- Email: [email protected]
Survival after Liver Resection for Metastatic Colorectal Carcinoma in a Large Population
Survival after Liver Resection for Metastatic Colorectal Carcinoma in a Large Population Shimul A Shah, MD, Rudy Bromberg, MD, Angela Coates, MEd, Eddy Rempel, BMath, Marko Simunovic, MD, Steven Gallinger, MD Previous reports of liver resection for metastatic colorectal cancer (CRC) are typically from single centers and cannot account for selection or referral bias. We measured longterm survival after liver resection for metastatic CRC in the province of Ontario, Canada (population 12 million). STUDY DESIGN: The Ontario Cancer Registry is an administrative database that links all hospital records, pathology reports, and vital statistics for patients with a diagnosis of cancer. We used the Registry to identify and obtain information on all patients who underwent liver resection for metastatic CRC in calendar years 1996 to 2004. Pathology reports of the original CRC resection and subsequent liver resections were individually reviewed. RESULTS: Eight hundred forty-one resections were performed at 43 centers across Ontario during the 9-year period, including wedge resection (n ⫽ 303; 36%); lobectomy (n ⫽ 466; 55%); and trisectionectomy (n ⫽ 72; 9%). Ninety-one percent and 54% of resections were performed at teaching and high-volume centers (⬎ 80 resections), respectively. Most liver resections were performed more than 120 days after original CRC operation (672 of 841; 80%). Perioperative mortality was 3%. Unadjusted 1-, 3-, and 5-year survival after liver resection was 88%, 59%, and 43%, respectively. Survival was improved when resection was performed for fewer than 2 tumor nodules, at high-volume centers, or in the years 2001 to 2004. CONCLUSIONS: Results in this population-based series are consistent with those of single-hospital series assessing longterm survival after liver resection for metastatic CRC. These findings support continued efforts to aggressively identify and resect CRC liver metastases. (J Am Coll Surg 2007;205: 676–683. © 2007 by the American College of Surgeons) BACKGROUND:
gundy, Melbourne, Auckland, and Caen.12-15 We are aware of only two studies that measured longterm survival after metastatic CRC liver resection in a large population.14,16 Numerous studies from US state and national databases have confirmed that hepatectomy for a variety of conditions has a low operative mortality rate.11,17-19 Increased hospital volume for advanced surgical oncology procedures appears to correlate with improved operative mortality, confirming a positive volume-to-outcomes relationship.19-25 Using the National Medicare database, Fong and colleagues25 found improved perioperative mortality (4% versus 9%) for liver resections performed for cancer at high-volume centers, defined as ⬎ 25 patients/ year. But they were unable to show a substantial improvement in longterm survival with liver resections performed at high-volume centers, indicating that the major effect of volume was on perioperative outcomes.25 The positive influence of hospital volume on longterm outcomes after operation has been shown in other cancer types, including breast, colon, and lung cancer.26-28 The number of liver resection performed for all indications is increasing in the
Hematogenous spread to the liver occurs in 40% to 60% of patients with colorectal cancer (CRC).1 Latest advances in medical and surgical treatments of liver masses have improved the diagnostic accuracy and surgical options for patients with metastatic CRC. This has resulted in 30% to 50% 5-year survival after resection for liver metastases.2-11 Most reports are from single institutions and cannot account for selection or referral bias. Registry studies focused on metastatic CRC have only been reported from BurCompeting Interests Declared: None. Presented at the 2007 Academic Surgical Congress, Phoenix, AZ, February 2007. Received April 28, 2007; Revised May 24, 2007; Accepted June 6, 2007. From the Department of Surgery, University of Massachusetts Medical School, Worcester, MA (Shah); University Health Network, University of Toronto, Toronto, Canada (Shah, Bromberg, Gallinger); and Hamilton Health Sciences, McMaster University, Hamilton, Canada (Coates, Rempel, Simunovic). Correspondence address: Shimul A Shah, MD, Division of Organ Transplantation, Department of Surgery, University of Massachusetts Medical School, 55 Lake Ave N, S3-838, Worcester, MA 01655. email: [email protected]
© 2007 by the American College of Surgeons Published by Elsevier Inc.
676
ISSN 1072-7515/07/$32.00 doi:10.1016/j.jamcollsurg.2007.06.283
Vol. 205, No. 5, November 2007
Shah et al
US,17 but we are unaware of any studies that have measured the relationship between hospital volume and patient longterm survival after hepatectomy for metastatic colorectal cancer. We aimed to assess longterm survival after liver resection for CRC in the province of Ontario, Canada (population approximately 12 million). This would allow us to compare population-based results in the province with previous registry and single-institution reports. In addition, we attempted to determine if a positive hospital volume-topatient survival relationship exists.
METHODS Data source and study cohort
We obtained data during the years 1996 to 2004 from the Ontario Cancer Registry (OCR). The OCR uses probabilistic linkage strategies to identify and collect information on every newly diagnosed case of cancer in the province. Data sources include the Canadian Institute of Health Information and the Ontario Registered Persons Database. The Canadian Institute of Health Information database contains information on every patient discharged from an Ontario hospital, including patient demographics, major diagnoses, and procedures. The Ontario Registered Persons Database contains information on all births and deaths in the province and can provide information on longterm survival. As well, by law, hospitals and laboratories in Ontario must forward copies of all pathology reports with a cancer diagnosis to the OCR, although data from these reports are not systematically abstracted. Operative notes were not available. OCR linkage strategies have been shown to have high rates of case capture.29 Databases used in this study have a high degree of accuracy when coding major procedures and diagnoses.30 Research Ethics Board approval was received through Cancer Care Ontario and University Health Network. We obtained data for all patients who underwent primary resection of CRC from calendar years 1996 to 2004. Patients in the resulting cohort were linked to any subsequent liver resection for metastatic CRC for the same time period. Only patients who underwent liver resection for CRC metastases were included in this cohort. Patients with carcinoma in situ, lymphoma, sarcoma, carcinoid tumor, or hepatocellular carcinoma were excluded. Patients who underwent liver biopsy only were also excluded. Patient, tumor, and hospital groups
From 1996 to 2004, 1,437 patients underwent colorectal resection and then underwent a subsequent liver procedure. Patients were eliminated if they underwent a diagnostic biopsy only and not a curative resection for a liver me-
Resection of Liver Metastases from Colorectal Cancer
677
tastasis (n ⫽ 382; 27%); if they had a simultaneous small diagnostic wedge resection performed for confirmation of the diagnosis of metastatic CRC at the time of initial colorectal resection (n ⫽ 70; 5%); when the diagnosis of CRC metastases could not be confirmed (n ⫽ 25; 2%); if the pathology of liver resection was for a benign condition (n ⫽ 16; 1%) or was for another malignancy, such as hepatocellular carcinoma (n ⫽ 10; 1%); or if the pathology of a liver procedure could not be found (n ⫽ 92; 6%). Demographics for all patients were collected through Cancer Care Ontario registry. Nationality was not available. Age was treated as a continuous variable. Tumor characteristics for both the primary colon resection and subsequent liver resection were abstracted from the respective pathology reports. Data collected on CRC included date of resection, (T) and node (N) category using TNM staging, and presence of synchronous liver metastases. Presence of synchronous liver metastases was confirmed by liver biopsy, where applicable. A synchronous lesion was defined as a lesion that presented within 4 months (120 days) of the primary CRC procedure.4 Timing of resection of metastases was divided into 3 groups in relation to the original colorectal resection: 0 to 120 days, 120 to 240 days, or more than 240 days. Carcinoembryonic antigen levels were not available. Data collected on operation for liver metastases included date of diagnosis and resection, type of resection, tumor number, tumor size, location of lesion(s), lymph node involvement in the porta hepatitis, if examined and resected, and margin status. Terminology with respect to liver resection is that proposed by the International Hepato-Pancreatico-Biliary Association, also known as the Brisbane terminology (www. ihpba.org). Tumor size was determined by the largest diameter of the largest lesion in the specimen. Small wedge resections of the liver were considered biopsies if the margins were positive and the specimen weighed ⬍ 100 g. Presence of tumor closer than 1 cm from the cut edge was considered a positive margin.31-33 Before observing any patient outcomes data, volume cut points were selected to create volume groups.34 Hospital volume was modeled as a dichotomous variable by dividing patients into 2 equal groups based on the number of hepatic resections performed at each center during the 9-year period.17 Centers that performed ⬎ 80 liver resections were considered high volume. Teaching hospitals were 1 of 21 institutions in the province affiliated with a medical school. Patients were also divided into 2 different groups based on the year of liver resection, 1996 to 2000 and 2001 to 2004, to reflect enhanced regionalization efforts.35,36
678
Shah et al
Resection of Liver Metastases from Colorectal Cancer
J Am Coll Surg
Table 1. Patient Demographic and Tumor Characteristics Characteristic
Figure 1. Number of liver resections performed in the province of Ontario for metastatic colorectal cancer has increased yearly from 1996 to 2004.
