Non-response to preoperative chemotherapy is a contraindication to hepatectomy plus radiofrequency ablation in patients with colorectal liver metastases

The long-term outcome of 228 patients with colorectal liver metastases (CRLM) who underwent preoperative chemotherapy followed by hepatectomy ± RFA were retrospectively analyzed. Stratified by chemotherapy response, patients were divided into responding (n=129) and non-responding groups (n=99). Patients who underwent hepatectomy-RFA had a greater number of metastases (median of 4 vs. 2, p=0.000), a higher incidence of bilobar involvement (66.7% vs. 49.1%, p=0.014) and longer chemotherapy cycles (median of 6 vs. 4, p=0.000). In the responding group, the median overall survival (OS) and recurrence free survival (RFS) of hepatectomy-RFA and the hepatectomy alone subgroups were comparable (38.6 months vs. 43.2 months, p=0.824; 8.2 months vs. 11.4 months, p=0.623). In the non-responding group, the median OS and RFS of patients treated with hepatectomy-RFA were significantly shorter (18.5 months vs. 34.2 months, p=0.000; 5.1 months vs. 5.9 months, p=0.002). RFA was identified as the unfavorable independent factor for both OS (HR=3.60, 95%CI=1.81-7.16, p=0.039) and RFS (HR=1.70, 95%CI=1.00-2.86, p=0.048) in non-responsive patients. Local recurrence rate after hepatectomy-RFA was higher in the non-responding group (48.1% vs. 23.6%, p=0.018). Non-response to preoperative chemotherapy may be a contraindication to hepatectomy-RFA in patients with CRLM.


INTRODUCTION
Liver is the most frequent site for metastasis from colorectal cancer, with more than 50% of patients developing hepatic metastases during the course of the disease [1,2]. Liver resection combined with modern chemotherapy is considered the treatment of choice for patients with CRLM [3,4]. For CRLM judged to be resectable, preoperative chemotherapy is considered a standard of care in most Western countries [5]. For patients with CRLM not eligible for resection at diagnosis due to risk of subsequent liver insufficiency or maldistribution patients treated with resection-RFA or resection alone were different for baseline tumor characteristics [15]. It is hard to determine whether RFA impacts survival negatively because of a more aggressive disease biology that makes treatment with resection alone impossible. Therefore, it is more reasonable to identify a group of patients who will benefit from this treatment modality.
Previous studies have demonstrated that good longterm outcomes can be achieved by a combination of prior chemotherapy and RFA with or without liver resection in patients with unresectable CRLM [16][17][18]. We postulate chemotherapy response could also predict therapeutic effect of this treatment. However, the proportions of patients receiving preoperative chemotherapy vary among studies, and the impact of response on outcome is seldom assessed. In current study, we compared the outcome of patients treated with hepatic resection alone or resection-RFA, grouped by response to preoperative chemotherapy. The clinical efficacy of hepatectomy-RFA was comparable to hepatectomy alone in patients who responded to preoperative chemothearpy. Whereas, the long-term results of non-responding patients who underwent hepatectomy-RFA were significantly worse.

Patient characteristics
The clinicopathologic features of the patients in the study are displayed in Table 1. The study population comprised 228 patients (145 men and 83 women; median age, 54 years). Most patients (83.8%) developed synchronous liver metastases. 75.9% of the patients had more than one metastases, with a median of 3 lesions, and a maximum of 9 lesions. The median diameter of the largest lesion was 2.8cm, and 49.6% of patients had lesions larger than 3 cm. Bilobar distribution of metastases was observed in 54.4% of the patients. Preoperative chemotherapy regimens included oxaliplatin-based (70.2%), irinotecan-based (17.1%), oxaliplatin and irinotecan-based (12.7%). Biological agents were included in preoperative regimens for 28.5% of the patients. The median number of preoperative chemotherapy cycles was 5, with 46 patients (20.2%) receiving at least 8 cycles. PR, SD and PD to preoperative chemotherapy were observed in 129 (56.3%), 68 (29.8%) and 31 ( Patients who underwent hepatectomy-RFA had a greater number of total liver lesions (median of 4 vs. 2, p=0.000). Bilobar distribution of liver lesions was also more common in patients who underwent hepatectomy-RFA (66.7% vs. 49.1%, p=0.014). The median cycle of preoperative in hepatectomy-RFA cohort was six, compared with the four cycles in hepatectomy alone group (p=0.000). The proportion of patients who developed major complications were 29.0% and 11.9% after hepatectomy-RFA or hepatectomy alone, respectively (p=0.002).

