Immune-Related Adverse Events of Anti-PD-(L)1 Immunotherapy With and Without Chemotherapy in Patients with Advanced Non-Small Cell Lung Cancer: A Real-World Evidence Study

Background: The tolerability and dose escalation of immune checkpoint inhibitors (ICI) such as anti-PD-(L)1 immunotherapy are often limited by immune-related adverse events (irAEs). It is uncertain whether chemotherapy amplifies or reduces the risk of irAEs when co-administered with ICIs.

Objectives: Evaluate whether the risk of treatment-emergent irAEs differs for patients with advanced non-small cell lung cancer (aNSCLC) who receive first-line (1L) treatment with anti-PD-(L)1 immunotherapy versus anti-PD-(L)1 immunotherapy plus chemotherapy (chemoimmunotherapy).

Methods: The study population consisted of adults (≥18 years) with aNSCLC identified in the ConcertAI Oncology dataset receiving 1L ICI between 10 May 2017 and 31 July 2022. Patients receiving ICI monotherapy constituted the immunotherapy cohort; patients receiving ICI with same-day initiation of chemotherapy constituted the chemoimmunotherapy cohort. Patient characteristics were balanced using inverse probability of treatment weighting. Patients were followed from treatment initiation until 90 days after therapy discontinuation, death, last EHR activity, initiation of a new line of therapy, or end of study (October 31, 2022). Composite irAEs refer to any incident irAEs occurring during follow-up. Cox regression was used to estimate the hazard ratio (HR) of irAE between the cohorts.

Results: Composite irAE tended to be more frequent in the chemoimmunotherapy cohort (N=3020, 79.0 per 1000 person-years, 95% CI, 67.9-92.0) versus the immunotherapy cohort (N=1607, 51.4 per 1000 person-years, 95% CI, 40.2-65.8).
Patients receiving chemoimmunotherapy had a 44% higher crude HR of composite irAE versus those on immunotherapy (HR, 1.44, 95% CI, 1.1-1.9), with an adjusted HR of 1.33 (95% CI, 1.0-1.8; P = 0.053). The adjusted HR for composite grade 3+ irAE was 1.06 (95% CI, 0.74 -1.52). Pneumonitis and acute kidney injury, the two most prevalent irAEs, had adjusted HRs of 1.01 (95% CI, 0.63-1.63) and 1.02 (95% CI, 0.45-2.29). Autoimmune disease (AID) and myalgia had higher risks in the chemoimmunotherapy group with adjusted HRs of 4.44 (95% CI, 0.18-109.7) and 10.43 (95% CI, 0.61-177.26), while colitis and encephalopathy showed lower risks with adjusted HRs of 0.72 (95% CI, 0.37-1.4) and 0.48 (95% CI, 0.18-1.23). Adjusted HR for irAE was 8.17 (95% CI, 1.02-65.22) among the 6% of patients with baseline AID and 1.22 (95% CI, 0.91-1.65) for patients without baseline AID.

Conclusions: In first-line anti-PD(L)1 immunotherapy for aNSCLC, IrAEs tended to occur more frequently with chemoimmunotherapy than immunotherapy, with considerable heterogeneity between individual irAEs. These data bear re-examination in prospective clinical trials.