Study outcomes and analyses
Patient outcomes of interest included longterm survival from date of liver resection. Followup data were available to December 31, 2005. All deaths within 30 days were considered surgical mortality. Median followup after initial colorectal and liver resection was 42 months and 24 months, respectively. We used descriptive, univariate, and multivariate analyses. To check the robustness of our results, we removed patients suffering 30-day mortality and reperformed the analysis. Actuarial methods were used to estimate survival after liver resection of CRC metastasis. A proportional hazards model was used to measure longterm survival after adjusting for potentially confounding variables. The model included the explanatory variables of hospital teaching status (yes or no), hospital volume group (high or low), era (1996 to 2000 or 2001 to 2004), tumor number, and patient characteristics of age and gender. Patient age was treated as a continuous variable. Variables that had missing values were not included in the multivariate model, including margin status and tumor size. Unadjusted and adjusted hazards ratios (HR) were estimated with 95% confidence intervals. A p value ⬍ 0.05 was considered significant. Statistical analyses were performed using SPSS software version 14.0 (SPSS, Inc).
RESULTS
N Age (y), median (range) Male gender CRC nodal status* Node-negative Node-positive Unknown CRC tumor thickness T1 T2 T3 T4 Unknown Timing of metastases (mo) ⬍4 4.1–12 ⬎ 12 Type of liver resection Wedge resection Lobectomy Trisectionectomy Negative margins Tumor number 1 2 3–5 ⬎6 Tumor size (cm) 0–2 2–5 ⬎5 Unknown Era 1996–2000 2001–2004
n
%
841 64 (23⫺90) 509
60
200 496 145
24 59 17
18 57 545 70 151
2 7 63 8 18
170 224 469
20 26 54
303 466 72 510
36 55 9 80†
653 146 30 12
78 17 4 1
155 420 235 31
18 50 28 4
328 513
39 61
*CRC stage not available for all cases. † Information was not available for 24% of patients (201 of 841). CRC, colorectal cancer; metachronous, presentation of liver metastases ⬎ 4 months; negative margin, ⬎ 1 cm margin free of tumor from resected edge; synchronous, presentation of liver metastases ⬍ 4 mo.
Demographics of liver resections for CRC metastases in Ontario
Patient and tumor characteristics
There were 841 patients who underwent potentially curative liver resection for CRC metastases in Ontario during the course of this study. The annual number of liver resections performed for metastatic CRC increased at a steady rate (Fig. 1) and were performed at 43 hospitals in Ontario during the 9-year period. Although teaching hospitals constituted only 49% of hospitals that performed at least 1 liver resection for metastatic CRC, 91% of total liver resections were performed at teaching centers.
Patient demographics and tumor characteristics are shown in Table 1. Most patients were men (61%), and median age at the time of liver resection was 64 years. Initial CRC resections were node-positive in 59% of patients, and most colorectal tumors penetrated the subserosa (73% T3 or T4 lesions). Eighty percent of liver metastases presented 4 months or more from initial CRC resection. Sixty-four percent of liver resections performed were major resections consisting of either
Vol. 205, No. 5, November 2007
Shah et al
Resection of Liver Metastases from Colorectal Cancer
679
Table 2. Hospital Characteristics Stratified by Volume
Characteristics
Figure 2. Number of liver resections performed at low- and highvolume centers in Ontario from 1996 to 2004.
lobectomy (55%) or trisectionectomy (9%). Most metastases were single (78%) and 29% were ⬎ 5 cm. The number of liver resections performed has increased (1996 to 2000, 39% versus 2001 to 2004, 61%). Volume effects
The 5 centers (12%) that qualified as high-volume centers performed 55% of the liver resections. All high-volume centers were teaching hospitals. Resections performed at high-volume centers increased in proportion to low-volume hospitals (Fig. 2). The percentage of resections done in highvolume centers increased significantly from the first to the second era (Table 2; 57% versus 49%; p ⫽ 0.02). The majority of major resections (lobectomy or greater) were performed at high-volume centers (60%; p ⬍ 0.001). Low-volume centers performed more wedge resections (45% versus 28%) and treated more synchronous lesions
Era 1996–2000 (n ⫽ 328) 2001–2004 (n ⫽ 513) Teaching hospital Synchronous resection Resection type Wedge Lobectomy Trisectionectomy Tumor size (cm) 0–2 2.1–5 ⬎5 Tumor no. 1 2 ⬎3 Negative margins* Perioperative mortality
Low volume (n ⴝ 390) n %
High volume (n ⴝ 451) n %
168 222 318 95
51 43 41 24
160 291 451 74
49 57 59 16
177 198 15
45 51 4
126 268 57
28 59 13
95 170 107
25 46 28
60 250 128
14 57 29
302 66 24 252 16
77 17 6 84 4
371 81 21 258 12
78 17 5 78 3
p Value
0.02
NA 0.004 ⬍ 0.001
0.01
0.52
0.19 0.15
*Information was not available for 24% of patients (201 of 841). NA, not applicable; negative margins, ⬎ 1 cm margins free of tumor; synchronous resection, resection within 4 months of original colorectal operation; wedge, includes left lateral segmentectomy, nonanatomical and anatomical wedge resection, and mesohepatectomy.
(56%). Tumors resected at high-volume centers were larger than those resected at low-volume centers, when stratified by groups (p ⫽ 0.01). Specifically, lesions ⬎ 2 cm were resected more commonly at high-volume centers (84% versus 76%). Survival outcomes and prognostic factors
Figure 3. Recurrence-free survival in this cohort of patients in whom liver metastases developed and who underwent resection in Ontario from 1996 to 2004.