Overall survival
The median follow-up of the entire study population was 32 months. A total of 98 patients (42.9%) died during follow-up. Median survival was 35.7 months. The 1-, 3-, and 5-year OS rates were 93.3%, 50.5%, and 20.8%, respectively. According to response to preoperative chemotherapy, the included patients were divided into responding group (PR, n=129) and non-responding group (SD/PD, n=99). Risk factors for decreased OS are displayed in Table 2.
The median follow-up of the responding group (129 patients) was 31.6 months. 5-year survival rate of hepatectomy-RFA and the hepatectomy alone subgroups were comparable (30.7% vs. 33.6%, p=0.824) ( Figure  1B). Multivariate analysis revealed that prolonged chemotherapy (p=0.01), and major complications (p=0.002) remained as significant predictive factors for unfavorable survival.
The median follow-up of the non-responding group (99 patients) was 32 months. The median survival time for 27 patients who received hepatectomy-RFA was 18.5 months, significantly shorter than 34.2 months observed in those treated with hepatectomy alone (p=0.000) ( Figure 1C). In multivariate analysis, intraoperative RFA (p=0.039) and bilobar distribution (p=0.001) were the predictors of worse OS.
In the responding group (129 patients), the median RFS of patients who underwent hepatectomy-RFA was 8.2 months, not significantly different from the 11.4 months after treatment of hepatectomy alone (P=0.623) ( Figure  2B). Multivariate analysis demonstrated that bilobar distribution (p=0.007), and prolonged chemotherapy (p=0.002) were independent predictors for recurrence.
In the non-responding group (99 patients), all patients who received hepatectomy-RFA experienced recurrence within one year of operation, while the 3-year RFS rate for patients after hepatectomy alone was 19.7% (p=0.002) ( Figure 2C). In multivariate analysis, the significant predictors of higher recurrence was intraoperative RFA (p= 0.048), and ≥4 hepatic metastases (p=0.000).

Patients who underwent hepatectomy-RFA
A separate analysis was performed in sixty-nine patients who underwent hepatectomy-RFA. The median survival of the responding group was 42.3 months, significantly longer than the 18.5 months in the nonresponding group (p=0.000) ( Figure 3A). A significant difference was also found in RFS between the two groups (8.2 months vs. 5.1 months, p=0.000) ( Figure 3B). Multivariate analysis demonstrated that non-responsive to chemotherapy (p=0.000) and ≥4 hepatic metastases (p=0.008) were independent predictors for unfavorable survival. The predictive factors for recurrence were non-responsive to chemotherapy (p=0.013), and major complications (p=0.000).

Patterns of recurrence
A total of 167 (73.2%) patients experienced recurrence, including intrahepatic, extrahepatic, and both intra-and extrahepatic recurrences in 128 (56.1%), 11 (4.8%) and 28 (12.3%) patients, respectively. Differences in recurrence patterns among the treatment groups is shown in Table 4. The overall recurrence rate after hepatectomy-RFA and hepatectomy alone were 76.8% and 71.7%, respectively (p=0.423). The dominant pattern of recurrence was intrahepatic (63.8% vs. 52.8%, p=0.126); local recurrence rate at the ablation site/resection margin was documented in 31.9% of patients who underwent hepatectomy-RFA, compared with 18.2% after hepatecomy alone (p=0.023). The proportion of patients in two treatment groups found to have extrahepatic or both intra-and extrahepatic recurrences was similar.
Recurrence pattern after surgery in the responding and non-responding groups is shown in Table 5. In the non-responding group, recurrence of any kind occurred more frequently after hepatectomy-RFA (92.6%), versus that after resection alone (76.4%, p=0.068). Both intrahepatic and local recurrence rates after hepatectomy-RFA were significantly higher than those in patients treated with resection alone (77.8% vs. 55.6%, p=0.043; 48.1% vs. 23.6%, p=0.018). Rate of recurrence at the ablation site was comparable to that at the margin after resection alone, whether in the entire group (17.4% vs. 18.2%, p=0.868) or both subgroups separately (responding group: 11.9% vs. 13.8%, p=0.766; non-responding group: 25.9% vs. 23.6%, p=0.8111). Significant differences in recurrence patterns between two treatment modalities were not observed in the responding group.

