We retrospectively examined the relationship of changes in body weight and nutritional status with the efficacy of nivolumab for AGC. To our knowledge, no previous study has shown that weight loss and worsening nutritional status from first-line chemotherapy to the start of nivolumab treatment might affect the efficacy of body weight maintenance. Moreover, we showed that good nutritional status improved OS after initiation of nivolumab. Our study suggested that if LBW can be stopped early and nutritional status can be maintained, it could greatly benefit OS after the initiation of treatment in AGC.
LBW has been reported to affect prognosis during other chemotherapy regimens for AGC [11]. However, in our study, we focused on the use of nivolumab and investigated the importance of LBW during treatment prior to nivolumab. Our results showed that LBW < 4.5% after the initiation of first-line therapy was associated with improvement in OS and PFS in AGC patients later receiving nivolumab treatment. We then investigated the effect of nutritional status before the initiation of nivolumab treatment. We evaluated the CAR, which is a simple nutritional index, and has previously been reported to correlate with prognosis in cancer patients. We found that minimal changes in the CAR from first-line chemotherapy to start of nivolumab (ΔCAR < 0.01) correlated with a better prognosis, and that improvement or maintenance of nutritional status before starting nivolumab had a positive impact on prognosis. Thus, this study suggested that maintaining nutritional status during treatment prior to nivolumab had a positive effect on patients during subsequent nivolumab treatment.
We found that the CAR is a promising indicator of nutritional status. Cancer induces inflammatory cytokines, such as IL-6 and TNF-α, and causes LBW [6, 12]. Therefore, nutritional indicators reflecting these cytokines may more sensitively indicate cancer cachexia. The CAR has been established as a prognostic indicator in patients with acute disease [13], and is a useful prognostic factor in various cancers, including AGC [9]. IL-6 induces CRP production and affects cancer cell proliferation, invasion, metastasis, angiogenesis, and resistance to treatment, via the JAK/STAT3 pathway [12]. In addition, Alb is not only a nutritional indicator, but also an indicator of inflammation in the presence of inflammatory cytokines, such as IL-8. Therefore, the CAR might be an effective marker for investigating both nutrition and inflammation. It has been reported to be a better prognostic indicator for ICI treatment than other inflammatory factors, because it reflects IL-6 [14].
Based on our results, it is important to maintain weight loss below 4.5% and change in the CAR value below 0.01. However, it is particularly difficult to improve the nutritional status of patients with gastrointestinal cancer who have poor oral intake, and consequently cachexia [15]. Moreover, in a previous report, the existence of cancer cachexia was associated with a poor clinical outcome after nivolumab treatment in AGC [5]. Anamorelin has been developed as a selective and novel oral ghrelin-like agonist [16]. Ghrelin is an endogenous peptide, secreted primarily from the stomach, which binds to its receptors and stimulates multiple pathways that regulate body weight, muscle mass, appetite, and metabolism. Anamorelin increases body weight, muscle mass, and appetite in AGC patients with cancer cachexia. While general enteral nutritional supplements and professional nutritional guidance by a dietitian are important, anamorelin may also become an important factor in future. It should be noted, however, that although use of anamorelin has shown improvement in nutritional status, there are few data on the contribution of anamorelin itself to survival; therefore, further studies are required.
In the univariate and multivariate analyses, the LBW < 4.5% and ΔCAR < 0.01 groups had a better prognosis after nivolumab treatment (median OS, 12.9 months). In addition, the LBW < 4.5% group demonstrated better nivolumab efficacy. Our results suggested that OS and PFS in the good nutritional status group tended to be superior to those reported in the ATTRACTION-2 trial. Therefore, our study suggests that maintaining nutritional status may influence the effects of ICIs.
Our study had some limitations. First, only a few patients were included in our retrospective study. Second, the timing of the CT scans varied from case to case.
In conclusion, our study suggested that maintaining nutritional status during previous treatment may improve the effectiveness of subsequent ICI treatment. Nevertheless, further research is required in this regard.