Cover StoryPublication Exclusive

Advances in melanoma prompt re-evaluation of SLN biopsy

Sentinel lymph node biopsy has long been the standard staging and prognostic modality for melanoma.

Final results from the Multicenter Selective Lymphadenectomy Trial (MSLT-1) — published in February in The New England Journal of Medicine — confirmed the prognostic value of sentinel lymph node biopsy (SLNB) for patients with intermediate-thickness melanomas.

Ten-year melanoma-specific survival was significantly worse among patients identified as having sentinel-node metastases compared with patients with tumor-free nodes (62.1% vs. 85.1%; HR=3.09; 95% CI, 2.12-4.49). Further, biopsy-based management was associated with improved 10-year melanoma-specific survival (HR=0.56; 95% CI, 0.37-0.84) and distant DFS (HR=0.62; 95% CI, 0.42-0.91) among patients with nodal metastases from intermediate-thickness melanomas.

Final results of the MSLT-1 trial have added to data that support the role and value of sentinel lymph node biopsy in patients with melanoma, according to Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member.

Final results of the MSLT-1 trial have added to data that support the role and value of sentinel lymph node biopsy in patients with melanoma, according to Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member.

Source: Photo courtesy of Moffitt Cancer Center

“I continue to believe very strongly in the value of sentinel node biopsy, and our patients continue to choose it as an ideal way to stage their disease and help them make treatment decisions,” Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member, said in an interview. “Accumulating data, including the final results of MSLT-1, have continued to support the role and value of sentinel node biopsy.”

However, results of MSLT-1 showed SLNB did not improve melanoma-specific survival — the trial’s primary endpoint — in the total study population (81.4% for biopsy vs. 78.3% for observation; HR=0.84; 95% CI, 0.64-1.09). Critics emphasize the melanoma-specific survival and DFS advantages observed in MSLT-1 were results of subgroup analyses.

Despite the broad acceptance of SLNB’s prognostic utility, the lack of a therapeutic advantage has prompted some clinicians to question whether the procedure — which comes at a considerable financial cost, and also carries risks for overtreatment and morbidity — is justified. Research into alternative prognostic molecular assays also is underway.

“This is an ideal time to revisit and reassess the role of SLNB, to see where we are and where we are going,” Sondak said.

HemOnc Today spoke with several surgical oncologists about the clinical utility of SLNB, the controversies surrounding its use, the procedure’s role in specific subsets of patients, and how the procedure’s benefits could be maximized in the era of increasingly effective targeted treatments.

A question of survival

The lack of a survival benefit observed in MSLT-1 fuels most of the controversy that surrounds SLNB.

Clinicians who support the procedure’s use suggest the trial’s mature results provide valuable information beyond prognosis.

“Time is even more important in this study than most because we only learn about patients who have hidden node-positive disease through the passage of time,” Sondak said. “Now we have years of information, and enough data to make long-range projections.”

Although the study defined intermediate-thickness melanomas as those 1.2 to 3.5 mm, supplementary material published with the study provided data using the American Joint Committee on Cancer (AJCC) staging system, which defines intermediate-thickness melanomas as those with Breslow thickness of 1 mm to 4 mm.

“The data using this definition really helps support our day-to-day discussions with patients,” Sondak said. “It has made our decisions more focused, and our patients’ decisions more confident.”

Despite these advantages, some researchers say the lack of a survival advantage — reconfirmed by the final results of MSLT-1 — overshadow any secondary outcomes of the trial.

Mary S. Brady, MD, FACS

Mary S. Brady

“We have always known that preemptive removal of regional nodes increases DFS. That has been re-demonstrated by the recent, final results from the trial,” Mary S. Brady, MD, FACS, a surgical oncologist at Memorial Sloan Kettering Cancer Center, told HemOnc Today. “The final results also show once again there is no improvement in melanoma-specific survival in the entire group of patients. So, it is a negative trial, because melanoma-specific survival was the primary endpoint. All of the other findings were secondary objectives, and for good reason.”

The procedure — by nature — is not intended to improve survival, Robert H.I. Andtbacka, MD, CM, associate professor of surgery at Huntsman Cancer Institute at the University of Utah and a HemOnc Today Editorial Board member, said in an interview.

