This multicenter, randomized, phase II/III study demonstrated that intraoperative administration of CS significantly reduced postoperative elevations of serum CRP and IL-6 levels after gastrectomy, without increasing the incidence of postoperative complications. Nevertheless, CS administration did not significantly prolong RFS after surgery for cStage II or III gastric cancer. However, among patients with differentiated histological-type or cStage III gastric cancer, the CS group had significantly better RFS than the control group. The survival benefit of CS in these subgroups should be reexamined in a future trial.

In Japan, intraoperative administration of CS is common during esophagectomy for esophageal cancer, and it has been weakly recommended by Japanese esophageal cancer practice guidelines since 2012 [20, 21]. It has been shown to reduce respiratory failure and in-hospital mortality rates, shorten hospital stays, and suppress postoperative CRP and cytokine elevations without increasing postoperative infectious complications [22, 23]. However, clinical studies on the intraoperative administration of CS for gastric cancer have not yet been reported, and its efficacy and safety remain unknown. A multicenter phase III trial (CORTIFRENCH trial) to assess the impact of intraoperative CS on both short- and long-term outcomes in digestive cancer was recently initiated [24]. The potential efficacy of intraoperative CS administration for various types of cancer is of considerable interest, and the current work can be considered a pioneering study with valuable findings.

We selected CRPmax as the primary endpoint for the phase II portion of the study, because our previous large-scale cohort study demonstrated that the postoperative CRP level was a good predictor of prognosis in patients with gastric cancer [9]. We considered CRPmax to be a more comprehensive marker of postoperative inflammation, as it captures not only the degree of surgical stress but also the additional inflammatory burden associated with postoperative complications, compared to CRP on postoperative day 1, day 3, or serum IL-6 on postoperative day 1. However, although intraoperative CS administration significantly decreased CRPmax, there was no significant improvement in prognosis. One possible reason for the negative effect of intraoperative CS administration on survival is that the extent of systemic inflammation in the two groups did not differ as significantly as expected. In our previous study, the difference in the median CRPmax between the low and high CRPmax groups was 9.54 mg/dL (7.95 mg/dL and 17.49 mg/dL, respectively), while in the present study it was 2.42 mg/dL (CS group, 9.78 mg/dL; control group, 12.2 mg/dL). This discrepancy may be due to an insufficient dose of CS in this study. In previous studies investigating the efficacy of intraoperative CS administration in gastrointestinal surgery, the flash dose of methylprednisolone generally ranged from 10 mg/kg to 30 mg/kg [22, 25, 26], which was more than double the dose used in the present study.

Additionally, the extent of systemic inflammation after gastrectomy may not be large enough for CS to achieve significant benefits. With recent advances in minimally invasive surgery and improvements in surgical techniques, postoperative inflammatory responses may have decreased compared to those observed in patients from our previous retrospective study [9]. Indeed, the subgroup consisting of cStage III patients, who were assumed to undergo relatively more invasive surgery than cStage II patients, showed a survival benefit from intraoperative CS administration. Moreover, patients who experienced postoperative complications, underwent open surgery, or received lymph node dissection of D2 or greater also showed a trend toward prolonged survival after intraoperative CS administration, although the effect was not statistically significant. Even when postoperative complications occur, intraoperative CS administration may suppress the release of cytokines and help prevent deterioration of long-term outcomes. These findings suggest that certain subgroups may potentially derive greater benefit from intraoperative CS administration, and further studies with larger sample sizes focusing on these subpopulations are warranted to validate this possibility.

Our subgroup analysis showed a significant survival benefit of CS administration among patients with the differentiated histological type of gastric cancer. This is probably because patterns of metastasis differ by histological type, i.e., the differentiated histological type is associated with hematogenous metastasis, while the undifferentiated histological type is associated with peritoneal metastasis [27]. Furthermore, CS administered into the peripheral blood is readily transferred to the lymphatic vessels, but less frequently to the peritoneum [28]. Our preclinical study revealed that CS administration reduced the incidence of liver metastases by suppressing the expression of E-selectin in the vascular endothelium of the portal vein [13]. Indeed, in the current study, the difference in the gastric cancer recurrence rate between the CS and control groups was larger in the lymph nodes (5.0 vs 8.3%, respectively) and liver (5.0 vs 6.4%, respectively) than in the peritoneum (16.3 vs 16.7%, respectively).

A negative aspect of steroid use is that it may promote tumorigenesis. A previous study reported that CS reduced T-cell activation by interfering with T-cell receptor signaling [29]. Additionally, CS influences the polarization of T helper (Th) cells, favoring the differentiation of regulatory T cells over that of Th1 cells, which leads to tumor promotion or metastasis [30, 31]. It is unclear whether an intraoperative flash dose of CS causes subsequent persistent changes in tumor immunity, but the aforementioned effects may have influenced our results. In future studies, the use of anti-inflammatory drugs targeting more specific cytokines, such as anti–IL-6 receptor antibodies, may be considered.

Another drawback of steroid use is the potential for side effects. However, no significant difference in the incidence of total postoperative complications or infectious complications was observed between the CS group and the control group in this study. Two systematic reviews and meta-analyses revealed that intraoperative CS administration decreased total complications in hepatic resection, whereas in colorectal surgery there was no difference in total complications, infectious complications, or anastomotic leakage [25, 32]. Since gastrectomy is more similar to colorectal surgery than to hepatic resection in terms of the extent of postoperative inflammation, we believe our results are consistent with those of previous studies. CS administration may have several side effects; among these, hyperglycemia is associated with a higher risk of anastomotic leakage and surgical site infections [33]. Indeed, our study showed a higher incidence of anastomotic leakage, although it was not significant (P = 0.062). However, in a nationwide database, preoperative prophylactic use of CS following esophagectomy did not increase anastomotic leakage [22]. Thus, this issue remains controversial and should be further investigated in future studies. Nevertheless, our findings suggest that intraoperative CS administration is safe and does not increase the incidence of steroid-associated side effects.

This study has several limitations. First and foremost, only 82% of the planned number of patients were enrolled. Therefore, the study was insufficiently powered to permit definitive conclusions about long-term outcomes. Second, this was an open-label randomized controlled trial, and the investigators were not blinded to the clinical data or postoperative complications. Third, we did not restrict the use of CS for anaesthesiological reasons such as postoperative nausea and vomiting prophylaxis. Indeed, six patients in the CS group and 21 in the control group received CS outside of protocol treatment for anaesthesiological reasons. However, the results of a sensitivity analysis conducted without these 27 patients were essentially unchanged (data not shown).

In conclusion, intraoperative administration of 5 mg/kg of CS mitigated postoperative CRP elevation; however, it did not provide a significant survival benefit after surgery for cStage II or III gastric cancer.