Median time to liver resection after initial colorectal resection was 13.6 months (range 12.4 to 14.8 months) and 20% of resections occurred within 120 days of the colorectal resection and could be considered synchronous (Fig. 3). Perioperative mortality was 3%. Sixteen of 390 patients (4%) who underwent liver resection at low-volume centers died perioperatively. At high-volume centers, 12,451 patients died, for an operative mortality rate of 3%. Removal of patients who died within 30 days did not change the significance of any of the results presented. Unadjusted 1-, 3-, and 5-year survival after liver resection was 88%, 59%, and 43%, respectively (Fig. 4). Median survival after liver resection was 48 months (range 41.2 to 54.5 months). Unadjusted patient survival was significantly better at high-volume centers after liver resection for CRC metastases (median 55 versus 41 months; p ⫽
680
Shah et al
Resection of Liver Metastases from Colorectal Cancer
J Am Coll Surg
Figure 4. Unadjusted survival for all patients who underwent liver resection for colorectal cancer metastases in Ontario from 1996 to 2004.
Figure 5. Unadjusted survival of patients who underwent liver resection for colorectal cancer metastases comparing high- and lowvolume centers (p ⫽ 0.002).
0.002) (Fig. 5). Results were equivalent if patients who suffered an operative mortality were omitted from the analysis (data not shown). In univariate analysis, advanced age (p ⬍ 0.001), 2 or more tumor nodules (p ⬍ 0.001), center volume (p ⫽ 0.002), era (p ⫽ 0.005), and margin status (p ⫽ 0.03) were significant prognostic factors. Gender, tumor size, teaching status of center, CRC stage, timing of metastasis, and type of resection were not found to be associated with survival (Table 3). Multivariate analysis (Table 4) demonstrated that survival was improved when resection was performed at a high-volume center (HR ⫽ 0.72; 95% CI, 0.56 to 0.91, p ⫽ 0.005) or in the years 2001 to 2004 (HR ⫽ 0.45;
95% CI, 0.35 to 0.57, p ⫽ 0.001). Survival was worse with resection of 2 or more lesions (HR ⫽ 1.6; 95% CI, 1.42 to 2.34, p ⫽ 0.001).
Table 3. Univariate Analyses of Prognostic Factors for Survival after Liver Resection for Colorectal Cancer Metastases Variables
Patient and tumor characteristics Gender Advanced age* (⫹) Nodal CRC stage Major liver resection Metachronous presentation ⱖ 2 tumor nodules Tumor size ⬎ 5 cm Negative margins Era 2001–2004 Hospital characteristics Teaching hospital High-volume center
p Value
0.61 ⬍ 0.001 0.63 0.35 0.34 0.001 0.64 0.03 0.005
*Age was considered as a continuous variable. (⫹), positive; major liver resection, lobectomy or trisectionectomy.
0.05 0.002
DISCUSSION We observed a 5-year survival rate of 43% after liver resection for CRC metastasis and improved patient outcomes in high-volume hospitals. These results from a large population parallel those of single-hospital series. Because of selection biases and differences in management and surgical technique between centers, population-based studies are essential for providing reliable and reproducible information on epidemiology and management of solid organ tumors.14,17 These findings support continued efforts to agTable 4. Cox Regression Survival Analyses after Liver Resection for Colorectal Cancer Metastases Variable
Patient characteristics Male gender Advanced age* Lobectomy Metachronous presentation ⱖ 2 tumor nodules Era 2001–2004 Hospital characteristics Teaching hospital High-volume center
Hazards ratio
95% CI
p Value
0.94 1.01 0.97 1.14 1.62 0.45
0.75–1.18 1.00–1.02 0.77–1.22 0.87–1.49 1.42–2.34 0.35–0.57
0.57 0.06 0.79 0.34 0.001 0.001
0.85 0.72
0.59–1.25 0.56–0.91
0.41 0.005
*Advanced age was regarded as a continuous variable. CRC, colorectal cancer.
Vol. 205, No. 5, November 2007
Shah et al
gressively identify liver metastases after CRC resection among appropriate patients. In Ontario, the number of resections performed and overall survival increased with time. Metastases were most commonly metachronous and treated at teaching centers; this reflected the specialized nature involved in the care of CRC metastases correlating with other series.4,9,14 Operative mortality was low (3%), consistent with other series.2,4,5,9 This study demonstrates that the survival advantage at high-volume centers was not lost after the perioperative period.20,37 Our 5-year survival of 43% compares favorably with a recent study published by the Surveillance, Epidemiology, and End Results—Medicare database by Cummings and colleagues.16 In their study of patients with CRC and liver metastases from 1991 to 2001, the 5-year survival was 32.8% for patients who underwent liver resection. Prognostic factors for improved survival included resection of a solitary nodule, operation performed at high-volume centers, and later era of liver resection (2001–2004). This suggests that positive volume-to-outcomes relationships can persist after successful efforts to regionalize particular cancer operations.24,38 Age was an important factor for survival in our results, similar to some surgical series.4,14,39 Although we did not find margins ⬍ 1 cm to be a prognostic factor, similar to other series,10,40 our data were limited by a large number of unreported margin status in pathology reports. Improved surgical technique and earlier detection of lesions because of improved diagnostic regimens can account for improved survival for liver resections performed in the later era. Perioperative survival has been shown to be improved after liver resections performed at high-volume centers, but the relationship with longterm survival is not as clear.11,19,20,22 Using the National Medicare database, Fong and colleagues19 identified hospital volume as an independent predictor of longterm survival after liver resection for cancer (relative risk 1.2, p ⫽ 0.02). Dimick and colleagues,11 with use of the Nationwide Inpatient Sample, also reported improved survival at high-volume centers after hepatectomy. Improved survival in patients undergoing liver resection at high-volume hospitals was confirmed in the multivariate analysis to ensure that volume was an independent factor of survival after liver resection. In this study, because the influence of operative mortality was removed from these analyses, processes of care provided outside the operating room were likely responsible for this, such as superior patient selection or followup strategies. Regionalization could result in more efficient use of resources, especially important for procedures with low annual patient volumes and where there is a shortage of skilled personnel, as in various regions of Canada.19,38
Resection of Liver Metastases from Colorectal Cancer
681
Low-volume centers were more likely to perform small liver resections, remove smaller and fewer liver metastases, and were more likely to perform synchronous colorectal and liver resections. These results suggest that small CRC metastases at low-volume centers were likely removed by wedge excision at the time of initial colorectal resection. Limitations of this study include use of retrospective administrative data, with its potential for unforeseen confounding. The operation is only one variable in patient management. We also could not isolate surgeon procedure volume, surgical intent at the time of operation (operative note), or use of adjuvant or neoadjuvant therapy, all of which can be important determinants of outcomes. Our study concluded in 2004, before new chemotherapy regimens changed practice in management of metastatic CRC in North America.41-43 Nonetheless, the influence of chemotherapy before or after liver resection could not be determined. We were also unable to estimate the total number of patients in Ontario with liver metastases, irrespective of whether they needed operation. Although we began with a cohort of ⬎ 1,400 patients, this volume did not include patients who might not have undergone any diagnostic or therapeutic intervention for their metastatic disease. Leporrier and colleagues14 reported resection rates around 10% for synchronous and 33% for metachronous lesions. Cummings and colleagues16 found that in the Surveillance, Epidemiology, and End Results—Medicare database, only 6.1% of patients underwent liver resection from all patients with colorectal cancer and liver metastases (833 of 13,599). Another limitation is that the Cancer Care Ontario registry is only able to capture overall mortality; disease-specific survival and recurrence rates are unknown. Potentially important prognostic factors, such as CEA, tumor differentiation, and use of ablative techniques, were not available in this database. Our results are important because they provide an accurate and modern assessment of the role of operation for liver metastases from CRC at a population level. In addition to survival data, we have also captured center-specific data to assess outcomes in teaching and high-volume institutions. Advances in chemotherapy regimens have altered the management of CRC metastases, resulting in the majority of patients in the US receiving chemotherapy before liver resection for resectable lesions.42,44,45 Results of this new practice should be compared with outcomes of this report as a benchmark of surgical resection alone. A change in practice toward neoadjuvant chemotherapy before liver resection can be justified if 5-year survival considerably higher than 40% is achieved. Surgical resection of liver metastases from CRC is safe and an effective treatment strategy for resectable lesions.
