Impact of body mass index on progression-free survival and endocrine immune-related adverse events in advanced NSCLC patients treated with nivolumab: A retrospective cohort study

Abstract

OBJECTIVE

Body mass index (BMI) is a well-established prognostic factor in many cancer types. While BMI has been paradoxically associated with improved outcomes under immune checkpoint inhibitors (ICIs), its role in toxicity risk remains underexplored. This study examines the relationship between BMI and both progression-free survival (PFS) and immune-related adverse events (irAEs) in patients with advanced non-small cell lung cancer (NSCLC) receiving second-line treatment with nivolumab.

METHODS

This retrospective study examined 305 patients with histologically confirmed stage IV NSCLC who experienced progression after platinum-based chemotherapy. Patients were grouped based on whether their BMI was above or below the median value of 24.8 kg/m². PFS was assessed using Kaplan-Meier analysis. Frequencies of endocrine, hematologic, gastrointestinal, hepatic, and pulmonary toxicities were compared between BMI groups. Statistical significance was set at p<0.05.

RESULTS

No statistically significant difference in PFS was observed between BMI groups (<24.8: 7.53 months vs. ≥24.8: 8.30 months, p>0.05). Endocrine irAEs, however, occurred more frequently in patients with a higher BMI (p=0.034). No significant associations were found between BMI and hematologic (p=0.637), hepatic, gastrointestinal, or pulmonary irAEs. BMI was not predictive of PFS but was associated with endocrine toxicity.

CONCLUSION

These findings suggest that BMI may serve as a simple indicator for toxicity risk stratification in immunotherapy settings.

Highlight key points

• Efficacy & Progression-Free Survival (PFS): There was no significant difference in progression-free survival (PFS) between patients with BMI <24.8 kg/m² and those with BMI ≥24.8 kg/m², with median PFS being 7.53 months and 8.30 months, respectively.

• Safety Profile & Immune-Related Adverse Events (irAEs): Immune-related adverse events occurred in 5.9% of patients overall. Notably, endocrine irAEs (e.g., thyroid dysfunction) were significantly more common in the higher-BMI group (8.5% vs 2.0%).

• Clinical Implication: While BMI does not affect PFS with nivolumab, it can help identify patients at higher risk for endocrine-related toxicities, suggesting the need for closer endocrine monitoring during ICI therapy.

Non-small cell lung cancer (NSCLC) accounts for approximately 85% of all lung cancer cases and continues to be the leading cause of cancer-related deaths worldwide [1]. Despite advances in diagnostic and therapeutic strategies, most patients present with disease at an advanced stage, resulting in a poor prognosis and limited long-term survival. The advent of immune checkpoint inhibitors (ICIs), especially those targeting the PD-1/PD-L1 axis, has revolutionized the treatment of advanced NSCLC, providing better survival rates for specific populations [2, 3]. Nivolumab, a fully human IgG4 monoclonal antibody targeting PD-1, has demonstrated significant clinical benefits in second-line settings following platinum-based chemotherapy failure [4–6].

Despite the benefits of ICIs, response rates are heterogeneous, and predictive biomarkers remain suboptimal. Traditional markers, including PD-L1 expression, tumor mutational burden (TMB), and microsatellite instability (MSI), have demonstrated potential but are inconsistent and not universally applicable [7, 8]. Consequently, interest has grown in easily measurable host-related biomarkers, such as body mass index (BMI) [4, 8].

BMI is widely accessible and cost-effective, reflecting a patient’s nutritional and metabolic status. Paradoxically, several studies have reported improved outcomes in overweight or obese cancer patients undergoing immunotherapy, a phenomenon termed the “obesity paradox” [9–11]. Potential mechanisms include adipose tissue-mediated modulation of immune function, altered cytokine profiles, and differential pharmacokinetics of therapeutic antibodies. Furthermore, obesity-related inflammation may enhance immune activation but can also predispose individuals to immune-related adverse events (irAEs), particularly endocrinopathies such as thyroiditis and hypophysitis [12, 13].

