Repeat hepatectomy for recurrent colorectal liver metastases: A comparative analysis of short- and long-term results

Pulo Figueiredo Cost , Friio Ferreir Coelho , , Vgner Birk Jeismnn , Jime Arthur Pirol Kruger , Gilton Mrques Fonse , Ivn Ceonello , Pulo Hermn

a Postgraduate Course in Digestive Surgery Brazilian College of Digestive Surgery (Colégio Brasileiro de Cirurgia Digestiva - CBCD), S?o Paulo, Brazil

b Digestive Surgery Division, Department of Gastroenterology, University of S?o Paulo School of Medicine, S?o Paulo, Brazil

c Instituto do Cancer do Estado de S?o Paulo (ICESP), S?o Paulo, Brazil

Keywords: Colorectal cancer Metastasis Hepatectomy Postoperative complications Survival analysis

ABSTRACT Background: Liver recurrence after resection of colorectal liver metastases (CRLM) is frequent. Repeat hepatectomy has been shown to have satisfactory perioperative results. However, the long-term outcomes and the benefits for patients with early recurrence have not been clarified. The aim of this study was to compare the short- and long-term outcomes of patients undergoing single hepatectomy and repeat hepatectomy for CRLM. Additionally, the oncological outcomes of patients with early ( ≤6 months) and late recurrence who underwent repeat hepatectomy were compared. Methods: Consecutive adult patients undergoing hepatectomy for CRLM between June 20 0 0 and February 2020 were included and divided into two groups: single hepatectomy and repeat hepatectomy. Results: A total of 709 patients were included: 649 in the single hepatectomy group and 60 in the repeat hepatectomy group. Patients in the repeat hepatectomy group underwent more cycles of preoperative chemotherapy [4 (3-6) vs. 3 (2-4), P = 0.003]. Patients in the single hepatectomy group more frequently underwent major hepatectomies (34.5% vs. 16.7%, P = 0.004) and had a greater number of lesions re- sected (2.9 ± 3.6 vs. 1.9 ± 1.8, P = 0.011). There was no increase in operative time, estimated blood loss, length of hospital stay, complications, or mortality in the repeat hepatectomy group. There were no dif- ferences in overall survival ( P = 0.626) and disease-free survival ( P = 0.579) between the two groups. Similarly, for patients underwent repeat hepatectomy, no difference was observed between the early and late recurrence groups in terms of overall survival ( P = 0.771) or disease-free survival ( P = 0.350). Conclusions: Repeat hepatectomy is feasible and safe, with similar short- and long-term outcomes when compared to single hepatectomy. Surgical treatment of early liver recurrence offers similar oncological outcomes to those obtained for late recurrence.

Introduction

Colorectal cancer is the third most common cancer worldwide, with approximately 1.85 million new cases per year. It is estimated that more than 0.88 million disease-related deaths occur annually, making colorectal cancer the second leading cause of cancer mor- tality worldwide [1] . Approximately 50% of patients with colorec- tal cancer will develop metastases, and liver is the most common site (up to 80% of patients) [2] . After diagnosis of the primary tu- mor, the incidence of colorectal liver metastases (CRLM) is approx- imately 4.3% in the first 1 year, 8.7% in the second year, 12.7% in the third year, and 16.5% in the fifth year [3] . Metastatic disease has a major prognostic impact on patients with colorectal cancer, accounting for two-thirds of the associated deaths [4] .

Currently, the mainstay of CRLM treatment is a combination of chemotherapy regimens and complete resection of the liver lesions with clear surgical margins (R0 resection), which provides a 5-year overall survival (OS) ranging from 40% to 60% [5-8] . However, re- lapse after hepatectomy is frequent, occurring in more than 50% of patients [ 9 , 10 ]. Therefore, most patients with recurrent CRLM will require new treatment strategies after undergoing initial hepate- ctomy, such as systemic chemotherapy, ablative therapy, and/or a repeat resection of liver lesions. The first report in the early 1990s, demonstrated the potential benefits of repeat hepatectomy in se- lected cases of recurrent CRLM [11] . With the development of sur- gical techniques, the use of parenchyma-sparing strategies, and the improvement in perioperative care, repeat hepatectomy has been systematically offered to patients with recurrent CRLM in special- ized centers [12] .

