Staging Laparoscopy in the Management of Gastric Cancer: A Population-Based Analysis

Transcription

1 Staging Laparoscopy in the Management of Gastric Cancer: A Population-Based Analysis Paul J Karanicolas, MD, PhD, Elena B Elkin, PhD, Lindsay M Jacks, MSc, Coral L Atoria, MPH, Vivian E Strong, MD, FACS, Murray F Brennan, MD, FACS, Daniel G Coit, MD, FACS BACKGROUND: STUDY DESIGN: RESULTS: CONCLUSIONS: Staging laparoscopy can detect radiographically occult peritoneal metastases and prevent futile laparotomy in patients with gastric adenocarcinoma. We sought to assess the use of staging laparoscopy for gastric adenocarcinoma in a cohort of older patients and to compare outcomes after laparoscopy alone with nontherapeutic laparotomy. Using Surveillance, Epidemiology and End Results (SEER) population-based cancer registry data linked with Medicare claims, we identified patients aged 65 or older diagnosed with gastric adenocarcinoma between 1998 and We defined staging laparoscopy as a laparoscopic procedure from 1 month before the date of diagnosis until death and futile laparotomy as a laparotomy in the absence of a therapeutic intervention. We examined trends in the use of staging laparoscopy and compared outcomes between patients who underwent staging laparoscopy alone and those who had a futile laparotomy. Of 11,759 patients with gastric adenocarcinoma, 6,388 (54.3%) had at least 1 surgical procedure. Staging laparoscopy was performed in 506 (7.9%) patients who had any surgery, and 151 (29.8%) of these patients did not have a subsequent therapeutic intervention. Patients who underwent staging laparoscopy alone had a significantly lower rate of in-hospital mortality (5.3% vs 13.1%, p 0.001) and shorter length of hospitalization (2 vs 10 days, p 0.001) than patients who had futile laparotomy. Our findings in this large, population-based cohort suggest that staging laparoscopy is used infrequently in the management of older patients with gastric adenocarcinoma. Increased use of staging laparoscopy could reduce the substantial morbidity and mortality associated with nontherapeutic laparotomy. (J Am Coll Surg 2011;213: by the American College of Surgeons) Disclosure Information: Authors have nothing to disclose. Timothy J Eberlein, Editor-in-Chief, has nothing to disclose. Abstract presented at the American College of Surgeons 96th Annual Clinical Congress, Surgical Forum, Washington, DC, October Received May 20, 2011; Revised July 20, 2011; Accepted July 20, From the Departments of Surgery (Karanicolas, Strong, Brennan, Coit) and Epidemiology and Biostatistics (Elkin, Jacks, Atoria), Memorial Sloan- Kettering Cancer Center, New York, NY. Correspondence address: Paul J Karanicolas, MD, PhD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY, paul.karanicolas@sunnybrook.ca More than 21,000 new cases of gastric cancer are diagnosed annually in the United States. 1 The overall prognosis for patients with advanced disease is poor. The extent of disease at presentation is the most important predictor of prognosis and is a key determinant of the appropriate therapeutic modality. In the absence of metastases, resection of all gross disease with negative microscopic margins offers the only chance for cure. 2,3 Preoperative staging typically includes a combination of endoscopy, cross-sectional imaging (CT scan or MRI), endoscopic ultrasound, and PET scan. Despite appropriate staging with these modalities, up to 30% of patients with no preoperative evidence of metastatic disease harbor occult intra-abdominal metastases (peritoneal, liver, or nonregional lymph nodes) at the time of operation (Table 1) In these patients, resection yields no improvement in survival and is rarely needed for palliation of symptoms (most commonly, obstruction or bleeding). 13 Staging laparoscopy may detect occult metastatic disease and spare the patient an unnecessary laparotomy, resulting in fewer complications, less operating room time, and shorter hospital stay. 14 Staging laparoscopy may be particularly advantageous in older patients, who are at higher risk of suffering complications from laparotomy. 15 Despite the demonstrated benefits of staging laparoscopy, the uptake of this management approach in practice is unclear. We sought to explore trends in the use and outcomes of staging laparoscopy for gastric adenocarcinoma in a population-based cohort of older adults by the American College of Surgeons ISSN /11/$36.00 Published by Elsevier Inc. 644 doi: /j.jamcollsurg