682
Shah et al
Resection of Liver Metastases from Colorectal Cancer
Survival rates after operation are improving, possibly because of improved patient selection, surgical techniques, and continued regionalization efforts. Efforts to improve the quality of surgical care, whether through increased regionalization or hospital-level projects, should include ongoing data monitoring at the population level. Author Contributions
Study conception and design: Shah, Simunovic, Gallinger Acquisition of data: Shah, Bromberg Analysis and interpretation of data: Shah, Coates, Simunovic Drafting of manuscript: Shah, Bromberg Critical revision: Simunovic, Gallinger Statistical expertise: Shah, Simunovic Supervision: Gallinger
REFERENCES 1. Scheele J, Stangl R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history. Br J Surg 1990;77:1241–1246. 2. Abdalla EK, Vauthey JN, Ellis LM, et al. Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases. Ann Surg 2004;239:818–825. 3. Shah SA, Haddad R, Al Sukhni W, et al. Surgical resection of hepatic and pulmonary metastases from colorectal carcinoma. J Am Coll Surg 2006;202:468–475. 4. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999;230: 309–318. 5. Wei AC, Greig PD, Grant D, et al. Survival after hepatic resection for colorectal metastases: a 10-year experience. Ann Surg Oncol 2006;13:668–676. 6. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002;236:397– 406. 7. Abdalla EK, Vauthey JN. Improved survival after resection of colorectal liver metastases. J Clin Oncol 2006;24:2679–2681. 8. Aloia TA, Vauthey JN, Loyer EM, et al. Solitary colorectal liver metastasis: resection determines outcome. Arch Surg 2006;141: 460–466. 9. Choti MA, Sitzmann JV, Tiburi MF, et al. Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 2002;235:759–766. 10. Pawlik TM, Scoggins CR, Zorzi D, et al. Effect of surgical margin status on survival and site of recurrence after hepatic resection for colorectal metastases. Ann Surg 2005;241:715–722. 11. Dimick JB, Wainess RM, Cowan JA, et al. National trends in the use and outcomes of hepatic resection. J Am Coll Surg 2004; 199:31–38. 12. Alley PG, McNee RK. Colorectal cancer in Auckland 1981– 1982: patients with liver metastases. N Z Med J 1985;98:697– 699.
J Am Coll Surg
13. Jaeck D, Bachellier P, Guiguet M, et al. Long-term survival following resection of colorectal hepatic metastases. Association Francaise de Chirurgie. Br J Surg 1997;84:977–980. 14. Leporrier J, Maurel J, Chiche L, et al. A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer. Br J Surg 2006;93:465–474. 15. Kune GA, Kune S, Field B, et al. Survival in patients with largebowel cancer. A population-based investigation from the Melbourne Colorectal Cancer Study. Dis Colon Rectum 1990;33: 938–946. 16. Cummings LC, Payes JD, Cooper GS. Survival after hepatic resection in metastatic colorectal cancer: a population-based study. Cancer 2007;109:718–726. 17. Dimick JB, Cowan JA Jr, Knol JA, et al. Hepatic resection in the United States: indications, outcomes, and hospital procedural volumes from a nationally representative database. Arch Surg 2003;138:185–191. 18. Glasgow RE, Showstack JA, Katz PP, et al. The relationship between hospital volume and outcomes of hepatic resection for hepatocellular carcinoma. Arch Surg 1999;134:30–35. 19. Fong Y, Gonen M, Rubin D, et al. Long-term survival is superior after resection for cancer in high-volume centers. Ann Surg 2005;242:540–544. 20. Choti MA, Bowman HM, Pitt HA, et al. Should hepatic resections be performed at high-volume referral centers? J Gastrointest Surg 1998;2:11–20. 21. Dimick JB, Staiger DO, Birkmeyer JD. Are mortality rates for different operations related?: implications for measuring the quality of noncardiac surgery. Med Care 2006;44:774–778. 22. Birkmeyer JD, Sun Y, Goldfaden A, et al. Volume and process of care in high-risk cancer surgery. Cancer 2006;106:2476–2481. 23. Bentrem DJ, Brennan MF. Outcomes in oncologic surgery: does volume make a difference? World J Surg 2005;29:1210–1216. 24. Simunovic M, Rempel E, Theriault ME, et al. Influence of hospital characteristics on operative death and survival of patients after major cancer surgery in Ontario. Can J Surg 2006;49:251– 258. 25. Fong Y, Gonen M, Rubin D, et al. Long-term survival is superior after resection for cancer in high-volume centers. Ann Surg 2005;242:540–544. 26. Roohan PJ, Bickell NA, Baptiste MS, et al. Hospital volume differences and five-year survival from breast cancer. Am J Public Health 1998;88:454–457. 27. Schrag D, Cramer LD, Bach PB, et al. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA 2000;284:3028–3035. 28. Bach PB, Cramer LD, Schrag D, et al. The influence of hospital volume on survival after resection for lung cancer. N Engl J Med 2001;345:181–188. 29. Robles SC, Marrett LD, Clarke EA, et al. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495–501. 30. Simunovic M, To T, Langer B. Influence of hospital volume on mortality following major cancer surgery. JAMA 1999;281: 1374–1375. 31. Cady B, Jenkins RL, Steele GD Jr, et al. Surgical margin in hepatic resection for colorectal metastasis: a critical and improvable determinant of outcome. Ann Surg 1998;227:566–571. 32. Cady B, Stone MD, McDermott WV Jr, et al. Technical and biological factors in disease-free survival after hepatic resection for colorectal cancer metastases. Arch Surg 1992;127:561–568.
Vol. 205, No. 5, November 2007
Resection of Liver Metastases from Colorectal Cancer
Shah et al
33. Cady B, McDermott WV. Major hepatic resection for metachronous metastases from colon cancer. Ann Surg 1985;201: 204–209. 34. Hong D, Tandan VR, Goldsmith CH, et al. Users’ guide to the surgical literature: how to use an article reporting populationbased volume-outcome relationships in surgery. Can J Surg 2002;45:109–115. 35. Langer B, Stern H. An integrated system-wide strategy for quality improvement in cancer surgery. Br J Surg 2007;94: 3–5. 36. Simunovic M, To T, Theriault M, et al. Relation between hospital surgical volume and outcome for pancreatic resection for neoplasm in a publicly funded health care system. CMAJ 1999; 160:643–648. 37. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med 2002; 346:1128–1137. 38. Birkmeyer JD, Siewers AE, Marth NJ, et al. Regionalization of high-risk surgery and implications for patient travel times. JAMA 2003;290:2703–2708. 39. Hughes KS, Simon R, Songhorabodi S, et al. Resection of the liver for colorectal carcinoma metastases: a multi-
40.
41.
42. 43.
44. 45.
