Association of Surgical Treatment, Systemic Therapy, and Survival in Patients With Abdominal Visceral Melanoma Metastases, 1965-2014: Relevance of Surgical Cure in the Era of Modern Systemic Therapy

Key Points

Question  What is the role of surgical resection of abdominal visceral melanoma metastases in the era of modern systemic therapy?

Findings  In this review of 1623 patients with abdominal visceral melanoma metastases, median overall survival was superior in surgical (18.0 months) vs nonsurgical (7.0 months) patients. The most favorable 1-year and 2-year overall survival was seen after surgery for gastrointestinal tract and liver metastases, respectively, and the era of systemic treatment did not have a significant effect.

Meaning  Surgical resection of abdominal visceral melanoma metastases continues to afford patients durable long-term survival irrespective of advances in systemic treatment.

Importance  Systemic therapy for metastatic melanoma has evolved rapidly during the last decade, and patient treatment has become more complex.

Objective  To evaluate the survival benefit achieved through surgical resection of melanoma metastatic to the abdominal viscera in patients treated in the modern treatment environment.

Design, Setting, and Participants  This retrospective review of the institutional melanoma database from the John Wayne Cancer Institute at Providence St Johns Health Center, a tertiary-level melanoma referral center, included 1623 patients with melanoma diagnosed as having potentially resectable abdominal metastases before (1969-2003) and after (2004-2014) advances in systemic therapy.

Main Outcomes and Measures  Overall survival (OS).

Results  Of the 1623 patients identified in the database with abdominal melanoma metastases, 1097 were men (67.6%), and the mean (SD) age was 54.6 (14.6) years. Of the patients with metastatic melanoma, 1623 (320 [19.7%] in the 2004-2014 period) had abdominal metastases, including 336 (20.7%) with metastases in the gastrointestinal tract, 697 (42.9%) in the liver, 138 (8.5%) in the adrenal glands, 38 (2.3%) in the pancreas, 109 (6.7%) in the spleen, and 305 (18.8%) with multiple sites. Median OS was superior in surgical (n = 392; 18.0 months) vs nonsurgical (n = 1231; 7.0 months) patients (P < .001). The most favorable 1-year and 2-year OS was seen after surgery for gastrointestinal tract (52% and 41%) and liver (51% and 38%) metastases, respectively. Multivariable analysis found increasing age (hazard ratio [HR], 1.01; 95% CI, 1.00-1.01; P = .02) and the presence of ulceration (HR, 1.21; 95% CI, 1.01-1.45; P = .04) were associated with a worse OS. Alternatively, treatment with metastasectomy (HR, 0.59; 95% CI, 0.46-0.74; P < .001) and metastases involving the gastrointestinal tract (HR, 0.65; 95% CI, 0.48-0.87; P = .004) were associated with a better OS. The systemic treatment era did not significantly affect outcomes (HR, 0.82; 95% CI, 0.67-1.02; P = .15). Overall, patients with gastrointestinal tract metastases undergoing complete, curative resection derived the greatest benefit, with a median OS of 64 months.

Conclusions and Relevance  To our knowledge, this series is the largest single-institution experience with abdominal melanoma metastases, demonstrating that surgical resection remains an important treatment consideration even in the systemic treatment era.

Metastasectomy is a proven treatment approach in controlling widespread tumor dissemination.¹ In certain malignancies, such as colorectal cancer, surgical treatment of metastases complements systemic chemotherapy and offers the best chance for cure in patients with isolated liver involvement.² Five-year survival rates in surgically treated patients with stage IV colorectal cancer have consistently been reported to range from 30% to 60%.^3-6 Similarly encouraging results have been achieved in highly selected patients with renal cell carcinoma,^7,8 with up to 50% experiencing no recurrence at a median of 43 weeks and 95% surviving at 109 weeks.⁹ To our knowledge, surgical outcomes in patients with metastatic melanoma have not been clearly defined in the era of evolving and effective systemic therapeutics.