2 Vol. 213, No. 5, November 2011 Karanicolas et al Staging Laparoscopy for Gastric Cancer 645 Table 1. Summary of Studies Examining the Rate of Intraabdominal Occult Metastases at the Time of Staging Laparoscopy in Patients with Gastric Cancer First author Year n Preoperative workup Occult metastases, % Possik US 34 Kriplani US CT 13 Lowy CT 23 Stell US,CT 34 Ascencio US,CT 41 D Ugo US,CT 21 Romijn US,CT 40 Yano US,CT 34 Sarela CT 31 US, ultrasound. METHODS Data The primary data source was the Surveillance, Epidemiology and End Results (SEER) cancer registry data linked with Medicare claims and enrollment records. SEER is the National Cancer Institute (NCI)-sponsored program of cancer registries in selected geographic regions covering about 25% of the US population. 16 The SEER registries collect data regarding site and extent of disease, first course of cancer-directed therapy, and sociodemographic characteristics, with active follow-up for date and cause of death. Medicare is the primary health insurer for 97% of the US population 65 years or older. Hospitalization information for those eligible for Medicare Part A is available from the Medicare Provider Analysis and Review files. Outpatient and physician or supplier Medicare files for services rendered in physicians offices and hospital outpatient departments are available for the 95% of Medicare beneficiaries who elect Part B coverage. Approximately 93% of SEER patients 65 years or older have been successfully linked with their Medicare claims. 16 The SEER-Medicare files were used in accordance with a data-use agreement from the NCI, and the study was approved by the Institutional Review Board at Memorial Sloan-Kettering Cancer Center. Cohort We identified all patients aged 65 or over with primary gastric adenocarcinoma diagnosed between January 1, 1998 and December 31, Based on Medicare claims for surgical procedures from 1 month before the date of diagnosis until death, we restricted the cohort to patients who had an operative procedure for gastric cancer (Appendix, online only, for procedure codes). We excluded patients diagnosed only at the time of death, who had a history of another malignancy, who lacked Parts A or B of Medicare, and those enrolled in a managed care plan. Outcomes and predictors The primary endpoint was type of gastric cancer surgery received from 1 month before diagnosis through the date of death or end of follow-up. Therapeutic intervention was defined by the presence of a claim for gastric resection or gastric bypass (Fig. 1). A patient was classified as having laparotomy alone if there was a claim for laparotomy in the absence of a therapeutic intervention on the same day or earlier. If a patient had a therapeutic intervention at a later date, we assumed that metastatic disease was identified at the initial laparotomy and the subsequent therapeutic intervention was needed for palliation of symptoms. Additional endpoints were in-hospital mortality (defined as discharge status of dead) and length of hospital stay (for the first chronological procedure in patients who had multiple operations). Several patient and disease characteristics were assessed. Demographic characteristics included patient age, race, marital status, geographic location, and residence in a metropolitan vs a nonmetropolitan county. Clinical characteristics included location of the cancer within the stomach, tumor stage, grade, and nodal involvement. Comorbidity was estimated using the Charlson comorbidity index based on inpatient claims in the 12 months before cancer diagnosis. 17 We assessed the use of neoadjuvant chemotherapy by identifying claims for chemotherapy between the month of cancer diagnosis and the date of resection. Analysis Associations between patient and disease characteristics and receipt of staging laparoscopy were assessed using chisquare tests. We used multivariable logistic regression analysis to evaluate potential predictors of staging laparoscopy, including only characteristics that would be known preoperatively: age, sex, race, marital status, comorbidity, geographic region, and site of the tumor within the stomach. Similar analyses were performed to assess predictors of resection. We explored time trends in the use of staging laparoscopy using the Cochran-Armitage test for trend. Differences in the rate of in-hospital mortality and median length of hospital stay were compared between those receiving staging laparoscopy alone and laparotomy alone using the chi-square test and Wilcoxon rank-sum test, respectively. All statistical analyses were performed using SAS version 9.2 (SAS Institute Inc) software. RESULTS During the 8-year study period, 11,759 patients were diagnosed with gastric adenocarcinoma and 6,388 patients underwent an operative procedure. Staging laparoscopy