683
institutional study of patterns of recurrence. Surgery 1986; 100:278–284. Hamady ZZ, Cameron IC, Wyatt J, et al. Resection margin in patients undergoing hepatectomy for colorectal liver metastasis: a critical appraisal of the 1cm rule. Eur J Surg Oncol 2006;32: 557–563. Saltz LB, Meropol NJ, Loehrer PJ Sr, et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22:1201–1208. Bilchik AJ, Poston G, Curley SA, et al. Neoadjuvant chemotherapy for metastatic colon cancer: a cautionary note. J Clin Oncol 2005;23:9073–9078. Abdalla EK, Eng C, Madary A, et al. Southwest Oncology Group 0408: Phase II trial of neoadjuvant capecitabine/ oxaliplatin/bevacizumab for resectable colorectal metastases in the liver. Clin Colorectal Cancer 2006;5:436–439. Fong Y, Bentrem DJ. CASH (Chemotherapy-Associated Steatohepatitis) costs. Ann Surg 2006;243:8–9. Abdalla EK, Adam R, Bilchik AJ, et al. Improving resectability of hepatic colorectal metastases: expert consensus statement. Ann Surg Oncol 2006;13:1271–1280.
JACS CME-1 PROGRAM 1.0 credit is earned for completing both questions for each article. Completion of all four articles (8 questions) earns 4 CME-1 credits each month.
www.jacscme.facs
gundy, Melbourne, Auckland, and Caen.12-15 We are aware of only two studies that measured longterm survival after metastatic CRC liver resection in a large population.14,16 Numerous studies from US state and national databases have confirmed that hepatectomy for a variety of conditions has a low operative mortality rate.11,17-19 Increased hospital volume for advanced surgical oncology procedures appears to correlate with improved operative mortality, confirming a positive volume-to-outcomes relationship.19-25 Using the National Medicare database, Fong and colleagues25 found improved perioperative mortality (4% versus 9%) for liver resections performed for cancer at high-volume centers, defined as ⬎ 25 patients/ year. But they were unable to show a substantial improvement in longterm survival with liver resections performed at high-volume centers, indicating that the major effect of volume was on perioperative outcomes.25 The positive influence of hospital volume on longterm outcomes after operation has been shown in other cancer types, including breast, colon, and lung cancer.26-28 The number of liver resection performed for all indications is increasing in the
Hematogenous spread to the liver occurs in 40% to 60% of patients with colorectal cancer (CRC).1 Latest advances in medical and surgical treatments of liver masses have improved the diagnostic accuracy and surgical options for patients with metastatic CRC. This has resulted in 30% to 50% 5-year survival after resection for liver metastases.2-11 Most reports are from single institutions and cannot account for selection or referral bias. Registry studies focused on metastatic CRC have only been reported from BurCompeting Interests Declared: None. Presented at the 2007 Academic Surgical Congress, Phoenix, AZ, February 2007. Received April 28, 2007; Revised May 24, 2007; Accepted June 6, 2007. From the Department of Surgery, University of Massachusetts Medical School, Worcester, MA (Shah); University Health Network, University of Toronto, Toronto, Canada (Shah, Bromberg, Gallinger); and Hamilton Health Sciences, McMaster University, Hamilton, Canada (Coates, Rempel, Simunovic). Correspondence address: Shimul A Shah, MD, Division of Organ Transplantation, Department of Surgery, University of Massachusetts Medical School, 55 Lake Ave N, S3-838, Worcester, MA 01655. email: [email protected]
© 2007 by the American College of Surgeons Published by Elsevier Inc.
676
ISSN 1072-7515/07/$32.00 doi:10.1016/j.jamcollsurg.2007.06.283
Vol. 205, No. 5, November 2007
Shah et al
US,17 but we are unaware of any studies that have measured the relationship between hospital volume and patient longterm survival after hepatectomy for metastatic colorectal cancer. We aimed to assess longterm survival after liver resection for CRC in the province of Ontario, Canada (population approximately 12 million). This would allow us to compare population-based results in the province with previous registry and single-institution reports. In addition, we attempted to determine if a positive hospital volume-topatient survival relationship exists.
METHODS Data source and study cohort
We obtained data during the years 1996 to 2004 from the Ontario Cancer Registry (OCR). The OCR uses probabilistic linkage strategies to identify and collect information on every newly diagnosed case of cancer in the province. Data sources include the Canadian Institute of Health Information and the Ontario Registered Persons Database. The Canadian Institute of Health Information database contains information on every patient discharged from an Ontario hospital, including patient demographics, major diagnoses, and procedures. The Ontario Registered Persons Database contains information on all births and deaths in the province and can provide information on longterm survival. As well, by law, hospitals and laboratories in Ontario must forward copies of all pathology reports with a cancer diagnosis to the OCR, although data from these reports are not systematically abstracted. Operative notes were not available. OCR linkage strategies have been shown to have high rates of case capture.29 Databases used in this study have a high degree of accuracy when coding major procedures and diagnoses.30 Research Ethics Board approval was received through Cancer Care Ontario and University Health Network. We obtained data for all patients who underwent primary resection of CRC from calendar years 1996 to 2004. Patients in the resulting cohort were linked to any subsequent liver resection for metastatic CRC for the same time period. Only patients who underwent liver resection for CRC metastases were included in this cohort. Patients with carcinoma in situ, lymphoma, sarcoma, carcinoid tumor, or hepatocellular carcinoma were excluded. Patients who underwent liver biopsy only were also excluded. Patient, tumor, and hospital groups
From 1996 to 2004, 1,437 patients underwent colorectal resection and then underwent a subsequent liver procedure. Patients were eliminated if they underwent a diagnostic biopsy only and not a curative resection for a liver me-
Resection of Liver Metastases from Colorectal Cancer
677
tastasis (n ⫽ 382; 27%); if they had a simultaneous small diagnostic wedge resection performed for confirmation of the diagnosis of metastatic CRC at the time of initial colorectal resection (n ⫽ 70; 5%); when the diagnosis of CRC metastases could not be confirmed (n ⫽ 25; 2%); if the pathology of liver resection was for a benign condition (n ⫽ 16; 1%) or was for another malignancy, such as hepatocellular carcinoma (n ⫽ 10; 1%); or if the pathology of a liver procedure could not be found (n ⫽ 92; 6%). Demographics for all patients were collected through Cancer Care Ontario registry. Nationality was not available. Age was treated as a continuous variable. Tumor characteristics for both the primary colon resection and subsequent liver resection were abstracted from the respective pathology reports. Data collected on CRC included date of resection, (T) and node (N) category using TNM staging, and presence of synchronous liver metastases. Presence of synchronous liver metastases was confirmed by liver biopsy, where applicable. A synchronous lesion was defined as a lesion that presented within 4 months (120 days) of the primary CRC procedure.4 Timing of resection of metastases was divided into 3 groups in relation to the original colorectal resection: 0 to 120 days, 120 to 240 days, or more than 240 days. Carcinoembryonic antigen levels were not available. Data collected on operation for liver metastases included date of diagnosis and resection, type of resection, tumor number, tumor size, location of lesion(s), lymph node involvement in the porta hepatitis, if examined and resected, and margin status. Terminology with respect to liver resection is that proposed by the International Hepato-Pancreatico-Biliary Association, also known as the Brisbane terminology (www. ihpba.org). Tumor size was determined by the largest diameter of the largest lesion in the specimen. Small wedge resections of the liver were considered biopsies if the margins were positive and the specimen weighed ⬍ 100 g. Presence of tumor closer than 1 cm from the cut edge was considered a positive margin.31-33 Before observing any patient outcomes data, volume cut points were selected to create volume groups.34 Hospital volume was modeled as a dichotomous variable by dividing patients into 2 equal groups based on the number of hepatic resections performed at each center during the 9-year period.17 Centers that performed ⬎ 80 liver resections were considered high volume. Teaching hospitals were 1 of 21 institutions in the province affiliated with a medical school. Patients were also divided into 2 different groups based on the year of liver resection, 1996 to 2000 and 2001 to 2004, to reflect enhanced regionalization efforts.35,36
678
Shah et al
Resection of Liver Metastases from Colorectal Cancer
J Am Coll Surg
Table 1. Patient Demographic and Tumor Characteristics Characteristic
Figure 1. Number of liver resections performed in the province of Ontario for metastatic colorectal cancer has increased yearly from 1996 to 2004.
