Background
Colorectal cancer (CRC) is the third most common cause of cancer-related mortality and the second most common malignancy in Japan [
1]. Among CRC patients, approximately 25 % present with distant metastases at initial diagnosis and almost 50 % will develop metastases during the disease course [
2]. Metastasis is prevalent in the liver. It is present in nearly 80 % of stage IV patients and is the sole site of disease in approximately 40 % of these patients [
3]. Complete resection of colorectal liver metastasis (CLM) is considered to be the only curative method of treatment. However, only 20 % of CLM patients are able to undergo resection of metastases; the remaining patients undergo systemic chemotherapy, radiotherapy, or hepatic arterial infusion [
4].
Chemotherapy for CRC has improved substantially over the last decade with the development of new cytotoxic drugs and molecular targeted agents, including anti-epidermal growth factor receptor (EGFR) antibody and anti-vascular endothelial growth factor (VEGF) antibody. As a result, systemic chemotherapy is now effective and subsequent surgical resection possible in increasing numbers for patients with initially unresectable CLM [
5,
6]. The term “conversion therapy” has been proposed to refer to the use of induction chemotherapy in patients with initially unresectable CLM. Adam et al. reported that among patients with initially unresectable CLM who underwent induction chemotherapy, 12.5 % were subsequently able to undergo liver resection, and of these patients, 16 % could achieve a cure [
4,
7]. However, no definitive survival advantage has yet been demonstrated for surgical resection after conversion therapy for initially unresectable CLM. Among the reasons for this are that the definitions of unresectable CLM used vary among institutions and surgeons. Additionally, physicians in different specialties have different treatment preferences for CLM: medical oncologists may favor systemic chemotherapy, whereas surgeons may prefer resection of CLM. Even in a single institution, the indications for surgery might not be consistent. It might also be unrealistic to conduct a clinical trial comparing the outcomes of patients undergoing surgical resection with those of patients who continue with systemic chemotherapy after unresectable CLM becomes resectable, because surgical resection is strongly recommended if CLM is resectable. Therefore, in an effort to clarify the clinical benefit of surgical resection after downsizing systemic chemotherapy for patients with unresectable CLM, we retrospectively assessed the survival and clinical characteristics of surgical patients with CLM and compared them between patients with initially resectable CLM and patients with initially unresectable CLM.
Methods
Study design
All patients who underwent liver resection for CLM between January 2001 and December 2013 in the Department of Surgery of Shiga Medical Center for Adults were identified from a prospective surgery database. All identified inpatient and outpatient medical records were then retrospectively reviewed. Colon cancers were diagnosed by colonoscopy with pathologic confirmation. All patients underwent abdominopelvic and thoracic contrast-enhanced computed tomography (CT) scan or contrast-enhanced abdominal magnetic resonance imaging (MRI) and optional positron emission tomography (PET) preoperatively or prior to induction chemotherapy. After liver resection, all patients received at least 3 months of follow-up. CT scans and carcinoembryonic antigen (CEA) and cancer antigen 19-9 (CA19-9) serum levels were assessed during follow-up. Survival was calculated from the date of liver surgery to the last date of follow-up. The Ethics Committee of our institution approved this study.
Definitions
Contraindications to liver resection for CLM at our institute are (1) remnant liver volume <30 % or estimated liver remnant plasma clearance rate of indocyanine green (ICG-K) <0.5 and (2) the need for reconstruction of a major vessel that needs to be preserved for proper liver function but is invaded by the tumor. Synchronous metastases were defined as CLM diagnosed before colorectal resection or at the time of surgery. In accordance with the guidelines of the International Union Against Cancer, R0 resection was defined by the absence of microscopic tumor invasion of the resection margins (tumor-free margin ≥1 mm for all detected lesions) and R1 resection as the presence of such invasion of the resection margins (tumor-free margin 0 mm) [
8]. R2 resection was defined by the presence of macroscopic tumor invasion of the resection margins.
Tumor response was assessed at least every 6 cycles (12 weeks) using the same methods as those performed at baseline (CT, MRI, or PET), according to Response Evaluation Criteria In Solid Tumors (RECIST) criteria [
9]. Tumors were assessed for resectability by a multidisciplinary team. Resection was performed within 8 weeks of the last treatment cycle.
Statistical analysis
Qualitative data were reported as the number of patients and were compared with either the Pearson χ2 test or Fisher’s exact test, as deemed appropriate. For qualitative data, the Wilcoxon rank-sum test was used when comparing two groups and the Kruskal–Wallis test when comparing three groups. Survival curves were plotted using the Kaplan–Meier method and compared with the log-rank test. Survival was measured from the date of diagnosis of the liver metastasis until death or last follow-up. Parameters that had an impact on survival were identified by univariate analysis, and those with p values less than 0.15 were entered into a Cox proportional hazards model to identify the independent prognostic factors for survival. Tests were always two-sided and the level of statistical significance was set at p < 0.05. All analysis was performed using JMP 10 software (SAS Institute Inc., Cary, NC).