3 646 Karanicolas et al Staging Laparoscopy for Gastric Cancer J Am Coll Surg Figure 1. Procedure definitions. was performed in 506 (8%) patients who had any operation (Fig. 2). Of these patients, 151 (30%) did not have a further operative intervention (ie, staging laparoscopy alone) and 306 (60%) proceeded to have a therapeutic intervention (resection in 97% of patients). Of the 5,882 patients who did not have laparoscopy as part of their management, 5,304 (90%) had a therapeutic intervention (resection in 94% of these patients). The use of staging laparoscopy increased over the study period from 5.5% in 1998 to 11.1% in 2005 (p 0.01, Fig. 3). The rate of staging laparoscopy alone also increased, from 1.7% to 3.1% (p 0.01), and the rate of laparotomy alone was stable over time (p 0.52). Patients who had staging laparoscopy were more likely to be young, white, married, have less comorbidity, be treated in the Northeast, and have proximal tumors, compared with those who did not have laparoscopy (Table 2). Only 4% of patients received neoadjuvant chemotherapy, and this proportion did not vary by use of staging laparoscopy (data not shown). In multivariable analysis, significant predictors of staging laparoscopy were younger age, white race, lower comorbidity score, Northeast region, and proximal tumors (Table 3). Significant predictors of resection compared with any other gastric operation were Asian race, being married, and having a distal tumor (Table 4). Among patients who had any therapeutic intervention the most Figure 2. Classification of cohort by type of gastric cancer procedures.

4 Vol. 213, No. 5, November 2011 Karanicolas et al Staging Laparoscopy for Gastric Cancer 647 Figure 3. Trends in use of staging laparoscopy, staging laparoscopy alone, and laparotomy alone. common procedure was distal gastrectomy, independent of the use of staging laparoscopy (Table 5). Patients who had laparoscopy alone had a shorter length of hospital stay (2 days vs 10 days, p 0.01) and a lower in-hospital mortality rate (5.3% vs 13.1%, p 0.01) compared with patients who had laparotomy alone. DISCUSSION This large, population-based analysis of patients over the age of 65 years who underwent gastric cancer surgery in the US highlights several important findings. Despite clear benefits of staging laparoscopy in patients without radiologic evidence of metastases, only 8% of patients had staging laparoscopy during the course of their management. 13,14 In patients who did undergo laparoscopy, 30% did not have a later therapeutic intervention, presumably due to occult metastases identified at the time of laparoscopy. In an additional 10% of patients a laparotomy was performed without therapeutic intervention, likely due to occult metastases or locally unresectable disease that was not identified at the time of laparoscopy. The low current use of staging laparoscopy in the US suggests an opportunity to improve patient selection and thereby reduce the morbidity of futile laparotomy in this group of patients. Interpreting the findings from patients who did not undergo a staging laparoscopy as part of their management is more difficult. Based on the proportion of patients who underwent staging laparoscopy or laparotomy alone after laparoscopy, one might expect 40% of patients who did not undergo staging laparoscopy to have a futile laparotomy. In actuality, 90% of these patients underwent a therapeutic intervention, with only 10% having a laparotomy alone. There are at least 2 possible explanations for this apparent discrepancy. It is possible that surgeons effectively stratified patient risk of occult metastases and appropriately selected patients at higher risk for metastatic disease to undergo staging laparoscopy. However, the similar tumor and nodal stages between patients who had laparoscopy and those who did not have laparoscopy suggest this was not the case. Another, more compelling explanation is that a substantial proportion of these patients underwent therapeutic intervention in the setting of metastatic disease. It is likely that in some cases, having subjected patients to the morbidity of laparotomy, surgeons chose to perform a palliative resection or bypass despite metastases. The higher ratio of bypass to resection in patients who did not undergo staging laparoscopy supports this hypothesis. Although some surgeons may justify laparotomy in the setting of metastatic disease to perform a bypass, in a series of 165 patients with occult metastases detected at the time of laparoscopy, only 12% subsequently required laparotomy for symptoms. 