The Impact of Margins on Outcome After Hepatic Resection for Colorectal Metastasis

Abstract

Objective:

To analyze the impact of margin width on long-term outcome after hepatic resection for colorectal metastasis.

Summary Background Data:

The optimal margin width and its influence on long-term outcome after hepatic resection for colorectal metastasis are unclear.

Methods:

All patients undergoing hepatic resection for colorectal metastasis from 1991 to 2003 were identified, and the prognostic influence of margin width and other clinicopathologic factors were analyzed.

Results:

A total of 1019 patients with a clear description of margin width were included. Analysis of margin width as a continuous variable suggested the following grouping: group I, involved (n = 112, 11%); group II, <1–10 mm (n = 563, 55%); and group III, >10 mm (n = 344, 33.7%). On univariate analysis, there was a statistically significant difference in median survival between all 3 groups: group II versus group I (42 vs. 30 months, P < 0.01) and group III versus group II (55 vs. 42 months, P < 0.01). Margin width >1 cm retained statistical significance (P < 0.01) on multivariate analysis after adjusting for established risk factors. After adjustment, survival in group III was significantly better than either group I or II (P < 0.01), but there was no difference between groups I and II (P = 0.31).

Conclusions:

This study provides evidence that margin width of >1 cm is optimal and is an independent predictor of survival after hepatic resection for colorectal metastasis. However, subcentimeter resections are also associated with favorable outcome and should not preclude patients from undergoing resection.

The prognostic significance of margin width on long-term survival after hepatic resection for colorectal metastasis is unclear. The results of this study demonstrate that a >1 cm margin is an independent predictor of improved outcome when analyzed in the context of other well-known prognostic factors.

Hepatic resection is the only treatment modality associated with long-term survival in patients with colorectal liver metastasis. Five-year survival rates ranging from 25% to 58% have been reported with hepatic resection.^1–3 With improved surgical techniques and chemotherapy, more patients are subjected to resection, including patients previously considered unresectable. Several clinicopathologic factors have been analyzed and are now considered to be independent prognostic factors associated with survival.^4–13 The role of resection margin status as a prognostic factor to predict survival after hepatic resection for colorectal metastasis is controversial. In addition, the optimal margin width, if any, is unclear.

Early documentation by some reports of a potential benefit associated with a 1-cm margin led to the general acceptance of this margin width to guide hepatic resections.¹⁴ Since that time, several studies addressing the issue of margin width have yielded conflicting results. Some authors have shown that a margin width >1 cm is associated with improved survival when compared with subcentimeter resections.^4,15–19 In contrast, other reports have documented that the margin width does not have any effect on survival as long as it is negative,^2,20,21 whereas some studies have noted that although a >1 cm margin is desirable, survival is not influenced by the width of the subcentimeter margin (0–9 mm).^3,22 Firm conclusions from these studies are limited by small numbers of patients and limited statistical power. The only reports with a large number of patients were multi-institutional studies,^2,7,11 in which institutional differences in the method of liver transection and pathologic analysis of the margin may confound the results. From these conflicting studies, the prognostic significance of margin status and the optimal margin width remain unclear. This is all the more important, since an increasing number of complex resections are performed in which the extent of hepatic involvement frequently mandates close resection margins.

The aim of this study was to determine the impact of margin width on long-term survival after hepatic resection for colorectal metastasis in a large, consecutive series of patients from a single institution. Second, the study hopes to determine the influence of margin width on survival in the context of well-established clinicopathologic prognostic factors.

PATIENTS AND METHODS

All patients who underwent hepatic resection for colorectal metastasis from 1991 to 2003 were entered into a prospective database and included in the study. The study was approved by the Institutional Review Board and conformed to the Health Insurance Portability and Accountability Act. Data pertaining to the primary colorectal tumor (site and nodal status), hepatic metastasis (size, number, prehepatectomy CEA level, disease-free interval, and synchronous or metachronous presentation), operative details (type and extent of hepatic resection), and histopathology (margin width in millimeter) were collected. Patients underwent cross-sectional imaging of the abdomen with contrast-enhanced computed tomography (CT) and/or magnetic resonance imaging (MRI). The chest was imaged with x-rays or with CT. All patients had a colonoscopy within 1 year of liver resection. The operations were performed by 5 different surgeons and their approach to hepatic resection for colorectal metastasis has been previously described.^1,23 Intraoperative ultrasound of the liver was carried out in all patients. A central venous pressure less than 5 mm Hg was maintained during parenchymal transection and monitored by central venous access. Anatomic, segmental-based resections were performed whenever feasible. In general, the liver was transected with the “Kelly clamp” crush technique.

