The significance of countable and treatable metastatic liver disease in patients with gallbladder carcinoma after curative‐intent surgery: A 10‐year experience in China

Abstract Objective Our study was performed to evaluate the significance of countable and treatable metastatic liver disease (CTMLD) in patients with gallbladder carcinoma (GBC) after curative‐intent surgery. Methods Resected GBC patients between September 2010 and January 2021 were reviewed. Comparative analyses between patients with CTMLD and those without it were performed. A propensity score matching analysis was conducted for further validation. Results A total of 326 resected GBC patients were identified (33 with CTMLD). A significantly higher preoperative CA199 level was detected in those with CTMLD (p = 0.0160). Significantly higher incidences of major hepatectomy (p = 0.0010), lymph node metastasis (p < 0.0001), direct liver invasion (p < 0.0001), moderate to poor differentiation status (p < 0.0001), and T3–4 disease (p < 0.0001) were detected in patients with CTMLD. Even sharing comparable surgical margin status, patients with CTMLD still shared a significantly higher recurrence rate (93.9% vs. 57.3%, p < 0.0001) as well as a significantly higher recurrence rate within 6 months after surgery (63.6% vs. 14.7%, p < 0.0001). A significantly worse overall survival (median survival time: 12 vs. 33 months, p < 0.0001) or disease‐free survival (median recurrence‐free time: 6 vs. 30 months, p < 0.0001) was acquired in patients with CTMLD. After matching, a significantly higher early recurrence rate was still detected. Adjuvant chemotherapy seemed to have survival benefit for patients with CTMLD. Conclusion CTMLD was an indicator of advanced disease and served as an independent predictor of early recurrence among resected GBC patients. Whether curative‐intent surgery is superior to nonsurgical treatment among GBC patients with CTMLD remains to be explored in future prospective studies.


| INTRODUCTION
Gallbladder carcinoma (GBC) is a rare but fatal biliary epithelium-originated malignancy. 1 Due to lack of symptomatic clinical manifestations in the early phase of the disease, the majority of cases with GBC were diagnosed in an advanced stage and therefore lost the opportunity of receiving curative surgery. 1 Simple cholecystectomy is especially effective in achieving tumor clearance among T1a GBC patients with a reported 5-year survival rate reaching 100%. 2 For patients with T1b or more advanced disease, the radical cholecystectomy, including simple cholecystectomy, partial hepatectomy, and a regional or more extended lymphadenectomy, has always been regarded as the gold standard in the management of patients with GBC.A significantly prolonged postsurgical survival period has been observed among resected GBC patients than those who failed to receive the radical cholecystectomy. 1 Tumor metastasis has long been considered as the major obstacle in the successful management of various malignancies and liver is one of the most common metastatic site among the majority of digestive malignancies. 3Great progress has been achieved in the treatment of patients with liver metastases of colorectal cancer and satisfactory long-term survival has been reported after a combination of medical and surgical therapies, 4 with a reported 5-year survival rate reaching 58%. 5 Liver has also been introduced as a frequently detected metastatic site of GBC. 6,7According to the latest 8th American Joint Committee on Cancer (AJCC) staging criteria, the common metastatic sites of patients with GBC are peritoneum, liver, and occasionally to the lungs or pleura, which has been validated in previous researches that liver is the leading metastatic site and could be detected in over half of GBC patients. 6,7Liver metastasis has been introduced as an indicator of advanced stage and predicted poor prognosis among GBC patients. 6,7Over 90% GBC patients with liver dissemination were considered to be inoperable and were observed to succumb quickly due to advanced tumorrelated cachexia. 8However, clinically, GBC patients with liver metastasis can be furtherly divided into two subtypes, including operable and inoperable types.Inoperable type often indicated uncountable or diffusely disseminated metastatic tumor nodules within the liver, which could not be totally removed via invasive surgical procedures, such as partial hepatectomy or ultrasound-guided percutaneous or intraoperative radiofrequency ablation (RFA).As for operable type, as presented in Figure 1, the number of metastatic lesions within the liver were countable and often concentrated in one segment or several adjacent liver segments (Figure 1A,B), which could be totally resected via a combined partial hepatectomy.Moreover, for cases with metastatic nodules located in a disseminated manner (Figure 1C), tumor clearance could also be achieved via a combination of RFA or minor hepatectomy.The countable and treatable metastatic liver disease (CTMLD) can be frequently detected among GBC patients but has been rarely reported.Currently, it has not reached a consensus on the optimal therapeutic regime for GBC patients with CTMLD.Besides, the prognostic significance of CTMLD among resected GBC patients has also been rarely explored.Whether CTMLD in patients with GBC should be regarded as an indicator of distant metastasis and therefore precludes curative surgery or not as an indicator of distant metastasis and curative surgery is feasible remains unknown.Furthermore, the effect of postoperative adjuvant chemotherapy has also been rarely explored among resected GBC patients with CTMLD.According to the latest National Comprehensive Cancer Network (NCCN) guidelines (www.nccn.org/guidelines/ guide lines -detai l?categ ory=1&id=1517), an up to 6-month postoperative adjuvant chemotherapy or chemoradiotherapy was associated with survival benefit among resected GBC patients, especially in those with lymph node metastasis, which has also been validated in a recently reported phase 3 clinical trial that the efficiency of S-1 based chemotherapy was revealed in biliary malignancies. 9However, whether postoperative adjuvant chemo or chemoradiotherapy would bring survival benefit for GBC patients with CTMLD after curative-intent surgery remains unknown.
Therefore, based on our own single-center experience, we aimed to explore the significance of CTMLD in patients with GBC after curative surgery in terms of tumorrelated clinicopathologic features and long-term survival.A propensity score matching (PSM) was also performed for further validation.

