Abstract
Background:
We evaluated the associations of surgical margin status and somatic mutations with the incidence of local recurrence (LR) and oncologic outcomes in patients undergoing R0-intent (microscopically negative margin) resection of colorectal liver metastases (CLM).
Methods:
Patients with CLM who underwent initial R0-intent resection and analysis of tumor tissue using next-generation sequencing during 2001-2018 were analyzed. Recurrences were classified as LR (at the resection margin), other intrahepatic recurrence, or extrahepatic recurrence. Predictors and survival effect of LR were evaluated using univariate and multivariate analysis.
Results:
Of 552 patients analyzed, 415 (75%) had R0 resection (margin width ≥ 1.0 mm), and 38 (7%) had LR. LR incidence was not affected by surgical margin width. RAS/TP53 co-mutation was associated with increased risk of intrahepatic recurrence (67% vs. 49%; p < 0.001) and overall recurrence (p < 0.001). However, incidence of LR did not differ significantly by RAS/TP53, BRAF, SMAD4, or FBXW7 mutation. Extrahepatic disease (hazard ratio [HR], 1.47; p = 0.034), > 8 cycles of preoperative chemotherapy (HR, 1.98; p = 0.033), tumor viability ≥ 50% (HR, 1.55; p = 0.007), RAS/TP53 co-mutation (HR, 1.69; p = 0.001), and SMAD4 mutation (HR, 2.44; p < 0.001) were independently associated with poor overall survival, but surgical margin status was not.
Conclusions:
Although somatic mutations were associated with overall recurrence, neither surgical margin width nor somatic mutations affected LR risk after R0-intent hepatectomy for CLM. LR and prognosis were likely driven by individual tumor biology rather than surgical margins.
Keywords: Colorectal liver metastases, Hepatectomy, Local Recurrence, Surgical margin, Somatic mutation
INTRODUCTION
Resection of colorectal liver metastases (CLM) with tumor-free surgical margins results in 5-year survival rates of up to 60%.1 For this reason, surgical resection has been established as the standard of care for patients with CLM, and achieving microscopically negative margins (i.e., R0 resection) has traditionally been thought to be essential.
Surgeons continue to believe that achieving R0 resection decreases the risk of local recurrence resulting in improved outcomes,2 but the optimal surgical margin width has been debated for several decades.3-5 Currently, a tumor-free margin at least 1 mm wide is thought to be the ideal definition of an R0 resection, with resections resulting in tumor within 1 mm of the specimen edge considered R1 resection.6
As advances in surgical technique and preoperative chemotherapy have expanded surgical indications for CLM, the benefit of an R0 resection has been questioned.7, 8 Recently, tumor biology of CLM has been reported to have a greater impact on postoperative prognosis than surgical margin status,9, 10 and favorable outcomes have been reported even after R1-intent resection for CLM.11 The incidence of local recurrence after CLM resection differs widely among studies, and the association of local recurrence with surgical margin status after R0-intent resection in the recent era is unclear.
With advances in genetic analysis through next generation sequencing (NGS), we are now beginning to understand the role of somatic mutations as independent prognostic factors for patients with CLM. This discovery began with work around RAS mutations,12 where our group has reported that the margin width necessary to ensure an appropriately low risk of local tumor progression after ablation therapy for CLM can differ according to the tumor’s RAS status.13, 14 Similarly, it has been reported that wider resection margins may be required for patients with RAS-mutant CLM than for those with RAS wild-type CLM in order to decrease intrahepatic recurrence.15-17 More recently, our group and others have shown that overall prognosis may also be influenced by other somatic mutations or co-mutations, for instance, co-mutation of RAS and TP53 and single mutations of SMAD4 and FBXW7.18-23 This work raises important questions about the impact of these somatic mutations on necessary margin width and the risk of local recurrence, but no study attempting to answer these questions has yet been reported.
In this study, we investigated the associations between surgical margin status and somatic mutations analyzed with NGS and local recurrence and oncologic outcomes after R0-intent resection for CLM, as well as the impact of local recurrence on survival outcomes.
METHODS
Patient Population and Data Collection
Demographic, clinical, and perioperative data were collected from a prospectively maintained database of all consecutive patients who underwent elective initial liver resection with curative (R0) intent for CLM at The University of Texas MD Anderson Cancer Center during January 2001 through December 2018. The Institutional Review Board approved this study (PA18-0400).
A total of 1865 patients were identified (Supplemental Figure 1). Patients were excluded when somatic gene mutations in primary tumors or metastases were not examined by NGS (n=1264) or postoperative follow-up images were unavailable (n=1). Patients who underwent hepatectomy concurrently with or following ablation therapy (n=48) were also excluded in order to avoid the influence of local tumor progression after ablation therapy. The remaining 552 patients were included in the analysis. Somatic gene mutations in primary tumors or metastases were analyzed by NGS with 50-gene or 134-gene panels as previously reported.24
Institutional Therapeutic Approach and Follow-up
Our standard strategy for patients diagnosed with CLM includes perioperative chemotherapy.25 In patients diagnosed with primary tumors and CLM simultaneously, combined resection or hepatectomy followed by resection of the primary tumor was performed unless the primary tumor was causing symptoms, in which case resection of primary tumor was performed at first. Even in patients with extrahepatic disease, liver resection was performed if the extrahepatic disease was considered resectable. Given that our institution is a tertiary referral center, some patients underwent resection of their primary tumor and/or chemotherapy at a local hospital before presentation to our institution. Two-stage hepatectomy combined with portal vein embolization was considered in patients with insufficient future liver remnant per our institution’s treatment algorithms.26 Major resection was defined as liver resection including 3 or more liver segments.27 Postoperatively, patients had dynamic computed tomography every 3 to 4 months to check for tumor recurrence. The therapeutic approach was discussed and finally decided on at a multidisciplinary conference including surgeons, medical oncologists, and interventional radiologists. When the recurrence including local recurrence was identified, the resectability was evaluated and the therapeutic approach was discussed at a multidisciplinary conference.