DISCUSSION
Improvements in preoperative chemotherapy and combination hepatectomy with RFA allow radical treatment for an increased number of patients with CRLM. This study suggested that the benefits associated with this treatment modality were strongly impacted by response to chemotherapy.
The median survival of the entire study population was 35.7 months in present study, inferior to the reported survival of 40.5-48.2 months after conversion therapy [19][20][21]. We attributed this to the relatively lower response rate (56.3%) compared to a mean value of 74% (range 60-100%) reported in previous studies [20]; and the variable definition of resectability. The median RFS after conversion therapy was 7.7 months, which is in line with previous results (3.2-10.6 months) [19][20][21], reaffirming the finding of relatively short RFS and a prolonged OS after conversion therapy followed by surgery.
Objective evaluation of hepatectomy-RFA compared to resection alone is difficult. Indeed, patients who underwent hepatectomy-RFA were characterized by greater number and more widespread distribution of liver lesions in current study. Besides, these patients seem to be less sensitive to chemotherapy (prolonged chemotherapy was required in 34.8% of patients). Recently, Imai et al used propensity score matching to overcome imbalance of background characteristics [22]. After matching, overall and disease free survival in hepatectomy-RFA group were not different from those of patients treated with hepatectomy alone. The results justified the adjunct use of RFA in hepatectomy for selected patients. Whereas, detailed indications and contraindications for this treatment is still unclear.
RFA is usually regarded as an alternative therapeutic option with poor disease control for unresectable CRLM [23][24]. Indeed, local recurrence rate following hepatectomy-RFA was significantly higher in the entire study population (31.9% vs. 18.2%, p=0.023) and non-responding group (48.1% vs. 23.6%, p=0.018). Nevertheless, this was not observed in chemotherapy-responsive patients (21.4% vs. 13.8%, p=0.271). Although patients who underwent hepatectomy-RFA were characterized by a larger tumor burden, the responding group showed similar long-term survival after  Several studies have demonstrated that progression while preoperative chemotherapy is not an absolute contraindication to liver resection for patients whose disease remains resectable [26][27]. The major concern is the treatment of patients with CRLM whose disease remains unresectable after preoperative chemotherapy.
Is liver resection in combination with intraoperative RFA indicated, or should further chemotherapy regimens be planned? Given the low objective response rates to second-line chemotherapy [28][29][30] and the lack of other effective treatment options for patients with no response to chemotherapy, rescue surgery consist of hepatectomy and RFA was sometimes introduced as an alternative therapeutic option in our hospital. However, whether this treatment modality could work as effectively as hepatectomy alone in non-responsive patients is largely unknown. To our knowledge, few studies have focused on this topic. Our results suggested that hepatectomy-RFA provided significantly worse outcome with 3-year OS rate of 12%, and 3-year RFS rate of 4.4% in patients with downstaging not achieved. This long term result even seems inferior to that of palliative chemotherapy [31][32], which seem to be a result of high major complications rate (29%) and overall recurrence rate (92.6%) following hepatectomy-RFA. Even though one might argue the outcome may be a result of a selection bias, intraoperative RFA showed an independent association with decreased OS and RFS after adjusting for other clinical and pathological factors.
Patients undergoing hepatectomy-RFA mostly had worse tumor biology than those undergoing hepatectomy     RFA=Radiofrequency Ablation. Local recurrence=ablation site and resection margin recurrence in the responding group, suggesting significant effects of the extent of chemo-sensitivity on reducing recurrence and prolonging survival. Sasaki et al [33] recently performed a retrospective study of 485 patients who underwent curative hepatectomy and identified primary tumor nodal metastases, KRAS mutation, and preoperative high CEA related to survival. For patients who underwent hepatectomy-RFA, those with no or one risk factor had a 5-year OS rate similar to patients treated with resection alone. In contrast, patients with two or more risk factors had a much worse prognosis. The findings imply that patients who would benefit the most from hepatectomy-RFA could be selected using tumor biology related factors, among which chemotherapy response may play an important role. This study had some limitations. First, it was a retrospective study with relatively small sample size. Second, operations were performed by different surgeons, with their own criteria regarding patients selection for resection or hepatectomy-RFA. What's more, some lesions judged to be unresectable in the past may now be treated by resection alone, with improved surgical techniques and skills, which may result in selection bias. Furthermore, only patients with no extrahepatic disease were eligible for the current study and these represent only a proportion of patients with CRLM. Last but not least, preoperative chemotherapy regimen was not standardized and this is also a major source of bias.

Inclusion criteria
Patients who underwent surgery for CRLM at the Department of hepatobiliary surgery, Cancer Hospital, Chinese Academy of Medical Sciences between January 1, 2004, and December 31, 2015, were identified from our prospective institutional database. In these patients, patients who met the following criteria were considered for further analysis. Inclusion criteria were: (1) histologically proven colorectal adenocarcinoma liver metastases; (2) preoperative chemotherapy was given, followed by hepatectomy with or without intraoperative RFA for curative intents. Excluion criteria were: (1) extrahepatic metastases detected on preoperative imaging or during surgery; (2) R2 resection; (3) postoperative deaths (noncancer-related 90-day mortality); or (4) a history of prior hepatectomy for CRLM. In total, 228 patients met the eligibility criteria.