| Patients
We retrospectively reviewed surgically treated GBC patients between September 2010 and January 2021 at West China Hospital, Sichuan University.Comparative analyses were performed to evaluate the prognostic significance of CTMLD in resected GBC patients and therefore the postoperative long-term survival is the primary endpoint of our study.In order to control various confounding factors which might lead to unpredictable survival influence to our results, patients who died within 6 months after surgery were excluded in order to guarantee the quality of our cohort.This study was approved by institutional ethics review board of West China Hospital (approval number: ChiCTR2300071754).

| The definition of CTMLD
CTMLD was defined as follows: 1. Regardless of direct liver invasion, an isolated metastatic liver nodule was detected radiologically in any lobe or segment within the whole liver rather than the primary site, which could be removed combined with minor hepatectomy or RFA.
2. Regardless of direct liver invasion, numerous countable metastatic nodules were detected within one segment of the liver or within two or three adjacent segments, which can be totally removed combined with partial hepatectomy or RFA. 3. Regardless of direct liver invasion, numerous countable metastatic nodules were detected in nonadjacent liver segments and these lesions can be furtherly removed combined with partial hepatectomy and RFA.The patient finally received a gallbladder excision, partial hepatectomy (SIVB + V), a regional lymphadenectomy, and a combined radiofrequency ablation (S7 metastatic lesion).Nearly 2 months after surgery, recurrent disease were detected within liver (G) and (H) (white arrow).CTMLD, countable and treatable metastatic liver disease; GBC, gallbladder carcinoma.
well as to exclude patients who were contraindicated to curative-intent surgery.Surgery candidates should fulfill the following criteria.First, patients with serious comorbidities, such as stroke or sever coagulation abnormalities, were not allowed.Second, an estimated functional liver remnant volume >40% after partial hepatectomy was always required.Third, cases with radiologically suspected distant metastasis (abdominal dissemination, para-aortic lymph node enhancement, or lung metastasis) were not allowed.All surgical candidates would receive a multidisciplinary discussion to find the optimal therapeutic regime.For patients with CTMLD, their families were informed of their disease status that liver metastasis has already occurred and the effect of curative-intent surgery might be less satisfactory than those without CTMLD.If they still insisted on performing curative-intent surgery, they were required to sign the surgery-related informed consent and therefore surgery would be performed at the premise of their strong desire and written request.The specific surgical procedures were in consistent with our recently published series 10 and partial hepatectomy combined with or without RFA would be performed in order to achieve tumor clearance.All patients were staged based on the 8th AJCC staging system.