Definitions of Types of Tumor Recurrence
Recurrence was defined as the first recurrence after hepatectomy. Local recurrence was defined as recurrence at the resection margin, and other intrahepatic recurrence was defined as any intrahepatic recurrence at a location other than the resection margin without local recurrence. (Figure 1) Patients were classified into 4 groups according to the site of recurrence: local recurrence, other intrahepatic recurrence, extrahepatic recurrence alone, and no recurrence. Patients who had local and/or other intrahepatic disease in addition to extrahepatic disease at the time of first recurrence were classified as local and/or other intrahepatic disease.
Figure 1.
Typical imaging of local recurrence and other intrahepatic recurrence. (A) A case with solitary colorectal liver metastases in segment 4 (arrowhead), who underwent left hepatectomy. (B) Local recurrence (arrowhead) was identified 9 months after the resection. (C) A case with two colorectal liver metastases in the right liver, who underwent right hepatectomy. (D) Intrahepatic but no local recurrence (arrowhead) was identified 29 months after the resection.
Measurement of Surgical Margins
All patients underwent hepatic resection using a 2-surgeon technique using the Cavitron Ultrasonic Dissector (CUSA) and saline-linked cautery (TissueLink) as previously described.28 The resection margins were immediately inked and grossly evaluated. The width of the tumor-free parenchymal margin was measured, and frozen section was used if needed. For patients who underwent 2-stage hepatectomy, the shortest distance from specimen edge to tumor among all the specimens resected in first- and second-stage hepatectomy was defined as the resection margin width.
Statistical Analysis
Categorical variables were compared using the chi-squared test or the Fisher exact test, as appropriate. Continuous variables were summarized using medians and interquartile ranges and were using compared using the Mann-Whitney U test. A logistic regression model was used to assess potential predictors of local recurrence. Cumulative incidence plots for hepatic recurrence was performed using the competing risk analysis and compared using Gray’s test. A Cox proportional hazards model was used to assess the association of several variables with survival outcome. Local recurrence–free survival and overall survival (OS) were calculated using the Kaplan-Meier method and compared by using the log-rank test. Variables with p <0.1 in univariable analysis were retained in the multivariable model. P < 0.05 was considered statistically significant. hazard ratio (HR), and 95% confidence interval (95% CI) were calculated for each variable. The statistical analysis was performed using JMP Pro software, version 15.0 (SAS Institute Inc., Cary, NC).
RESULTS
Patient Characteristics
Patient characteristics are summarized in Table 1. A total of 129 patients (23%) had right-sided primary tumors, and 384 patients (72%) had lymph node metastases. A total of 121 patients (22%) had extrahepatic disease, of whom 82 patients (68%) underwent resection of the extrahepatic disease. A total of 481 patients (87%) underwent preoperative chemotherapy, of whom 470 received fluorouracil and oxaliplatin or irinotecan with bevacizumab or cetuximab. Regarding the chemotherapy regimen, 67% of patients received preoperative chemotherapy using 5-Fluorouracil and oxaliplatin and 62% of patients underwent preoperative chemotherapy including bevacizumab as our institutional policy. Cetuximab or other EGFR targeted agents were administered in 35 patients (12%). Seventy-four patients (13%) underwent portal vein embolization prior to hepatectomy, and 211 patients (38%) underwent major hepatectomy. A total of 524 patients (95%) had at least 1 somatic gene mutation. BRAF mutation was confirmed in 17 patients (3%), RAS/TP53 co-mutation in 182 patients (33%), SMAD4 mutation in 70 patients (13%), and FBXW7 mutation in 40 patients (7%).
Table 1.
Patient Characteristicsa
| Characteristic | All patients (N = 552) |
Local recurrence (N = 38) |
No local recurrence (N = 514) |
P value |
|---|---|---|---|---|
| Age, median (IQR) | 56 (46-63) | 56 (50-61) | 56 (46-63) | 0.729 |
| Sex, male | 312 (57) | 19 (50) | 293 (57) | 0.403 |
| Right-sided primary tumor | 129 (23) | 9 (24) | 120 (23) | 0.962 |
| Primary LN metastases * | 384 (72) | 26 (70) | 358 (72) | 0.789 |
| Synchronous CLM | 411 (74) | 28 (74) | 383 (75) | 0.910 |
| Extrahepatic disease | 121 (22) | 11 (29) | 110 (22) | 0.302 |
| Chemotherapy before hepatectomy | 0.989 | |||
| None | 71 (13) | 5 (13) | 66 (13) | |
| ≤ 8 cycles | 382 (69) | 26 (68) | 356 (70) | |
| > 8 cycles | 97 (18) | 7 (18) | 90 (18) | |
| Chemotherapy regimen | 0.524 | |||
| 5-FU + Oxaliplatin | 324 (67) | 22 (67) | 302 (67) | |
| 5-FU alone | 11 (2) | 2 (6) | 9 (2) | |
| 5-FU + Irinotecan | 81 (17) | 6 (18) | 75 (17) | |
| 5-FU + Oxaliplatin + Irinotecan | 25 (5) | 2 (6) | 23 (5) | |
| Addition of Bevacizumab | 375 (68) | 26 (68) | 349 (68) | 0.947 |
| Addition of anti-EGFR agent | 35 (6) | 3 (8) | 32 (6) | 0.693 |
| Chemotherapy after hepatectomy | 395 (74) | 26 (72) | 369 (74) | 0.791 |
| Tumor viability < 50% ** | 251 (50) | 17 (47) | 234 (50) | 0.748 |
| Multiple CLM | 345 (63) | 28 (74) | 317 (62) | 0.131 |
| Largest metastasis ≥ 5 cm | 76 (14) | 9 (24) | 67 (13) | 0.088 |
| Portal vein embolization | 74 (13) | 3 (8) | 71 (14) | 0.270 |
| Major hepatectomy | 211 (38) | 17 (45) | 194 (38) | 0.396 |
| Any somatic gene mutation | 524 (95) | 26 (93) | 488 (93) | 0.956 |
| BRAF mutation | 17 (3) | 1 (3) | 16 (3) | 0.865 |
| RAS/TP53 co-mutation | 184 (33) | 15 (39) | 169 (33) | 0.411 |
| SMAD4 mutation | 70 (13) | 5 (13) | 65 (13) | 0.927 |
| FBXW7 mutation | 40 (7) | 1 (3) | 39 (8) | 0.198 |
Abbreviations: LN, lymph node; CLM, colorectal liver metastases; 5-FU, 5-Fluorouracil; EGFR, epidermal growth factor receptor.