“This is a prognostic test, and its role really is to evaluate regional lymph node metastases,” Andtbacka said. “We have other tests that serve similar roles in evaluating distant metastases, such as CT scans, that do not necessarily improve survival.”

However, improved survival should be the primary goal, James Grichnik, MD, PhD, professor in the department of dermatology and cutaneous surgery at the University of Miami Miller School of Medicine and director of the Anna Fund Melanoma Program at the Sylvester Comprehensive Cancer Center, said in an interview.

“I think we all wanted this procedure to improve OS,” Grichnik said. “We would like to think that the lymph node chain is a filter, and that by removing it we are going to save lives. But sentinel node biopsy and completion lymph node dissection do not appear to do that.”

Subgroup analyses that demonstrated a survival benefit in limited populations should be interpreted with caution, researchers said.

Giorgos C. Karakousis, MD

Giorgos C. Karakousis

“As with any trial, there is skepticism as to the truth behind subgroup analyses,” Giorgos C. Karakousis, MD, assistant professor of surgery at the Hospital of the University of Pennsylvania, told HemOnc Today. “Several statistical analyses were performed in the latest presentation of the MSLT-1 data to try to demonstrate that the subgroup analyses were valid, and in truth there is not really a better way to get at the question at present, because you do not know upfront who is going to have a positive sentinel lymph node.”

How these data are presented is key.

“We need to take the data as it stands and not how it has been filtered or interpreted post-hoc,” Brady said. “I think the way it has been presented has led a lot of clinicians to overselling sentinel node biopsy.”

Despite the lack of a survival advantage, the data provided by MSLT-1 are critical to the current standards in melanoma, Sondak said.

“It is puzzling at times that the vitriol gets so focused on this particular endpoint, because in evidence-based medicine, we don’t say, ‘Here’s a negative endpoint. Throw all the data away, pay no attention, and lose all the information from treating and following 2,000 patients for decades’ simply because you can stamp a big red ‘negative’ on the cover of the study,” Sondak said. “That is not how we practice modern medicine.”

Cost and morbidity

Detractors of SLNB warn that the lack of survival benefit is accompanied by morbidity and costs.

A study by Wrightson and colleagues, published in 2003 in Annals of Surgical Oncology, demonstrated a 4.6% risk for major and minor complications with SLNB. Complications include lymphedema, chronic axillary pain and hematoma infection. Other studies have shown similar rates for harm.

“The morbidity from the procedure is extremely low, and it is very well tolerated,” Andtbacka said. “Nevertheless, we use 5% as the cutoff for potential morbidity from the procedure. If the risk of lymph node involvement is 5% or greater, then we would recommend sentinel node biopsy.

Others contend SLNB can lead to unnecessary surgery.

“Benign melanocytic lesions such as blue nevi, Spitz nevi and congenital nevi can release cells into lymph nodes, so just because there’s a cell in the lymph node does not necessarily mean it is malignant,” Grichnik said. “There is no guarantee that a lymph node with a few melanocytic cells in it is going to end up being clinically palpable and lead to fully metastatic disease.”

However, mature data from MSLT-1 eased concerns about false-positive SLNB results, Sondak said.

“Patients who have positive sentinel nodes are patients who, if nothing else is done, sooner or later will have a lymph node that shows up as a palpable node,” Sondak said. “In patients with thick melanoma on the trial, every single case on the sentinel node biopsy arm that was positive was matched with a patient on the observation arm who had a positive node develop clinically.”

Questions also persist about SLNB’s cost-effectiveness.

“There are certainly not insignificant costs associated with the procedure,” Karakousis said. “In this health care climate — in which costs become more important — the cost-to-benefit ratio is something that is going to have to be carefully examined.”

The fact that the procedure cannot be completed in-office adds to the cost.

“You have tremendous cost issues, because wide excision alone — which for most patients could have been completed in-office — now has to be performed in the operating room under general anesthesia to allow for sentinel node staging,” Brady said. “It’s a real commitment socially to pay for a prognostic tool in the absence of any level one data that using it will improve survival.”

