Impact of Anti‐Inflammatory Drug Therapy on the Efficacy of First‐Line Atezolizumab Treatment in Patients With Advanced Non‐Small Cell Lung Cancer: An Integrated Analysis of Five Randomized Controlled Trials Authors: Kazuki Takada, Alessio Cortellini, Shinkichi Takamori, Fumitaka Mizuki, David J Pinato, Mototsugu Shimokawa Published: 2026 ABSTRACT ABSTRACTIntroductionAnti‐inflammatory agents including acetaminophen (APAP) and non‐steroidal anti‐inflammatory drugs (NSAIDs) may blunt anti‐tumor immunity.MethodsWe performed a patient‐level cross‐trial analysis of five randomized controlled studies (IMpower110, IMpower130, IMpower131, IMpower132, and IMpower150) that assessed the efficacy of first‐line atezolizumab therapy in advanced non‐small cell lung cancer (NSCLC). We investigated whether APAP/NSAIDs exposure influenced progression‐free survival (PFS) and overall survival (OS) in both atezolizumab‐treated and non‐atezolizumab‐treated cohorts using forest plot analysis.ResultsIn 3781 patients, APAP exposure was 15.4% (583/3781) and NSAIDs use was 33.9% (1282/3781). APAP use was associated with worse PFS and OS in both the atezolizumab and non‐atezolizumab groups (PFS with atezolizumab: hazard ratio [HR] = 1.18, 95% confidence interval [CI]: 0.97–1.43; PFS without atezolizumab: HR = 1.23, 95% CI: 1.05–1.43; OS with atezolizumab: HR = 1.13, 95% CI: 0.98–1.30; OS without atezolizumab: HR = 1.31, 95% CI: 1.04–1.65). NSAIDs use was not associated with differences in PFS or OS in the atezolizumab group but was linked to worse PFS and OS in the non‐atezolizumab group (PFS with atezolizumab: HR = 0.98, 95% CI: 0.89–1.08; PFS without atezolizumab: HR = 1.18, 95% CI: 1.03–1.34; OS with atezolizumab: HR = 1.01, 95% CI: 0.91–1.13; OS without atezolizumab: HR = 1.27, 95% CI: 1.10–1.46). After excluding patients on baseline proton pump inhibitors, a significant interaction between NSAIDs use and treatment modality was observed in the multivariable analysis of OS (p = 0.0140), with no interaction for APAP use (p = 0.6707).ConclusionNeither APAP nor NSAIDs exposure influenced atezolizumab efficacy.Keywords: acetaminophen, atezolizumab, non‐small cell lung cancer, non‐steroidal anti‐inflammatory drug Introduction Non‐small cell lung cancer (NSCLC) is one of the most common malignant diseases and remains a leading cause of cancer‐related deaths worldwide [1]. However, the emergence of immune checkpoint inhibitors (ICIs) has revolutionized the treatment landscape of NSCLC, resulting in improved survival outcomes for many patients. Among these, atezolizumab, an anti‐programmed death‐ligand 1 (PD‐L1) antibody, has become a key first‐line treatment option for patients with advanced or metastatic NSCLC [2, 3, 4, 5, 6]. However, the efficacy of anti‐PD‐L1 immunotherapy can be influenced by a number of host factors, including the use of concomitant medications. Previous studies have shown that commonly used drugs such as antibiotics and proton pump inhibitors (PPIs) may negatively affect the efficacy of atezolizumab therapy and worsen patient outcomes [7, 8, 9, 10]. These findings suggest that the interaction between atezolizumab and concomitant medications should be carefully considered in the clinical setting. Anti‐inflammatory agents, especially acetaminophen (APAP) and non‐steroidal anti‐inflammatory drugs (NSAIDs), are frequently used in patients with cancer to manage symptoms such as fever and pain. Recent studies showed the potential impact of APAP on ICI efficacy, suggesting that higher circulating levels of APAP might be associated with poorer prognosis in patients receiving ICIs [11, 12]. However, these studies did not specifically focus on patients with advanced and unresectable NSCLC, including resectable stage IIIA NSCLC or solid tumors. There are opposing trends for NSAIDs use and anti‐cancer immunotherapy efficacy. Some studies have suggested that patients exposed to NSAIDs have a better prognosis compared to those not taking NSAIDs [13, 14]: a finding that is opposed by other studies that reported a worse prognosis for NSAIDs users [15, 16]. These inconsistencies underscore the need for further research into the impact of NSAIDs on clinical outcomes, particularly in relation to the efficacy of ICI therapy. To address this uncertainty, we conducted an integrated, patient‐level analysis of IMpower110, IMpower130, IMpower131, IMpower132, and IMpower150: five landmark randomized controlled trials (RCTs) that evaluated the efficacy of first‐line atezolizumab therapy in advanced NSCLC. This analysis aimed to assess whether APAP and NSAIDs exposure was associated with worse oncological outcomes of patients receiving first‐line atezolizumab therapy. Materials and Methods Study Design and Population The five RCTs included in this analysis are IMpower110 (NCT02409342), IMpower130 (NCT02367781), IMpower131 (NCT02367794), IMpower132 (NCT02657434), and IMpower150 (NCT02366143) [2, 3, 4, 5, 6]: IMpower110, atezolizumab vs. cisplatin/carboplatin plus pemetrexed/gemcitabine in patients with non‐squamous and squamous NSCLC [4]; IMpower130, atezolizumab plus carboplatin and nab‐paclitaxel vs. carboplatin and nab‐paclitaxel in non‐squamous NSCLC [3]; IMpower131, atezolizumab plus carboplatin and paclitaxel vs. atezolizumab plus carboplatin and nab‐paclitaxel vs. carboplatin and nab‐paclitaxel in squamous NSCLC [5]; IMpower132, atezolizumab plus cisplatin/carboplatin and pemetrexed vs. cisplatin/carboplatin and pemetrexed in non‐squamous NSCLC [6]; IMpower150, atezolizumab plus carboplatin and paclitaxel plus bevacizumab vs. atezolizumab plus carboplatin and paclitaxel vs. carboplatin and paclitaxel plus bevacizumab in non‐squamous NSCLC [2]. Eligibility, baseline characteristics and outcomes for each clinical trial can be found in the primary study publications [2, 3, 4, 5, 6]. Patients included in this analysis were those who had been treated according to their randomized allocation in the respective trials. Patients with driver mutations such as epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK) mutations, those with incomplete baseline characteristics or missing data on concomitant medications, were excluded from the analysis. In this study, we assessed the use of APAP, NSAIDs, and proton pump inhibitors (PPIs). PPIs were included in the analysis because they are frequently co‐administered with APAP and NSAIDs in clinical practice. Medication use was defined as any administration within 30 days prior to the initiation of atezolizumab or chemotherapy. Patients were included irrespective of the dosing, timing, duration, and classes of APAP/NSAIDs and indication for APAP/NSAIDs prescription. Therefore, single use and chronic use were analyzed within the same category in this study. Statistical Analysis Descriptive statistics and contingency tables were used to summarize patient demographics and baseline characteristics, with standardized mean differences (SMDs) calculated to assess baseline associations. Progression‐free survival (PFS) and overall survival (OS) were defined as the time from treatment initiation to disease progression/death and death/last follow‐up, respectively. Kaplan–Meier survival curves were constructed for both PFS and OS, and the log‐rank test was used to compare survival distributions. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using Cox proportional hazards regression models. The prognostic associations of APAP or NSAIDs use with PFS and OS were evaluated on a trial‐by‐trial basis, separately for patients in the atezolizumab arms and those in the non‐atezolizumab arms. Additionally, the impact of atezolizumab compared to non‐atezolizumab treatment on PFS and OS was assessed separately for APAP/NSAIDs users and non‐users. Given the well‐established negative prognostic impact of PPIs on clinical outcomes in patients treated with ICIs [8], and their frequent co‐prescription with APAP/NSAIDs, we conducted an exploratory analysis of OS only excluding patients on baseline PPIs. HRs and 95% CIs were estimated using Cox proportional hazards models for each trial, and the results were summarized using forest plots with random‐effects models. These analyses were adjusted for various factors, including age, sex, race, smoking history/status, Eastern Cooperative Oncology Group performance status (ECOG PS), histology, liver metastasis, PD‐L1 expression, PPI use, and APAP/NSAIDs use. Chi‐square heterogeneity tests were used to test for statistical heterogeneity among trials. We also calculated the I 2 statistic expressing the proportion of variability in the results attributable to heterogeneity with an I 2 value below 30% considered indicative of low heterogeneity. The interaction between APAP/NSAIDs use and treatment modality (atezolizumab vs. non‐atezolizumab) was also assessed using Cox regression analysis. All requested studies were integrated into the Vivli research