Values in table are number of patients (percentage) unless otherwise indicated.
Pathological evaluation was unavailable in 20 patients because primary tumor was in place
Data missing in 48 patients
Association of Patterns of Recurrence with Surgical Margin Width and Somatic Mutation Status
At a median follow up of 58.4 months (range, 3.2-169), 435 patients (79%) had recurred. The recurrence patterns were as follows: local recurrence, 38 patients (6.9%); other intrahepatic recurrence, 265 patients (48%); extrahepatic recurrence alone, 132 patients (24%); and no recurrence, 117 patients (21%).
Surgical margin width was < 1.0 mm (i.e., R1 resection) in 137 patients (25%); 1.0 to 4.9 mm in 153 patients (28%); 5.0 to 9.9 mm in 121 patients (22%); and ≥ 10 mm in 141 patients (26%). Recurrence patterns by surgical margin width are summarized in Table 2. Surgical margin width had no association with any pattern of recurrence.
Table 2.
Patterns of recurrence by surgical margin width
| Margin < 1.0 mma (n = 137) |
Margin 1.0-4.9 mm (n = 153) |
Margin 5.0-9.9 mm (n = 121) |
Margin ≥10 mm (n = 141) |
P value | |
|---|---|---|---|---|---|
| Any recurrence, No. (%) | 112 (82) | 119 (78) | 95 (79) | 109 (77) | 0.795 |
| Local recurrence, No. (%) | 11 (8) | 12 (8) | 7 (6) | 8 (6) | 0.787 |
| Other intrahepatic recurrence, No. (%) | 73 (53) | 73 (48) | 57 (47) | 62 (44) | 0.476 |
| Extrahepatic recurrence alone, No. (%) | 28 (20) | 34 (22) | 31 (26) | 39 (28) | 0.488 |
| No recurrence, No. (%) | 25 (18) | 34 (22) | 26 (21) | 32 (23) | 0.795 |
R0 resection.
The relationships between somatic gene mutations and surgical margin width are shown in Table 3. Among the 184 patients with RAS/TP53 co-mutation, surgical margin width was <1.0 mm in 39 patients (21%), 1.0 to 4.9 mm in 54 (29%), 5.0 to 9.9 mm in 38 (21%), and ≥10 mm in 53 (29%). This pattern of surgical margin widths was similar to the one in patients without RAS/TP53 co-mutation (p = 0.378). Patterns of surgical margin widths were also similar between patients with BRAF mutation and wild-type BRAF, SMAD4 mutation and wild-type SMAD3, and FBXW7 mutation and wild-type FBXW7.
Table 3.
Relationships between somatic gene mutations and surgical margin width
| Mutation and status | Total no. of patients |
No. (%) of patients with | P value | |||
|---|---|---|---|---|---|---|
| Margin < 1.0 |
Margin 1.0-4.9 |
Margin 5.0-9.9 |
Margin ≥ 10 |
|||
| RAS/TP53 | ||||||
| Co-mutant | 184 | 39 (21) | 54 (29) | 38 (21) | 53 (29) | 0.378 |
| Others | 368 | 98 (27) | 99 (27) | 83 (23) | 88 (24) | |
| BRAF | ||||||
| Mutant | 17 | 3 (18) | 4 (24) | 5 (29) | 5 (29) | 0.795 |
| Wild-type | 535 | 134 (25) | 149 (28) | 116 (22) | 136 (25) | |
| SMAD4 | ||||||
| Mutant | 70 | 13 (19) | 22 (31) | 16 (23) | 19 (27) | 0.602 |
| Wild-type | 482 | 124 (26) | 131 (27) | 105 (22) | 122 (25) | |
| FBXW7 | ||||||
| Mutant | 40 | 9 (23) | 10 (25) | 12 (30) | 9 (23) | 0.674 |
| Wild-type | 512 | 128 (25) | 143 (28) | 109 (21) | 132 (26) | |
| Any mutation | ||||||
| Positive | 524 | 132 (25) | 144 (27) | 113 (22) | 135 (26) | 0.663 |
| Negative | 28 | 5 (18) | 9 (32) | 8 (29) | 6 (21) | |
Recurrence patterns by somatic mutation status are summarized in Figure 2. RAS/TP53 co-mutation was associated with increased risk of overall recurrence (88% vs. 74%; p < 0.001), but there was no difference in the rate of overall recurrence between patients with and without BRAF mutation (71% vs. 79%; p = 0.418), SMAD4 mutation (86% vs. 78%; p = 0.115), or FBXW7 mutation (75% vs. 79%; p = 0.549). RAS/TP53 co-mutation was also associated with increased risk of intrahepatic recurrence (i.e., local or other intrahepatic recurrence) (67% vs. 49%; p < 0.001), but the incidence of local recurrence was similar in patients with and without RAS/TP53 co-mutation (8.2% and 6.3%, respectively; p = 0.411). Rates of local recurrence were also similar in patients with the mutant and wild-type forms of BRAF (5.9% and 6.9%, respectively; p = 0.865), SMAD4 (7.1% and 6.9%, respectively; p = 0.927), and FBXW7 (2.5% and 7.2%, respectively; p = 0.198).