Given these conflicting implications, this study aimed to investigate the role of BMI in predicting treatment effectiveness and toxicity in a real-world group of patients with advanced NSCLC receiving second-line nivolumab.

Materials and Methods

This retrospective cohort study was carried out at Kartal Dr. Lutfi Kirdar City Hospital in Istanbul, Turkiye. It included 305 consecutive patients with stage IV NSCLC who received nivolumab as second-line treatment from January 2017 to December 2023. All patients had histologically confirmed NSCLC and experienced progression following prior platinum-based chemotherapy.

Demographic and clinical variables, including age, sex, smoking status, Eastern Cooperative Oncology Group (ECOG) performance status, tumor histology, and metastasis status, were collected.

BMI was calculated from baseline measurements taken before nivolumab treatment by dividing weight in kilograms by the square of height in meters. Patients were stratified into two groups based on the cohort’s median BMI (24.8 kg/m²) to ensure balanced group sizes and improve internal statistical power. Although the World Health Organization (WHO) BMI classification is widely used in clinical practice, the distribution in our cohort did not permit robust subgrouping across standard categories (e.g., underweight, overweight, obese), as this would have led to considerable imbalance and underpowered comparisons.

Due to data unavailability, PD-L1 expression levels were not included in the analysis, a limitation noted in the discussion. Nivolumab was administered as a second-line treatment during a period when its use was not contingent on PD-L1 status. Therefore, PD-L1 testing was not routinely performed for all patients, and these data were not uniformly available in the medical records. Also, data on organ-specific metastatic involvement (e.g., liver, brain, bone) and M substage classification (M1a, M1b, M1c) were retrospectively reviewed from available medical records. However, due to inconsistent documentation and incomplete radiologic staging details across the study period, these variables were not uniformly available for all patients and were therefore excluded from the final statistical analysis.

irAEs were identified retrospectively by reviewing electronic medical records, oncology progress notes, laboratory results, and endocrinology consultations. Only clinically significant irAEs (Grade ≥2) or those requiring therapeutic intervention (e.g., corticosteroids, endocrinology referral, or treatment delay/discontinuation) were included in the analysis. Because this was a retrospective chart review, low-grade (Grade 1) irAEs may have been underreported or inconsistently documented.

Endocrine irAEs were identified based on documented clinical diagnoses, laboratory abnormalities (e.g., TSH, free T4, cortisol), and endocrinology consultations. Events were categorized using CTCAE v5.0 criteria when sufficient documentation was available. However, due to the retrospective design, complete grading and temporal data (e.g., time to onset, reversibility) were not uniformly documented and were therefore not analyzed in detail.

Progression-free survival (PFS) refers to the period from the initial dose of nivolumab until either disease progression confirmed by radiology or death from any cause. Response assessment is conducted using RECIST v1.1 criteria [14].

Adverse events were documented and categorized following the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 [15]. Toxicities were grouped by system: endocrine (such as thyroiditis, hypophysitis, diabetes mellitus), hematologic (including anemia, leukopenia, thrombocytopenia), gastrointestinal (like colitis, diarrhea), hepatic (elevations in serum transaminases and bilirubin), and pulmonary (pneumonitis).

Cases with missing data on key variables were excluded from the relevant statistical analyses, and no imputation was performed.

PFS was calculated using the Kaplan-Meier method, and group differences were assessed with the log-rank test. The correlation between BMI and categorical toxicity events was analyzed using either the chi-square or Fisher’s exact test. A p-value below 0.05 was regarded as statistically significant. To evaluate the independent association of BMI with PFS, a multivariable Cox proportional hazards regression model was fit. Covariates included age, sex, ECOG performance status, tumor histology, and BMI category (<24.8 vs. ≥24.8 kg/m²). Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported. The proportional hazards assumption was tested and found to hold.

