Role of staging laparoscopy for gastric cancer patients

Abstract Staging laparoscopy (SL) is frequently carried out in patients with advanced gastric cancer. However, some clinical questions are being debated and consensus must be obtained. With this aim, a literature search of PubMed/MEDLINE was carried out using the keywords “gastric cancer,” “SL,” and “diagnostic laparoscopy”. Articles published online up to February 2019 were analyzed, focusing on the following questions. (i) What is an adequate indication for SL? (ii) How do you carry out SL? (iii) Does SL provide accurate information about peritoneal dissemination? (iv) Is the yield of SL different by tumor location? (v) Is SL a safe procedure? (vi) Is “repeat SL” needed? (vii) Does SL provide oncological benefit? Results provided the following responses: (i) In Western countries, clinically resectable advanced tumor is an indication for SL. Terms to be introduced for adequate indication include “location,” “type 4 (linitis feature),” “large tumor,” “equivocal computed tomography (CT] findings,” and “lymph node swelling”. (ii) Exploration of the entire peritoneal cavity is preferable. (iii) Detection rate of peritoneal disease is 43%‐52% in Japanese institutions and 7.8%‐40% in other countries. False‐negative findings during SL were 0%‐17%, and 10%‐13% when limited to cytology. (iv) Yield of SL was higher in gastric cancer compared with esophagogastric junctional tumor. (v) SL‐related complications were estimated to occur in 0.4%. (vi) Repeat SL is important after treatment. (vii) If the efficacy of neoadjuvant chemotherapy for patients with P0CY1 is established, SL can provide oncological benefit. SL can be carried out safely and effectively. Considering the prevalence of neoadjuvant treatment, the role of SL will become more important.

| 497 FUKAGAWA metastases) than diagnosis by imaging. Peritoneal dissemination (P) may be diagnosed using a computed tomography (CT) scan with findings of ascites and multiple mesenteric or omental nodules, but diagnostic accuracy is not high. 7 Peritoneal dissemination is sometimes detected only during laparotomy, without definitive radiological findings. The presence of peritoneal dissemination, regarded as stage IV GC, is a poor prognostic factor. Patients have no indication for gastrectomy except for bleeding or obstruction, and undergo systemic chemotherapy. 8 If those patients undergo chemotherapy without palliative gastrectomy, laparotomy would be non-therapeutic. SL can provide accurate information about peritoneal dissemination and lavage cytology (CY) with less surgical invasiveness and an appropriate therapeutic strategy. 9 It is beneficial to avoid useless laparotomies and shorten the time between diagnosis and the initiation of chemotherapy. However, some problems regarding this procedure should be discussed.

| ME THODS
The present review was based on articles from PubMed and MEDLINE and carried out in February 2019. "Gastric cancer", "staging laparoscopy" and "diagnostic laparoscopy" were used as search

| RE SULTS
Results provided responses to the seven questions based on the database search.

| What is an adequate indication for SL?
The main purpose of SL is to detect occult peritoneal disease (P and/ or CY) that cannot be definitively diagnosed using imaging examinations. 10 In Western countries where advanced gastric cancer is common, a reported indication for SL was "resectable GC and EGJ without definite distant metastases". In many Japanese institutions, SL is carried out based on the clinical trials of the Japan Clinical Oncology Group (JCOG). JCOG0501 11 is a randomized controlled trial evaluating the efficacy of neoadjuvant chemotherapy (NAC) for patients with large type 3 (≥8 cm) and type 4 advanced gastric cancer. JCOG 0405 12 is a phase II clinical trial evaluating the efficacy of NAC for GC patients with bulky lymph node metastases (≥3 cm) or para-aortic node metastases (PAN). In both clinical trials, SL was mandatory for the confirmation of eligibility criteria. Therefore, "large type 3 and type 4" or "bulky N/PAN" are frequently adopted in practice as indications for SL.
Indications for SL are discussed in two ways (Table 1). First, some clinical factors that lead to a high incidence of peritoneal disease among SL cases are evaluated using multivariate analysis.
Sarela et al 13 reported the incidence of peritoneal dissemination by clinical findings: EGJ (42%), whole stomach (66%), poorly differentiated adenocarcinoma (36%), age ≤70 (34%), lymphadenopathy ≥1 cm by CT scan (49%), and depth of tumor was T3 or T4 (63%). Among these factors, "location (EGJ or whole stomach)" and "lymphadenopathy by CT scan" were significant predictive factors by multivariate analysis of 65 SL cases. Ikoma et al 14 reported similar results for location: fundus/body/antrum (38%), poorly differentiated (38%), signet ring cell morphology (41%), linitis feature (66%), and equivocal CT findings (65%). Among them, "poorly differentiated", "linitis feature" and "equivocal CT findings" were significant by multivariate analysis. The second way of discussing indications for SL is a validation method using a large number of cases, including patients who did not undergo SL. Tsuchida et al 15 determined that "three portions (=whole stomach)," "type 3/4/5" and "lymph node metastases by CT scan" were significant predictive factors for peritoneal disease by multivariate analysis of 31 SL cases. If the indication for SL was defined as two or three factors among these, sensitivity, specificity, positive predictive factor (PPV), negative predictive factor (NPV) and accuracy for peritoneal disease were 91.9%, 37.9%, 46.7%, 88.7% and 58.0%, respectively, using a total of 231 cases limited to c T3/T4. The F I G U R E 1 Peritoneal dissemination during staging laparoscopy. Small nodules of peritoneal dissemination on the surface of diaphragm study of Hu et al 16 used a similar method to Tsuchida. 15 The significant predictive factors for peritoneal disease were "≥4 cm", "T4b" and "type 3 or 4". If the indication for SL was defined as two or three factors among them, sensitivity, specificity, PPV, NPV and accuracy for peritoneal disease were 85%, 69%, 43%, 94% and 72%, respectively, using a total of 582 cases (c T2-4b). The report by Hur et al, 17 however, was not an analysis of SL cases. "Type 3 or 4", "T3 or T4" and "≥4 cm" were significant by multivariate analysis using 589 clinically advanced GC cases. If the indication for SL was defined as "all three factors", 42.4% of all cases were expected to be indicated for SL, and sensitivity, specificity, PPV, NPV and accuracy for peritoneal dissemination were 83.3%, 63.2%, 24.0%, TA B L E 1 Indications for staging laparoscopy (SL) for gastric cancer patients  96.4% and 65.7%, respectively, using the same series. In the report of Irino et al, 18 the indication for SL was defined as "large type 3 (≥8 cm)" or "type 4" or "bulky N" or "PAN" or "suspicious findings of peritoneal disease". Validation analysis using 721 cases (c T3/4) in the same period showed that sensitivity, specificity, PPV, NPV