Since 1991, uniform standards in the final histopathologic description of liver tumors were established at our institution and have been followed since. This includes number of tumors, size of the largest tumor, and a numerical description of the margin width in millimeters. Patients without an accurate numerical description of the margin width were excluded from the analysis. In addition, patients in whom the margins could not be assessed by the pathologist due to the effect of the method of resection (such as cryoassisted resection) were also excluded. In the remaining patients, the distance (in millimeter) of the lesion closest to the transected liver surface was judged to be the final margin width. In the case of multiple tumors, the closest margin was recorded as the final margin.

Margin status (categorized) was compared with other categorical variables using the χ² test and continuous variables using the t test. Survival was measured from time of resection to death or last follow-up. Categorization of the margin width was based on a plot of median survival versus margin width. Survival probabilities were estimated using the Kaplan-Meier method. Individual groups were compared using log-rank test and the multivariate analysis used stepwise Cox regression.

RESULTS

Demographics

A total of 1119 patients underwent hepatic resection for colorectal metastasis from 1991 to 2003. Of these, 1019 (91%) had a clear numerical description of the margin width (in millimeter) and were included in the study. One hundred patients (9%) were excluded due to the absence of a clear numerical description of margin width or the inability to assess margins secondary to the method of liver transaction. The median age was 63 years (range, 23–89) and 571 patients (56%) were male. The 30-day operative mortality rate was 2%, and the incidence of postoperative morbidity was 44%. With a median follow-up of 42 months for the entire cohort of 1019 patients, the median survival was 44 months and overall 1-, 3-, and 5-year survival rates were 90%, 59%, and 37%, respectively.

Clinicopathologic Characteristics of the Primary and Metastatic Tumors

The primary tumor was in the colon in 850 patients (76%). In 355 patients (34%), the presentation of metastasis was synchronous with the primary and in 564 patients (55%) the disease-free interval was less than 12 months. Five hundred ninety patients (58%) had involved regional lymph nodes associated with the primary tumor. The mean number of hepatic metastases was 2.4 ± 2.3 and 506 patients (50%) had ≥2 tumors. The mean size of the largest hepatic metastasis was 4.8 ± 3.2 cm, and 679 patients (67%) had tumors measuring <5 cm. Preoperative chemotherapy was administered to 302 patients (33%) before liver resection.

Operative Details

Six hundred and seventy-eight (66.5%) patients underwent a hemihepatectomy and of these, 287 (28%) were subjected to an extended hepatectomy (trisegmentectomy). In 441 patients (43%), hepatic resection involved both lobes. The average number of segments resected was 3.5 ± 1.5, and the distribution of segmental resection was as follows: 1 segment 13%; bisegmentectomy 19%; 3 segments 8%; 4 and above 60%.

Results With Historical Margin Width Criteria

Based on previous publications, the margin width was initially analyzed at a 1 cm cut-off, with subcentimeter margins arbitrarily divided into 4 categories: involved, <1 mm, 1–5 mm, and 5–10 mm. With increasing margin width, the survival rates increased proportionately when compared with that of patients in the reference group (involved margin) (Table 1). When compared with patients with involved margins, this increase in survival was not statistically significant in patients with 1–10 mm margins and assumed statistical significance at >1 cm (P < 0.01).

TABLE 1. Initial Analysis of the Impact of Margins on Survival (n = 1019)

Analysis of Margin Width as a Continuous Variable

The influence of margins on survival was then analyzed as a continuous variable. This analysis revealed a survival benefit that increased incrementally with the increasing width of the margin with a plateau effect in the later part of the curve (Fig. 1). In addition, there were 2 inflection points on the curve that were associated with more significant differences in survival. The 2 significant cut-off points were at approximately 0 and 10 mm. Based on these 2 cut-off points (as suggested by the statistical analysis), the patients were divided into 3 groups: groups I (involved), II (<1–10 mm), and III (>10 mm), respectively.

FIGURE 1. Impact of margin width on survival analyzed as a continuous variable.

The influence of margin width on overall survival in patients in these 3 groups was analyzed and is graphically displayed in Figure 2. On univariate analysis, there was a statistically significant difference in survival between all 3 groups. Patients in group II had an improved median survival when compared with those in group I (42 vs. 30 months, P < 0.01). Similarly, patients in group III had a better median survival when compared with those in both groups II and I (55 vs. 42 and 30 months, P < 0.01) (Table 2).