Figure 2.
Pattern of recurrence by somatic gene mutation status in colorectal liver metastases. LR, local recurrence; OHR, other intrahepatic recurrence; EHR, extrahepatic recurrence alone; NR, no recurrence.
Regarding the correlation between tumor viability and surgical margin width, surgical margin width was <1.0mm in 73 patients (29%), 1.0 – 4.9 mm in 75 patients (30%), 5.0 -9.9 mm in 52 patients (21%), and ≥ 10 mm in 53 patients (21%) in patients with minor response, while <1.0mm in 54 patients (22%), 1.0 – 4.9 mm in 65 patients (26%), 5.0 -9.9 mm in 59 patients (24%), and ≥ 10 mm in 73 patients (29%) in those with major response (p = 0.066). When we focused on R0 status, there was no significant difference in patients with minor response compared to those with major response (22% vs. 29%, p = 0.057).
Incidence of and Risk Factors for Local Recurrence
Local recurrence occurred earlier than other intrahepatic recurrence (median time to recurrence, 7.1 months vs. 10.3 months; p = 0.009) (Figure 3).
Figure 3.
Cumulative incidences of local recurrence and other hepatic recurrence.
The results of a Cox proportional hazards model analysis for local recurrence are shown in Table 4. For largest CLM ≥ 5 cm, with the HR for early local recurrence 2.11 (95% CI 1.00-4.46; p = 0.050); however, none of the clinicopathologic factors examined, including surgical margin width and somatic mutation status, was significantly associated with poor local recurrence.
Table 4.
Univariate analysis of risk for local recurrence-free survival (N=552)
| No. (%) of patients with LR |
Univariate Analysis | ||
|---|---|---|---|
| HR (95% CI) | P value | ||
| Sex | |||
| Male | 19 (6.1%) | 0.71 (0.38-1.34) | 0.288 |
| Female | 19 (7.9%) | ||
| Primary Location | |||
| Right Sided | 9 (7.0%) | 1.11 (0.53-2.35) | 0.779 |
| Left Sided | 29 (6.9%) | ||
| Primary LN metastases | |||
| Positive | 26 (6.8%) | 1.01 (0.50-2.04) | 0.984 |
| Negative | 11 (7.4%) | ||
| Timing of diagnosis as CLM | |||
| Synchronous | 28 (6.8%) | 0.97 (0.47-2.00) | 0.938 |
| Metachronous | 10 (7.1%) | ||
| Extrahepatic Disease | |||
| Positive | 11 (9.1%) | 1.52 (0.75-3.06) | 0.258 |
| Negative | 27 (6.3%) | ||
| Chemotherapy before Hepatectomy | |||
| None | Reference | ||
| ≤ 8 cycles | 33 (6.9%) | 1.10 (0.42-2.86) | 0.850 |
| > 8 cycles | 5 (6.0%) | 1.30 (0.41-4.11) | 0.651 |
| Postoperative Chemotherapy | |||
| Yes | 26 (6.6%) | 0.88 (0.42-1.82) | 0.729 |
| No | 10 (7.3%) | ||
| Tumor viability after chemotherapy | |||
| ≥ 50% | 19 (7.5%) | 1.19 (0.62-2.29) | 0.600 |
| < 50% | 17 (6.8%) | ||
| Tumor Number | |||
| Multiple | 28 (8.1%) | 1.95 (0.95-4.02) | 0.070 |
| Solitary | 10 (4.8%) | ||
| Size of Largest CLM (mm) | |||
| ≥ 5cm | 9 (11.8%) | 2.11 (1.00-4.46) | 0.050 |
| < 5cm | 29 (6.1%) | ||
| Portal vein embolization | |||
| Yes | 3 (4.1%) | 0.49 (0.15-1.61) | 0.242 |
| No | 35 (7.3%) | ||
| Major hepatectomy | |||
| Yes | 17 (8.1%) | 1.23 (0.65-2.33) | 0.530 |
| No | 21 (6.2%) | ||
| BRAF status | |||
| Mutant | 1 (5.9%) | 0.81 (0.11-5.88) | 0.832 |
| Wild | 37 (6.9%) | ||
| RAS/TP53 status | |||
| Co-mutant | 15 (8.2%) | 1.39 (0.73-2.67) | 0.318 |
| Others | 23 (6.3%) | ||
| SMAD4 status | |||
| Mutant | 5 (7.1%) | 1.14 (0.44-2.92) | 0.785 |
| Wild | 33 (6.9%) | ||
| FBXW7 status | |||
| Mutant | 1 (2.5%) | 0.37 (0.05-2.67) | 0.322 |
| Wild | 37 (7.2%) | ||
| Surgical Margin Status | |||
| < 1.0 mm | 11 (8.0%) | Reference | |
| 1.0 - 4.9 mm | 12 (7.8%) | 0.84 (0.37-1.91) | 0.684 |
| 5.0-9.9 mm | 7 (5.8%) | 0.54 (0.21-1.40) | 0.207 |
| > 10 mm | 8 (6.7%) | 0.50 (0.20-1.25) | 0.138 |
Abbreviations: HR, hazard ratio; LN, lymph node; CLM, colorectal liver metastases.