13 Given the substantially increased morbidity and mortality from laparotomy compared with laparoscopy, this approach should be discouraged. Although the stage of disease was similar between groups, there were some differences in the characteristics of patients who underwent staging laparoscopy compared with those who did not. Patients who had staging laparoscopy were generally younger and had less comorbidity than patients who did not. This trend is counterintuitive be-

5 648 Karanicolas et al Staging Laparoscopy for Gastric Cancer J Am Coll Surg Table 2. Characteristics of Study Cohort by Use of Staging Laparoscopy Characteristic All patients (n 6,388) Staging laparoscopy (n 506) No staging laparoscopy (n 5,882) n % n % n % Age at diagnosis, y , , , , , , , , Sex Male 3, , Female 2, , Race White 4, , Black Asian Hispanic Other/unknown Marital status Married 3, , Not married 2, , Unknown Charlson Comorbidity Score 0 3, , , , , , Region West 3, , Midwest Northeast 1, , South Location in the stomach Cardia/fundus 1, , Body Antrum/pylorus 2, , Lesser curve Greater curve Overlapping/other 1, , T Stage T0/Tis/T1 1, , T2a T2b 1, , T T TX 1, Lymph nodes Negative 2, , Positive 3, , Unknown Grade Well differentiated Moderately differentiated 1, , Poor/undifferentiated 3, , Unknown p Value

6 Vol. 213, No. 5, November 2011 Karanicolas et al Staging Laparoscopy for Gastric Cancer 649 Table 3. Predictors of Staging Laparoscopy Adjusted Predictor odds ratio 95% CI p Value Age at diagnosis, y Ref Sex Male Ref 0.56 Female Race White Ref 0.01 Black Asian Hispanic Other/unknown Marital Status Married Ref 0.06 Not married Unknown Charlson Comorbidity Score 0 Ref Region West Ref 0.01 Midwest Northeast South Location in the stomach Antrum/pylorus Ref 0.01 Cardia/fundus Body Lesser curve Greater curve Overlapping/other Ref, reference group. cause elderly patients with comorbidities are likely to benefit most from avoidance of nontherapeutic laparotomy. Surgeons selected patients with more proximal tumors for laparoscopy, which is appropriate given the increased incidence of occult metastases in patients with proximal tumors. 10 Patients with distal tumors would also be more amenable to palliative bypass, which could influence surgeons away from performing staging laparoscopy. Greater use of laparoscopy in married patients and white patients suggests that social support and socioeconomic status may also influence surgical decisions. Despite the overall low use of staging laparoscopy in this population, there is reason for optimism: the rate of staging laparoscopy doubled between 1998 and This rise was mirrored by an increase in the rate of staging laparoscopy alone, suggesting that despite increasing use of laparoscopy occult metastases were identified in a similar proportion of patients. Even in the most recent year the majority of patients still did not undergo staging laparoscopy, and the rate of futile laparotomy was stable. Enhanced education and outreach to surgeons may help further increase the use of staging laparoscopy in practice. The findings from this population-based study are supported by previous institutional series in which the rate of occult metastases ranged from 13% to 41% Despite advances in cross-sectional imaging, the ability of staging laparoscopy to identify peritoneal-based disease remains unsurpassed, particularly with adjuncts for micrometasta- Table 4. Predictors of Resection Adjusted Predictor odds ratio 95% CI p Value Age at diagnosis, y Ref Sex Male Ref 0.13 Female Race White Ref 0.01 Black Asian Hispanic Other/unknown Marital status Married Ref 0.01 Not married Charlson Comorbidity Score 0 Ref Region West Ref 0.61 Midwest Northeast South Location in the stomach Antrum/pylorus Ref 0.01 Cardia/fundus Body Other Ref, reference group.

