Body Fat Indices and Survival in Immunotherapy-Treated Patients With Cancer

We were intrigued to read the report from Martini et al,¹ a timely study that found a survival advantage in immunotherapy-treated patients with cancer with an elevated sub-cutaneous fat index (SFI), a high body mass index (BMI), and a low intramuscular fat index (IFI). The prevalence of obesity and the burgeoning use of immune checkpoint inhibitors (ICIs) in patients with cancer warrant the investigation of this critical question. Two recent reports found improved survival outcomes in patients with cancer and obesity after treatment with ICIs^2,3; however, both studies assessed adiposity by using the BMI, a readily available yet flawed obesity metric.⁴ We, therefore, applaud Martini et al for their use of computed tomography scanning and appreciate their findings, which suggest that an assessment of patient adiposity via computed tomography is plausible and correlates with patient outcomes. Moreover, this work by Martini et al is thought-provoking and raises several important questions and counterpoints for consideration in future studies.

First, we question the decision to pool all cancer types, given that obesity is associated with some of these cancers (ovarian, gastrointestinal, and breast cancers) but not others (skin and lung cancers).⁵ It is likely that excess body weight differentially affects obesity-associated and non–obesity-associated malignancies; thus, grouping these patients may obscure the results and dilute clinically relevant nuanced trends. For example, it has been reported that a high BMI is associated with favorable outcomes in ICI-treated men with melanoma.² Thus, it is possible that the trends observed by Martini et al¹ are driven by the predominance of patients with melanoma in this cohort. Separating patients by tumor type will determine whether distinct trends exist for cancers that are considered obesity related, an important question in the field.

Also concerning was the grouping of normal-weight BMI patients with underweight BMI patients. The World Health Organization Consultation on Obesity and the National Heart, Lung, and Blood Institute suggest the use of predefined standard BMI groupings (underweight, BMI < 18.5 kg/m²; normal weight, BMI = 18.5 to <25 kg/m²; overweight, BMI = 25 to <30 kg/m²; and obesity, BMI ≥ 30 kg/m²).^6,7 The use of nonstandard BMI categories may alter findings and complicate comparisons among studies.⁸ Therefore, we seek clarification on the decision to group normal-weight and underweight BMI patients, particularly if there were more patients with an underweight BMI versus a normal-weight BMI. If so, these analyses may be comparing cachexic and overweight/obese patients. This is concerning because cachexia is a known prognostic factor associated with shorter survival.⁹ To inform decision making in the clinical setting, individual mortality data for overweight and obesity in comparison with normal weight is needed.

Next, we seek clarification on the use of recursive partitioning to determine SFI-IFI risk groups. Although these methods enabled straightforward modeling of complex measures, the generalizability of the defined SFI-IFI thresholds is questionable. Quantifying the SFI, IFI, and visceral fat index (VFI) is a major strength of the study by Martini et al¹; however, grouping these measures into a regression tree to create risk scores may nullify known individual contributions of each adiposity type.¹⁰ For example, risk scoring may lead to the exclusion of the VFI from survival analysis. Including the VFI when one is analyzing immunotherapy-treated patients is important because of the known immunological consequences and inflammatory nature of visceral adipose tissue.^4,11 Importantly, other studies have reported detrimental effects of visceral adipose tissue on survival outcomes in patients with cancer.¹² We encourage separate analyses for the SFI, IFI, and VFI to determine individual contributions to survival outcomes.

Finally, we were intrigued by the finding that a high IFI correlated with worse survival outcomes in patients with a high SFI. The authors noted that the IFI measures myosteatosis as an alternative to assessing sarcopenia. However, sarcopenia was independently associated with a higher risk of death in patients with breast cancer,¹³ and this illustrates the importance of quantifying muscle mass. We, therefore, ask whether the authors quantified sarcopenia in addition to the IFI and whether this measure independently correlated with survival outcomes. A comprehensive analysis of both host adiposity and muscle mass is needed to fully understand the impact of excess body weight on survival outcomes in immunotherapy-treated patients with cancer.

We thank the authors for their contribution to the field of obesity and cancer and for providing an analysis that will surely lead to more studies. In designing future studies and prospective clinical trials, however, we urge caution when pooling patients of various malignancies, advise separate analyses of each BMI category and fat type, and recommend quantification of muscle mass in addition to adiposity. Further research encompassing these measures is needed to determine the impact of host adiposity and/or obesity on survival outcomes in patients with cancer treated with immunotherapy.