Discussion
This comparative study of the outcomes of liver resection for initially resectable and unresectable CLM found that OS was comparable between patients who underwent liver resection for initially resectable CLM and those who underwent liver resection after downsizing systemic chemotherapy for initially unresectable CLM. The converted patients tended to have recurrence shortly after liver resection; however, their survival might be improved through repeat hepatectomy or resection of metastases at other sites. These results support the clinical efficacy of liver resection after downsizing chemotherapy for patients with unresectable CLM.
The criteria for resectability not only differed between previous studies but also were often poorly defined [
10]. Although general principles were applied to the criteria for resectability in our institution, the decision regarding resectability could be made based on personal judgment in the clinical setting. Not surprisingly, judgment of conversion from unresectable to resectable status is somewhat subjective. Therefore, since palliative hepatectomy without curative intent might have been provided for some patients with aggressive symptomatic liver metastases in addition to uncontrolled macroscopic extrahepatic disease, patients with macroscopic extrahepatic disease at the time of liver resection and those with R2 liver resection were excluded to enable comparison of patients who underwent treatment with curative intent alone.
Although no prospective randomized controlled trials have demonstrated the efficacy of conversion therapy, some surgical series have reported favorable results of liver resection after downsizing chemotherapy for initially unresectable CLM. A systematic review reported a median OS of 45 months (range, 36–60 months) and DFS of 14 months, although the majority of the reviewed studies were published a decade ago [
10]. Recent studies have reported a median OS of 40.5–57.6 months and DFS of 3.2–10.6 months, indicating significantly shorter median DFS in the conversion group than in the initially resectable group but a comparable median OS [
11,
12]. The present study reaffirms the finding of prolonged survival of patients in the conversion group despite their relatively shorter DFS, which we attribute to repeat hepatectomy or resection of metastases at other sites.
Multivariate analysis identified two significant factors of OS (age >65 years and synchronous disease). A number of previous reports have shown that several prognostic factors affect patient outcome, and similar factors were noted in this study [
13‐
18]. The presence of multiple hepatic metastases has been reported to be a very important risk factor for recurrence after liver resection, and the results in this study showed a similar trend but did not reach statistical significance [
19,
20]. In the present study, R1 resection was not a significant risk factor that affected both OS and DFS. Although many publications have reported microscopic involvement of surgical resection margins as a significant poor prognostic factor, whether R1 resection is contraindicated remains a controversial issue [
13,
20‐
25]. Some authors have justified intended R1 resection combined with increasingly efficient chemotherapy regimens [
21,
26,
27]. Our study also suggests that R1 resection is not a contraindication for resection and that resection of all liver metastasis in converted patients may be effective even if the resection margin becomes microscopically positive.
In previous studies, the addition of molecular targeted drugs, anti-epidermal growth factor receptor (EGFR) agents such as cetuximab or panitumumab, or the anti-vascular endothelial growth factor (VEGF) agent bevacizumab to doublet chemotherapy such as FOLFOX or FOLFIRI achieved a higher response than doublet chemotherapy alone, resulting in a higher R0 resection rate for patients with initially unresectable CLM [
5,
6]. In our institution, KRAS wild-type patients who had unresectable CLM alone or unresectable CLM with controllable extrahepatic metastases were to receive FOLFOX6 plus anti-EGFR agents in principle in the second half of the study period. All of the converted patients received doublet chemotherapy in combination with molecular-targeted drugs as first-line chemotherapy. Of these 16 patients, 11 had KRAS wild-type tumors and received FOLFOX6 or FOLFIRI plus anti-EGFR agents and 5 had KRAS mutant-type tumors and received FOLFOX6 plus bevacizumab. Although limited findings are available so far on the potency of these new regimes as conversion therapy, modern systemic chemotherapy has definitely increased conversion rates. Future studies should address which regimens are optimal for achieving conversion of patients with initially unresectable CLM.
This study is not without limitations. First, the conversion group was small. To determine the clinical efficacy of liver resection after downsizing chemotherapy, we must accumulate data from more cases to achieve adequate statistical power. Second, this study was not a comparison of outcomes between surgical patients and patients who continued chemotherapy after unresectable CLM was rendered resectable. In assessing the value of resection after “conversion therapy,” appropriate control patients might be converted patients who continued chemotherapy without liver resection. However, it is unlikely that a prospective randomized trial comparing these outcomes will be conducted since surgical resection is strongly recommended if CLM is deemed resectable. Third, this was a retrospective study and selection bias is possible. The patients in the conversion group were younger and might have been selected because of their good general condition aside from their cancer. Fourth, the chemotherapy regimen used in the study period has changed and the impact of modern chemotherapy on survival could not be determined. Finally, we did not collect data from patients who had unresectable CLM with or without extrahepatic metastases, so the true number of patients with unresectable disease and the conversion rate are unclear.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JK participated in the study design and coordination, carried out the data gathering and analysis, and drafted the manuscript. TY participated in the study design and coordination, carried out the data gathering, and helped to draft the manuscript. KS, YK, RO, MT, JK, MY, HH, and HY carried out the data gathering and helped to draft the manuscript. MZ participated in the study design and coordination and revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.