Historically, metastasectomy was the sole modality offering long-term survival in patients with stage IV melanoma. Several reports demonstrated acceptable outcomes after surgical treatment of metastases in select melanoma cases involving the gastrointestinal (GI) tract, liver, adrenal, pancreas, and spleen, with median overall survival (OS) of more than 1 year and potentially up to 27 months.^1,10-13 The most effective systemic chemotherapy (ie, dacarbazine) and immune modulating agents (ie, interleukin-2) prior to the modern era produced response rates of 15% to 20%, in exchange for significant toxicity.¹⁴ These older systemic treatments have given way to a newer class of immune and targeted therapies that have shown lasting results.^15-22 While many systemic therapies, like ipilimumab, started receiving US Food and Drug Administration approval in 2011, they have been available through clinical trials since the early 2000s. For example, clinical trials investigating the efficacy of ipilimumab, a monoclonal antibody that blocks cytotoxic T-lymphocyte antigen-4 and causes a release of the inhibition against the antitumor T-cell response, became available to patients with metastatic melanoma in 2004. These landmark studies demonstrated a durable response to treatment in 20% to 30% of patients with stage IV melanoma,¹⁶ thus ushering in a new era of systemic treatment for melanoma

Given the paradigm shift regarding the treatment of patients with stage IV melanoma, the role of surgical resection for abdominal visceral melanoma metastases remains unclear. Our aim was to evaluate the survival benefit achieved through surgical resection of melanoma metastatic to the abdominal viscera in patients treated in the modern treatment environment.

We used our institutional melanoma database to prospectively collect data on database-identified patients with stage IV melanoma with potentially resectable abdominal metastases before (1969-2003) and after (2004-2014) advances in systemic therapies. Because our institution is a tertiary referral center for melanoma, all patients eligible for systemic therapy were enrolled in clinical trials testing the efficacy of immune and targeted therapy starting in 2004. As a result, this year was chosen to define the beginning of the modern era of systemic treatment, as represented in our database. Ipilimumab was the first immunotherapy available to the patients on trial, but toward the end of the period, several patients gained access to PD-1 inhibitors (ie, nivolumab and pembrolizumab). The timing of systemic therapy varied from patient to patient, although patients received neoadjuvant therapy as a general rule, and metastasectomy was considered for sites of resistance. In some cases, patients who were symptomatic underwent surgical resection prior to initiating systemic therapy. Abdominal metastases were defined as disease involving the GI tract (ie, the stomach, small intestine, large intestine, and rectum), liver (including biliary tract), adrenal glands, pancreas, or spleen. Patients with multiple abdominal sites of disease were grouped into a separate cohort. Patient data (age at stage IV disease and sex) and primary tumor data (Breslow thickness, location, and ulceration) as well as data on metastatic disease (prior stage III, isolated vs systemic metastases, abdominal metastases-free survival, and abdomen as first site of stage IV disease) and treatment information (surgery and immunotherapy) were collected. Surgery was defined as curative or palliative based on intent. Patients with incomplete therapy or survival data were excluded. Subgroup analysis by abdominal disease site was performed. This study of deidentified data was exempted from review by the Western Institutional Review Board. Informed consent was waived, as all data was deidentified and contained in a secure, password-protected database on the institution’s server.

Descriptive statistics with the χ² test (for categorical variables) and t test (for continuous variables) were used to compare variables between the 2 treatment periods. The primary end point was OS from date of stage IV diagnosis. Kaplan-Meier curves were calculated, and differences in treatment period were compared using the log-rank test. Univariable and multivariable analysis using the Cox proportional hazards model was performed to evaluate the influence of various factors, including metastatic site, surgery, and treatment era, on OS. A 2-sided P value less than .05 was considered significant. All analyses were performed using SPSS statistical software version 17.0 (IBM Corporation).

Of 5668 patients identified within the database with stage IV melanoma, 1623 had abdominal metastases, including 336 patients (20.7%) with metastases in the GI tract, 697 (42.9%) in the liver, 138 (8.5%) in the adrenal glands, 38 (2.3%) in the pancreas, 109 (6.7%) in the spleen, and 305 (18.8%) with multiple sites. Patients with GI tract involvement had the best OS outcomes, with 1-year, 2-year, 5-year, and 10-year survivals of 37%, 30%, 23%, and 11%, respectively. Patients with abdominal metastases undergoing surgical resection had a superior median survival compared with patients treated medically (18 months vs 7 months; P < .001). Significant variability was apparent when subdividing patients with stage IV melanoma by site of abdominal metastasis. The most favorable outcomes were observed in patients undergoing surgery for GI tract metastases, with greater than 50% surviving 1 year and almost one-third surviving 5 years. Patients with liver, pancreas, spleen, and even multiple abdominal sites undergoing metastasectomy also exhibited a significant survival benefit vs those treated nonsurgically (Table 1).