However, SLNB adds minimal costs to wide excision, Andtbacka said.

“If you did not complete sentinel node biopsy, 21% of patients would recur based on MSLT-1 data,” Andtbacka said. “What is the cost and morbidity of treating patients at that time? A study on the full cost value has not been completed, but you have to assess this from multiple perspectives.”

Patients’ fears about whether their disease has spread to the lymph nodes is an indirect cost of forgoing SLNB, Andtbacka said. Karakousis expressed a similar sentiment.

“It is hard to put a price on prognostic information, or the ability to tell patients whether you know if they have a favorable tumor,” Karakousis said. “It is not something that can readily be quantified.”

Thin melanomas

A thorough evaluation of the risk–benefit ratio is especially pertinent for thin melanomas, or those less than 1 mm in Breslow thickness. An estimated 60% to 70% of patients with melanoma fall in this category. Although SLNB is not universally recommended for patients with thin melanomas, the MSLT-1 trial contained inadequate data to draw any conclusions in this area.

Guidelines for thin melanomas also are scant. The joint ASCO and Society of Surgical Oncology guidelines state “there is insufficient evidence to support routine SLNB” in this setting. National Comprehensive Cancer Network guidelines state: “Because patients with thin melanoma have a generally favorable prognosis, the role of SLNB in this cohort is unclear.”

Clinicians take several clinical and pathologic factors into account to determine a patient’s risk for positive nodes, Karakousis said. These factors help clinicians weigh the likelihood of finding nodal metastasis against the potential risks of the procedure. Common pathologic factors considered include the depth of the thin lesion, the mitotic index, lymphovascular invasion, Clark level, ulceration status and deep margin status.

“If you have a patient with a melanoma thicker than .75 mm that’s Clark level IV/V with mitoses, for instance, their risk for sentinel node positivity may be on the order of 8% to 10%, and that would be in the range of some patients whose melanoma is just over 1 mm in thickness for whom we routinely recommend sentinel node biopsy,” Karakousis said. “If we are going to offer it to that group, it seems justified that we would offer it to carefully selected patients with thin lesions who have characteristics associated with an increased risk of sentinel lymph node metastasis.”

A study published last year in Journal of Clinical Oncology suggests that SLNB may be indicated for patients with melanomas as thin as 0.75 mm. In the study, Han and colleagues found Breslow thickness ≥0.75 mm was significantly associated with sentinel node positivity (P=.03), and 6.3% of these patients had node-positive disease.

However, Brady said she tends not to recommend SLNB for patients with thin melanomas.

“In most patients with thin melanoma, the risk of a positive node is 5% to 7%,” Brady said. “This means you are undertaking the risks of the surgical procedure in a setting where you have a 93% to 95% chance of having negative nodes.”

Because of this lack of clarity, the decision to use SLNB in thin melanomas often comes down to patient preferences.

“We need to avoid major blanket statements in this group, and take everything into consideration about the individual tumor and patient to fine-tune the decision,” Sondak said. “The decision-making regarding thin melanomas is one of the most personalized decisions we have to make.”

Pediatric melanomas

The role of SLNB in pediatric melanoma also is a subject of intense debate, and questions surrounding the use of the procedure in this patient population are particularly pertinent due to the increasing incidence of pediatric melanoma in the United States.

A study by Wong and colleagues, published in 2013 in Pediatrics, showed incidence of pediatric melanoma has increased an average of 2% (95% CI, 1.4-2.7) each year from 1973 to 2009.

The challenges clinicians face during treatment of adults with melanoma are magnified in the pediatric setting. For example, a paper by Niebling and colleagues published in July in Annals of Surgical Oncology showed 5.1% of melanoma diagnoses changed after pathologic review. The rate for discrepancy often is much higher in pediatric patients, Sondak said.

“Sometimes the pathologist has a high degree of uncertainty about the exact pathology that adds to the problematic nature of having a young child with a cancer diagnosis,” Sondak said. “We have found that SLNB is helpful in dealing even with those cases in which the malignant diagnosis is uncertain because it can eliminate remaining uncertainty.”