Prognostic Impact of Surgical Margin Width and Local Recurrence
The 5-year OS rate in the entire cohort was 65.2%. Patients with local recurrence had significantly worse OS than those with no recurrence or extrahepatic recurrence alone. However, there was no difference in OS between patients with local recurrence and other intrahepatic recurrence (p = 0.089) (Figure 4).
Figure 4.
Overall survival by pattern of recurrence. NR, no recurrence; EHR, extrahepatic recurrence alone; OHR, other intrahepatic recurrence; LR, local recurrence.
The results of a Cox proportional hazards model analysis for OS are shown in Table 5. In the multivariate analysis, presence of extrahepatic disease (HR, 1.47, 95% CI, 1.03-2.11; p = 0.034), > 8 cycles of chemotherapy before hepatectomy (vs. no chemotherapy; HR, 1.98, 95% CI, 1.06-3.71; p = 0.033), tumor viability ≥ 50% (HR, 1.55, 95% CI, 1.13-2.14; p = 0.007), RAS/TP53 co-mutation (HR, 1.69, 95% CI, 1.22-2.34; p = 0.001), and SMAD4 mutation (HR, 2.44, 95% CI, 1.65-3.63; p < 0.001) were independently associated with poor OS. However, surgical margin width < 1 mm was not an independent risk factor for OS. Patients with no recurrence (HR, 0.11, 95% CI, 0.05-0.28; p < 0.001) and extrahepatic recurrence (HR, 0.23, 95% CI, 0.12-0.42; p < 0.001) had significantly better OS than those with local recurrence; however, there was no difference in OS between patients with other intrahepatic recurrence and local recurrence (HR, 0.62, 95% CI, 0.37-1.03; p = 0.066).
Table 5.
Univariate and multivariate analysis of risk for OS (N=552)
| Univariate Analysis | Multivariate Analysis | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Sex | ||||
| Male | 0.86 (0.65-1.15) | 0.320 | ||
| Female | ||||
| Primary Location | ||||
| Right Sided | 1.27 (0.92-1.77) | 0.151 | ||
| Left Sided | ||||
| Primary LN metastases | ||||
| Positive | 1.33 (0.95-1.86) | 0.095 | 1.34 (0.94-1.92) | 0.107 |
| Negative | ||||
| Timing of diagnosis as CLM | ||||
| Synchronous | 1.10 (0.79-1.54) | 0.571 | ||
| Metachronous | ||||
| Extrahepatic Disease | ||||
| Positive | 1.43 (1.03-1.98) | 0.034 | 1.47 (1.03-2.11) | 0.034 |
| Negative | ||||
| Chemotherapy before Hepatectomy | ||||
| None | Reference | Reference | ||
| ≤ 8 cycles | 1.52 (0.97-2.40) | 0.070 | 1.18 (0.67-2.07) | 0.567 |
| > 8 cycles | 2.47 (1.47-4.15) | <0.001 | 1.98 (1.06-3.71) | 0.033 |
| Regimen of Preoperative Chemotherapy | 0.122 | |||
| 5-FU + Oxaliplatin | Reference | |||
| 5-FU alone | 0.76 (0.28-2.09) | 0.599 | ||
| 5-FU + Irinotecan | 1.24 (0.82-1.86) | 0.314 | ||
| 5-FU + Oxaliplatin + Irinotecan | 1.57 (0.82-3.01) | 0.172 | ||
| Postoperative Chemotherapy | ||||
| Yes | 1.03 (0.73-1.46) | 0.861 | ||
| No | ||||
| Tumor viability after chemotherapy | ||||
| ≥ 50% | 1.50 (1.11-2.03) | 0.009 | 1.55 (1.13-2.14) | 0.007 |
| < 50% | ||||
| Tumor Number | ||||
| Multiple | 1.65 (1.21-2.25) | 0.002 | 1.33 (0.93-1.91) | 0.121 |
| Solitary | ||||
| Size of Largest CLM (mm) | ||||
| ≥ 5cm | 1.22 (0.81-1.83) | 0.342 | ||
| < 5cm | ||||
| Portal vein embolization | ||||
| Yes | 1.27 (0.87-1.86) | 0.221 | ||
| No | ||||
| Major hepatectomy | ||||
| Yes | 1.35 (1.02-1.80) | 0.038 | 1.14 (0.84-1.56) | 0.408 |
| No | ||||
| BRAF status | ||||
| Mutant | 1.69 (0.83-3.45) | 0.149 | ||
| Wild | ||||
| RAS/TP53 status | ||||
| Co-mutant | 1.97 (1.48-2.64) | <0.001 | 1.69 (1.22-2.34) | 0.001 |
| Others | ||||
| SMAD4 status | ||||
| Mutant | 1.88 (1.31-2.70) | <0.001 | 2.44 (1.65-3.63) | <0.001 |
| Wild | ||||
| FBXW7 status | ||||
| Mutant | 1.37 (0.81-2.32) | 0.245 | ||
| Wild | ||||
| Surgical Margin Status | ||||
| < 1.0 mm | Reference | Reference | ||
| 1.0 - 4.9 mm | 0.7 (0.47-1.06) | 0.091 | 0.83 (0.52-1.31) | 0.427 |
| 5.0 - 9.9 mm | 0.98 (0.66-1.45) | 0.908 | 1.51 (0.98-2.33) | 0.063 |
| ≥ 10 mm | 0.61 (0.40-0.92) | 0.018 | 0.88 (0.60-1.40) | 0.600 |
| Pattern of Recurrence | ||||
| Local recurrence | Reference | Reference | ||
| Other intrahepatic recurrence | 0.65 (0.39-1.07) | 0.089 | 0.62 (0.37-1.03) | 0.066 |
| Extrahepatic recurrence alone | 0.28 (0.16-0.50) | <0.001 | 0.23 (0.12-0.42) | <0.001 |
| No recurrence | 0.08 (0.03-0.19) | <0.001 | 0.11 (0.05-0.28) | <0.001 |
Abbreviations: HR, hazard ratio; LN, lymph node; CLM, colorectal liver metastases; 5-FU, 5-Fluorouracil; EGFR, epidermal growth factor receptor.