7 650 Karanicolas et al Staging Laparoscopy for Gastric Cancer J Am Coll Surg Table 5. First Operative Procedure in Patients who Received Therapeutic Intervention by Use of Staging Laparoscopy Any therapeutic intervention (n 5,610) Staging laparoscopy (n 306) No staging laparoscopy (n 5,304) Procedure n % n % n % Total gastrectomy 1, , Distal gastrectomy 2, , Proximal gastrectomy Other resection Bypass ses such as peritoneal cytology and reverse transcriptase polymerase chain reaction analysis for tumor markers. 18,19 The optimal study design to address the benefits of staging laparoscopy would be a randomized controlled trial of staging laparoscopy compared with no staging laparoscopy. To our knowledge no such trial has been conducted and given the existing evidence, such a trial is unlikely to be conducted in the future. However, the consistent findings in previous single-institutional studies of frequent occult metastases (Table 1), the strong data suggesting that patients with metastatic disease (occult or otherwise) do not benefit from resection, 13 and the minimal morbidity of staging laparoscopy argue strongly in favor of its widespread adoption in the management of patients with gastric cancer. The length of hospitalization (median 2 days) and inhospital mortality rate (5.3%) after staging laparoscopy alone appear remarkably high at first glance. Indeed, staging laparoscopy is frequently performed as an outpatient procedure and the anticipated mortality rate should be low. However when this is compared with the corresponding data for laparotomy alone (10 days and 13.1%, respectively), the advantages are clear. Furthermore, it is important to consider that this represents a group of elderly patients (over age 65) with metastatic gastric cancer. The relatively poor outcomes in these patients reflect the underlying disease process and highlight the importance of avoiding nontherapeutic interventions when possible. The use of staging laparoscopy for gastric cancer in routine practice in the US has not been well described. Coburn and colleagues 20 recently assessed patterns of gastric cancer management in Ontario, Canada. Among 2,399 patients who underwent gastric cancer surgery between 2000 and 2005, 308 (12.8%) had staging laparoscopy. Interestingly, despite the similar overall rate of staging laparoscopy, 66.6% of patients who underwent staging laparoscopy in the Canadian analysis did not have a subsequent therapeutic intervention. This remarkably high rate of laparoscopy alone suggests that Ontario surgeons applied laparoscopy more selectively to patients at high risk of metastatic disease. Several limitations of our study warrant mention. Our findings may not be applicable to patients younger than 65 years. It is conceivable that surgeons perform staging laparoscopy more frequently in younger patients, although elderly patients and those with comorbidities are the most likely to benefit due to the increased risk of complications from laparotomy. Although we were able to control for potential confounding by a number of important sociodemographic and disease characteristics, other factors that may be associated with surgical decision making, such as the patient s functional status and patient and physician preferences, are not available in the SEER-Medicare dataset, so could not be evaluated. Surgical intent (palliative or curative operation) was inferred retrospectively based on the timing of procedures relative to each other. Therefore, we cannot precisely distinguish elective futile laparotomies from urgent exploratory laparotomies in patients who presented with gastric outlet obstruction or bleeding. Despite this limitation, our results support the primary conclusion that staging laparoscopy is infrequently used in the management of patients with gastric adenocarcinoma. CONCLUSIONS In summary, staging laparoscopy is performed infrequently in older patients with gastric adenocarcinoma in the US. Our findings, in combination with those from other studies, suggest that up to 30% of patients could