Association Between Pre-Treatment Biological Indicators and Compliance to Neoadjuvant/Perioperative Chemotherapy in Operable Gastric Cancer.

Background and aims: Perioperative treatment is currently the gold standard approach in Europe for locally advanced gastric cancer (GC). Unfortunately, the phenomenon of patients dropping out of treatment has been frequently observed. The primary aims of this study were to verify if routine blood parameters, inflammatory response markers, sarcopenia, and the depletion of adipose tissues were associated with compliance to neoadjuvant/perioperative chemotherapy. Methods and study design: Blood samples were considered before the first and second cycles of chemotherapy. Sarcopenia and adipose indices were calculated with a CT scan before starting chemotherapy and before surgery. Odds ratios (OR) from univariable and multivariable models were calculated with a 95% confidence interval (95% CI). Results: A total of 84 patients with locally advanced GC were identified between September 2010 and January 2021. Forty-four patients (52.4%) did not complete the treatment according to the number of cycles planned/performed. Eight patients (9.5%) decided to suspend chemotherapy, seven patients (8.3%) discontinued because of clinical decisions, fourteen patients (16.7%) discontinued because of toxicity and fifteen patients (17.9%) discontinued for miscellaneous causes. Seventy-nine (94%) out of eighty-four patients underwent gastrectomy, with four patients having surgical complications, which led to a suspension of treatment. Sarcopenia was present in 38 patients (50.7%) before chemotherapy began, while it was present in 47 patients (60%) at the CT scan before the gastrectomy. At the univariable analysis, patients with basal platelet to lymphocyte ratio (PLR) ≥ 152 (p = 0.017) and a second value of PLR ≥ 131 (p = 0.007) were more frequently associated with an interruption of chemotherapy. Patients with increased PLR (p = 0.034) compared to the cut-off were associated with an interruption of chemotherapy, while patients with increased monocytes between the first and second cycles were associated with a lower risk of treatment interruption (p = 0.006); patients who underwent 5-fluorouracil plus cisplatin or oxaliplatin had a higher risk of interruption (p = 0.016) compared to patients who underwent a 5-fluorouracil plus leucovorin, oxaliplatin and docetaxel (FLOT) regimen. The multivariable analysis showed a higher risk of interruption for patients who had higher values of PLR compared to the identified cut-off both at pretreatment and second-cycle evaluation (OR: 5.03; 95% CI: 1.34–18.89; p = 0.017) as well as for patients who had a lower PLR than the identified cut-off at pretreatment evaluation and had a higher PLR value than the cut-off at the second cycle (OR: 4.64; 95% CI: 1.02–21.02; p = 0.047). Becker regression was neither affected by a decrease of sarcopenia ≥ 5% (p = 0.867) nor by incomplete compliance with chemotherapy (p = 0.281). Conclusions: Changes in PLR values which tend to increase more than the cut-off seem to be an immediate indicator of incomplete compliance with neoadjuvant/perioperative treatment. Fat loss and sarcopenia do not appear to be related to compliance. More information is needed to reduce the causes of interruption.