Case series and retrospective cohort studies have shown that repeat hepatectomy leads to satisfactory perioperative outcomes with low morbidity and mortality [ 12 , 13 ]. In addition, some au- thors have shown that reoperation provides long-term survival in patients with recurrent CRLM [ 14 , 15 ]. However, few studies have compared the outcomes of patients undergoing repeat hepatec- tomy with those who underwent single hepatectomy; therefore, the oncological benefits have not been clarified. Another contro- versial issue is the selection of the best candidates for repeat hep- atectomy, especially whether patients with early recurrence would benefit from the procedure.

The aim of this study was to compare the short- and long-term outcomes of patients undergoing repeat hepatectomy with those of contemporary patients who underwent single hepatectomy for CRLM. In addition, the long-term survival of patients with early and late recurrence who underwent repeat hepatectomy was com- pared.

Methods

This study was approved by the Institutional Ethics Committee (19476) and was conducted following the Strengthening the Re- porting of Observational Studies in Epidemiology (STROBE) recom- mendations [16] .

From a prospective database, consecutive adult patients under- going hepatectomy for CRLM between June 20 0 0 and February 2020 were evaluated. The exclusion criteria were as follows: pa- tients who underwent a two-stage hepatectomy or associating liver partition and portal vein ligation for staged hepatectomy (ALPPS), patients who underwent palliative resection (R2 resection), and pa- tients with incomplete data. The indication of the surgical proce- dure was carried out after discussion in a multidisciplinary meet- ing. Briefly, single and repeat hepatectomies were indicated if a complete resection could be achieved on all CRLM with an ade- quate future liver remnant ( ＞ 25% in healthy livers and ＞ 30% after long-term exposure to chemotherapy). Patients with initially un- resectable CRLM or with worrisome prognostic features [i.e., syn- chronous metastases, multiple nodules, and elevated serum car- cinoembryonic antigen (CEA) levels] were initially managed with systemic treatment. As a rule, in our institution we use the re- sponse to chemotherapy as a guide to indicate surgery in patients with liver recurrence after single hepatectomy. Limited extrahep- atic disease, whenever resectable, was not considered a contraindi- cation for hepatectomy.

After inclusion, patients were divided into two groups: single hepatectomy group and repeat hepatectomy group. Recurrence oc- curring within 6 months of the procedure was defined as early re- currence [ 10 , 17 ].

Hepatectomies were defined according to Brisbane terminol- ogy [18] . Major hepatectomy was defined as resection of at least three contiguous liver segments. CRLM were defined as syn- chronous when diagnosed at the same time as the primary tumor, early metachronous when diagnosed within 12 months of colorec- tal cancer, and as late metachronous when diagnosed after this period, as proposed by Adam et al. [19] . For all hepatectomies, including repeat hepatectomies, we used the initial diagnosis of liver metastases (before single hepatectomy) to classify patients as having synchronous or metachronous disease. In the cases of syn- chronous metastases, we also evaluated the timing of liver/primary tumor resection (primary tumor resection first, simultaneous re- section, or liver first approach).

The following preoperative parameters were analyzed: age, sex, BMI, preoperative laboratory tests, American Society of Anesthesi- ologists (ASA) physical status score, size and location of lesions, in- terval between primary tumor treatment and initial hepatectomy, primary tumor staging, percentage of patients who underwent pre- operative chemotherapy, and the number of cycles of chemother- apy. Preoperative chemotherapy was defined as any chemotherapy treatment before hepatectomy in the presence of liver metastases. Extrahepatic disease was defined as any colorectal metastases out- side the liver, excluding cases of contiguous invasion of adjacent organs/structures to the liver. For the intra- and postoperative pe- riods, the following data were retrieved: type of hepatectomy, as- sociated ablation of liver metastases (at the same time of hepate- ctomy), en-bloc resection of adjacent organs, operative time, esti- mated blood loss, transfusion rate, length of stay in ICU, length of hospital stay, postoperative complications, and perioperative mor- tality. The specimens obtained were assessed for free margin fre- quency and smaller distances in millimeters (mm).