All analyses were performed using SPSS software version 25.0 (IBM Corp., Armonk, NY).

This study was carried out in accordance with the principles of the Declaration of Helsinki and received approval from the Ethics/Institutional Review Board of Kartal Lutfi Kirdar City Hospital (approval date: April 30, 2025; approval number: 2025/010.99/15/10).

Results

The study encompassed 305 patients, with 84.6% being male, and the average age recorded at 63.0±8.6 years (ranging from 24 to 88). Adenocarcinoma emerged as the most common histologic subtype at 43.3%, followed by squamous cell carcinoma at 40.0%. Detailed data on organ-specific metastatic distribution and M substage were not consistently available in the medical records and therefore could not be reliably analyzed across the entire cohort. The median BMI was 24.8 kg/m². Among the participants, 152 individuals had a BMI below 24.8 kg/m², whereas 153 individuals presented with a BMI of 24.8 kg/m² or higher (Table 1).

Table 1.

Patient’s characteristics

Characteristic	n	(%)
Age
Mean±SD	63.0±8.6
Range	24–88
ECOG performance status
0	153	50.1
1	152	49.9
Sex
Female	47	15.4
Male	258	84.6
BMI Category
<24.8	152	49.9
≥24.8	153	50.1
Histopatholgy
Adenocarcinoma	132	43.3
Squamous cell carcinoma	122	40.0
Large cell	2	0.7
Mixed type	42	13.7
Other	41	13.4

SD: Standard deviation; ECOG PS: Eastern Cooperative Oncology Group performance status; BMI: Body mass index. “Mixed-type” histology refers to tumors that demonstrate features of both adenocarcinoma and squamous cell carcinoma, as documented in the original pathology reports.

The median PFS was 7.53 months (95% CI: 5.05–10.01) for the group with a BMI below 24.8, and 8.30 months (95% CI: 6.03–10.56) for those with a BMI of 24.8 or higher. There was no statistically significant difference between the groups (p>0.05) (Fig. 1).

Figure 1.

Progression-free survival by BMI categories.

BMI: Body mass index.

The overall incidence of irAEs was 5.9%. Among these, endocrine irAEs were significantly more prevalent in the group with higher BMI (p=0.034). Specific endocrine events encompassed hypothyroidism, hyperthyroidism, and suspected hypophysitis. Hematologic toxicities, such as anemia and neutropenia, were observed in 9.8% of patients, with no significant correlation to BMI (p=0.637). Gastrointestinal toxicities were reported in 4.3%, hepatic toxicities in 3.6%, and pneumonitis in 2.3%, with no statistically significant differences across BMI categories (Table 2). Although no statistically significant differences were observed between BMI groups for hematologic (p=0.637), hepatic, gastrointestinal, or pulmonary irAEs, the very low incidence of these events limited our ability to conduct meaningful statistical comparisons. Therefore, the absence of significance should be interpreted with caution given the limited power.

Table 2.

Frequency of endocrine irAEs by BMI category

EndocrineEvent	BMI <24.8 (n=152)%	BMI ≥24.8 (n=153)%	p
Hypothyroidism	1.3	3.9	>0.05
Hyperthyroidism	0.7	2.6	>0.05
Suspected Hypophysitis	0.0	2.0	>0.05
Total Endocrineir AEs	2.0	8.5	0.034

BMI: Body mass index; irAEs: immune related advers events.

Among endocrine irAEs, most were Grade 2 according to available documentation. The majority were managed with hormone replacement therapy (levothyroxine, hydrocortisone) and did not lead to permanent treatment discontinuation. However, precise time-to-onset and reversibility data were inconsistently recorded and thus not analyzed. Suspected hypophysitis was defined as a clinical diagnosis by the treating oncologist or endocrinologist in the presence of fatigue, hyponatremia, hypotension, and low cortisol levels, with or without confirmatory imaging.