Still, there is an increased risk for false-positive SLNB in pediatric patients given that benign nevi also can seed nodes.

“As patients get older, the number of real melanomas goes up, and the number of growing moles goes down,” Grichnik said. “But in children and young adults, the number of growing moles is huge and the number of melanomas is rare.”

The incidence of lymph node metastases is more prevalent in younger patients, even though survival often is favorable. A study by Averbrook and colleagues, published in 2012 in Cancer, showed 30% of pediatric patients who underwent SLNB had node-positive disease. However, 10-year OS ranged from 100% of those aged birth to 10 years, to 79.5% of patients aged older than 15 years.

“It is a little counterintuitive, in that younger patients appear to have a higher incidence of nodal metastases but a better OS compared to older patients with similar clinical stage tumors,” Karakousis said. “Moreover, the value of SLN biopsy in pediatric patients with melanocytic tumors of uncertain malignant potential — in which a pathologic diagnosis of melanoma cannot be firmly established — is less clear. These patients, when atypical cells are identified in their lymph nodes, tend to have a very favorable prognosis making it challenging to interpret the clinical significance of their SLN findings.”

Yet, SLNB can help provide some clarity, Sondak said.

“This is one of the real gray areas of modern medicine, and we are using SLNB as one piece of evidence among many that influences how we treat our children,” Sondak said.

The future of SLNB

Anti-CTLA-4 and PD-1 antibodies, along with BRAF- and MEK-targeted inhibitors, are poised to change treatment standards in melanoma.

SLNB can help researchers identify high-risk patients eligible to participate in clinical trials designed to assess the efficacy of these novel agents.

“There are many newer treatments that are being evaluated, and knowing the exact tumor burden and whether there is presence of lymph node metastases is extremely important,” Andtbacka said. “The role of a sentinel lymph node biopsy in that setting has become absolutely essential so that we can have better-defined patient populations in adjuvant studies.”

Although the therapeutic value of these targeted agents in the adjuvant setting is yet unknown, these therapies have the potential to improve OS after SLNB, Karakousis said.

The identification of biologic markers to predict recurrence — particularly among patients who have a negative sentinel node — also could help improve the prognostic utility of SLNB, Andtbacka said.

“It is critical that we have more data and better markers to determine the patients who will and will not benefit from adjuvant therapy,” Andtbacka said. “We are getting a much better understanding of the biology of the disease, and I am very hopeful that within the next 5 years, the way we look at melanoma from risk-of-recurrence and staging perspectives will change and incorporate more biological prognostic factors.”

Molecular diagnostic tools or circulating tumor cell assays may capture some of these biological risk factors, Grichnik said.

Data presented at ASCO by Lawson and colleagues showed that a gene-expression profile test and SLNB demonstrated comparable positive predictive values for distant metastases in node-positive patients (50% vs. 55%), yet the negative predictive value was greater with the gene assay (82% vs. 67%).

Further, patients who had a low-risk gene test result and negative SLNB demonstrated increased 5-year distant metastasis-free survival than patients with a negative SLNB alone (86% vs. 64%).

Some researchers say identification of high-risk markers in the primary tumor may eliminate the need for SLNB.

“I foresee a future in which we can remove a lesion with pathologic characteristics suggestive of melanoma and, through molecular analyses, determine the lethal (or non-lethal) potential of that lesion and best therapies for that patient,” Grichnik said. “It may very well be that if they have a high-risk lesion, we should immediately start them on an adjuvant protocol without having to check the sentinel node.”

Any attempt to eliminate SLNB from the prognostic paradigm would be met with considerable resistance.

“There is a lot of surgical history in removing lymph nodes from melanoma, and so a lot of people are holding on to concepts that in the past they have felt very strongly about,” Brady said. “It is very difficult to get people to rethink whether this is a valuable procedure.”

Yet, frequent re-evaluation is necessary to ensure progress, Grichnik said.

“SLNB currently has a role, and it clearly has prognostic and staging value,” Grichnik said. “We just have to be careful deciding something is the gold standard, and then never challenging it.”

