A large retrospective research supported that melanoma patients with a BMI of 18.5–24.9 had a significantly shorter PFS than those with a BMI of 25.0–29.9 or ≥ 30 after therapy with pembrolizumab, nivolumab, or atezolizumab (median PFS: 19.9 vs. 27.2 or 28.8 months) [23]. However, one recent study showed that among 287 melanoma patients treated by ICIs, BMI was not related to clinical benefit or toxicity [13]. Another retrospective research indicated that solid malignant tumor patients, including NSCLC, melanoma, and renal cell carcinoma, with a BMI of ≥ 25.0 had a significantly longer PFS after ICIs treatment, compared with a BMI of <25.0 (11.7 vs. 3.7 months; HR: 0.46; 95% CI: 0.39–0.54; p < 0.0001) [24]. So far, it remains to be seen whether BMI was related to clinical benefit in lung cancer patients who had received ICIs. In this study, we found that lung cancer patients with a BMI of≥25 had a longer PFS compared with a BMI of<25. There was a positive association with overweight and better clinical outcomes with ICIs.
Overweight and obese patients have improved survival outcomes when comparing with patients with a normal body weight, which is known as “obesity paradox” [9]. At present, the mechanism of the influence of BMI on survival outcomes after ICIs therapy is just beginning to be understood. Obesity causes dysregulation of the immune response by promoting the formation of a systemic meta-inflammation may be the potential interpretation. A recent research indicated that adipose cells in the human obese subcutaneous adipose tissues could secrete a few of pro-inflammatory cytokines and chemokines, which contribute to establish and maintain inflammation, and consequence may enhance influence on immune checkpoint inhibitors [25]. Furthermore, part of the explanation of how BMI impact the efficacy of ICIs is that obesity increases T cell aging leading to higher PD-1 expression and dysfunction or the PD-1-mediated T cell dysfunction in obesity significantly leaves tumors markedly more responsive to ICIs according to basic experimental study [26].
Body composition is complicated, only BMI may not be enough to fully reflect it. BMI is not an accurate indicator of lean figure or adiposity [27]. In clinical practice, serum albumin level is usually chosen as an indicator for patients’ nutritional status. One retrospective study was pointed out that serum albumin level was not an independent predictable marker for overall response, but it was an important predictive and prognostic marker for anti-PD-1 treatment with NSCLC patients [20]. In our research, we found that the PFS of high ALB group was much longer than that of low ALB group, and the difference was significant. When renal function is considered, selecting serum creatinine as an indirect assessment of skeletal muscle mass represents a simple selection [16,17,18]. Cancers are highly proliferating and energy-demanding tissues. Especially in advanced malignant tumor, the increasing metabolic needs result in nutrient mobilization from skeletal muscle [28]. Low level of serum creatinine (< 0.7 mg/dL) is an indicator for weakness and sarcopenia especially for the older people; it is a powerful predictor of mortality in patients with chronic diseases who has a normal BMI [22]. Cachexia and sarcopenia are negative prognostic factors [14,15]. Our results also confirm this point, it showed that PFS time of low creatinine group was shorter than that of high creatinine group, and the difference was statistically significant. Low muscle mass is related to poor immunologic function, because skeletal muscle provides essential nutrients for the function of lymphocytes and monocytes [29,30;31] which probably related to the setting of immunotherapy based on checkpoint [32].
According to multivariate analyses including sex, age, histology, treatment line, ICIs type, serum ALB, BMI and creatinine, the results showed that serum ALB and creatinine were independent on influence factors for PFS, while sex, age, histology, treatment line, ICIs type and BMI were not significantly related to PFS. We suspect that the effects of BMI on PFS may be related to levels of serum albumin and creatinine. The above results indicated that nutritional status may be an important predictor of immunotherapy for the lung cancer patients. Nutrition is regarded as an important determining factor of immunoreaction, and dystrophy is the most common reason of immunodeficiency. The nutritional status of patients may influence the tumor microenvironment.
Myeloid-derived suppressor cells (MDSCs) are a marker of tumor-related inflammation and mediate the inhibition of T cell responses in lymphoma [33]. MDSCs are viewed as a heterogeneous population of cells at different differentiation stages. MDSCs can be differentiated into polymorphonuclear and monocyte MDSCs, which are respectively similar in morphologically and phenotypically to neutrophils and monocytes [34]. Studies showed that accumulate monocyte MDSCs resulted in reduced tumor infiltrating lymphocytes and increased tumorigenicity, aggravating immunosuppression [35]. Additionally, increasing evidence shows that there is a negative correlation between raised lymphocyte counts and tumor proliferation and invasion [36]. Platelets induce the migration of circulating cancer cells from epithelium to mesenchyme and promote the extravasation and metastasis of tumor cells [37,38]. It was found that PLR was significantly connected with the appearance of irAEs in NSCLC [39]. Thus, we evaluated the influence of LMR and PLR on overall survival in lung cancer patients with ICIs therapy. The results suggest that patients with baseline LMR≥2.12 had a longer PFS than those with LMR<2.12, while no significant difference was found between patients with PLR<135 and PLR≥135. LMR could serve as predictive biomarker for the efficacy of anti-PD-1 therapy to advanced lung cancer.
At present, one of the ’hottest topic’ is about the complex relationship between body composition and immune reaction, and some studies have attempted to explain that point [40]. The biological basis remains indistinct, therefore further researches are needed to illustrate these mechanisms. Furthermore, we consider that in prospective randomized research with non-ICIs control arm, BMI should be regarded as a stratification factor to better define its role in the treatment of checkpoint inhibitors. At the same time, further subgroup analysis is needed to confirm whether the impact of BMI on ICIs therapeutic effect is determined by serum albumin and creatinine levels.
This study is a real-word clinical study. Of course, there are several shortcomings in this study. As a retrospective study, there are inevitable selection bias. What’s more, a total of 66 patients were enrolled in the study, and small sample sizes may lead to the deviation of results. Further prospective researches on larger queues are needed to verify these results.