FIGURE 2. Kaplan–Meier survival plot comparing 3 different margin width groupings.

TABLE 2. Impact of Margins on Survival Based on Grouping Suggested by Analysis of Margins as a Continuous Variable

On univariate analysis, several other factors were associated with worse long-term survival and are summarized in Table 3. These include resection greater than a hemihepatectomy, bilateral resection, resection of ≥2 hepatic tumors, size >5 cm, perioperative transfusion, preoperative CEA >200 ng/mL, node-positive primary, and margin width. On multivariate analysis, bilateral resection, resection of ≥2 hepatic tumors, size >5 cm, perioperative transfusion, node-positive primary and margin width >1 cm (group III) retained significance as independent predictors of outcome. On multivariate analysis, margin width was a significant factor when comparing group III with groups I and II (P < 0.01), but there was no significance between groups II and I (P = 0.31).

TABLE 3. Univariate and Multivariate Analysis of the Impact of Margins and Other Clinicopathologic Factors on Survival

Factors Associated With Close Margins

Factors potentially associated with a margin width of <1 cm were also analyzed and are summarized in Table 4. The factors associated with a narrower margin width were bilateral resection and the presence of ≥2 hepatic tumors. Size >5 cm and CEA >200 ng/mL were marginally associated with closer margins (P = 0.09 and 0.08). Notably, the use of preoperative chemotherapy was not associated with wider margins.

TABLE 4. Influence of Tumor-Related and Technical Factors on Margin Width

DISCUSSION

Several factors have been documented to be independent prognostic factors associated with outcome after hepatic resection for colorectal metastasis.1,3,5,8,14 These include nodal status of the primary colorectal tumor, disease-free interval, size of the largest hepatic tumor, prehepatectomy CEA level, number and distribution of metastatic lesions, and extent of resection. The role of margin status as an independent predictor of survival and the optimal margin width after hepatic resection for colorectal metastasis is unclear. This is compounded by the fact that there is a paucity of reports that have addressed this issue,2,4,14–18,20,21 most of which are limited by small numbers of patients and limited statistical power.4,16,17,20,21 The studies that included adequate numbers of patients were multi-institutional,^2,15 and in such studies, institutional differences in hepatic transection techniques and histologic analysis of margins make interpretation of results difficult.

Despite the lack of firm evidence, there has been a general consensus since the early 1980s that a 1 cm margin is optimal, with some authors suggesting that inability to obtain such a margin should preclude hepatic resection.^7,14,19 For example, in their series of 72 resections, Ekberg et al concluded that, due to the poor outcomes associated with subcentimeter resections, liver resection should not be performed if a margin width of 1 cm is unobtainable.¹⁴ Since then, most studies that have assessed the impact of margins relied on this 1 cm margin cut-off for analysis and have reported conflicting results.2,3,17,21 In their series of 112 patients, Wray et al found that subcentimeter resections were associated with increased local and distant recurrence, and concluded that a >1 cm margin should be attempted in all cases.¹⁷ Conversely, by assessing specimens for K-ras and p53 mutations in 58 patients with colorectal liver metastasis, Kokudo et al²¹ noted that micrometastasis are very uncommon and most are confined to within <2 mm of the tumor. On the basis of these findings, the authors concluded that a 1 cm margin is not essential.

It is critical to analyze the effect of margin status on survival in the context of well-established clinicopathologic factors in a multivariate analysis. This is the best way to ensure that margin status, as interpreted by the pathologist, is not simply a surrogate for other parameters that impact on survival. Shirabe et al¹⁶ found that a resection margin of <10 mm was independently associated with poor prognosis on multivariate analysis; however, their study consisted of only 31 patients with limited statistical power to support such conclusions. In contrast, other reports have found that the margin width does not influence long-term outcome on multivariate analysis.^2,3 Scheele et al³ analyzed the effect of margins in 414 patients who underwent complete gross resection. Forty-eight (12%) of these 414 patients had positive margins and were not included in the analysis. Margin width had an influence on overall and disease-free survival on univariate analysis, but not on multivariate analysis. It is possible, however, that if the patients with positive margins (and likely the worst outcomes) were also included for comparison, there may have been an independent survival benefit associated with margin width. In a multicenter retrospective study of 557 patients, Pawlik et al² found that a positive margin was associated with an increased rate of surgical margin recurrence and survival on univariate analysis. Interestingly, outcome was the same for patients with negative margins regardless of margin width. When subjected to multivariate analysis, margin status was not an independent predictor of outcome. Definitive conclusions from this study are limited by the relatively small number of patients with positive margins (n = 45) and the potentially confounding issue of different parenchymal transection techniques and histologic analysis in the different institutions.