Sondak agreed that the knowledge gained through molecular analyses can improve upon what researchers have learned with SLNB.

“If somebody comes along tomorrow with a new test for prognosis and prediction of outcome in melanoma, they need to show us how it adds to what we already know from sentinel node biopsy,” Sondak said. “It was a good platform, and just as Clark level and Breslow thickness got us to one place, sentinel node biopsy got us to the next. We certainly hope the future will show molecular analyses can get us to a new place that is even further along for our patients.” — by Alexandra Todak

References:

Averbrook BJ. Cancer. 2014;doi:10.1002/cncr.28289.

Coit DG. J Natl Compr Canc Netw. 2013;11:395-407.

Han D. J Clin Oncol. 2013;31:4387-4393.

Lawson DH. Abstract #9022. Presented at: ASCO Annual Meeting; May 30-June 3, 2014; Chicago.

Morton DL. New Engl J Med. 2014;370:599-609.

Niebling MG. Ann Surg Oncol. 2014;21:2245-2251.

Wong JR. Pediatrics. 2013;131:846-854.

Wong SL. J Clin Oncol. 2012;30:2912-2918.

Wrightson WR. Ann Surg Oncol. 2003;10:676-680.

For more information:

Robert H.I. Andtbacka, MD, CM, can be reached at Huntsman Cancer Institute, Patient Care Center, 2000 Circle of Hope, Salt Lake City, UT 84112; email: robert.andtbacka@hci.utah.edu.

Mary S. Brady, MD, FACS, can be reached at Memorial Sloan Kettering Cancer Center, 1275 York Ave., New York, NY 10065; email: bradym@mskcc.org.

James Grichnik, MD, PhD, can be reached at Sylvester Comprehensive Cancer Center, 1475 NW 12th Ave., Outpatient Clinic, Miami, FL 33136; email: jgrichnik@med.miami.edu.

Giorgos C. Karakousis, MD, can be reached at Hospital of the University of Pennsylvania, Department of Surgery, 3400 Spruce St., 4 Silverstein, Philadelphia, PA 19104; email: giorgos.karakousis@uphs.upenn.edu.

Vernon K. Sondak, MD, can be reached at Moffitt Cancer Center, 12902 Magnolia Drive, Tampa, FL 33612; email: vernon.sondak@moffitt.org.

Disclosure: Grichnik reports consultant roles with Amgen, Castle Biosciences and Novartis. Karakousis reports consultant fees from Amgen. Sondak reports personal fees from and consultant/advisory roles with Amgen, Bristol-Myers Squibb, GlaxoSmithKline, OncoSec, MabVax, Merck, Novartis and Polynoma. Andtbacka and Brady report no relevant financial disclosures. 

 

 POINTCOUNTER

Should all patients with microscopic positive nodes undergo completion lymph node dissection?

POINT

Additional surgery is the standard for sentinel node-positive melanoma.

Mark B. Faries, MD, FACS

Mark B. Faries

Most stage III melanoma is now diagnosed early, through sentinel lymph node biopsy (SLNB). When it was originally conceived, SLNB was intended only to determine the status of basins at risk, and not to render any benefit beyond staging. However, it has become evident that most patients with sentinel node metastases have all of their nodal disease removed with the SLNB, which has led to the question of whether the larger, more morbid completion dissection is warranted. The equipoise on this issue provided justification for the second Multicenter Selective Lymphadenectomy Trial (MSLT-II), in which patients will be randomly assigned to completion dissection or to clinical follow-up with nodal ultrasound. That trial has completed accrual and will provide the answer to this question.

In the meantime, completion lymph node dissection (CLND) remains the “standard” option, though as is pointed out in the accompanying piece, there are reasons to question that standard. There are clear reasons, though, for continuing this practice, at least until we know the result of MSLT-II. The first is completeness of staging. The status of non-sentinel lymph nodes provides important prognostic information beyond that of the sentinel node, which has been demonstrated in several retrospective series. Patients with non-SLN metastases have a significantly worse long-term outlook, which appears similar to patients with clinically detectable disease (stage IIIB/IIIC). This added information may be helpful as patients weigh options for currently available toxic adjuvant therapies with modest clinical activity.