DISCUSSION
In this study, local recurrence was confirmed in 6.9% of patients who underwent R0-intent resection for CLM, which was consistent with previous reports.3, 4 While RAS/TP53 co-mutation was significantly associated with higher incidence of intrahepatic and overall recurrence, neither surgical margin status nor somatic mutation status was associated with the incidence of local recurrence. Pathological tumor viability ≥ 50%, RAS/TP53 co-mutation, and SMAD4 mutation were identified as independent risk factors for poor OS. Patients with early recurrence were more likely to have local recurrence than other intrahepatic recurrence; however, there was no difference in OS between patients with local and other intrahepatic recurrence. This is the first study focusing on the impact of detailed genetic profile analyzed by NGS on the incidence of local recurrence after R0-intent resection for CLM. These results suggest that surgeons performing R0-intent hepatectomy should not adjust surgical margin width based on tumor biology in an effort to reduce the risk of local recurrence.
Previous studies have reported that resection margin < 1 mm in patients undergoing hepatectomy for CLM could lead to a higher risk of local recurrence as well as poor postoperative survival.3, 4 However, most of these studies were published over 10 years ago, in an era predating the use of perioperative modern chemotherapy and advanced precise parenchymal resection technique. Our data would suggest that surgical margin status or width does not affect the rates of local recurrence or other hepatic recurrence after R0 intent hepatectomy, but caution must be advised when applying these findings to other practices where perioperative chemotherapy and precise parenchymal transection techniques are not used.
Similarly, given that all the patients in our study underwent R0-intent resection, it is important to note that our findings do not justify performing an R1-intent resection for CLM. In a recent report by Ardito et al., R1 resection was associated with a higher risk of local recurrence and poor postoperative survival in the era of modern chemotherapy.7 However, the reported rate of local recurrence in their study (12.8%) was higher than the rates across other reports, including ours.8 Their study included patients with a risk of R1 resection anticipated preoperatively, which may have led to these divergent outcomes and conclusions. It is also important to distinguish R1 vascular resection from R1 parenchymal resection. Viganò et al. proposed that R1 vascular resection resulted in a much lower risk of local recurrence (4.3%) than R1 parenchymal resection (19.6%),11 which could also suggest that R1-intent resection, particularly at a parenchymal transection line, may cause a much higher risk of local recurrence than R0-intent resection.
An impact of tumor somatic mutations on surgical margin width has also been suggested in previous studies; however, our study showed no impact of RAS/TP53 co-mutation or BRAF, SMAD4, or FBXW7 mutations on either surgical margin width or the incidence of local recurrence. Our group previously reported that RAS mutation in CLM was a predictor of closer surgical margin,15 but we did not analyze the impact of RAS mutation on local recurrence. In addition, in our previous study, only patients who had intrahepatic recurrence were included, and the effect of TP53 gene mutation was not considered. RAS mutation in CLM has been reported to be a predictor of micrometastases surrounding CLM,17 and studies have indicated conflicting results regarding whether the prognostic effect of resection margin may differ by RAS mutation status.10, 16 However, none of these studies reported on local recurrence. Additionally, RAS/TP53 co-mutation and SMAD4 mutation have recently been reported to be better prognostic factors than RAS mutation alone.18-20 Our current study suggests that recurrence site and survival are driven by tumor biology rather than surgical margins.
In our analysis of local recurrence, we found that local recurrences occurred earlier than other intrahepatic recurrences, but this did not translate into a difference in OS between patients with local and other intrahepatic recurrences. Our group previously reported the safety and efficacy of repeat hepatectomy for recurrent CLM and the negative prognostic impact of RAS mutation in patients who underwent repeat hepatectomy.29 Our previous findings suggest that early detection and liver-directed therapy for recurrent disease could offer favorable outcome irrespective of pattern of hepatic recurrence, while OS mainly depends on tumor biology. Additionally, in our study reported here, the analysis for OS stratified by the pattern of recurrence showed that patients with hepatic recurrence had significantly worse OS than patients with extrahepatic recurrence irrespective of the site of hepatic recurrence. These findings reinforce the importance of hepatic control of CLM and justify resection of CLM in patients with unresectable or unresected concurrent lung metastases, as Mise et al. suggested.30
Limitations of this study included its retrospective nature and the lack of external validation. However, this study was based on a prospectively collected database that included a large number of patients who underwent liver resection with similar surgical technique and standardized pathological evaluation. Additionally, the observation period of this study was longer than that for previous studies investigating the prognostic impact of resection margin status. The tumor biology, including the tumor viability after chemotherapy and somatic gene alterations, could be considered in selecting the therapeutic approach in marginally resectable recurrence disease or patients with short recurrence-free survival, and this may affect overall survival. Finally, there is no difference in surgical approach/decision between patients who enrolled in the current study and those who were excluded. However, comprehensive multigene testing was not commonly performed before 2012, and there is possibly a selection bias because of the time trend.
CONCLUSION
Neither surgical margin width nor somatic mutations affected the risk of local recurrence after R0-intent hepatectomy for CLM, although RAS/TP53 co-mutation was associated with higher risk of intrahepatic and overall recurrence, as well as worse OS. Recurrence and prognosis were likely driven by individual tumor biology rather than surgical margins.
Supplementary Material
Supplemental Figure 1. Patient selection. CLM, colorectal liver metastases; NGS, next-generation sequencing; CT, computed tomography.