| How do you carry out SL?
Surgical procedure for SL has been established, and it was similar in each study. It must be determined as to whether exploration inside the omental bursa is mandatory. If GC was located at the posterior side of the stomach, there could be peritoneal dissemination inside the omental bursa only. Several studies described inspection inside the omental bursa, 13,15,19-23 but the incremental detection of peritoneal dissemination was not described. We also questioned whether the total length of mesentery must be inspected. Ishigami et al 20 and Miki et al 24 included "the surface of the entire bowel" and "from the oral to anal side "in the exploration area of SL, but there was no description in other reports. Definite answers could not be drawn from the reported articles, but wide exploration may be needed to reduce false-negative results.

| Does SL provide accurate information about peritoneal dissemination?
Reports on the detection of P1 and/or CY1 that were not seen by imaging examinations are listed in Table 2. Ikoma et al 14 defined "the yield of SL" as the proportion of patients among all patients who underwent laparoscopy for staging whose laparoscopy showed positive findings, including those with macroscopic carcinomatosis, positive cytology, or other clinically important findings. The main purpose of SL was to find disseminated nodules that could not be detected by imaging, but other clinical findings that could change the therapeutic strategy (such as liver metastases or invasion to adjacent organs) may be detected at SL. Irino et al, 18 Hosogi et al, 25 Miki et al, 24 Ishigami et al, 20  the rate of false negativity was lower than that from Japan. 15,18,19,24,25 This is also due to differences in the indication for SL.
CY0 at initial SL may change to CY1 at the following laparotomy.
Yamagata et al 19

| Is the yield of SL different by tumor location of either esophagogastric junctional cancer or gastric cancer?
Some studies about SL from Western countries included not only GC, but also EGJ cancer or lower esophageal cancer (including squamous cell carcinoma) for the indication of SL. Yield of SL was higher in cases with GC compared with EGJ in all studies except one (Table 4).

| Is SL a safe procedure?
Staging laparoscopy is considered a safe procedure carried out within one hour under general anesthesia. Some reported intestinal injury during SL. 28,29,39,41 In many reports, there were no SL-related complications ( Table 5), but the estimated SL-related complication rate was 0.4% based on total accumulated data.

| Is "Repeat SL" after chemotherapy needed?
"Repeat SL" refers to a second SL carried out after chemotherapy. Thiels et al 42

| Does SL provide oncological benefit?
Detection of occult peritoneal disease can avoid non-therapeutic laparotomy and shorten the interval to induction of chemotherapy.
However, the survival benefit was obscured because the therapeu-    (Figure 2), even if initial SL may be avoided.

| D ISCUSS I ON
"False-negative SL" is a problem. In many studies from Japanese institutions, the proportion of "false negative" rates was reportedly over 10%. The indication for SL in those studies was large type 3 and type 4 that had a high potential of peritoneal disease, so the incidence of "false negative" rates was high. Hato et al 52 reported a high incidence of "false negative SL" in JCOG0501 targeting the same population. Careful exploration seems to be important in cases with suspicion of peritoneal dissemination.
In many reports, patients with P0CY1 have a poor prognosis, along with patients with macroscopic peritoneal dissemination. 53  In this therapeutic strategy, selecting patients with POCY1 at first SL and confirming a good response to NAC at repeat SL is very important.
In this sense, SL can provide oncological benefit. However, the prognosis for all patients with POCY1 who undergo NAC (including responders and non-responders) may still be poor. In JCOG0501, where eligibility criteria were large type 3 and type 4, including P0CY1 and localized P1, the survival efficacy of NAC was not justified compared with immediate surgery followed by postoperative chemotherapy. 54 In a large-scale retrospective cohort, there was no significant survival difference between NAC and postoperative chemotherapy for patients with P0CY1. 55 From those reports, the rationale that patients with P0CY1 should undergo NAC may not be justified at this time. If the efficacy of intraperitoneal chemotherapy for peritoneal disease, 56 including both macroscopic carcinomatosis and positive cytology is established, the importance of SL will be further advanced.

| CON CLUS ION
A current literature review suggests that staging laparoscopy is very important for determining the correct therapeutic strategy for the treatment of advanced gastric cancer. (i) Indication for SL is patients with some of the following: "whole stomach," "type 4 (linitis feature)," "large tumor," "equivocal CT findings," and "lymph node swelling". (ii) The exploration of the entire peritoneal cavity is pref-