In the multivariable Cox regression analysis (Table 3), BMI ≥24.8 kg/m² was not independently associated with improved PFS compared with BMI <24.8 kg/m² (HR: 0.91; 95% CI: 0.72–1.14; p=0.39). ECOG performance status 1 (vs. 0) was associated with inferior PFS (HR: 1.45; 95% CI: 1.10–1.91; p=0.008). Age, sex, and histology were not significantly associated with PFS.

Table 3.

Multivariable cox proportional hazards analysis for progression-free survival

Variables	HR	95% CI	p
BMI ≥24.8 vs<24.8	0.91	0.72–1.14	0.39
Age (per year increase)	1.01	0.99–1.03	0.27
Male vs Female	1.12	0.83–1.51	0.47
ECOG 1 vs ECOG 0	1.45	1.10–1.91	0.008
Adenocarcinoma vs others	0.97	0.75–1.26	0.83

HR: Hazard ratio; CI: Confidence interval; BMI: Body mass index; ECOG: Eastern Cooperative Oncology Group performance status. “BMI ≥24.8” refers to patients with BMI equal to or above the cohort median; “others” in histology includes squamous cell carcinoma, large cell carcinoma, mixed histology, and other NSCLC subtypes.

Discussion

Our analysis revealed that while BMI was not significantly associated with PFS in NSCLC patients receiving nivolumab, those with higher BMI experienced more endocrine-related irAEs. This finding reinforces the emerging notion that obesity modulates immune responses and toxicity profiles under ICI therapy [13, 16, 17].

The higher rate of endocrine irAEs observed in patients with elevated BMI may reflect underlying immunometabolic differences. Prior studies have hypothesized that obesity-related chronic inflammation—characterized by elevated leptin, IL-6, and TNF-a—can modulate immune activation and potentially increase susceptibility to autoimmune phenomena, including endocrine toxicities. Leptin, for instance, has been shown to promote T-cell proliferation and pro-inflammatory cytokine production, which could plausibly contribute to an increased risk of immune-related adverse events. However, because our study did not directly measure these biomarkers, these mechanistic interpretations remain speculative and should be interpreted with caution [18, 19]. These changes may account for the higher rate of endocrinopathies observed in our higher BMI group [3, 20].

Despite theoretical benefits, our findings did not demonstrate a significant PFS advantage in patients with higher BMI, which contrasts with other studies [4, 16] that reported improved survival in overweight patients with NSCLC and melanoma, respectively. These discrepancies may stem from differences in patient populations, sample size, and the retrospective nature of most data [21, 22]. Additionally, BMI may not adequately capture relevant body composition details such as sarcopenia and visceral adiposity, both of which are known to affect ICI efficacy [23, 24].

In the multivariable Cox regression model, BMI was not an independent predictor of progression-free survival after adjusting for relevant clinical covariates, including ECOG performance status, histology, age, and sex. These findings further support the limited prognostic value of BMI alone in the immunotherapy setting.

Real-world heterogeneity, comorbidities, and the lack of PD-L1 data may also explain the observed absence of a survival benefit. Our findings align with other reports indicating no association between PFS and BMI-related toxicity [22, 25, 26].

Importantly, our study emphasizes the necessity of close endocrine monitoring in overweight and obese patients undergoing ICI therapy. Given the potential for irreversible damage, early detection and intervention are critical [16, 24]. The increase in endocrine irAEs in higher BMI patients aligns with the idea that obesity contributes to enhanced immunogenicity and autoimmune susceptibility via leptin-driven Th1 polarization and elevated IL-6/TNF-a expression. Our findings agree with prior research on the prognostic and toxicological importance of BMI in patients with advanced NSCLC who are receiving ICIs. Several studies describe the “obesity paradox,” where higher BMI is unexpectedly linked to better survival rates among cancer patients receiving ICIs. Conversely, elevated BMI has also been linked to a higher rate of irAEs, especially endocrine toxicities like hypothyroidism and hypophysitis. In our study, although no statistically significant difference in PFS was observed across BMI groups, patients with higher BMI had a significantly higher rate of endocrine irAEs. This supports mechanistic research indicating that obesity-related inflammation—driven by increased leptin, IL-6, and TNF-a—creates a proinflammatory immune environment that may improve therapeutic effects but also increases susceptibility to autoimmune toxicities [11, 27, 28].