avoid futile laparotomy if staging laparoscopy was performed. Unequivocally, futile laparotomy is associated with substantially longer hospitalization and higher perioperative mortality than staging laparoscopy. Surgeons should be encouraged to offer patients with gastric adenocarcinoma staging laparoscopy before initiating laparotomy. Author Contributions Study conception and design: Karanicolas, Elkin, Jacks, Brennan, Coit Acquisition of data: Karanicolas, Elkin, Jacks Analysis and interpretation of data: Karanicolas, Elkin, Jacks, Atoria, Strong, Brennan, Coit Drafting of manuscript: Karanicolas, Elkin Critical revision: Karanicolas, Elkin, Jacks, Atoria, Strong, Brennan, Coit Acknowledgment: The authors gratefully acknowledge the Applied Research Program, NCI; the Office of Information Services and Office of Strategic Planning, Centers for Medicare & Medicaid Services (CMS); Information Management Services, Inc; and the SEER Program tumor registries for cre-

8 Vol. 213, No. 5, November 2011 Karanicolas et al Staging Laparoscopy for Gastric Cancer 651 ation of the SEER-Medicare dataset. The authors acknowledge Nicole M Ishill, MSc, for assistance with statistical programming. REFERENCES 1. Jemal A, Siegel R, Ward E, et al. Cancer Statistics, CA Cancer J Clin 2008;58: Brennan M. Benefit of aggressive multimodality treatment for gastric cancer. Ann Surg Oncol 1995;2: Martin RI, Jacques D, Brennan M, Karpeh M. Achieving R0 resection for locally advanced gastric cancer: is it worth the risk of multiorgan resection? J Am Coll Surg 2002;194: Asencio F, Aguilo J, Salvador JL, et al. Video-laparoscopic staging of gastric cancer. A prospective multicenter comparison with noninvasive techniques. Surg Endosc 1997;11: D Ugo DM, Persiani R, Caracciolo F, et al. Selection of locally advanced gastric carcinoma by preoperative staging laparoscopy. Surg Endosc 1997;11: Kriplani AK, Kapur BM. Laparoscopy for pre-operative staging and assessment of operability in gastric carcinoma. Gastrointest Endosc 1991;37: Lowy AM, Mansfield PF, Leach SD, Ajani J. Laparoscopic staging for gastric cancer. Surgery 1996;119: Possik RA, Franco EL, Pires DR, et al. Sensitivity, specificity, and predictive value of laparoscopy for the staging of gastric cancer and for the detection of liver metastases. Cancer 1986; 58: Romijn MG, van Overhagen H, Spillenaar Bilgen EJ, et al. Laparoscopy and laparoscopic ultrasonography in staging of oesophageal and cardial carcinoma. Br J Surg 1998;85: Sarela A, Lefkowitz R, Brennan M, Karpeh M. Selection of patients with gastric adenocarcinoma for laparoscopic staging. Am J Surg 2006;191: Stell DA, Carter CR, Stewart I, Anderson JR. Prospective comparison of laparoscopy, ultrasonography and computed tomography in the staging of gastric cancer. Br J Surg 1996;83: Yano M, Tsujinaka T, Shiozaki H, et al. Appraisal of treatment strategy by staging laparoscopy for locally advanced gastric cancer. World J Surg 2000;24: ; discussion Sarela A, Miner T, Karpeh M, et al. Clinical outcomes with laparoscopic stage M1, unresected gastric adenocarcinoma. Ann Surg 2006;243: Burke E, Karpeh M, Conlon K, Brennan M. Laparoscopy in the managment of gastric adenocarcinoma. Ann Surg 1997;225: Karanicolas PJ, Dubois L, Colquhoun PH, et al. The more the better?: the impact of surgeon and hospital volume on inhospital mortality following colorectal resection. Ann Surg 2009;249: Warren JL, Klabunde CN, Schrag D, et al. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care 2002;40:IV Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol 2000;53: Bentrem D, Wilton A, Mazumdar M, et al. The value of peritoneal cytology as a preoperative predictor in patients with gastric carcinoma undergoing a curative resection. Ann Surg Oncol. 2005;12: Dalal K, Woo Y, Kelly K, et al. Detection of micrometastases in peritoneal washings of gastric cancer patients by the reverse transcriptase polymerase chain reaction. Gastric Cancer 2008;11: Coburn NG, Lourenco LG, Rossi SE, et al. Management of gastric cancer in Ontario. J Surg Oncol 2010;102:54 63.