environment, where all statistical analyses were conducted using R software. Statistical significance was defined as a p value of less than 0.05. Results Patient Selection A total of 4097 patients were enrolled and randomized across the five IMpower studies, including 572 patients in IMpower110, 724 in IMpower130, 1021 in IMpower131, 578 in IMpower132, and 1202 in IMpower150. Of these, 91 patients did not receive treatment, 213 had driver mutations (EGFR or ALK), and 12 had unavailable data on baseline characteristics or concomitant medication use. After excluding these patients, a total of 3781 patients were included in the final analysis (Figure S1). Baseline Characteristics The baseline characteristics of the patients included in each of the IMpower studies are summarized in Table S1. The exposure to APAP and NSAIDs across the IMpower studies was 15.4% (583/3781) for APAP and 33.9% (1282/3781) for NSAIDs. Table 1 shows the baseline characteristics of patients by APAP use. The factors with SMDs of 0.10 or below were as follows: allocated atezolizumab, age, sex, race, smoking history, histology, and liver metastasis, while the factors with SMDs greater than 0.10 were as follows: ECOG PS, PD‐L1 expression, PPI use, and NSAIDs use. Table 2 provides the baseline characteristics of patients by NSAIDs use. The factors with SMDs of 0.10 or below were as follows: allocated atezolizumab, sex, race, ECOG PS, histology, liver metastasis, and PD‐L1 expression, while the factors with SMDs greater than 0.10 were as follows: age, smoking history, PPI use, and APAP use. TABLE 1Summary of the patients' baseline characteristics by APAP use.OverallWithout APAPWith APAPSMD N 37813198583Allocated atezolizumab (%)0.03Yes2373 (62.8)2015 (63.0)358 (61.4)No1408 (37.2)1183 (37.0)225 (38.6)Study (%)0.31IMpower110528 (14.0)483 (15.1)45 (7.7)IMpower130669 (17.7)583 (18.2)86 (14.8)IMpower131997 (26.4)849 (26.5)148 (25.4)IMpower132558 (14.8)439 (13.7)119 (20.4)IMpower1501029 (27.2)844 (26.4)185 (31.7)Age (median [IQR])64.0 [58.0, 70.0]64.0 [58.0, 70.0]65.0 [57.5, 70.0]0.02Sex (%)0.08Male2586 (68.4)2205 (68.9)381 (65.4)Female1195 (31.6)993 (31.1)202 (34.6)Race (%)0.10White3157 (83.5)2676 (83.7)481 (82.5)Asian422 (11.2)362 (11.3)60 (10.3)Other/Unknown202 (5.3)160 (5.0)42 (7.2)ECOG PS (%)0.2201471 (38.9)1296 (40.5)175 (30.0)12310 (61.1)1902 (59.5)408 (70.0)Smoking (%)0.10Never435 (11.5)378 (11.8)57 (9.8)Previous2376 (62.8)1986 (62.1)390 (66.9)Current970 (25.7)834 (26.1)136 (23.3)Histology (%)0.09Non‐squamous2629 (69.5)2203 (68.9)426 (73.1)Squamous1152 (30.5)995 (31.1)157 (26.9)Liver metastasis (%)0.03Yes584 (15.4)500 (15.6)84 (14.4)No3197 (84.6)2698 (84.4)499 (85.6)PD‐L1 expression a (%)0.18TC0 and IC0 (negative)1496 (39.6)1239 (38.7)257 (44.1)TC1/2 or IC1/2 (low)1355 (35.8)1175 (36.7)180 (30.9)TC3 or IC3 (high)706 (18.7)609 (19.0)97 (16.6)Unknown224 (5.9)175 (5.5)49 (8.4)PPI users (%)0.23Yes1070 (28.3)853 (26.7)217 (37.2)No2711 (71.7)2345 (73.3)366 (62.8)NSAIDs users (%)0.17Yes1282 (33.9)1043 (32.6)239 (41.0)No2499 (66.1)2155 (67.4)344 (59.0)Abbreviations: APAP, acetaminophen; ECOG PS, Eastern Cooperative Oncology Group performance status; IQR, interquartile range; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor; SMD, standardized mean difference. a PD‐L1 expression was assessed by immunohistochemical analysis. Tumor cell (TC) or tumor‐infiltrating immune cell (IC) group levels (TC PD‐L1 expression on < 1% [TC0] or ≥ 1% to < 5% [TC1] or ≥ 5% to < 50% [TC2] or ≥ 50% [TC3]; IC PD‐L1 expression on < 1% [IC0] or ≥ 1% to < 5% [IC1] or ≥ 5% to < 10% [IC2] or ≥ 10% [IC3]). Summary of the patients' baseline characteristics by APAP use. Abbreviations: APAP, acetaminophen; ECOG PS, Eastern Cooperative Oncology Group performance status; IQR, interquartile range; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor; SMD, standardized mean difference. PD‐L1 expression was assessed by immunohistochemical analysis. Tumor cell (TC) or tumor‐infiltrating immune cell (IC) group levels (TC PD‐L1 expression on < 1% [TC0] or ≥ 1% to < 5% [TC1] or ≥ 5% to < 50% [TC2] or ≥ 50% [TC3]; IC PD‐L1 expression on < 1% [IC0] or ≥ 1% to < 5% [IC1] or ≥ 5% to < 10% [IC2] or ≥ 10% [IC3]). TABLE 2Summary of the patients' characteristics by NSAIDs use.OverallWithout NSAIDsWith NSAIDsSMD N 378124991282Allocated atezolizumab (%)0.04Yes2373 (62.8)1553 (62.1)820 (64.0)No1408 (37.2)946 (37.9)462 (36.0)Study (%)0.28IMpower110528 (14.0)414 (16.6)114 (8.9)IMpower130669 (17.7)382 (15.3)287 (22.4)IMpower131997 (26.4)655 (26.2)342 (26.7)IMpower132558 (14.8)384 (15.4)174 (13.6)IMpower1501029 (27.2)664 (26.6)365 (28.5)Age (median [IQR])64.0 [58.0, 70.0]64.0 [58.0, 69.0]65.0 [58.0, 71.0]0.13Sex (%)0.02Male2586 (68.4)1702 (68.1)884 (69.0)Female1195 (31.6)797 (31.9)398 (31.0)Race (%)0.08White3157 (83.5)2070 (82.8)1087 (84.8)Asian422 (11.2)280 (11.2)142 (11.1)Other/Unknown202 (5.3)149 (6.0)53 (4.1)ECOG PS (%)0.0801471 (38.9)1005 (40.2)466 (36.3)12310 (61.1)1494 (59.8)816 (63.7)Smoking (%)0.22Never435 (11.5)344 (13.8)91 (7.1)Previous2376 (62.8)1522 (60.9)854 (66.6)Current970 (25.7)633 (25.3)337 (26.3)Histology (%)0.04Non‐squamous2629 (69.5)1721 (68.9)908 (70.8)Squamous1152 (30.5)778 (31.1)374 (29.2)Liver metastasis (%)0.03Yes584 (15.4)394 (15.8)190 (14.8)No3197 (84.6)2105 (84.2)1092 (85.2)PD‐L1 expression a (%)0.07TC0 and IC0 (negative)1496 (39.6)964 (38.6)532 (41.5)TC1/2 or IC1/2 (low)1355 (35.8)913 (36.5)442 (34.5)TC3 or IC3 (high)706 (18.7)466 (18.6)240 (18.7)Unknown224 (5.9)156 (6.2)68 (5.3)PPI users (%)0.52Yes1070 (28.3)510 (20.4)560 (43.7)No2711 (71.7)1989 (79.6)722 (56.3)APAP users (%)0.13Yes583 (15.4)344 (13.8)239 (18.6)No3198 (84.6)2155 (86.2)1043 (81.4)Abbreviations: APAP, acetaminophen; ECOG PS, Eastern Cooperative Oncology Group performance status; IQR, interquartile range; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor; SMD, standardized mean difference. a PD‐L1 expression was assessed by immunohistochemical analysis. Tumor cell (TC) or tumor‐infiltrating immune cell (IC) group levels (TC PD‐L1 expression on < 1% [TC0] or ≥ 1% to < 5% [TC1] or ≥ 5% to < 50% [TC2] or ≥ 50% [TC3]; IC PD‐L1 expression on < 1% [IC0] or ≥ 1% to < 5% [IC1] or ≥ 5% to < 10% [IC2] or ≥ 10% [IC3]). Summary of the patients' characteristics by NSAIDs use. Abbreviations: APAP, acetaminophen; ECOG PS, Eastern Cooperative Oncology Group performance status; IQR, interquartile range; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor; SMD, standardized mean difference. PD‐L1 expression was assessed by immunohistochemical analysis. Tumor cell (TC) or tumor‐infiltrating immune cell (IC) group levels (TC PD‐L1 expression on < 1% [TC0] or ≥ 1% to < 5% [TC1] or ≥ 5% to < 50% [TC2] or ≥ 50% [TC3]; IC PD‐L1 expression on < 1% [IC0] or ≥ 1% to < 5% [IC1] or ≥ 5% to < 10% [IC2] or ≥ 10% [IC3]). Impact of APAP Use on PFS and OS Median follow‐up time in the overall population was 13.9 months, as shown in Table S2. To first assess the overall association between APAP use and clinical outcomes, Kaplan–Meier curves for PFS and OS according to APAP use and treatment arm are shown in Figure S2. Patients using APAP had a worse prognosis compared to non‐users, with a median PFS of 5.5 months in the APAP group versus 6.7 months in the non‐APAP group (HR = 1.22, 95% CI: 1.11–1.33, p < 0.0001; Figure S2A). This pattern was observed in both the atezolizumab‐treated and non‐atezolizumab‐treated subgroups (Figure S2A). For OS, APAP users had a shorter MST of 12.2 months compared to 16.3 months in non‐users (HR = 1.27, 95% CI: 1.14–1.40, p < 0.0001; Figure S2B). A similar pattern was also seen across both treatment subgroups (Figure S2B). We next evaluated the prognostic impact of APAP use separately within each treatment arm using forest plot analyses. Unadjusted results are presented in Figure S3, and adjusted results are shown in Figure 1. APAP use was associated with worse PFS and OS in both the atezolizumab and non‐atezolizumab groups (PFS with atezolizumab, HR = 1.18, 95% CI: 0.97–1.43; PFS without atezolizumab, HR = 1.23, 95% CI: 1.05–1.43; OS with atezolizumab, HR = 1.13, 95% CI: 0.98–1.30; OS without atezolizumab, HR = 1.31, 95% CI: 1.04–1.65; Figure 1). We then examined whether the treatment effect of atezolizumab versus chemotherapy differed according to APAP use. Unadjusted analyses are shown in Figure S4, and adjusted analyses are shown in Figure 2. The benefit of atezolizumab treatment compared to chemotherapy was similar among both APAP users and non‐users (PFS without APAP: HR = 0.70, 95% CI: 0.64–0.75; PFS with APAP: HR = 0.79, 95% CI: 0.66–0.96; OS without APAP: HR = 0.85, 95% CI: 0.78–0.93; OS with APAP: HR = 0.82, 95% CI: 0.66–1.01; Figure 2). Forest plot of APAP efficacy (versus without APAP) according to with or without atezolizumab by study using adjusted analyses. (A) Progression‐free survival and (B) overall survival. Analyses adjusted for age, sex, race, ECOG PS, smoking, histology, liver metastasis, PD‐L1 expression, PPI use, and NSAIDs use. MST: months. APAP, acetaminophen; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor. Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without APAP by study using adjusted analyses. (A) Progression‐free survival and (B) overall survival. Analyses adjusted for age, sex, race, ECOG PS, smoking, histology, liver metastasis, PD‐L1 expression, PPI use, and NSAIDs use. MST: months. APAP, acetaminophen; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor. Impact of NSAIDs Use on PFS and OS To first assess the overall association between NSAIDs use and clinical outcomes, Kaplan–Meier curves for PFS and OS according to NSAIDs use and treatment arm are shown in Figure S5. NSAIDs use did not significantly affect PFS, with a MST of 6.0 months in the NSAIDs group and 6.5 months in the non‐NSAIDs group (HR = 1.04, 95% CI: 0.97–1.12, p = 0.3115; Figure S5A). No difference was observed in the atezolizumab group, whereas NSAIDs users had a poorer prognosis in the non‐atezolizumab group (Figure S5A). For OS, NSAIDs users had a shorter MST of 14.5 months compared to 16.1 months in non‐users (HR = 1.10, 95% CI: 1.01–1.19, p = 0.0212; Figure S5B). As with PFS, this difference was mainly driven by poorer outcomes in the non‐atezolizumab group (Figure S5B). We next evaluated the prognostic impact of NSAIDs use separately within each treatment arm using forest plot analyses. Unadjusted results are presented in Figure S6, and adjusted results are shown in Figure 3. NSAIDs use was not associated with differences in PFS or OS in the atezolizumab group, but was associated with worse PFS and OS in the non‐atezolizumab group (PFS with atezolizumab: HR = 0.98, 95% CI: 0.89–1.08; PFS without atezolizumab: HR = 1.18, 95% CI: 1.03–1.34; OS with atezolizumab: HR = 1.01, 95% CI: 0.91–1.13; OS without atezolizumab: HR = 1.27, 95% CI: 1.10–1.46; Figure 3). We then examined whether the treatment effect of atezolizumab versus chemotherapy differed according to NSAIDs use. Unadjusted analyses are shown in Figure S7, and adjusted analyses are shown in Figure 4. The benefit of atezolizumab over chemotherapy was greater in NSAIDs users compared to non‐users (PFS without NSAIDs: HR = 0.75, 95% CI: 0.68–0.81; PFS with NSAIDs: HR = 0.64, 95% CI: 0.56–0.72; OS without NSAIDs: HR = 0.89, 95% CI: 0.80–0.98; OS with NSAIDs: HR = 0.76, 95% CI: 0.66–0.87; Figure 4). Forest plot of NSAIDs efficacy (versus without NSAIDs) according to with or without atezolizumab by study using adjusted analyses. (A) Progression‐free survival and (B) overall survival. Analyses adjusted for age, sex, race, ECOG PS, smoking, histology, liver metastasis, PD‐L1 expression, PPI use, and APAP use. MST: months. APAP, acetaminophen; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor. Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without NSAIDs by study using adjusted analyses. (A) Progression‐free survival and (B) overall survival. Analyses adjusted for age, sex, race, ECOG PS, smoking, histology, liver metastasis, PD‐L1 expression, PPI use, and APAP use. MST: months. APAP, acetaminophen; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs; PD‐L1, programmed death‐ligand 1; PPI, proton pump inhibitor. Interaction Between APAP/NSAIDs Use and Treatment Modality After PPI Exclusion Finally, we performed a sensitivity analysis after excluding PPI users to further explore the interaction between APAP/NSAIDs use and treatment modality. In the analysis excluding PPI users, APAP had a negative prognostic impact in both the atezolizumab and non‐atezolizumab groups (Figure S8A,B), while NSAIDs use was negatively associated with OS only in patients treated with non‐atezolizumab therapy and had no impact in those treated with atezolizumab therapy (Figure S8C,D). The pooled multivariable interaction p value between APAP use and treatment modality was 0.6707, while the interaction between NSAIDs use and treatment modality was statistically significant (p = 0.0140). Discussion Anti‐inflammatory drug use is highly prevalent in patients with cancer to treat pain and fever and preliminary evidence has suggested APAP and NSAIDs use to correlate with worse prognosis [11, 12, 15, 16]. However, a clear mechanistic understanding of this association is lacking and the risk of associative bias with cancer‐related pro‐inflammatory response [17], leading to prescription of these therapies, is