Study outcomes and analyses
Patient outcomes of interest included longterm survival from date of liver resection. Followup data were available to December 31, 2005. All deaths within 30 days were considered surgical mortality. Median followup after initial colorectal and liver resection was 42 months and 24 months, respectively. We used descriptive, univariate, and multivariate analyses. To check the robustness of our results, we removed patients suffering 30-day mortality and reperformed the analysis. Actuarial methods were used to estimate survival after liver resection of CRC metastasis. A proportional hazards model was used to measure longterm survival after adjusting for potentially confounding variables. The model included the explanatory variables of hospital teaching status (yes or no), hospital volume group (high or low), era (1996 to 2000 or 2001 to 2004), tumor number, and patient characteristics of age and gender. Patient age was treated as a continuous variable. Variables that had missing values were not included in the multivariate model, including margin status and tumor size. Unadjusted and adjusted hazards ratios (HR) were estimated with 95% confidence intervals. A p value ⬍ 0.05 was considered significant. Statistical analyses were performed using SPSS software version 14.0 (SPSS, Inc).
RESULTS
N Age (y), median (range) Male gender CRC nodal status* Node-negative Node-positive Unknown CRC tumor thickness T1 T2 T3 T4 Unknown Timing of metastases (mo) ⬍4 4.1–12 ⬎ 12 Type of liver resection Wedge resection Lobectomy Trisectionectomy Negative margins Tumor number 1 2 3–5 ⬎6 Tumor size (cm) 0–2 2–5 ⬎5 Unknown Era 1996–2000 2001–2004
n
%
841 64 (23⫺90) 509
60
200 496 145
24 59 17
18 57 545 70 151
2 7 63 8 18
170 224 469
20 26 54
303 466 72 510
36 55 9 80†
653 146 30 12
78 17 4 1
155 420 235 31
18 50 28 4
328 513
39 61
*CRC stage not available for all cases. † Information was not available for 24% of patients (201 of 841). CRC, colorectal cancer; metachronous, presentation of liver metastases ⬎ 4 months; negative margin, ⬎ 1 cm margin free of tumor from resected edge; synchronous, presentation of liver metastases ⬍ 4 mo.
Demographics of liver resections for CRC metastases in Ontario
Patient and tumor characteristics
There were 841 patients who underwent potentially curative liver resection for CRC metastases in Ontario during the course of this study. The annual number of liver resections performed for metastatic CRC increased at a steady rate (Fig. 1) and were performed at 43 hospitals in Ontario during the 9-year period. Although teaching hospitals constituted only 49% of hospitals that performed at least 1 liver resection for metastatic CRC, 91% of total liver resections were performed at teaching centers.
Patient demographics and tumor characteristics are shown in Table 1. Most patients were men (61%), and median age at the time of liver resection was 64 years. Initial CRC resections were node-positive in 59% of patients, and most colorectal tumors penetrated the subserosa (73% T3 or T4 lesions). Eighty percent of liver metastases presented 4 months or more from initial CRC resection. Sixty-four percent of liver resections performed were major resections consisting of either
Vol. 205, No. 5, November 2007
Shah et al
Resection of Liver Metastases from Colorectal Cancer
679
Table 2. Hospital Characteristics Stratified by Volume
Characteristics
Figure 2. Number of liver resections performed at low- and highvolume centers in Ontario from 1996 to 2004.
lobectomy (55%) or trisectionectomy (9%). Most metastases were single (78%) and 29% were ⬎ 5 cm. The number of liver resections performed has increased (1996 to 2000, 39% versus 2001 to 2004, 61%). Volume effects
The 5 centers (12%) that qualified as high-volume centers performed 55% of the liver resections. All high-volume centers were teaching hospitals. Resections performed at high-volume centers increased in proportion to low-volume hospitals (Fig. 2). The percentage of resections done in highvolume centers increased significantly from the first to the second era (Table 2; 57% versus 49%; p ⫽ 0.02). The majority of major resections (lobectomy or greater) were performed at high-volume centers (60%; p ⬍ 0.001). Low-volume centers performed more wedge resections (45% versus 28%) and treated more synchronous lesions
Era 1996–2000 (n ⫽ 328) 2001–2004 (n ⫽ 513) Teaching hospital Synchronous resection Resection type Wedge Lobectomy Trisectionectomy Tumor size (cm) 0–2 2.1–5 ⬎5 Tumor no. 1 2 ⬎3 Negative margins* Perioperative mortality
Low volume (n ⴝ 390) n %
High volume (n ⴝ 451) n %
168 222 318 95
51 43 41 24
160 291 451 74
49 57 59 16
177 198 15
45 51 4
126 268 57
28 59 13
95 170 107
25 46 28
60 250 128
14 57 29
302 66 24 252 16
77 17 6 84 4
371 81 21 258 12
78 17 5 78 3
p Value
0.02
NA 0.004 ⬍ 0.001
0.01
0.52
0.19 0.15
*Information was not available for 24% of patients (201 of 841). NA, not applicable; negative margins, ⬎ 1 cm margins free of tumor; synchronous resection, resection within 4 months of original colorectal operation; wedge, includes left lateral segmentectomy, nonanatomical and anatomical wedge resection, and mesohepatectomy.
(56%). Tumors resected at high-volume centers were larger than those resected at low-volume centers, when stratified by groups (p ⫽ 0.01). Specifically, lesions ⬎ 2 cm were resected more commonly at high-volume centers (84% versus 76%). Survival outcomes and prognostic factors
Figure 3. Recurrence-free survival in this cohort of patients in whom liver metastases developed and who underwent resection in Ontario from 1996 to 2004.
Median time to liver resection after initial colorectal resection was 13.6 months (range 12.4 to 14.8 months) and 20% of resections occurred within 120 days of the colorectal resection and could be considered synchronous (Fig. 3). Perioperative mortality was 3%. Sixteen of 390 patients (4%) who underwent liver resection at low-volume centers died perioperatively. At high-volume centers, 12,451 patients died, for an operative mortality rate of 3%. Removal of patients who died within 30 days did not change the significance of any of the results presented. Unadjusted 1-, 3-, and 5-year survival after liver resection was 88%, 59%, and 43%, respectively (Fig. 4). Median survival after liver resection was 48 months (range 41.2 to 54.5 months). Unadjusted patient survival was significantly better at high-volume centers after liver resection for CRC metastases (median 55 versus 41 months; p ⫽
680
Shah et al
Resection of Liver Metastases from Colorectal Cancer
J Am Coll Surg
Figure 4. Unadjusted survival for all patients who underwent liver resection for colorectal cancer metastases in Ontario from 1996 to 2004.
Figure 5. Unadjusted survival of patients who underwent liver resection for colorectal cancer metastases comparing high- and lowvolume centers (p ⫽ 0.002).