Perioperative management
All patients were evaluated preoperatively, including carcinoembryonic antigen (CEA) levels; abdominopelvic computed tomography (CT) and magnetic resonance imaging (MRI); and chest radiography or chest CT to determine the disease stage. Preoperative chemotherapy was mainly recommended to patients with initially unresectable liver metastases; or to patients with multiple high-risk factors: synchronous metastases, ≥4 hepatic lesions, primary tumor invading nearby tissues/organs and imaged mesenteric nodal disease. The definitions of unresectability were as follows: multiple liver metastases that required resection of more than 70 % of non-tumor liver for removal of all tumors, tumors invading all three hepatic veins, tumors invading both the left and right branches of the hepatic artery or portal vein, and unresectable extrahepatic metastases. Chemotherapy was composed of a combination of 5-fluorouracil/capecitabine and oxaliplatin/irinotecan with or without bevacizumab and cetuximab. Tumor response was assessed according to the Response Evaluation Criteria In Solid Tumors criteria (RECIST, version 1.1) every two cycles. Complete response (CR) was defined as the disappearance of all signs of the current disease recorded from the start of the treatment. Partial response (PR) was defined as a decrease in the size of the tumor of ~30%. Stable disease (SD) was defined as no disease progression or regression. Progressive disease (PD) was defined as an increase in the tumor size of >20% during or after treatment. All imaging studies were reviewed by at least two independent radiologists until the final conclusion was drawn. Surgery with curative intent was performed for hepatic lesions considered to be treatable by hepatecomy ± RFA, with the residue liver volume of at least 30%. The decision to undertake surgery for controversial cases was reached by consensus of a multidisciplinary team (MDT) including surgeons, oncologists and radiologists.
During surgery, the peritoneal cavity was inspected to exclude previously undetected extrahepatic lesions. Manual liver palpation and intra-operative ultrasound were used to rule out occult lesions and confirm the number, size and location of the liver metastases. If all the lesions were deemed resectable, hepatectomies alone were undertaken. Neither the number nor the size of metastases excluded any patients from resection. Patients were treated with intraoperative RFA when one hepatic lesion was considered unresectable because of deep location or proximity to major vascular structures, so as to avoid extensive hepatectomy, especially for lesions less than 3 cm. The principle of surgery was to remove all detectable lesions with a tumorfree margin. After resection of the resectable metastases, intraoperative ultrasound was performed to guide placement of the RFA needle into the remaining lesions and then ablation was started. Successful ablation was defined as the complete destruction of the tumor with at least a 1-cm zone of normal liver parenchyma in real-time ultrasound. All specimens were subjected to histological evaluation to confirm the pathological diagnosis, number and size of liver lesions, and the width of the surgical margin. In case of multiple liver metastases, the diameter of the largest lesion was defined as the final size, and the closest margin was recorded. R1 resection was defined with a distance from the metastasis edge to the transection line of less than 1 mm.
Postoperative complications were graded according to the Clavien system, and major complications were defined as any grade III or IV complication. After discharge, adjuvant chemotherapy was recommended to most patients.

Follow-up
After surgery, patients were followed up at regular intervals. Serum CEA and imaging studies were performed to detect any intrahepatic recurrence or distant metastases. The first follow-up occurred one month post-surgery, with subsequent ones every 3 months for up to 2 years, and every 6 months thereafter.

Statistical analysis
Continuous variables are expressed as the median (range). Continuous and categorical variables were compared by Mann-Whitney U test and Chi square test or Fisher's exact test respectively. Survival analyses were carried out using the Kaplan-Meier method, with group comparisons by the log rank test. Multivariate models were used for OS and RFS using the Cox proportional hazard method. Clinicopathological factors were included in each model if they achieved a p < 0.1 for significance in univariate regression analysis. Use of RFA was included in each model, irrespective of statistical significance. P < 0.05 was considered to indicate statistical significance. OS was estimated from liver resection to death; RFS was from liver resection to the first documented disease recurrence. Statistical analyses were performed using the SPSS, version 22, Armonk NV, USA.

CONCLUSIONS
This study showed that hepatectomy-RFA in patients who respond to preoperative chemotherapy is feasible and may be associated with a long-term outcome similar to that after hepatectomy alone. However, the poor results obtained by hepatectomy-RFA in patients with tumor stablization/progression suggest that non-response to preoperative chemotherapy is a contraindication to this treatment in patients with CRLM, and further chemotherapy may be a more reasonable choice for these patients. Chemotherapy response could be a useful tool for selecting patients for hepatectomy-RFA.