high. In this study, we performed an integrated, patient‐level analysis of five RCTs to evaluate the impact of APAP and NSAIDs use on the efficacy of first‐line atezolizumab therapy in patients with advanced NSCLC. APAP exposure carried a negative prognostic effect on both PFS and OS in the atezolizumab and non‐atezolizumab groups, with no significant interaction between APAP use and treatment modality. This suggests that the negative impact of APAP is likely driven by associative bias, as APAP use is often linked to poorer baseline conditions or advanced symptoms such as pain and fever, rather than any direct immune‐modulating mechanism. In contrast, NSAIDs use was not associated with differences in PFS or OS in the atezolizumab group but was linked to worse outcomes in the non‐atezolizumab group. Additionally, the benefit of atezolizumab over chemotherapy was greater in NSAIDs users, and a significant interaction was observed between NSAIDs use and treatment modality. This suggests that potential pharmacological interactions between NSAIDs and atezolizumab could mitigate the negative impact seen in chemotherapy‐only treatments, possibly through mechanisms related to immune activation or inflammation control which warrant further investigation. Some studies have suggested that patients using NSAIDs treated with ICI have a better prognosis compared to non‐users [13, 14], while others have reported a worse prognosis for NSAIDs users [15, 16], or no significant difference [18]. By analyzing both the control arms and the atezolizumab arms in this large dataset, our study provides evidence that the interaction between atezolizumab and NSAIDs mitigates the negative bias between worse clinical conditions and NSAIDs use observed in chemotherapy‐only treatments. However, we cannot conclude that NSAIDs enhance the efficacy of atezolizumab. Recent studies suggest that APAP may affect the efficacy of ICIs [11, 12]. In patients with resectable stage IIIA NSCLC undergoing neoadjuvant therapy (ICI with chemotherapy), higher serum APAP was associated with a lower pathological response rate [12]. In a Lewis lung carcinoma model, mice treated with both anti‐programmed death‐1 and APAP showed faster tumor growth, increased neutrophil infiltration, and elevated neutrophil extracellular traps, which are linked to immunotherapy resistance [12, 19, 20]. Another study found higher serum APAP levels correlated with poorer outcomes in patients with solid tumors treated with ICIs and increased regulatory T cell infiltration in a colorectal cancer mouse model [11]. However, our forest plot analysis found no impact of APAP on outcomes or atezolizumab efficacy, likely due to differences in measuring serum APAP versus any use within 30 days before treatment. Given APAP's short half‐life, the timing and duration relative to ICI therapy may warrant further investigation. When interpreting the results on NSAIDs, it is important to consider both their potential benefits and drawbacks in the context of ICI therapy. NSAIDs can inhibit cyclooxygenase‐2 (COX‐2), which might enhance the anti‐tumor immune response [21, 22], but they may also affect the gut microbiome, which plays a crucial role in cancer immunotherapy [23, 24, 25]. COX‐2 is often overexpressed in cancers and associated with poor prognosis [26, 27]. It promotes the production of prostaglandin E2 (PGE2), a critical inflammatory factor in the tumor microenvironment, which aids cancer survival, proliferation, and immune evasion [28]. By reducing PGE2 levels, NSAIDs may help counteract immunosuppression and boost the efficacy of ICIs [21, 22]. However, NSAIDs might also negatively impact ICI therapy by influencing the gut microbiome, similar to antibiotics and PPIs, which have been linked to poorer outcomes in ICI‐treated patients [7, 8, 9, 10, 29]. NSAIDs are known to affect the gut microbiome, suggesting this as a potential mechanism for diminished ICI efficacy [30, 31, 32]. Thus, NSAIDs have both beneficial and detrimental effects in the context of ICI therapy, making their role complex. In this study, the interaction between NSAIDs and atezolizumab counteracted the negative biases associated with NSAIDs use seen in chemotherapy‐only treatments. NSAIDs did not enhance atezolizumab's efficacy, but they did not adversely affect outcomes either, suggesting they can be safely prescribed to patients receiving atezolizumab. The strengths of this study include the