AUTHORS
Matthew Roderick Osborn, MD, MS, Ying Liu, MD2, Rachel Abramova3, Caitlin Rogers4, Nora Chokr, MD5, Dimitrios Papaioannou, MD6, Marc Schwartz, MD7, Jingmei Hsu, MD8, Oscar Lahoud, MD9, Faith Davies, MD10, Maher Abdul-Hay, MD11, Samuel Yamshon, MD12
AFFILIATIONS
1 NYU Langone Health, New York, New York, United States
2 NYU Langone Health, New York, New York, United States
3 NYU Langone Health, New York, New York, United States
4 NYU Langone Health, New York, New York, United States
5 NYU Langone Health, New York, New York, United States
6 NYU Langone Health, New York, New York, United States
7 NYU Langone Health, New York, New York, United States
8 NYU Langone Health, New York, New York, United States
9 NYU Langone Health, New York, New York, United States
10 NYU Langone Health, New York, New York, United States
11 NYU Langone Health, New York, New York, United States
12 NYU Langone Health, New York, New York, United States
ABSTRACT
Pre-CAR T cell therapy PFTs are often required for insurance authorization, but their utility in predicting outcomes and toxicity related to CAR T cell therapy is unclear. There is heterogeneous practice in requiring screening PFTs in all patients versus obtaining them only in patients with pulmonary comorbidities.
Screening PFTs are required in all patients receiving CAR T cells at the NYU Langone Perlmutter Cancer Center; patients from January 2021 to April 2026 receiving commercial CAR T cells for any indication were included in the study. Demographics, CAR indication, performance status, number of prior lines of therapy, CAR product, product co-stimulatory domain, smoking history in pack-years, and presence of pulmonary comorbidities (such as asthma or chronic obstructive pulmonary disease) were collected. Forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC), and diffusion capacity for carbon monoxide (DLCO) were recorded. DLCO values were adjusted for hemoglobin concentration. Abnormal PFTs were defined as FEV1, FVC, or adjusted DLCO < 5th percentile using the Global Lung Function Initiative reference equations. Fisher’s exact tests were used to assess if predictor variables were associated with abnormal PFTs. The highest grades of cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), the best response within 100 days of CAR T cell administration, and mortality data were collected for each patient as dependent variables. Multivariable logistic regression models incorporating predictor variables were fit to predict dependent variables.
125 patients (91 male, 34 female) received CAR T cell therapy and completed pretherapy PFTs. The median age at administration was 66 years and the median Karnofsky score was 80%. Diffuse large B cell lymphoma (n=60) was the most common indication for therapy, followed by multiple myeloma (n=31), acute lymphoblastic leukemia (n=13), mantle cell lymphoma (n=11), chronic lymphocytic leukemia (n=4), follicular lymphoma (n=4), and marginal zone lymphoma (n=2). The CAR T cell products were axicabtagene ciloleucel (n=58), ciltacabtagene autoleucel (n=30), brexucabtagene autoleucel (n=17), lisocabtagene maraleucel (n=15), obecabtagene autoleucel (n=4), and idecabtagene vicleucel (n=1).
Smoking history was reported by 42% (53/125) of patients, and 15% (19/125) had a known pulmonary comorbidity before undergoing PFTs. 41% (51/125) of patients had abnormal PFTs, 78% (40/51) of whom had no pulmonary comorbidity. Age > 65 years (50% abnormal versus 31% with age ≤ 65, p = 0.04), Karnofsky score < 90% (50% abnormal versus 24% with score ≥ 90%, p = 0.006), and > 8 pack-years smoking history (61% abnormal versus 35% with ≤ 8 pack-years, p = 0.02) were significantly associated with abnormal PFTs. Only 14% of patients with none of these criteria had abnormal PFTs.
After CAR T cell therapy, CRS of any grade and grade 3+ occurred in 81% (101/125) and 7.2% (9/125) of patients, respectively. ICANS of any grade and grade 3+ occurred in 41% (51/125) and 8.8% (11/125) of patients, respectively. Within 100 days of CAR T cell administration, 80% (100/125) of patients achieved at least a partial response.
In the multivariable logistic regression model, z-scores for FEV1 (odds ratio (OR) 0.84, 95% confidence interval (CI) 0.25-2.8), FVC (OR 0.69, 95% CI 0.21-2.2), and DLCO (OR 1.2, 95% CI 0.77-1.9) were not associated with any-grade CRS. z-scores for FEV1 (OR 1.4, 95% CI 0.50-3.8), FVC (OR 0.49, 95% CI 0.18-1.3), and DLCO (OR 0.84, 95% CI 0.57-1.2) were not associated with any-grade ICANS. The only variable associated with CRS or ICANS was receipt of a CAR product with a CD28 co-stimulatory domain, which was significantly associated with the presence of any-grade and grade 3+ CRS and ICANS. No variable was significantly associated with patients developing relapsed/refractory disease 100 days post-treatment.
Baseline PFT abnormalities are present in 41% of patients undergoing CAR T cell therapy for hematologic malignancies. However, abnormal PFTs are uncommon among patients ≤ 65 years old, with Karnofsky score ≥ 90%, and with ≤ 8 pack-years of smoking history. FEV1, FVC, and DLCO are not significantly associated with CRS, ICANS, or developing relapsed/refractory disease 100 days post-treatment. Requiring PFT measurement before CAR T cell therapy in all patients may be of limited utility and testing targeted to patients with specific risk factors.
DISCLOSURES
Matthew Roderick Osborn, MD, MS: No relevant conflicts of interest to declare.