Approximately 10% to 20% of patients with sentinel node metastases will have additional non-SLN involvement, which is likely to result in regional nodal recurrences over time. These recurrences, and the distress they induce, can be substantially reduced by immediate completion dissection. Although nodal disease can frequently be removed at the time of nodal recurrence, data from MSLT-I suggest that the morbidity of those procedures, specifically lymphedema, may be increased if the operation is delayed.

Finally, at present, eligibility for adjuvant clinical trials generally requires CLND. Although one may question this requirement, it remains common.

Overall, completion dissection remains the standard, though an informed, reasonable patient may elect to choose a different course. Discussion of options, though, should include clear enumeration of the established benefits of dissection on staging, regional disease control and clinical trial eligibility.

Mark B. Faries, MD, FACS, is director of the Donald L. Morton Melanoma Research Program at John Wayne Cancer Institute.  He can be reached at mark.faries@jwci.org.  Disclosure: Faries reports consultant/advisory board roles with Amgen, Astellas and Genentech.

References:

Bamboat ZM. Ann Surg Oncol. 2014;21:3117-3123.

Faries MB. Ann Surg Oncol. 2010;17:3324-3329.

Leung AM. JAMA Surg. 2013;148:879-884.

Reintgen M. Ann Surg Oncol. 2013;20:668-674.

COUNTER

It may be very reasonable to offer observation as an alternative to complete lymph node dissection in selected patients with microscopic positive sentinel lymph nodes.

Michael S. Sabel, MD

Michael S. Sabel

Our tradition in oncology, particularly surgical oncology, is often to offer the most aggressive therapy as “standard of care,” and then require prospective, randomized data to pull back from that. The logic behind this is clear, but the cost remains the exposure of a large group of patients to the morbidity of therapy to benefit what might be a small minority of patients.

Despite a growing desire to create narrow guidelines, well-defined standards and strict quality measures, the true role of the surgeon is to educate patients about the risks and benefits of a procedure so they can choose what is appropriate for them. It cannot be considered inappropriate for a patient to consider a small — and perhaps unknown — benefit against a considerable impact on quality of life, and conclude that the treatment is not for them.

Although it is certainly possible that the survival benefits of CLND for a positive SLN outweigh the potential risks, it is unclear to me how we can be so dogmatic in that belief given our difficulty in clearly demonstrating a survival benefit to either elective node dissection or SLNB as compared with observation alone. Although we actively await the results of a randomized trial, prompted by the strong possibility that completion node dissection offers no significant benefit to survival, we have only retrospective data to consider. Those few studies demonstrate no difference in survival, and perhaps regional recurrence, between those patients who undergo CLND and those who do not. And, although the criticism of these studies is the selective bias inherent to retrospective review, this does not invalidate the results, but rather supports the idea that with proper selection, we can avoid surgery in a subset of patients.

The remaining questions are: How large is that subset and how is it defined? By age, co-morbidities, risk of distant disease, tumor burden within the SLN or a combination of factors? Stratifying the potential benefit is critical, including not just OS but also regional control and how the information may influence adjuvant therapy decisions. These must then be weighed against the varying risks, which — like the benefits — are different from patient to patient, based on their co-morbidities and the site of the node dissection. The conversation must therefore be different, for example, between a 70-year-old obese patient facing an inguinal lymph node dissection for a few immunohistochemistry-positive cells and a 40-year-old patient considering axillary lymph node dissection for a SLN with 50% replacement.

It is disingenuous to have the equipoise to offer no CLND on a clinical trial but then not consider it a viable option off of a clinical trial, or to consider it a violation of quality care. Conversely, given the high number of patients not undergoing CLND for a positive SLN, it also is wrong to inform SLN-positive patients they don’t need a CLND, rather than presenting the potential risks and benefits and letting them choose. The art of oncology lies in a balanced presentation based on a thorough understanding of the available data, and letting patients maintain autonomy. Just as it is not unreasonable for a patient to either decline or desire adjuvant systemic therapy based on a risk-to-benefit analysis, it is not unreasonable for any individual patient to conclude that the present level of evidence does not support CLND for a positive SLN in their particular situation. As researchers, it becomes even more incumbent upon us to continue to generate data, both prospective and retrospective, to improve our ability to target our surgical therapies where they will do the most good.