These conflicting results leave the question of margins unanswered. The aim of this study, therefore, was to determine whether margin status is an independent predictor of long-term outcome following hepatic resection for colorectal metastasis in a large consecutive series of patients from a single institution with a uniform approach to parenchymal transection (Kelly clamp crush technique) and strict standards of histopathologic analysis.

In contrast to the previous reports, we analyzed the influence of margin width on survival as a continuous variable. This analysis showed an incremental increase in survival as the margin width increased with 2 inflection points on the curve at approximately 0 and 10 mm. It is interesting to note a small, but statistically insignificant, increase in survival beyond 10 mm with an ultimate plateau at 25 mm. On the basis of this analysis, the patients in our study were therefore divided into 3 groups based on this statistical suggestion (involved, <1–10 mm, and >10 mm). On univariate analysis, there was a statistically significant difference in survival between all 3 groups. On multivariate analysis, however, patients with a >1 cm margin had a statistically significant improvement in long-term outcome when compared with the remaining patients (P < 0.01). Although there was a difference in survival between subcentimeter resections and positive margins, this did not reach significance on multivariate analysis (P = 0.31). This analysis confirmed the independent prognostic importance of a 1 cm margin in the context of other well-established clinicopathologic prognostic factors. It is important to interpret these results in the context of our parenchymal transection technique. The Kelly clamp crush technique may create a margin of different character than other well-known techniques and the margin, as determined by the pathologist, may have different clinical relevance. Nonetheless, the linear and independent relationship between margin and survival is strong evidence that the margin width is an important factor.

Our study is the only report to date to document an association of a 1 cm margin width with improved long-term outcome on multivariate analysis in a large series of patients. The large number of patients provided significant statistical power and afforded us the ability to detect relatively small differences, and it is important to interpret these data carefully. Only 9% of patients were excluded from this analysis largely due to inadequate pathologic descriptions or, less commonly, techniques in which the margin was not measurable (such as cryoassisted resection). It is unlikely that these patients would have altered the outcome since they are a small percentage of the overall group and are generally related to the earlier time period in which pathologic reporting standards were changing at our institution. It must be emphasized that, even in patients with subcentimeter resections, the long-term survival is still favorable (median survival of 30–42 months). Although this is inferior to patients with >1 cm margins, it is better than in patients who do not undergo hepatic resection or are unresectable for any reason.^24–26 The outcome after subcentimeter resection is also better than any other single modality of treatment that is currently available, including the current chemotherapy regimens (median survival of 20–22 months).^27–29 Furthermore, chemotherapy “alone” rarely produces 5-year survivors or durable complete responders.^27–29 Therefore, inability to obtain >1 cm margin based on preoperative imaging should not preclude hepatic resection.

Although the authors’ general strategy is to achieve >1 cm margins, this was not possible in all cases due to anatomic constraints or simply misjudgment of the actual margin. In our study, we found the only factors that were associated with a decrease in the rate of obtaining >1 cm margin were bilateral resection and resection of ≥2 hepatic tumors. This finding is logical since in many of these patients with multiple bilobar tumors, wide margins are not technically possible. Tumor size >5 cm and CEA level >200 ng/mL were associated with a reduction in the rate of >1 cm margins, but this did not reach statistical significance. It has been suggested that administration of preoperative chemotherapy can increase the likelihood of clear margins.³⁰ In a series of 108 patients, Parikh et al³⁰ found that the 61 patients who were administered 5FU based chemotherapy with or without CPT 11 tended to have a higher rate of negative margins that did not reach statistical significance. We did not find that preoperative chemotherapy increased the incidence of >1 cm margins in the 302 patients that received it.

In summary, this study provides evidence that margin status is an independent predictor of long-term outcome after hepatic resection for colorectal metastasis. A margin width of >1 cm is independently associated with the best long-term outcome and should be attempted whenever possible. Despite the statistical difference, subcentimeter resections are also associated with a favorable outcome when compared with any other single modality of treatment that is currently available. In patients undergoing hepatic resection for colorectal metastasis, a >1 cm margin should be attempted whenever possible. However, inability to achieve this should not preclude hepatic resection.