Postoperative morbidity was defined as any complication occur- ring in the first 90 postoperative days and was stratified according to the Clavien-Dindo classification [20] . Perioperative mortality was defined as death within the first 90 days of hepatectomy.

Patients were followed up according to the institutional proto- col with cross-sectional imaging and CEA levels obtained every 4-6 months for the first 2 years and screening and consultations every 6 months thereafter for up to five years [7] . For both groups, OS was defined as the time interval between the initial hepatectomy and the date of death or the most recent follow-up, and disease- free survival (DFS) was defined as the time interval between the initial hepatectomy and diagnosis of relapse or the most recent date of follow-up.

Continuous data were expressed as median and interquartile range (IQR) or mean and standard deviation (SD); categorical vari- ables were expressed as percentage. Quantitative data were com- pared using the Mann-Whitney U test. For the categorical variables the Fisher’s exact test or the Chi-square test was used. Survival curves were calculated using the Kaplan-Meier method and com- pared using the log-rank test. A P value ＜ 0.05 was defined as sta- tistically significant.

Results

During the study period, from 1465 patients submitted to hep- atectomy at our institution, 709 patients who underwent hepa- tectomy for CRLM were included: 649 in the single hepatectomy group and 60 in the repeat hepatectomy group.

The clinical and surgical characteristics of the groups are shown in Table 1 . The groups were homogeneous in terms of age, sex, BMI, ASA classification, primary tumor staging, and preopera- tive laboratory test results (including CEA level). There was a higher frequency of synchronous and early metachronous metas- tases in the single hepatectomy group and a predominance of late metachronous metastases in the repeat hepatectomy group. In addition, patients who underwent repeat hepatectomy also un- derwent more cycles of preoperative chemotherapy [4 (3-6) vs. 3 (2-4), P = 0.003]. Patients in the single hepatectomy group were more frequently subjected to major hepatectomies (34.5% vs. 16.7%, P = 0.004), especially right hepatectomies (22.0% vs. 6.7%, P = 0.004), with a greater number of resected lesions per patient (2.9 ± 3.6 vs. 1.9 ± 1.8, P = 0.011).

The perioperative results are shown in Table 2 . There was no in- crease in the operative time, estimated blood loss, length of ICU or hospital stay, and complications or perioperative mortality among patients who underwent repeat hepatectomy compared with pa- tients who underwent the single hepatectomy for CRLM.

Table 1 Patients’ demographic and surgical data.

Table 2 Perioperative results.

The OS of patients in the single hepatectomy group was 88.4% at 1 year, 64.0% at 3 years, 49.0% at 5 years, and 31.5% at 10 years, which was similar to that of patients in the repeat hepatectomy group (95.0% at 1 year, 68.0% at 3 years, 39.0% at 5 years, and 24.5% at 10 years;P= 0.626; Fig. 1 A). Similarly, no difference in DFS was observed between the two groups (53.3% vs. 50.8% at 1 year, 33.9% vs. 30.5% at 3 years, 30.4% vs. 21.8% at 5 years, and 21.8% vs. 21.8% at 10 years;P= 0.579; Fig. 1 B).

No differences were observed between the patients who un- derwent repeat hepatectomy for early recurrence ( ≤6 months,n= 14) versus late recurrence (＞6 months,n= 46) in terms of OS (P= 0.771, Fig. 2 A) and DFS (P= 0.350, Fig. 2 B).

Fig. 1. Overall survival ( A ) and disease-free survival ( B ) of patients who underwent single hepatectomy and repeat hepatectomy.