Michael S. Sabel, MD, is associate professor of surgery at the University of Michigan. He can be reached at University of Michigan Health Systems, 1500 E. Medical Center Drive, 3303 Cancer Geriatric Center SPC 5932, Ann Arbor, MI 48109; email: msabel@umich.edu. Disclosure:  Sabel reports an advisory board role with Merck.

References:

Balch CM. Ann Surg. 1996;224:255-266.

Bilimoria KY. Ann Surg Oncol. 2008;15:683-690.

Cascinelli N. Lancet. 1998;351:793-796.

Frankel TL. Ann Surg Oncol. 2008;15:2403-2411.

Gershenwald JE. J Clin Oncol. 2008;26:4296-4303.

Kingham TP. Ann Surg Oncol. 2010;17:514-520.

Morton DL. N Engl J Med. 2014;370:599-609.

van Akkooi AC. Ann Surg. 2008; 248: 949-955.

Wong SL. Ann Surg Oncol. 2006;13:809-816.

Sentinel lymph node biopsy has long been the standard staging and prognostic modality for melanoma.

Final results from the Multicenter Selective Lymphadenectomy Trial (MSLT-1) — published in February in The New England Journal of Medicine — confirmed the prognostic value of sentinel lymph node biopsy (SLNB) for patients with intermediate-thickness melanomas.

Ten-year melanoma-specific survival was significantly worse among patients identified as having sentinel-node metastases compared with patients with tumor-free nodes (62.1% vs. 85.1%; HR=3.09; 95% CI, 2.12-4.49). Further, biopsy-based management was associated with improved 10-year melanoma-specific survival (HR=0.56; 95% CI, 0.37-0.84) and distant DFS (HR=0.62; 95% CI, 0.42-0.91) among patients with nodal metastases from intermediate-thickness melanomas.

Final results of the MSLT-1 trial have added to data that support the role and value of sentinel lymph node biopsy in patients with melanoma, according to Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member.

Final results of the MSLT-1 trial have added to data that support the role and value of sentinel lymph node biopsy in patients with melanoma, according to Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member.

Source: Photo courtesy of Moffitt Cancer Center

“I continue to believe very strongly in the value of sentinel node biopsy, and our patients continue to choose it as an ideal way to stage their disease and help them make treatment decisions,” Vernon K. Sondak, MD, chair of cutaneous oncology at Moffitt Cancer Center and a HemOnc Today Editorial Board member, said in an interview. “Accumulating data, including the final results of MSLT-1, have continued to support the role and value of sentinel node biopsy.”

However, results of MSLT-1 showed SLNB did not improve melanoma-specific survival — the trial’s primary endpoint — in the total study population (81.4% for biopsy vs. 78.3% for observation; HR=0.84; 95% CI, 0.64-1.09). Critics emphasize the melanoma-specific survival and DFS advantages observed in MSLT-1 were results of subgroup analyses.

Despite the broad acceptance of SLNB’s prognostic utility, the lack of a therapeutic advantage has prompted some clinicians to question whether the procedure — which comes at a considerable financial cost, and also carries risks for overtreatment and morbidity — is justified. Research into alternative prognostic molecular assays also is underway.

“This is an ideal time to revisit and reassess the role of SLNB, to see where we are and where we are going,” Sondak said.

HemOnc Today spoke with several surgical oncologists about the clinical utility of SLNB, the controversies surrounding its use, the procedure’s role in specific subsets of patients, and how the procedure’s benefits could be maximized in the era of increasingly effective targeted treatments.

A question of survival

The lack of a survival benefit observed in MSLT-1 fuels most of the controversy that surrounds SLNB.

Clinicians who support the procedure’s use suggest the trial’s mature results provide valuable information beyond prognosis.

“Time is even more important in this study than most because we only learn about patients who have hidden node-positive disease through the passage of time,” Sondak said. “Now we have years of information, and enough data to make long-range projections.”

