to determine if pts with baseline lung abnormalities are at increased risk for development of pneumonitis from T-DXd. Methods: This retrospective cohort study included all pts treated with T-DXd for MBC at an academic institution from 1/2020 to 2/2023 for more than 1 cycle (>21 days). Outcomes were evaluated from treatment initiation until 2/1/2023. Two board-certified radiologists blinded to outcomes reviewed staging chest CTs performed before initiating T-DXd and graded each scan on 16 parenchymal abnormalities, including presence of interstitial lung abnormalities (ILA). Pneumonitis was retrospectively adjudicated by an independent oncologist using chart review and graded according to CTCAE v5.0 and compared to the treating physician’s assessment. The difference of proportions was analyzed with the Pearson Chi-Square test. Significance was declared at the 0.05 type I error threshold. Results: A total of 68 pts (median age 57.3 years) were included (30 = HER2+, 24 = HR+/HER2 low, 14 = TNBC/HER2 low). By independent assessment, 16 pts (23.5%) had grade 1 pneumonitis and 3 pts (4.4%) grade 2 pneumonitis. Median time from treatment initiation to pneumonitis was 83 days (range 35-266). By treating physician assessment, 5 pts (7.4%) developed grade 1 pneumonitis and 3 pts (4.4%) developed grade 2 pneumonitis. 16 of 68 pts (23.5%) had baseline ILA and 62 pts (91.2%) had at least one other parenchymal abnormality. As assessed by an independent oncologist, of pts with baseline ILA (N = 16), 12.5% developed T-DXd related pneumonitis, compared to 32.7% of pts without pre-existing ILA (N=52), p=0.16. There was no statistical difference in T-DXd related pneumonitis assessed by treating physician between pts with baseline ILA (2/16 patients, 12.5%) compared to pts without baseline ILA (7/52 pts, 13.5%), p=0.68. Conclusions: In this study, presence of baseline ILA was not associated with subsequent pneumonitis from T-DXd. Notably, the rate of independently assessed pneumonitis was higher than treatment-attributed pneumonitis as defined by the treating physician. Further research is needed to validate these findings and define predictors of treatment-associated pneumonitis. [Table: see text]