0.002) (Fig. 5). Results were equivalent if patients who suffered an operative mortality were omitted from the analysis (data not shown). In univariate analysis, advanced age (p ⬍ 0.001), 2 or more tumor nodules (p ⬍ 0.001), center volume (p ⫽ 0.002), era (p ⫽ 0.005), and margin status (p ⫽ 0.03) were significant prognostic factors. Gender, tumor size, teaching status of center, CRC stage, timing of metastasis, and type of resection were not found to be associated with survival (Table 3). Multivariate analysis (Table 4) demonstrated that survival was improved when resection was performed at a high-volume center (HR ⫽ 0.72; 95% CI, 0.56 to 0.91, p ⫽ 0.005) or in the years 2001 to 2004 (HR ⫽ 0.45;
95% CI, 0.35 to 0.57, p ⫽ 0.001). Survival was worse with resection of 2 or more lesions (HR ⫽ 1.6; 95% CI, 1.42 to 2.34, p ⫽ 0.001).
Table 3. Univariate Analyses of Prognostic Factors for Survival after Liver Resection for Colorectal Cancer Metastases Variables
Patient and tumor characteristics Gender Advanced age* (⫹) Nodal CRC stage Major liver resection Metachronous presentation ⱖ 2 tumor nodules Tumor size ⬎ 5 cm Negative margins Era 2001–2004 Hospital characteristics Teaching hospital High-volume center
p Value
0.61 ⬍ 0.001 0.63 0.35 0.34 0.001 0.64 0.03 0.005
*Age was considered as a continuous variable. (⫹), positive; major liver resection, lobectomy or trisectionectomy.
0.05 0.002
DISCUSSION We observed a 5-year survival rate of 43% after liver resection for CRC metastasis and improved patient outcomes in high-volume hospitals. These results from a large population parallel those of single-hospital series. Because of selection biases and differences in management and surgical technique between centers, population-based studies are essential for providing reliable and reproducible information on epidemiology and management of solid organ tumors.14,17 These findings support continued efforts to agTable 4. Cox Regression Survival Analyses after Liver Resection for Colorectal Cancer Metastases Variable
Patient characteristics Male gender Advanced age* Lobectomy Metachronous presentation ⱖ 2 tumor nodules Era 2001–2004 Hospital characteristics Teaching hospital High-volume center
Hazards ratio
95% CI
p Value
0.94 1.01 0.97 1.14 1.62 0.45
0.75–1.18 1.00–1.02 0.77–1.22 0.87–1.49 1.42–2.34 0.35–0.57
0.57 0.06 0.79 0.34 0.001 0.001
0.85 0.72
0.59–1.25 0.56–0.91
0.41 0.005
*Advanced age was regarded as a continuous variable. CRC, colorectal cancer.
Vol. 205, No. 5, November 2007
Shah et al
gressively identify liver metastases after CRC resection among appropriate patients. In Ontario, the number of resections performed and overall survival increased with time. Metastases were most commonly metachronous and treated at teaching centers; this reflected the specialized nature involved in the care of CRC metastases correlating with other series.4,9,14 Operative mortality was low (3%), consistent with other series.2,4,5,9 This study demonstrates that the survival advantage at high-volume centers was not lost after the perioperative period.20,37 Our 5-year survival of 43% compares favorably with a recent study published by the Surveillance, Epidemiology, and End Results—Medicare database by Cummings and colleagues.16 In their study of patients with CRC and liver metastases from 1991 to 2001, the 5-year survival was 32.8% for patients who underwent liver resection. Prognostic factors for improved survival included resection of a solitary nodule, operation performed at high-volume centers, and later era of liver resection (2001–2004). This suggests that positive volume-to-outcomes relationships can persist after successful efforts to regionalize particular cancer operations.24,38 Age was an important factor for survival in our results, similar to some surgical series.4,14,39 Although we did not find margins ⬍ 1 cm to be a prognostic factor, similar to other series,10,40 our data were limited by a large number of unreported margin status in pathology reports. Improved surgical technique and earlier detection of lesions because of improved diagnostic regimens can account for improved survival for liver resections performed in the later era. Perioperative survival has been shown to be improved after liver resections performed at high-volume centers, but the relationship with longterm survival is not as clear.11,19,20,22 Using the National Medicare database, Fong and colleagues19 identified hospital volume as an independent predictor of longterm survival after liver resection for cancer (relative risk 1.2, p ⫽ 0.02). Dimick and colleagues,11 with use of the Nationwide Inpatient Sample, also reported improved survival at high-volume centers after hepatectomy. Improved survival in patients undergoing liver resection at high-volume hospitals was confirmed in the multivariate analysis to ensure that volume was an independent factor of survival after liver resection. In this study, because the influence of operative mortality was removed from these analyses, processes of care provided outside the operating room were likely responsible for this, such as superior patient selection or followup strategies. Regionalization could result in more efficient use of resources, especially important for procedures with low annual patient volumes and where there is a shortage of skilled personnel, as in various regions of Canada.19,38
Resection of Liver Metastases from Colorectal Cancer
681
Low-volume centers were more likely to perform small liver resections, remove smaller and fewer liver metastases, and were more likely to perform synchronous colorectal and liver resections. These results suggest that small CRC metastases at low-volume centers were likely removed by wedge excision at the time of initial colorectal resection. Limitations of this study include use of retrospective administrative data, with its potential for unforeseen confounding. The operation is only one variable in patient management. We also could not isolate surgeon procedure volume, surgical intent at the time of operation (operative note), or use of adjuvant or neoadjuvant therapy, all of which can be important determinants of outcomes. Our study concluded in 2004, before new chemotherapy regimens changed practice in management of metastatic CRC in North America.41-43 Nonetheless, the influence of chemotherapy before or after liver resection could not be determined. We were also unable to estimate the total number of patients in Ontario with liver metastases, irrespective of whether they needed operation. Although we began with a cohort of ⬎ 1,400 patients, this volume did not include patients who might not have undergone any diagnostic or therapeutic intervention for their metastatic disease. Leporrier and colleagues14 reported resection rates around 10% for synchronous and 33% for metachronous lesions. Cummings and colleagues16 found that in the Surveillance, Epidemiology, and End Results—Medicare database, only 6.1% of patients underwent liver resection from all patients with colorectal cancer and liver metastases (833 of 13,599). Another limitation is that the Cancer Care Ontario registry is only able to capture overall mortality; disease-specific survival and recurrence rates are unknown. Potentially important prognostic factors, such as CEA, tumor differentiation, and use of ablative techniques, were not available in this database. Our results are important because they provide an accurate and modern assessment of the role of operation for liver metastases from CRC at a population level. In addition to survival data, we have also captured center-specific data to assess outcomes in teaching and high-volume institutions. Advances in chemotherapy regimens have altered the management of CRC metastases, resulting in the majority of patients in the US receiving chemotherapy before liver resection for resectable lesions.42,44,45 Results of this new practice should be compared with outcomes of this report as a benchmark of surgical resection alone. A change in practice toward neoadjuvant chemotherapy before liver resection can be justified if 5-year survival considerably higher than 40% is achieved. Surgical resection of liver metastases from CRC is safe and an effective treatment strategy for resectable lesions.
