Although surgery represents the cornerstone of AGC treatment, it is mainly indicated in the absence of metastatic disease. Despite the proven value of SL in the diagnosis of PM, there is no consensus yet on its routine use in clinical practice, and it remains underutilized in the management of GC, with reported frequencies of less than 25%.[22–26] Thus, the present study was developed to further address this issue and to highlight the importance of SL in the diagnosis of PM and treatment strategy. In our study, of the 130 patients submitted to SL, PM was found in 66 (50.76%). Additionally, our results regarding the accuracy of CT scan were consistent with previous studies and confirmed its limitations in detecting PM, with 51.5% of sensitivity, 87.5% of specificity and 69.2% of accuracy.
Even with technological advances in imaging diagnostics, SL still provides a superior ability to inspect the peritoneal surface. Currently, several guidelines include its use in staging.[27–29] The main objective is to detect occult PM and other factors that can change the therapeutic strategy, such as the invasion of adjacent structures and liver metastasis.[8] However, the appropriate selection of patients who are candidates for SL is still controversial and differs among various institutions. Some recommend it for all resectable GC (stage IB–III), especially for those who are being considered for neoadjuvant treatment.[28]
Based on previous literature, it is difficult to define the most appropriate indications for SL, notably due to the differences in studied populations and inclusion criteria. In Western series, where most cases are diagnosed in advanced stages, the tendency was to perform it in all patients with AGC. While in Eastern series, where most patients present with early-stage disease, the recommendation was to perform it selectively in individuals at high risk for PM.[12, 30, 31] A systematic review carried out by Fukugawa et al. found that, in Japanese institutions, SL evidenced positive findings in 42.7% to 53.4% of the cases, higher than in other countries (7.8% to 40%). This discrepancy was partly because they had a greater selection of patients at risk for PM.[8] In our study, PM was found in 66 patients (50.8%) submitted to SL, which is also a higher number compared to previous studies, mainly due to the same selection bias of high risk cases.
Some factors are known to increase the poor prognosis of GC and to be related to the presence of PM, as large Bormann type 3 and Bormann type 4.[13, 32–34] In a prospective cohort study, with pre-specified indications for SL, Irino et al. found a greater impact, achieving accuracy of 91.5%. Their indications were: large ≥ 8cm Borrmann type 3, Borrmann type 4, bulky lymph nodes or paraaortic lymph node involvement and suspicion of PM on CT scan.[12] In our study, presence of ascites (p=0.001) and suspected PM on CT scan (p=0.007), were independent risk factors for PM (P1 group), suggesting that patients with these characteristics should always be considered for SL.
In previous systematic review, the use of SL provided a benefit by changing the treatment in 8.5–59.6% of cases, sparing patients from unnecessary laparotomy in 8.5–43.8% of cases.[25] These discrepancies are also partly due to the fact that the studies used different indications for performing SL. Our study revealed a considerably high percentage of therapeutic strategy change after SL (30.8%), sparing laparotomy in 24,6% of the cases and offering surgery to 6,1% of the patients who were previously considered stage IVb by CT scan.
Difficulties related to the widespread adoption of SL are related to cost, and its invasive risk-prone procedure.[13, 14, 35–37]. Increasingly restricted availability of scheduling in the operating rooms for a procedure that requires general anesthesia may also play a role. Regarding the cost-effectiveness of SL, Kevin et al. (2017) reported that the expected benefit from avoiding unnecessary laparotomies may be low compared to the cost of routine use of SL. Nevertheless, it can be good if the procedure yield is high, especially in those with a high risk of occult PM suggesting a more selective practice.[37]
When comparing survival rates of patients who underwent SL to those staged by tomography, a difference was observed in the survival curves of M0 patients by tomography versus M0 patients by SL. Despite the absence of statistical significance, this could represent the impact of under-staging of CT scan on prognosis. Better selection of patients who are candidates for curative treatment is achieved with SL and some patients considered M0 by tomography could show occult carcinomatosis, and consequently worse prognosis and survival. A more significant number of patients should be studied to assess the statistical difference between the groups.
Despite the high accuracy of SL, of the 64 patients classified as P0 by SL, 8 (12.5%) patients presented with PM during the second surgery. Some authors reported this outcome as “false negative for SL'', with rates ranging from 0% to 17.2%.[8, 12, 32, 33, 38] While this corresponds to a failure of SL, it could also represent a disease progression between SL and the second surgery. The answer, however, remains unknown.
Our study had some limitations. First, being a retrospective study it carries an inherent selection bias in indicating patients with a greater suspicion of PM for SL. Additionally, we do not perform second SL after neoadjuvant chemotherapy. Thus, the percentage of change of conduct is higher than that found in some other studies. Also, we did not include in the study the evaluation of peritoneal lavage by cytology, since we do not have the results available for all patients. In addition, in our service, the evaluation of peritoneal lavage is performed using cytological and immunohistochemical techniques. Until now, molecular techniques such as RT-PCR for investigating tumor cells in the lavage are not part of our diagnostic routine. Indeed, previous studies have demonstrated improvements in the sensitivity of peritoneal washing using molecular diagnosis via reverse transcription-polymerase chain reaction (RT-PCR), being also useful for predicting the peritoneal recurrence and prognosis.[39, 40]
Notwithstanding the limitations, our study has also important strengths. It represents a 11-year period Western tertiary single-center experience. Our findings demonstrated that SL had a significant impact in GC staging, especially in the diagnosis of PM, along with its superiority regarding conventional imaging tests. Patients with peritoneal spread still represent a major challenge in oncology. Currently, clinical researches related to effective strategies to improve long-term survival are underway and novel intraperitoneal chemotherapies have emerged as potential treatment options.[41, 42] According to this perspective, recognizing patients with metastasis limited only to the peritoneum, becomes even more important for best management.[43, 44]