| Follow-up
Postoperative patients reexamined their disease status at the outpatient of our hospital every 1-2 months in the first 2 years after surgery and every 6 months thereafter.Their survival status was updated once a year via telephone number recorded in their electronic database.The most recent update was performed on September 1, 2022.

| Statistical analysis
IBM SPSS version 25.0 and Graph-Pad Prism 7 were used for statistical analysis.Categorical data were presented as numbers (percentages).Categorical variables were compared through chi-squared or Fisher's exact tests.overall survival (OS) indicated the survival period from the date of surgery to the date of death or last follow-up.Diseasefree survival (DFS) indicated the survival time from the date of surgery to the date of radiologically confirmed disease recurrence.Kaplan-Meier curves were applied in depicting survival outcomes and log-rank test was used in evaluating survival differences.Univariate and multivariate Cox regression model were used to evaluate potential prognostic factors and independent prognostic factors for OS and DFS of the entire cohort.Results were recorded as hazard ratio (HR) with its 95% confidence interval (CI).
A p value < 0.05 indicated statistical difference.Binary logistic regression was furtherly applied to identify the potential predictors of early recurrence.A PSM was also performed to control bias and for further validation (ratio 1:1, standard deviation 0.1).

| Baseline characteristics of the entire cohort
Initially, a total of 329 resected GBC patients with adequate follow-up and clinical data were identified.In order to guarantee the quality of our cohort as previously descripted, three cases who died within 6 months after surgery were excluded.Two cases died perioperatively due to the severe abdominal hemorrhage derived from the spontaneous rupture of pseudoaneurysm.Another patient died 3 months after surgery due to unpredicted and explosive liver failure after major hepatectomy.Finally, 326 resected GBC patients were included in our study, 33 (10.1%) of whom were diagnosed with CTMLD.The specific process of patient selection and identification was presented in Figure 2. The overall sex ratio was nearly 2:1 (female: male, 221: 105).The majority of cases were diagnosed order than 60 (171, 52.5%).Forty patients (12.3%) were presented with obstructive jaundice preoperatively and six (1.8%) of them received preoperative biliary drainage (PTCD or ENBD).Sixty-four (19.6%) patients received combined extra-hepatic bile duct resection, 35 patients (10.7%) received major hepatectomy, 39 (12.0%) patients received combined multi-visceral resection, and 14 (4.3%) patients received combined major vascular resection and reconstruction.The overall incidence of lymph node metastasis was 38% (124/326) and the overall incidence of direct liver invasion was 32.6% (142/326).The overall R0 rate was 91.4% (298/326).Eighty-six patients have received at least four cycles of postoperative intravenous-mediated adjuvant systematic chemotherapy.Two chemotherapy Flowchart of the specific process of patient selection and identification.GBC, gallbladder carcinoma.
agents, Gemcitabine plus Cisplatin, served as the corner stone of the postoperative chemotherapy.The median survival time of the entire cohort was 28.5 months, ranging from 6 to 124 months.

| Comparative analysis between patients with CTMLD and those without CTMLD before PSM
As is summarized in  3A) as well as DFS than those without CTMLD after surgery (median recurrence-free time: 6 vs. 30 months, p < 0.0001; Figure 3B).

| Propensity score matching analysis
Owing to the retrospective nature of our cohort, many uncontrolled factors hindered us from further exploration on the significance of CTMLD in resected patients with GBC.Age, sex, and numerous independent prognostic factors (lymph node metastasis, T stage, CA199, postoperative chemotherapy, surgical margin) were controlled between CTMLD (+) and CTMLD (−) groups.After PSM, as is summarized in Table 3, a total of 66 patients were identified (ratio 1:1).Patients of two groups were comparable in all these matched factors.However, the incidence of moderate to poor differentiation status (100.0%vs. 84.8%,p = 0.0270) were still significantly higher in patients with CTMLD.Although comparable OS (p = 0.4146) (Figure 3C) and DFS (p = 0.3107) (Figure 3D) were acquired between two groups, a significantly higher recurrence rate within 6 months after surgery was still detected in patients with CTMLD (63.6% vs. 27.3%,p = 0.0030).