large sample size and the use of individual patient data from five RCTs. However, several limitations should be acknowledged. Observational and retrospective studies on the impact of concomitant medications are inherently subject to associative bias, which stems from the underlying conditions that lead to the prescription of these medications, limiting the ability to draw definitive conclusions. Additionally, the dosing, timing, duration, and classes of APAP and NSAIDs were not strictly controlled, which may have influenced the results. In this study, exposure to APAP/NSAIDs initiated after treatment onset was not taken into account; rather, the analysis was limited to baseline exposure. This approach may introduce time‐dependent bias; however, baseline exposure was selected to minimize immortal time bias and ensure comparability across groups. Moreover, information regarding the indication for APAP/NSAIDs use was available only for a subset of the included RCTs. Therefore, we were unable to comprehensively assess or report the reasons for APAP/NSAIDs use across the entire pooled study population. Another limitation is that this study pooled data from trials with different treatment backbones, which may have introduced clinical heterogeneity. Nevertheless, the forest plots demonstrated that the overall results were broadly consistent across trials despite these differences. Finally, whether similar findings would be observed in patients treated with anti–PD‐1 antibodies, especially in a monotherapy setting, remains uncertain. Thus, the generalizability of our findings beyond atezolizumab‐based combination regimens warrants further investigation. In conclusion, our data do not support a detrimental effect for either APAP or NSAIDs exposure over the efficacy of atezolizumab. APAP/NSAIDs, especially APAP, use may reflect a higher tumor burden or greater symptom severity, and therefore may be interpreted more as a prognostic factor rather than a predictor of the efficacy of immunotherapy. Although mechanistic evidence is required to assess the impact of these concomitant therapies on the tumor microenvironment of patients with NSCLC, APAP, and NSAIDs therapy should be considered safe in patients with NSCLC receiving atezolizumab therapy. Author Contributions Kazuki Takada: conceptualization, data curation, resources, project administration, supervision, writing – original draft. Alessio Cortellini: formal analysis, investigation, methodology, writing – review and editing. Shinkichi Takamori: writing – review and editing. Fumitaka Mizuki: writing – review and editing. David J. Pinato: writing – review and editing. Mototsugu Shimokawa: formal analysis, investigation, methodology, writing – review and editing. Funding The authors have nothing to report. Ethics Statement The study was conducted in accordance with the Declaration of Helsinki and ethical approval was obtained at each participating site, with participants providing written informed consent. Consent The authors have nothing to report. Conflicts of Interest Alessio Cortellini received grants for consultancies/advisory boards: MSD, OncoC4, Roche, Regeneron, BMS, Amgen, Daiichi Sankyo, Astrazeneca, Access Infinity, Ardelis Health, Alpha Sight, Capvision, Techspert. He also received speaker fees from Astrazeneca, Roche, Pierre‐Fabre, MSD, SANOFI/Regeneron; compensation for writing/editorial activity: BMS, MSD; travel support from Sanofi, MSD, Roche and funding (to institution) from the International Association for the Study of Lung Cancer. David J. Pinato received lecture fees from Bayer Healthcare, EISAI, BMS, Roche, Boston Scientific, travel expenses from BMS and Bayer Healthcare; consulting fees for Mina Therapeutics, DaVolterra, Mursla, IPSEN, Exact Sciences, Avamune, EISAI, Roche, Starpharma, LiFT biosciences and Astra Zeneca; received research funding (to institution) from MSD, GSK, and BMS. All other authors declare no conflicts of interest associated with the present study. Supporting information Figure S1: CONSORT diagram of this study. ALK, anaplastic lymphoma kinase; EGFR, epidermal growth factor receptor. Figure S2: The Kaplan–Meier curves of (A) progression‐free survival and (B) overall survival according to the use of APAP (with or without APAP) and treatment arm (with or without atezolizumab) in overall population. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S3: Forest plot of APAP efficacy (versus without APAP) according to with or without atezolizumab by study. (A) Progression‐free survival and (B) overall survival. MST: months. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S4: Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without APAP by study. (A) Progression‐free survival and (B) overall survival. MST: months. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S5: The Kaplan–Meier curves of (A) progression‐free survival and (B) overall survival according to the use of NSAIDs (with or without NSAIDs) and treatment arm (with or without atezolizumab) in overall population. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S6: Forest plot of NSAIDs efficacy (versus without NSAIDs) according to with or without atezolizumab by study. (A) Progression‐free survival and (B) overall survival. MST: months. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S7: Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without NSAIDs by study. (A) Progression‐free survival and (B) overall survival. MST: months. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S8: The Kaplan–Meier curves of overall survival in patients excluding PPI users. (A) APAP use or non‐use in without‐atezolizumab cohort. (B) APAP use or non‐use in with‐atezolizumab cohort. (C) NSAIDs use or non‐use in without‐atezolizumab cohort. (D) NSAIDs use or non‐use in with‐atezolizumab cohort. APAP, acetaminophen; NSAIDs, non‐steroidal anti‐inflammatory drugs; PPI, proton pump inhibitor. Table S1: Summary of the patients' baseline characteristics by IMpower study. Table S2: Summary of the observation time by IMpower study. Acknowledgments This publication is based on research using data from Roche that has been made available through Vivli Inc. Vivli has not contributed to or approved, and is not in any way responsible for the contents of this publication. David J. Pinato would like to acknowledge infrastructural support provided by the NIHR Imperial Biomedical Research Centre and Imperial Experimental Cancer Medicine Centre (ECMC). The views expressed are those of the author(s) and not necessarily those of the funder, the NIHR or the Department of Health and Social Care. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Figure S1: CONSORT diagram of this study. ALK, anaplastic lymphoma kinase; EGFR, epidermal growth factor receptor. Figure S2: The Kaplan–Meier curves of (A) progression‐free survival and (B) overall survival according to the use of APAP (with or without APAP) and treatment arm (with or without atezolizumab) in overall population. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S3: Forest plot of APAP efficacy (versus without APAP) according to with or without atezolizumab by study. (A) Progression‐free survival and (B) overall survival. MST: months. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S4: Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without APAP by study. (A) Progression‐free survival and (B) overall survival. MST: months. APAP, acetaminophen; CI, confidence interval; HR, hazard ratio; MST, median survival time. Figure S5: The Kaplan–Meier curves of (A) progression‐free survival and (B) overall survival according to the use of NSAIDs (with or without NSAIDs) and treatment arm (with or without atezolizumab) in overall population. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S6: Forest plot of NSAIDs efficacy (versus without NSAIDs) according to with or without atezolizumab by study. (A) Progression‐free survival and (B) overall survival. MST: months. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S7: Forest plot of atezolizumab efficacy (versus without atezolizumab) according to with or without NSAIDs by study. (A) Progression‐free survival and (B) overall survival. MST: months. CI, confidence interval; HR, hazard ratio; MST, median survival time; NSAIDs, non‐steroidal anti‐inflammatory drugs. Figure S8: The Kaplan–Meier curves of overall survival in patients excluding PPI users. (A) APAP use or non‐use in without‐atezolizumab cohort. (B) APAP use or non‐use in with‐atezolizumab cohort. (C) NSAIDs use or non‐use in without‐atezolizumab cohort. (D) NSAIDs use or non‐use in with‐atezolizumab cohort. APAP, acetaminophen; NSAIDs, non‐steroidal anti‐inflammatory drugs; PPI, proton pump inhibitor. Table S1: Summary of the patients' baseline characteristics by IMpower study. Table S2: Summary of the observation time by IMpower study. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.