682
Shah et al
Resection of Liver Metastases from Colorectal Cancer
Survival rates after operation are improving, possibly because of improved patient selection, surgical techniques, and continued regionalization efforts. Efforts to improve the quality of surgical care, whether through increased regionalization or hospital-level projects, should include ongoing data monitoring at the population level. Author Contributions
Study conception and design: Shah, Simunovic, Gallinger Acquisition of data: Shah, Bromberg Analysis and interpretation of data: Shah, Coates, Simunovic Drafting of manuscript: Shah, Bromberg Critical revision: Simunovic, Gallinger Statistical expertise: Shah, Simunovic Supervision: Gallinger
REFERENCES 1. Scheele J, Stangl R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history. Br J Surg 1990;77:1241–1246. 2. Abdalla EK, Vauthey JN, Ellis LM, et al. Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases. Ann Surg 2004;239:818–825. 3. Shah SA, Haddad R, Al Sukhni W, et al. Surgical resection of hepatic and pulmonary metastases from colorectal carcinoma. J Am Coll Surg 2006;202:468–475. 4. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999;230: 309–318. 5. Wei AC, Greig PD, Grant D, et al. Survival after hepatic resection for colorectal metastases: a 10-year experience. Ann Surg Oncol 2006;13:668–676. 6. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002;236:397– 406. 7. Abdalla EK, Vauthey JN. Improved survival after resection of colorectal liver metastases. J Clin Oncol 2006;24:2679–2681. 8. Aloia TA, Vauthey JN, Loyer EM, et al. Solitary colorectal liver metastasis: resection determines outcome. Arch Surg 2006;141: 460–466. 9. Choti MA, Sitzmann JV, Tiburi MF, et al. Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 2002;235:759–766. 10. Pawlik TM, Scoggins CR, Zorzi D, et al. Effect of surgical margin status on survival and site of recurrence after hepatic resection for colorectal metastases. Ann Surg 2005;241:715–722. 11. Dimick JB, Wainess RM, Cowan JA, et al. National trends in the use and outcomes of hepatic resection. J Am Coll Surg 2004; 199:31–38. 12. Alley PG, McNee RK. Colorectal cancer in Auckland 1981– 1982: patients with liver metastases. N Z Med J 1985;98:697– 699.
J Am Coll Surg
13. Jaeck D, Bachellier P, Guiguet M, et al. Long-term survival following resection of colorectal hepatic metastases. Association Francaise de Chirurgie. Br J Surg 1997;84:977–980. 14. Leporrier J, Maurel J, Chiche L, et al. A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer. Br J Surg 2006;93:465–474. 15. Kune GA, Kune S, Field B, et al. Survival in patients with largebowel cancer. A population-based investigation from the Melbourne Colorectal Cancer Study. Dis Colon Rectum 1990;33: 938–946. 16. Cummings LC, Payes JD, Cooper GS. Survival after hepatic resection in metastatic colorectal cancer: a population-based study. Cancer 2007;109:718–726. 17. Dimick JB, Cowan JA Jr, Knol JA, et al. Hepatic resection in the United States: indications, outcomes, and hospital procedural volumes from a nationally representative database. Arch Surg 2003;138:185–191. 18. Glasgow RE, Showstack JA, Katz PP, et al. The relationship between hospital volume and outcomes of hepatic resection for hepatocellular carcinoma. Arch Surg 1999;134:30–35. 19. Fong Y, Gonen M, Rubin D, et al. Long-term survival is superior after resection for cancer in high-volume centers. Ann Surg 2005;242:540–544. 20. Choti MA, Bowman HM, Pitt HA, et al. Should hepatic resections be performed at high-volume referral centers? J Gastrointest Surg 1998;2:11–20. 21. Dimick JB, Staiger DO, Birkmeyer JD. Are mortality rates for different operations related?: implications for measuring the quality of noncardiac surgery. Med Care 2006;44:774–778. 22. Birkmeyer JD, Sun Y, Goldfaden A, et al. Volume and process of care in high-risk cancer surgery. Cancer 2006;106:2476–2481. 23. Bentrem DJ, Brennan MF. Outcomes in oncologic surgery: does volume make a difference? World J Surg 2005;29:1210–1216. 24. Simunovic M, Rempel E, Theriault ME, et al. Influence of hospital characteristics on operative death and survival of patients after major cancer surgery in Ontario. Can J Surg 2006;49:251– 258. 25. Fong Y, Gonen M, Rubin D, et al. Long-term survival is superior after resection for cancer in high-volume centers. Ann Surg 2005;242:540–544. 26. Roohan PJ, Bickell NA, Baptiste MS, et al. Hospital volume differences and five-year survival from breast cancer. Am J Public Health 1998;88:454–457. 27. Schrag D, Cramer LD, Bach PB, et al. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA 2000;284:3028–3035. 28. Bach PB, Cramer LD, Schrag D, et al. The influence of hospital volume on survival after resection for lung cancer. N Engl J Med 2001;345:181–188. 29. Robles SC, Marrett LD, Clarke EA, et al. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495–501. 30. Simunovic M, To T, Langer B. Influence of hospital volume on mortality following major cancer surgery. JAMA 1999;281: 1374–1375. 31. Cady B, Jenkins RL, Steele GD Jr, et al. Surgical margin in hepatic resection for colorectal metastasis: a critical and improvable determinant of outcome. Ann Surg 1998;227:566–571. 32. Cady B, Stone MD, McDermott WV Jr, et al. Technical and biological factors in disease-free survival after hepatic resection for colorectal cancer metastases. Arch Surg 1992;127:561–568.
Vol. 205, No. 5, November 2007
Resection of Liver Metastases from Colorectal Cancer
Shah et al
33. Cady B, McDermott WV. Major hepatic resection for metachronous metastases from colon cancer. Ann Surg 1985;201: 204–209. 34. Hong D, Tandan VR, Goldsmith CH, et al. Users’ guide to the surgical literature: how to use an article reporting populationbased volume-outcome relationships in surgery. Can J Surg 2002;45:109–115. 35. Langer B, Stern H. An integrated system-wide strategy for quality improvement in cancer surgery. Br J Surg 2007;94: 3–5. 36. Simunovic M, To T, Theriault M, et al. Relation between hospital surgical volume and outcome for pancreatic resection for neoplasm in a publicly funded health care system. CMAJ 1999; 160:643–648. 37. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med 2002; 346:1128–1137. 38. Birkmeyer JD, Siewers AE, Marth NJ, et al. Regionalization of high-risk surgery and implications for patient travel times. JAMA 2003;290:2703–2708. 39. Hughes KS, Simon R, Songhorabodi S, et al. Resection of the liver for colorectal carcinoma metastases: a multi-
40.
41.
42. 43.
44. 45.
683
institutional study of patterns of recurrence. Surgery 1986; 100:278–284. Hamady ZZ, Cameron IC, Wyatt J, et al. Resection margin in patients undergoing hepatectomy for colorectal liver metastasis: a critical appraisal of the 1cm rule. Eur J Surg Oncol 2006;32: 557–563. Saltz LB, Meropol NJ, Loehrer PJ Sr, et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22:1201–1208. Bilchik AJ, Poston G, Curley SA, et al. Neoadjuvant chemotherapy for metastatic colon cancer: a cautionary note. J Clin Oncol 2005;23:9073–9078. Abdalla EK, Eng C, Madary A, et al. Southwest Oncology Group 0408: Phase II trial of neoadjuvant capecitabine/ oxaliplatin/bevacizumab for resectable colorectal metastases in the liver. Clin Colorectal Cancer 2006;5:436–439. Fong Y, Bentrem DJ. CASH (Chemotherapy-Associated Steatohepatitis) costs. Ann Surg 2006;243:8–9. Abdalla EK, Adam R, Bilchik AJ, et al. Improving resectability of hepatic colorectal metastases: expert consensus statement. Ann Surg Oncol 2006;13:1271–1280.
JACS CME-1 PROGRAM 1.0 credit is earned for completing both questions for each article. Completion of all four articles (8 questions) earns 4 CME-1 credits each month.
www.jacscme.facs