Fig. 2. Overall survival ( A ) and disease-free survival ( B ) of patients with early and late recurrence who underwent repeat hepatectomy.

Discussion

Currently, the treatment for CRLM is multidisciplinary, in- cluding the use of modern combination chemotherapy regimens (based on oxaliplatin and irinotecan, associated or not to mon- oclonal antibodies) and the radical resection and/or ablation of liver lesions [ 6 , 21 , 22 ]. Recent studies have shown that the sur- vival rates after treatment for CRLM have increased, reaching 40%-60% at 5 years [ 6 , 7 ]. However, despite the use of the best available treatment, the recurrence rate can reach 50%-78% at 5 years [ 9 , 10 ], and in approximately 30% of cases, the recurrence is limited to the liver [14] . Therapeutic options for the treat- ment of recurrent CRLM are chemotherapy and/or locoregional treatment, with ablative methods and surgical resection being the most common. A recent study by Dupré et al. [23] com- pared radiofrequency to reoperation for treating recurrence of liver metastases and showed that despite a higher rate of periopera- tive complications, resection offered better progression-free sur- vival.

In recent years, reoperation for CRLM in specialized centers has been shown to be feasible and safe [ 12 , 24 ]. However, repeat hep- atectomy is possible in only a small proportion of patients. In a recent study, Park et al. [25] reported that repeat hepatectomies were performed on only 31% of patients who developed CRLM re- currence. The main factors implicated in the non-indication of re- operation were extensive liver relapse, extrahepatic disease, and insufficient patient’s clinical performance [25] . In our study, 207 patients (29.2%) had liver recurrence, of which 60 (29.0%) under- went repeat hepatectomy.

Our results showed that repeat hepatectomy was indicated more frequently for patients with late metachronous disease, who had undergone more cycles of chemotherapy and those with a limited number of liver nodules. These data show the profile of patients that can benefit from a repeat hepatectomy. Kulaylat et al. [26] studied the factors influencing the indication for repeat hepatectomy and observed that response to chemotherapy [odds ratio (OR) = 15.5; 95% CI: 1.5-157.6;P= 0.02] and a single metas- tasis (OR = 13.5; 95% CI: 2.98-61.23;P= 0.001) were indepen- dent factors for a second resection. Despite a lack of statistical sig- nificance in the multivariate analysis, a disease-free interval＞12 months was a potential positive factor associated with second re- section (OR = 2.5; 95% CI: 0.89-7.07;P= 0.08).

Currently, major hepatectomies are performed less frequently for CRLM treatment. Parenchyma-sparing resections are associated with increased survival due to the increased possibility of reop- eration in liver recurrence [ 7 , 27 ]. Mise et al. [27] retrospectively compared patients who underwent parenchyma-sparing resections with those who underwent non-sparing resections and observed that minor resections did not negatively impact OS or DFS. In fact, they allowed more patients to undergo a repeat hepatectomy (68% vs. 24%,P＜0.01) and were associated with a higher OS after liver recurrence (73.6% vs. 30.1%,P= 0.018). The multivariate analy- sis showed that the need for an extended resection was an inde- pendent predictive factor for not performing a repeat hepatectomy [hazard ratio (HR) = 8.18; 95% CI: 1.89-45.7;P＜0.01]. In our cen- ter, the use of parenchyma-sparing techniques was analyzed in a recent study comparing two different eras: 20 0 0-20 09 and 2009- 2014. Across these institutional eras we observed an increase in frequency of patients with bilobar disease who underwent resec- tion (23.2% vs. 37.1%,P= 0.014), in non-anatomic parenchyma pre- serving resections (13.4% vs. 31.4%,P= 0.006) and in minimally in- vasive procedures (1.2% vs. 20.1%,P＜0.001) without impairment in OS and DFS [7] .

