Patient characteristics
A total of 83 patients were enrolled. Patient characteristics are shown in Table 1. The median age was 72 (range: 44–86) and 63% (n = 52) of the patients were male. Fifty-five patients (66%) had good performance status (ECOG performance status; 0–1). Twenty-six patients (31%) had liver metastases and 30 patients (36%) had ascites resulting from peritoneal dissemination.
Of the 83 patients, the NLRs at 1st-line initiation ranged from 0.72 to 48.9 (the median NLR was 3.00). Based on the definitive cut-off value for NLR, 42 patients were categorized into an NLR-low group (NLR < 3.00) and 41 patients into an NLR-high group (NLR > 3.00, Fig. 1).
At the time of the analysis, 12 patients were still undergoing 1st-line treatment, 15 patients had been transferred to best supportive care (BSC) after 1st-line treatment, and 56 patients (79%) had received 2nd-line treatment. Of the 56 patients, 38 patients were classified into the NLR-low group and 18 patients into the NLR-high group according to their blood NLRs evaluated prior to the start of 2nd-line treatment.
Among the 56 2nd-line treatment patients, 4 were still undergoing 2nd-line treatment and 34 (65%) had received 3rd-line treatment. Of the 34 3rd-line treatment patients, 23 were categorized into the NLR-low group and 11 patients into the NLR-high group according to blood NLR values prior to 3rd-line treatment. Finally, of the 34 patients treated with 3rd-line chemotherapy, 2 were still under 3rd-line treatment, 21 (66%) received 4th-line chemotherapy, and the remaining 11 patients were transferred to BSC.
Clinical outcome in relation to NLR values prior to each chemotherapy line
Table 2 presents clinicopathological characteristics of GC patients at each chemotherapy line in relation to the NLR-high and NLR-low groups. Regarding the patient characteristics, the NLR-high group had significantly more patients with a poor performance status and ascites throughout the course of chemotherapy.
Of the 83 patients who were analyzed, the median OS was 13.2 months (95% CI: 9.1–17.9). When we divided the patients into two groups based on NLR value evaluated prior to 1st-line chemotherapy, the median OS was shorter in the NLR-high group compared with the NLR-low group (OS: 8.0 months [95% CI: 5.7–13.2] vs. 18.1 months [11.9–20.9], p = 0.06, Fig.2, Panel A). Of 56 patients who received 2nd-line chemotherapy, the median OS was significantly shorter in the NLR-high group compared with the NLR-low group (OS: 4.5 months [95% CI: 3.4–6.0] vs. 10.7 months [95% CI: 7.4–13.7], p < 0.05, Fig.2, Panel B). Of the 34 patients who received 3rd-line chemotherapy, the median OS was significantly shorter in the NLR-high group compared with the NLR-low group (OS: 4.7 months [95% CI: 0.8–7.3] vs. 8.7 months [95% CI: 5.6–14.0], p < 0.05, Fig.2, Panel C).
Next, we focused on the 24 patients who received nivolumab monotherapy during 3rd-line treatment. According to the NLR value prior to nivolumab therapy, 14 patients (55%) were stratified into the NLR-low group and 10 patients into the NLR-high group. Patient characteristics are shown in Table 3. The median OS was shorter in the NLR-high group compared with that in the NLR-low group (median OS: 6.6 months [95% CI: 0.8–7.3 months] vs. 8.3 months [2.6–10.3 months], p = 0.06, Fig.2, Panel D).
Changes in NLR value throughout chemotherapeutic drug treatment in unresectable GC patients
We examined fluctuations of pretreatment NLR values in each case throughout the course of chemotherapy (Fig. 3). Among 35 NLR-low patients at 1st-line who were eligible for 2nd-line treatment and categorized into the NLR-low group by pretreatment NLR values at 1st-line, 23 (66%) retained their NLR within 3.0 and only 4 cases (13%) showed an increase in NLR by 3.0 or more prior to 2nd-line therapy. In contrast, of the 35 NLR-high patients at 1st-line who were transferred to 2nd-line therapy, although 15 patients (43%) recovered, their NLR was within 3.0, and 14 (40%) retained an NLR 3.0 value or more prior to 2nd-line therapy (p = 0.026, Fig. 4).