| The effect of postsurgical adjuvant chemotherapy in patients with CTMLD
We subsequently explored the efficiency of postoperative adjuvant systematic chemotherapy in patients with CTMLD.Adjuvant chemotherapy was associate with a more favorable OS (median survival time: 31.5 vs. 12 months, p = 0.0349) among resected GBC patients with CTMLD (Figure 4A).Moreover, a more favorable DFS with a borderline p value (median survival time: 20 vs. 6 months, p = 0.0616) was also observed in patients with CTMLD who received adjuvant chemotherapy (Figure 4B).

| DISCUSSION
To our knowledge, few researchers insisted on performing curative-intent surgery among GBC patients with liver metastasis.Systematic chemotherapy, radiotherapy, RFA, targeted therapy, or other less invasive therapeutic regimes are clinically preferred in GBC patients with metastatic liver disease. 11However, we cannot deny the fact that for some experienced and aggressive centers, the curative surgery has been attempted to be applied in a small portion of patients with liver metastasis, especially among those with CTMLD.3][14][15] Our final results indicated that although GBC patients with CTMLD seemed to be operable, significantly worse OS and DFS were detected even patients were actively treated.Moreover, patients with CTMLD still shared a much higher early recurrence rate even after various independent prognostic factors were matched between two groups.CTMLD was an independent predictor of early recurrence.However, postoperative adjuvant chemotherapy seemed to have significantly prolonged the survival period of patients with CTMLD after curative surgery, which might provide clinical reference on the indication of postoperative adjuvant chemotherapy in GBC patients to some extent.
Liver metastasis has been a long-standing but debating issue which greatly hindered the successful management of various cancers, particularly for cancers with great potential to develop metastatic liver disease, such as pancreatic cancer. 16However, with the evolution of hepatectomyrelated surgical techniques, ultrasound-guided RFA, as well as progress in perioperative management, liver has never been considered as the restricted area in the surgical management of various primary or metastatic liver malignancies.RFA has been widely used as a minimally invasive surgical procedure for patients with primary or secondary liver malignancies.Ultrasound-guided RFA combined with experienced manipulators could achieve similar oncological outcomes and long-term survival as conventional partial hepatectomy in cases with small liver tumors, avoiding unnecessary surgery-related trauma to human body, which has been validated by Zheng et al. 17 Moreover, due to the survival benefit brought by adjuvant systematic chemotherapy, the prognosis of patients with colorectal metastatic liver disease has greatly improved, with a reported 5-year survival rate reaching 60%. 5 However, as for GBC patients with liver metastasis, their prospect still seems to be vague.
Relevant studies regarding GBC patients with liver metastasis are quite rare either clinically or basically, which might be partially due to the rarity of the disease.According to the latest 8th AJCC staging criteria, liver metastasis was just mentioned as an indicator of distant metastasis but was ill-defined.In the latest NCCN guidelines, it has also not reached a consensus regarding the optimal therapeutic regime for GBC patients with metastatic liver disease, especially among those with CTMLD.Whether GBC patients with CTMLD should be treated aggressively via surgical procedures or treated conservatively via nonsurgical procedures remains unexplored.In order to supply the treatment reference, we have attempted to compare the overall prognosis between patients with CTMLD who received surgery and those who received nonsurgical treatment.However, having rescreened our cancer database, almost all patients with CTMLD have received curativeintent surgery and the intended comparison was not feasible.Therefore, the absence of comparison between patients with CTMLD who received curative-intent surgery and patients with CTMLD who received nonsurgical treatment might be a major limitation within our paper, which was expected to be explored in future prospective studies.
Clinically, in the most Asian centers, especially Japanese centers, surgeons have always been aggressive in the management of biliary malignancies. 18,19][21] For example, one previous meta-analysis has systematically evaluated the significance of hepatopancreatoduodenectomy among GBC patients with advanced T4 disease and revealed that although high rates of morbidities and mortalities were detected, long-term survival could also be achieved among patients for whom curative-intent surgery were feasible. 22Moreover, from our own center's experience, surgeons' decisions on the optimal therapeutic regime of GBC with liver metastasis were mostly driven by the operability of the metastatic tumor.In other words, more specifically, whether the surgery would be performed or not was mainly decided by the clinicians' subjective judgment.If there are too many metastatic nodules within the liver or there are extensive abdominal dissemination, tumor clearance is often unfeasible via invasive surgical procedures.In this case, curative surgery would not be performed and systematic chemotherapy, chemoradiotherapy or targeted therapy would be preferred. 23,24owever, at least in our center, clinicians were more likely to perform invasive surgical procedures for GBC patients with CTMLD.Unfortunately, although these patients with CTMLD were actively treated and tumor clearance seemed to have been achieved via invasive surgery, patients with CTMLD still shared a worse prognosis even when tumor stage was matched.Moreover, even various vital prognostic factors have been matched, patients with CTMLD still shared more aggressive tumor biological features, including more awful tumor differentiation status, a higher frequency of liver invasion, and a significantly higher early recurrence rate.Therefore, CTMLD was a signal of advanced disease and patients with CTMLD were often presented with extensive metastatic dissemination, that is, the disease has evolved into a systematic disease rather than an isolated disease.This might partially explain the worse survival even being actively treated as well as the high recurrence rate after matching.Additionally, the treatment received after recurrence been detected might also influence the overall prognosis to some extent.The overall recurrence rate was 61.0% in our cohort and the majority of cases were presented with obstructive jaundice upon recurrence, combined with multifocal spread identified in the contrast-enhanced CT scan.In their electronic medical record, most cases with recurrence only received palliative treatment, such as PTCD for patients with biliary obstruction and jejunal nutrition tube for patients with ileus.A few patients received palliative biliaryenteric anastomosis or gastrointestinal anastomosis.Only a limited number of patients with favorable health conditions received systematic chemotherapy or targeted therapy.However, only 3-4 months extension of survival period was detected and the majority of cases with recurrence died within 6 months.
Another major finding of our study is the survival benefit brought by postoperative adjuvant chemotherapy in GBC patients with CTMLD.Accumulating evidence has demonstrated the survival benefit brought by postoperative adjuvant chemotherapy among resected biliary cancers. 9,25,26In consistent with previous findings, our data also indicated that postoperative adjuvant chemotherapy  was a favorable prognostic factor with an obviously improved postoperative survival versus those who did not receive systematic adjuvant chemotherapy.Adjuvant systematic chemotherapy served as an independent prognostic factor for both OS and DFS.Furthermore, among patients with CTMLD, a significantly prolonged survival has also been detected after the application of adjuvant chemotherapy.Currently, as previously described, the application of adjuvant postoperative chemotherapy was especially recommended in GBC patients with lymph node metastasis.CTMLD might also be an indicator of adjuvant chemotherapy if an extended survival was observed among patients with CTMLD who received curative surgery than those who received nonsurgical treatment.