Patients were next split into 2 cohorts based on the era of systemic treatment (1969-2003 and 2004-2014). In the earlier period, 1303 patients were evaluated for abdominal metastases vs 320 patients during the more contemporary era. The patients seen in the modern treatment era were younger (mean, 53.1 years vs 60.5 years; P < .001) with a higher rate of head and neck and extremity primary melanomas (43% vs 33%, respectively). Significantly fewer patients were diagnosed as having stage III disease, with longer mean metastasis-free intervals (85.8 months vs 62.4 months) prior to identification of abdominal metastases (Table 2). Patients with abdominal metastases diagnosed during the 2004-2014 systemic treatment era had improved median OS compared with stage IV patients with melanoma diagnosed prior to 2004 (11 months vs 8 months, respectively; P = .003). Surgically treated patients showed a trend toward better survival during the modern era.

Multivariable Cox regression analysis controlled for differences between the 2 systemic treatment eras. Older age at stage IV diagnosis (hazard ratio [HR], 1.01; 95% CI, 1.00-1.01; P = .02) and the presence of ulceration (HR, 1.21; 95% CI, 1.01-1.45; P = .04) were associated with poorer survival. Conversely, GI tract metastases (HR, 0.65; 95% CI, 0.48-0.87; P = .004) and treatment with metastasectomy (HR, 0.59; 95% CI, 0.46-0.74; P < .001) were associated with better OS. Importantly, treatment during the more contemporary systemic treatment era did not have a significant effect on survival (Table 3).

A subgroup analysis was performed on patients with only GI tract metastases because these patients showed the best outcomes of all intra-abdominal sites investigated. Four hundred sixty-one patients with GI tract involvement (336 with GI tract only plus 125 with GI tract and another abdominal site) were analyzed and found to have similar demographic and clinical characteristics compared with the overall cohort. However, a greater proportion (259 of 461 [56.2%]) were treated surgically. Patients undergoing combined treatment with surgery and immunotherapy had a median OS of 23 months, which was similar to those treated with surgery alone (21 months) and significantly longer than those patients receiving either immunotherapy alone (6 months) or no therapy (5 months; P < .001) (Figure 1). Patients whose GI tract metastases were treated surgically with curative intent had a median OS of 64 months. Alternatively, patients receiving palliative surgery or no surgery demonstrated significantly inferior survival (Figure 2).

Interestingly, patients found to have GI tract metastases incidentally on routine surveillance imaging had an improved median OS compared with patients presenting with symptoms (39 months vs 19 months; P < .001). While patients with abdominal pain (yes, 12 months vs no, 26 months; P = .01) or obstructive symptoms (yes, 12 months vs no, 31 months; P < .001) on presentation did worse overall, the presence or absence of anemia did not predict any differences in OS (yes, 26 months vs no, 17 months; P = .28). There was no significant difference in outcomes between metastases discovered on computed tomography compared with positron emission tomography (28 months vs 32 months; P = .69), as well as by laparoscopic vs open surgical approach (32 months vs 22 months; P = .51).

Surgical resection of stage IV disease has increased substantially across several common cancer types, including melanoma, despite the increasing use and efficacy of modern systemic therapy.²³ A 2012 retrospective analysis of data from a multicenter randomized clinical trial highlights the value of metastasectomy in patients with m1c disease.²⁴ In 2014, our institution reported that stabilization on systemic therapy was the most important factor affecting survival in surgically treated patients with melanoma with liver metastases.²⁵ Further, our institution demonstrated that adrenalectomy, either as a complete metastasectomy or targeted to lesions resistant to systemic therapy, has been associated with improved long-term survival in metastatic melanoma.²⁶ In a 2015 study by He et al,²⁷ patients with metastatic melanoma treated with BRAF inhibitor therapy and metastasectomy (14% of whom underwent small bowel resection) demonstrated durable postoperative control of disease in select patients undergoing resection of oligometastatic disease.²⁷