Conflicts of Interest and Source of Funding:
This article was supported in part by the National Institutes of Health through grant T32 CA 009599 and grant P30CA016672, which supports the MD Anderson Cancer Center Clinical Trials Office.
Footnotes
This study was presented at the presidential plenary session of the 62nd annual meeting of the Society for Surgery of the Alimentary Tract on May 21st, 2021.
REFERENCES
- 1.Kopetz S, Vauthey JN. Perioperative chemotherapy for resectable hepatic metastases. Lancet (London, England). 2008;371(9617):963–5. doi: 10.1016/s0140-6736(08)60429-8. [DOI] [PubMed] [Google Scholar]
- 2.Vigano L, Costa G, Cimino MM, Procopio F, Donadon M, Del Fabbro D et al. R1 Resection for Colorectal Liver Metastases: a Survey Questioning Surgeons about Its Incidence, Clinical Impact, and Management. Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract. 2018;22(10):1752–63. doi: 10.1007/s11605-018-3820-z. [DOI] [PubMed] [Google Scholar]
- 3.Kokudo N, Miki Y, Sugai S, Yanagisawa A, Kato Y, Sakamoto Y et al. Genetic and histological assessment of surgical margins in resected liver metastases from colorectal carcinoma: minimum surgical margins for successful resection. Archives of surgery (Chicago, Ill : 1960). 2002;137(7):833–40. doi: 10.1001/archsurg.137.7.833. [DOI] [PubMed] [Google Scholar]
- 4.Pawlik TM, Scoggins CR, Zorzi D, Abdalla EK, Andres A, Eng C et al. Effect of surgical margin status on survival and site of recurrence after hepatic resection for colorectal metastases. Annals of surgery. 2005;241(5):715–22, discussion 22–4. doi: 10.1097/01.sla.0000160703.75808.7d. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Holdhoff M, Schmidt K, Diehl F, Aggrawal N, Angenendt P, Romans K et al. Detection of tumor DNA at the margins of colorectal cancer liver metastasis. Clinical cancer research : an official journal of the American Association for Cancer Research. 2011;17(11):3551–7. doi: 10.1158/1078-0432.CCR-10-3087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hamady ZZ, Lodge JP, Welsh FK, Toogood GJ, White A, John T et al. One-millimeter cancer-free margin is curative for colorectal liver metastases: a propensity score case-match approach. Annals of surgery. 2014;259(3):543–8. doi: 10.1097/SLA.0b013e3182902b6e. [DOI] [PubMed] [Google Scholar]
- 7.Ardito F, Panettieri E, Vellone M, Ferrucci M, Coppola A, Silvestrini N et al. The impact of R1 resection for colorectal liver metastases on local recurrence and overall survival in the era of modern chemotherapy: An analysis of 1,428 resection areas. Surgery. 2019;165(4):712–20. doi: 10.1016/j.surg.2018.09.005. [DOI] [PubMed] [Google Scholar]
- 8.Andreou A, Knitter S, Schmelzle M, Kradolfer D, Maurer MH, Auer TA et al. Recurrence at surgical margin following hepatectomy for colorectal liver metastases is not associated with R1 resection and does not impact survival. Surgery. 2020. doi: 10.1016/j.surg.2020.11.024. [DOI] [PubMed] [Google Scholar]
- 9.Truant S, Sequier C, Leteurtre E, Boleslawski E, Elamrani M, Huet G et al. Tumour biology of colorectal liver metastasis is a more important factor in survival than surgical margin clearance in the era of modern chemotherapy regimens. HPB : the official journal of the International Hepato Pancreato Biliary Association. 2015;17(2):176–84. doi: 10.1111/hpb.12316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Margonis GA, Sasaki K, Kim Y, Samaha M, Buettner S, Amini N et al. Tumor Biology Rather Than Surgical Technique Dictates Prognosis in Colorectal Cancer Liver Metastases. Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract. 2016;20(11):1821–9. doi: 10.1007/s11605-016-3198-8. [DOI] [PubMed] [Google Scholar]
- 11.Viganò L, Procopio F, Cimino MM, Donadon M, Gatti A, Costa G et al. Is Tumor Detachment from Vascular Structures Equivalent to R0 Resection in Surgery for Colorectal Liver Metastases? An Observational Cohort. Annals of surgical oncology. 2016;23(4):1352–60. doi: 10.1245/s10434-015-5009-y. [DOI] [PubMed] [Google Scholar]
- 12.Vauthey JN, Zimmitti G, Kopetz SE, Shindoh J, Chen SS, Andreou A et al. RAS mutation status predicts survival and patterns of recurrence in patients undergoing hepatectomy for colorectal liver metastases. Annals of surgery. 2013;258(4):619–26; discussion 26–7. doi: 10.1097/SLA.0b013e3182a5025a. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Odisio BC, Yamashita S, Huang SY, Harmoush S, Kopetz SE, Ahrar K et al. Local tumour progression after percutaneous ablation of colorectal liver metastases according to RAS mutation status. The British journal of surgery. 2017;104(6):760–8. doi: 10.1002/bjs.10490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Calandri M, Yamashita S, Gazzera C, Fonio P, Veltri A, Bustreo S et al. Ablation of colorectal liver metastasis: Interaction of ablation margins and RAS mutation profiling on local tumour progression-free survival. European radiology. 2018;28(7):2727–34. doi: 10.1007/s00330-017-5273-2. [DOI] [PubMed] [Google Scholar]