| CONCLUSION
Current study first explored the significance of CTMLD among GBC patients after curative-intent surgery.Even tumor stage was matched between two groups, patients with CTMLD still shared more aggressive tumor biological features and developed recurrences much earlier than those without CTMLD after curative-intent surgery.CTMLD could be regarded as an indicator of advanced disease as well as an independent predictor of early recurrence for patients with GBC after curative-intent surgery.Future prospective well-designed studies are demanded for further exploration on the efficiency of curative-intent surgery versus nonsurgical treatment among GBC patients with CTMLD.

F I G U R E 1
Representative figures of GBC patients with CTMLD.(A) primary gallbladder lesion (blue arrow); (B) regional metastatic liver lesions (blue arrow); (C) distant isolated metastatic liver lesion; (D) and (E) metastatic liver lesions detected by intraoperative ultrasound; (F) intraoperative radiofrequency ablation.

F I G U R E 3
KM curves presenting survival outcomes before and after PSM.(A) OS before PSM; (B) DFS before PSM; (C) OS after PSM; (D) DFS after PSM.CTMLD, countable and treatable metastatic liver disease; DFS, diseasefree survival; OS, overall survival; PSM, propensity score matching.

T A B L E 4
Logistic regression on the potential predictors of early recurrence for patients with GBC.F I G U R E 4 KM curves presenting survival outcomes between patients with CTMLD with adjuvant chemotherapy and those without adjuvant chemotherapy.(A) OS; (B) DFS.CTMLD, countable and treatable metastatic liver disease; DFS, disease-free survival; OS, overall survival.

Table 2
Baseline characteristics of the entire cohort before PSM.
T A B L E 2 T A B L E 3