Readministration of nivolumab for patients with advanced gastric cancer: a case series.

Introduction
Gastric cancer is the fifth most frequently diagnosed cancer and the fourth leading cause of cancer-related mortality, with especially high incidence and death rates in East Asia.1 Although chemotherapy extends survival in patients with advanced gastric cancer (AGC), median overall survival (OS) remains poor at ∼15 months,2 underscoring the need for improved therapeutic strategies.
Nivolumab, an immune checkpoint inhibitor (ICI) targeting programmed cell death protein 1 (PD-1), has demonstrated efficacy across several malignancies, including gastric cancer. The phase III ATTRACTION-2 trial showed that nivolumab monotherapy significantly prolonged median OS compared with placebo in later-line treatment for AGC.3 Further studies revealed that combining nivolumab or pembrolizumab with chemotherapy as first-line therapy for human epidermal growth factor receptor 2 (HER2)-negative AGC improved survival over chemotherapy alone.4, 5, 6 Consequently, nivolumab is now a standard treatment for patients with AGC.
In later-line AGC treatment, agents such as trifluridine/tipiracil hydrochloride (FTD/TPI) and irinotecan are frequently used; however, their survival benefit remains limited.7, 8, 9, 10 Some research indicates that readministration of ICIs following disease progression or recovery from immune-related adverse events (irAEs) may be both feasible and effective in certain cancers, including melanoma and lung cancer.11,12 Despite these findings, evidence regarding ICI readministration in patients with AGC is scarce, making the feasibility and efficacy of ICI readministration challenging to assess. This study evaluates the efficacy and safety of nivolumab readministration in patients with AGC.

Case presentation
We retrospectively analyzed the clinical records of eight patients with histologically confirmed unresectable or recurrent gastric cancer treated at Aichi Cancer Center Hospital between September 2017 and December 2022. All patients had previously received nivolumab, either as monotherapy or in combination with chemotherapy, and subsequently underwent nivolumab readministration—either as a rechallenge following disease progression or as a reintroduction after irAEs. The study was approved by the Institutional Review Board of Aichi Cancer Center Hospital (No. IR 051109) and conducted in accordance with the Declaration of Helsinki. Due to the retrospective nature of the analysis and absence of direct intervention, written informed consent was waived; patients were provided with an opt-out option via the institutional website.
During readministration, all patients received nivolumab monotherapy at 240 mg/body biweekly or 480 mg/body every 4 weeks until disease progression or unacceptable adverse effects. Rechallenge was defined as nivolumab readministration following disease progression, whereas reintroduction referred to resumption after recovery from irAE. Ascites were classified using computed tomography into four categories: no detectable ascites, mild (limited to the upper or lower abdomen), moderate, and massive (diffuse throughout the abdominal cavity). Tumor programmed death-ligand 1 (PD-L1) expression was assessed using the combined positive score (CPS), evaluated via immunohistochemistry using the 28-8 pharmDx kit (Dako, Santa Clara, CA).
Time to treatment failure (TTF) was defined as the duration from nivolumab readministration to treatment discontinuation due to any cause, including disease progression, toxicity, or death. Progression-free survival (PFS) and OS were measured from the date of readministration to disease progression, death, or last follow-up. TTF, PFS, and OS were estimated using the Kaplan–Meier method. Tumor response was assessed according to RECIST version 1.1, with no confirmation requirements. Objective response rate (ORR) was defined as the proportion of patients achieving a complete response (CR) or partial response (PR), while disease control rate (DCR) included CR, PR, or stable disease (SD). Adverse events, including irAEs, were graded using the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. Data analyses were conducted with JMP version 15.1.0 (SAS Institute Japan Ltd., Tokyo, Japan).
Eight patients were included in this analysis. Baseline patient characteristics are summarized in Table 1. Six patients discontinued the initial nivolumab course due to disease progression and later underwent rechallenge; two discontinued and underwent reintroduction. The treatment regimen and responses are summarized in Figure 1A and Table 2.

Patients with nivolumab rechallenge
Among six patients, three patients had initially received nivolumab in combination with chemotherapy, while three had been treated with monotherapy. Elevated alkaline phosphatase (ALP) levels above the upper limit of normal (ULN) were observed in four patients. Three had ascites. The median number of prior treatment lines before nivolumab rechallenge was 5 (range 2-9); four patients had received irinotecan and two had received FTD/TPI as prior treatments.
During the initial nivolumab treatment, the median TTF and PFS were 6.9 months [95% confidence interval (CI) 1.4-11.7 months] and 6.9 months (95% CI 1.9-22.4 months), respectively. Two patients achieved PR, and one had SD. The median interval between initial treatment and rechallenge was 13.9 months (range 5.1-23.5 months). Following rechallenge, all patients discontinued treatment due to disease progression and subsequently died. The median PFS and OS were 1.8 months (95% CI 0.6-19.3 months) and 17.9 months (95% CI 1.8-42.0 months), respectively (Figure 2A and B). Two patients achieved SD during rechallenge, with an ORR of 0% and a DCR of 33%, respectively. Waterfall plots for five patients with measurable lesions are shown in Figure 1B. Patients with ALP levels below the ULN were more likely to achieve SD, while those with elevated ALP predominantly exhibited progressive disease (PD). Three of the four patients with PD had bulky liver or peritoneal metastases. The presence of ascites and an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 1 were associated with shorter PFS and OS. No other clinicopathological features were identified as predictors of response. Regarding safety, one patient developed a grade 1 rash and grade 1 arthralgia, which were considered irAE-related. No additional treatment-related adverse events or deaths occurred.