Our study builds on these reports, which together provide solid evidence and emphasize the need for a randomized clinical trial evaluating a combination of modern systemic therapy and selective metastasectomy. Surgical resection was associated with improved median OS at all abdominal sites, with the best results occurring in patients with GI tract–only disease (32% 5-year survival). Surprisingly, the 2004-2014 systemic treatment era did not have a significant effect on survival outcomes.

Patients with metastases to the GI tract undergoing complete, curative resection derived the greatest benefit from metastasectomy, affording patients a 64-month median OS. This could in part be owing to improved imaging techniques (ie, thin-slice computed tomography and positron emission tomography), allowing asymptomatic patients with GI tract involvement to have an earlier diagnosis and surgical treatment.

This study, while comprehensive, was not without limitations. First, surgical intent had to be gleaned from the medical records owing to the study’s retrospective nature. Further, complete data on the specific immunotherapy regimens administered to patients was not available. However, because this study used data from a database from a tertiary referral melanoma center, most patients during the modern era of systemic treatment were evaluated for and treated with modern therapy. Patients were treated according to clinician preference, and treatment decisions were not based on a specific protocol. As a result, the patients within our cohort were likely carefully selected. For example, our institutional approach has been aggressive resection in younger patients resulting in longer disease-free intervals from diagnosis to the appearance of abdominal metastases. Because the study patients received their care during a 40-year period, it is possible that survival may have been affected by improvements in surgical technology, in addition to other refinements that we were unable to capture. With the advent and widespread use of more effective systemic immune and targeted therapy, we believe there is significant potential for a paradigm shift in the treatment of patients with stage IV melanoma with abdominal disease.

To our knowledge, this is the largest single-institution experience with abdominal melanoma metastases, and it demonstrates that surgical resection continues to afford patients durable long-term survival irrespective of advances in systemic treatments. Through application of a multidisciplinary approach, the combination of metastasectomy and cutting-edge systemic therapies could further enhance disease response and its durability.

Corresponding Author: Gary B. Deutsch, MD, MPH, Division of Surgical Oncology, Department of Surgery, Hofstra Northwell School of Medicine, 450 Lakeville Rd, Lake Success, NY 11042 (gdeutsch@northwell.edu).

Accepted for Publication: January 14, 2017.

Correction: This article was corrected on August 29, 2018, to correct pervasive errors in Table 1.

Published Online: April 5, 2017. doi:10.1001/jamasurg.2017.0459

Author Contributions: Dr Deutsch had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Deutsch, Kirchoff, Faries, Bilchik.

Acquisition, analysis, or interpretation of data: Deutsch, Flaherty, Kirchoff, Bailey, Vitug, Foshag, Faries.

Drafting of the manuscript: Deutsch, Flaherty, Bailey, Vitug.

Critical revision of the manuscript for important intellectual content: Deutsch, Kirchoff, Foshag, Faries, Bilchik.

Statistical analysis: Deutsch, Bailey, Vitug.

Obtained funding: Faries.

Administrative, technical, or material support: Deutsch, Kirchoff, Foshag.

Supervision: Deutsch, Foshag, Faries, Bilchik.

Conflict of Interest Disclosures: Dr Faries consults for Genentech Inc and advises for Astellas Pharmaceuticals, Amgen Inc, and Myriad Genetics Inc. No other disclosures are reported.

Funding/Support: Supported by funding provided by the Donald L. Morton, MD, Melanoma Research Program and Gastrointestinal Research Program at the John Wayne Cancer Institute. This study was supported by grant R01 CA189163 from the National Cancer Institute. Funding also provided by the California Oncology Research Institute, Los Angeles, California.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Disclaimer: The content of this report is solely the responsibility of the authors and does not necessarily represent the official view of the National Cancer Institute or the National Institutes of Health.

Previous Presentation: Presented as an oral presentation at the American College of Surgeons Clinical Congress; October 8, 2015; Chicago, Illinois.