Although repeat hepatectomies are increasingly performed in specialized centers, the evidence supporting their use is mainly drawn from case series and non-comparative cohorts [ 10 , 13-15 ]. A recent meta-analysis, including 34 non-comparative studies (3039 patients), showed that the median morbidity rate associ- ated with repeat hepatectomy was 23% (8%-71%) and that the me- dian mortality was 0% (0%-6%) [14] , which is consistent with our results.

Comparative studies evaluating the outcomes after reoperation for CRLM are scarce [ 24 , 28 ]. A few studies have shown poorer pe- rioperative results associated with repeat hepatectomies, with in- creased blood loss and operative time [ 13 , 24 ]. However, Wurster et al. [15] in their meta-analysis including eight comparative stud- ies (450 repeat hepatectomies vs. 2669 single hepatectomies) showed no differences in morbidity and perioperative mortality between the groups, which is similar to the findings of the present study.

The median OS at 5 years in patients who underwent repeat hepatectomy was 42% (17%-73%) in non-comparative studies [14] . This result is similar to that obtained in our study. Moreover, there was no difference in OS compared to the single hepatectomy group, which is also consistent with previous studies [13-15] . The main factors that negatively affect OS after repeat hepatectomy are the following: primary tumor stage T3/T4, multiple metastases, le- sion size ≥5 cm, positive margin in the initial hepatectomy, in- creased CEA level, disease-free interval ≤12 months, bilobar dis- ease, and extrahepatic disease [ 14 , 29 ]. An important finding of our study was the significant percentage of patients who remained dis- ease free after undergoing repeat hepatectomy, which was similar to that of patients who underwent single hepatectomy, in line with some recent series [ 14 , 25 ].

Few studies have evaluated the role of repeat hepatectomy in patients with early relapse ( ≤ 6 months) after initial hepatec- tomy [ 10 , 17 ]. We found no difference in OS and DFS between patients who underwent resection after early or late recurrence. This result indicates the potential role of reoperation for early re- currence; however, it should be interpreted with caution because of the limited number of patients in this subgroup. In other se- ries addressed to this issue, early recurrence was associated with a worse prognosis [ 10 , 17 ]. Inoue et al. [10] showed that patients with early relapse had worse survival, even after surgical treat- ment with curative intent. Similarly, Watanabe et al. [17] observed that patients with early relapse had worse OS at 5 years compared with patients with late relapse (24.0% vs. 57.7%,P＜0.01); how- ever, patients with early relapse undergoing rehepatectomy had increased survival rates compared with those who did not un- dergo surgery (HR = 6.48; 95% CI: 3.22-13.10;P= 0.01). In this context, preoperative chemotherapy may play an important role in the selection of the best candidates for reoperation. In our institution, we systematically use the response to chemotherapy as a guide to refer patients with early relapse to surgery, which can explain the absence of differences in survival observed in our study.

In conclusion, repeat hepatectomy is feasible and safe for re- current CRLM, with similar perioperative and long-term outcomes compared to single hepatectomy. Surgical treatment of early liver recurrence offers oncological outcomes similar to those obtained for late recurrence and is an important therapeutic option in this subgroup of patients.

Acknowledgments

None.

CRediTauthorshipcontributionstatement

PauloFigueiredoCosta:Conceptualization, Formal analysis, Writing - original draft, Writing - review & editing.FabricioFer-reiraCoelho:Conceptualization, Formal analysis, Writing - original draft, Writing - review & editing.VagnerBirkJeismann:Concep- tualization, Data curation, Writing - review & editing.JaimeArthurPirolaKruger:Conceptualization, Data curation, Writing - review & editing.GiltonMarquesFonseca:Conceptualization, Data cu- ration, Writing - review & editing.IvanCecconello:Supervision, Writing - review & editing.PauloHerman:Conceptualization, Su- pervision, Writing - original draft, Writing - review & editing.

Funding

None.

Ethicalapproval

This study was approved by the Ethics Committee of the Hos- pital das Clinicas, University of S?o Paulo School of Medicine, S?o Paulo, Brazil (19476).

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.