Patients with nivolumab reintroduction
During initial nivolumab treatment, both patients diagnosed with grade 2 pneumonitis associated with irAE were started on steroid therapy at a dose of 0.5-1 mg/kg. The interval between full recovery from pneumonitis and nivolumab reintroduction was 8.5 months in case 7 and 19.7 months in case 8, respectively. Following reintroduction, the patient of case 7 achieved PR and waterfall plot for the two patients is shown in Figure 1B. Regarding safety, the patient in Case 7 experienced a grade 2 rash, which was considered irAE-related. However, no additional treatment-related adverse events including pneumonitis or deaths occurred.

Discussion
In clinical practice, ICI readministration is occasionally pursued for selected patients with AGC. To our knowledge, this is the largest retrospective series to evaluate nivolumab readministration in patients with AGC. Although the cohort size was limited, our findings suggest that nivolumab readministration may provide clinical benefit in patients with low ALP levels, good ECOG PS, and absence of ascites or bulky metastases. Importantly, no recurrence of pneumonitis—the cause of discontinuation during initial treatment in two reintroduction cases—was observed, and all irAEs during readministration were grade ≤2. Furthermore, this study is the first to examine the efficacy and safety of ICI readministration after initial treatment involving combination chemotherapy.
Treatment options in the later-line setting for patients with HER2-negative AGC are limited. Previously reported ORR ranged from 0% to 22%, DCR from 40% to 85%, median PFS from 1.6 to 5.9 months, and median OS from 4.0 to 9.4 months.7, 8, 9, 10,13, 14, 15 Additionally, only 53% and 30% of patients have been reported to retain adequate oral intake during second-line and third-line therapy, respectively,16 limiting the feasibility of agents like FTD/TPI. Irinotecan use is also constrained by its toxicity profile. In comparison, patients with nivolumab rechallenge in our study yielded similar median PFS and OS, and showed promising trends in DCR, despite most patients being heavily pretreated with a median of five prior regimens.
The rationale for ICI readministration is based on the hypothesis that tumor sensitivity may be restored following a drug-free interval.17,18 Although ICIs often induce durable responses even after discontinuation, the longevity of this effect and its impact on subsequent ICI efficacy remain unclear. Systematic reviews suggest that ICI readministration can offer outcomes comparable to those of initial therapy. However, many of these rechallenge or reintroduction studies include heterogeneous ICI regimens, such as switching to cytotoxic T-lymphocyte-associated protein 4 inhibitors or combining agents, and do not focus exclusively on anti-PD-(L)1 monotherapy.11,12 We extracted data from reports specifically involving only anti-PD-1 or anti-PD-L1 rechallenge. Across these studies, median PFS ranged from 1.5 to 4.4 months, median OS from 2.7 to 15.7 months, and ORR from 0% to 32%. In our rechallenge cohort, median PFS of 1.8 months, OS of 17.9 months, and DCR of 33% were similar to those seen in initial nivolumab monotherapy trials for later-line AGC.3 Notably, no relationship was observed between the ICI-free interval length and treatment efficacy. In contrast, a phase II study (WJOG9616L) evaluating nivolumab retreatment in patients with non-small-cell lung cancer reported that a longer ICI-free interval was associated with improved PFS.19
Several clinical features have been associated with limited benefit from nivolumab in AGC, including poor ECOG PS, low serum albumin, high ALP levels, diffuse-type histology, peritoneal metastasis, and ascites.20 Our data corroborate these findings: patients with liver or peritoneal metastases and impaired PS exhibited shorter PFS and minimal tumor response. Currently, there is insufficient evidence regarding biomarkers that might guide patient selection for ICI rechallenge. Future studies are needed to address this gap.
Regarding safety, no severe irAEs were observed during readministration, even in patients with reintroduction who had previously discontinued nivolumab due to pneumonitis. A prior systematic review noted that anti-PD-(L)1 reintroduction has a safety profile comparable to that of treatment-naive use.11,12 Our findings support the notion that nivolumab can be safely reintroduced, even in patients with a history of irAE.
This study has several limitations. Firstly, it was a retrospective analysis conducted at a single institution, with a small sample size that limits the generalizability of the findings. Secondly, among patients who initially received nivolumab with chemotherapy, it remains difficult to determine whether observed efficacy during rechallenge was attributable solely to nivolumab. Thirdly, data on predictive biomarkers such as CPS or microsatellite instability status were incomplete. Fourthly, although pseudoprogression during initial nivolumab treatment cannot be excluded due to unconfirmed response assessment, it is unlikely because all such patients had PD after rechallenge. To address these limitations, we initiated a prospective observational study of nivolumab rechallenge for AGC (NIVO RETURNS study) in Japan in April 2023 (UMIN000050515).

Conclusions
In conclusion, nivolumab readministration may offer clinical benefit to select patients with AGC, particularly those with favorable clinical characteristics—including low ALP levels, good ECOG PS, and absence of ascites or bulky metastases—while demonstrating an acceptable toxicity profile. Future studies are essential to validate these findings and identify biomarkers that can better predict treatment response.