- 15.Brudvik KW, Mise Y, Chung MH, Chun YS, Kopetz SE, Passot G et al. RAS Mutation Predicts Positive Resection Margins and Narrower Resection Margins in Patients Undergoing Resection of Colorectal Liver Metastases. Annals of surgical oncology. 2016;23(8):2635–43. doi: 10.1245/s10434-016-5187-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Xu D, Wang HW, Yan XL, Li J, Wang K, Xing BC. Sub-millimeter surgical margin is acceptable in patients with good tumor biology after liver resection for colorectal liver metastases. European journal of surgical oncology : the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology. 2019;45(9):1551–8. doi: 10.1016/j.ejso.2019.03.010. [DOI] [PubMed] [Google Scholar]
- 17.Zhang Q, Peng J, Ye M, Weng W, Tan C, Ni S et al. KRAS Mutation Predicted More Mirometastases and Closer Resection Margins in Patients with Colorectal Cancer Liver Metastases. Annals of surgical oncology. 2019. doi: 10.1245/s10434-019-08065-5. [DOI] [PubMed] [Google Scholar]
- 18.Chun YS, Passot G, Yamashita S, Nusrat M, Katsonis P, Loree JM et al. Deleterious Effect of RAS and Evolutionary High-risk TP53 Double Mutation in Colorectal Liver Metastases. Annals of surgery. 2019;269(5):917–23. doi: 10.1097/sla.0000000000002450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kawaguchi Y, Lillemoe HA, Panettieri E, Chun YS, Tzeng CD, Aloia TA et al. Conditional Recurrence-Free Survival after Resection of Colorectal Liver Metastases: Persistent Deleterious Association with RAS and TP53 Co-Mutation. Journal of the American College of Surgeons. 2019;229(3):286–94.e1. doi: 10.1016/j.jamcollsurg.2019.04.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kawaguchi Y, Kopetz S, Newhook TE, De Bellis M, Chun YS, Tzeng CD et al. Mutation Status of RAS, TP53, and SMAD4 is Superior to Mutation Status of RAS Alone for Predicting Prognosis after Resection of Colorectal Liver Metastases. Clinical cancer research : an official journal of the American Association for Cancer Research. 2019;25(19):5843–51. doi: 10.1158/1078-0432.Ccr-19-0863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lang H, Baumgart J, Heinrich S, Tripke V, Passalaqua M, Maderer A et al. Extended Molecular Profiling Improves Stratification and Prediction of Survival After Resection of Colorectal Liver Metastases. Annals of surgery. 2019;270(5):799–805. doi: 10.1097/sla.0000000000003527. [DOI] [PubMed] [Google Scholar]
- 22.Datta J, Smith JJ, Chatila WK, McAuliffe JC, Kandoth C, Vakiani E et al. Coaltered Ras/B-raf and TP53 Is Associated with Extremes of Survivorship and Distinct Patterns of Metastasis in Patients with Metastatic Colorectal Cancer. Clinical cancer research : an official journal of the American Association for Cancer Research. 2020;26(5):1077–85. doi: 10.1158/1078-0432.ccr-19-2390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kawaguchi Y, Newhook TE, Tran Cao HS, Tzeng CD, Chun YS, Aloia TA et al. Alteration of FBXW7 is Associated with Worse Survival in Patients Undergoing Resection of Colorectal Liver Metastases. Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract. 2021;25(1):186–94. doi: 10.1007/s11605-020-04866-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Loree JM, Pereira AAL, Lam M, Willauer AN, Raghav K, Dasari A et al. Classifying Colorectal Cancer by Tumor Location Rather than Sidedness Highlights a Continuum in Mutation Profiles and Consensus Molecular Subtypes. Clinical cancer research : an official journal of the American Association for Cancer Research. 2018;24(5):1062–72. doi: 10.1158/1078-0432.ccr-17-2484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kawaguchi Y, Kopetz S, Panettieri E, Hwang H, Wang X, Tran Cao HS et al. Improved Survival Over Time after Resection of Colorectal Liver Metastases and Clinical Impact of Multigene Alteration Testing in Patients with Metastatic Colorectal Cancer. Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract. 2021;in press. [DOI] [PubMed] [Google Scholar]
- 26.Passot G, Chun YS, Kopetz SE, Zorzi D, Brudvik KW, Kim BJ et al. Predictors of Safety and Efficacy of 2-Stage Hepatectomy for Bilateral Colorectal Liver Metastases. Journal of the American College of Surgeons. 2016;223(1):99–108. doi: 10.1016/j.jamcollsurg.2015.12.057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Strasberg SM, Belghiti J, Clavien PA, Gadzijev E, Garden JO, Lau WY et al. The Brisbane 2000 Terminology of Liver Anatomy and Resections. Hpb. 2000;2(3):333–9. doi: 10.1016/s1365-182x(17)30755-4. [DOI] [Google Scholar]
- 28.Aloia TA, Zorzi D, Abdalla EK, Vauthey JN. Two-surgeon technique for hepatic parenchymal transection of the noncirrhotic liver using saline-linked cautery and ultrasonic dissection. Annals of surgery. 2005;242(2):172–7. doi: 10.1097/01.sla.0000171300.62318.f4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Denbo JW, Yamashita S, Passot G, Egger M, Chun YS, Kopetz SE et al. RAS Mutation Is Associated with Decreased Survival in Patients Undergoing Repeat Hepatectomy for Colorectal Liver Metastases. Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract. 2017;21(1):68–77. doi: 10.1007/s11605-016-3189-9. [DOI] [PubMed] [Google Scholar]
- 30.Mise Y, Kopetz S, Mehran RJ, Aloia TA, Conrad C, Brudvik KW et al. Is complete liver resection without resection of synchronous lung metastases justified? Annals of surgical oncology. 2015;22(5):1585–92. doi: 10.1245/s10434-014-4207-3. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Figure 1. Patient selection. CLM, colorectal liver metastases; NGS, next-generation sequencing; CT, computed tomography.