This study’s limitations are significant, including its retrospective design, single-center environment, and lack of inflammatory or metabolic biomarkers. An important limitation of this study is the absence of uniformly documented data on metastatic organ involvement and M substage classification (M1a–M1c). The extent and distribution of metastatic disease may influence both survival outcomes and immune-related toxicity risk, and could potentially act as confounding factors. Due to the retrospective design and incomplete staging documentation, these variables could not be included in the statistical models. Future prospective studies with detailed stratification of metastatic burden are warranted to clarify the independent role of BMI in this setting. Data on PD-L1 expression and systemic inflammatory or metabolic biomarkers (e.g., CRP, IL-6, neutrophil-to-lymphocyte ratio) are lacking. These parameters are known to influence both the efficacy and toxicity of immune checkpoint inhibitors, and their absence may confound observed associations with BMI. For instance, obesity is associated with chronic low-grade inflammation, which may overlap with immune activation mechanisms triggered by ICIs. Without adjustment for such biomarkers, it is difficult to determine whether BMI directly impacts outcomes or merely reflects underlying immunometabolic states. Future prospective studies incorporating these biomarkers are essential to better understand the complex interactions between host factors and immunotherapy responses.

A key limitation of our study is the incomplete characterization of endocrine irAEs. Because the data were retrospective, detailed information on toxicity grades, time to onset, reversibility, and long-term endocrine follow-up was often unavailable. Additionally, the definition of “suspected hypophysitis” relied on clinical judgment and laboratory findings rather than uniform imaging confirmation. Future prospective studies are needed to better define the spectrum and clinical implications of endocrine toxicities in this context. Another important limitation of this study is the very low incidence of non-endocrine irAEs, including hepatic, gastrointestinal, and pulmonary toxicities. As a result, our study lacked the statistical power to detect associations between BMI and these toxicities. Therefore, conclusions regarding the absence of a relationship between BMI and non-endocrine irAEs should be interpreted with caution, and further studies with larger event numbers are warranted. The relatively low overall incidence of irAEs (5.9%) in our cohort likely reflects the retrospective nature of the study and the selective inclusion of clinically significant or treatment-modifying toxicities. Mild irAEs (e.g., asymptomatic thyroid dysfunction or low-grade rash) may not have been consistently documented in medical records and were therefore not captured in this analysis. This limitation may have led to an underestimation of the overall irAE frequency compared with prospective clinical trials. To better elucidate the role of BMI in immunotherapy, future prospective studies should include thorough body composition analyses and immunophenotyping. These enhancements will help validate BMI’s influence in this context.

Conclusions

Although BMI was not predictive of progression-free survival, its significant association with endocrine irAEs suggests its potential as a simple, cost-effective tool for toxicity risk assessment in NSCLC patients receiving ICI therapy. These findings require further validation in prospective multicenter studies.

Ethics Committee Approval

The Kartal Lutfi Kirdar City Hospital Ethics Committee granted approval for this study (date: 30.04.2025, number: 2025/010.99/15/10).

Informed Consent

Written informed consents were obtained from patient and his family.

Conflict of Interest

None declared.

Financial Disclosure

The authors declared that this study received no financial support.

Use of AI for Writing Assistance

Not declared.

Authorship Contributions

Concept – SO; Design – SO; Supervision – NT, HO, NB; Fundings – SO; Materials – SY, HSY; Data collection and/or processing – SO, SY, HSY.

Peer-review

Externally peer-reviewed.