Although the study defined intermediate-thickness melanomas as those 1.2 to 3.5 mm, supplementary material published with the study provided data using the American Joint Committee on Cancer (AJCC) staging system, which defines intermediate-thickness melanomas as those with Breslow thickness of 1 mm to 4 mm.

“The data using this definition really helps support our day-to-day discussions with patients,” Sondak said. “It has made our decisions more focused, and our patients’ decisions more confident.”

Despite these advantages, some researchers say the lack of a survival advantage — reconfirmed by the final results of MSLT-1 — overshadow any secondary outcomes of the trial.

Mary S. Brady, MD, FACS

Mary S. Brady

“We have always known that preemptive removal of regional nodes increases DFS. That has been re-demonstrated by the recent, final results from the trial,” Mary S. Brady, MD, FACS, a surgical oncologist at Memorial Sloan Kettering Cancer Center, told HemOnc Today. “The final results also show once again there is no improvement in melanoma-specific survival in the entire group of patients. So, it is a negative trial, because melanoma-specific survival was the primary endpoint. All of the other findings were secondary objectives, and for good reason.”

PAGE BREAK

The procedure — by nature — is not intended to improve survival, Robert H.I. Andtbacka, MD, CM, associate professor of surgery at Huntsman Cancer Institute at the University of Utah and a HemOnc Today Editorial Board member, said in an interview.

“This is a prognostic test, and its role really is to evaluate regional lymph node metastases,” Andtbacka said. “We have other tests that serve similar roles in evaluating distant metastases, such as CT scans, that do not necessarily improve survival.”

However, improved survival should be the primary goal, James Grichnik, MD, PhD, professor in the department of dermatology and cutaneous surgery at the University of Miami Miller School of Medicine and director of the Anna Fund Melanoma Program at the Sylvester Comprehensive Cancer Center, said in an interview.

“I think we all wanted this procedure to improve OS,” Grichnik said. “We would like to think that the lymph node chain is a filter, and that by removing it we are going to save lives. But sentinel node biopsy and completion lymph node dissection do not appear to do that.”

Subgroup analyses that demonstrated a survival benefit in limited populations should be interpreted with caution, researchers said.

Giorgos C. Karakousis, MD

Giorgos C. Karakousis

“As with any trial, there is skepticism as to the truth behind subgroup analyses,” Giorgos C. Karakousis, MD, assistant professor of surgery at the Hospital of the University of Pennsylvania, told HemOnc Today. “Several statistical analyses were performed in the latest presentation of the MSLT-1 data to try to demonstrate that the subgroup analyses were valid, and in truth there is not really a better way to get at the question at present, because you do not know upfront who is going to have a positive sentinel lymph node.”

How these data are presented is key.

“We need to take the data as it stands and not how it has been filtered or interpreted post-hoc,” Brady said. “I think the way it has been presented has led a lot of clinicians to overselling sentinel node biopsy.”

Despite the lack of a survival advantage, the data provided by MSLT-1 are critical to the current standards in melanoma, Sondak said.

“It is puzzling at times that the vitriol gets so focused on this particular endpoint, because in evidence-based medicine, we don’t say, ‘Here’s a negative endpoint. Throw all the data away, pay no attention, and lose all the information from treating and following 2,000 patients for decades’ simply because you can stamp a big red ‘negative’ on the cover of the study,” Sondak said. “That is not how we practice modern medicine.”

Cost and morbidity

Detractors of SLNB warn that the lack of survival benefit is accompanied by morbidity and costs.

A study by Wrightson and colleagues, published in 2003 in Annals of Surgical Oncology, demonstrated a 4.6% risk for major and minor complications with SLNB. Complications include lymphedema, chronic axillary pain and hematoma infection. Other studies have shown similar rates for harm.

“The morbidity from the procedure is extremely low, and it is very well tolerated,” Andtbacka said. “Nevertheless, we use 5% as the cutoff for potential morbidity from the procedure. If the risk of lymph node involvement is 5% or greater, then we would recommend sentinel node biopsy.

Others contend SLNB can lead to unnecessary surgery.

“Benign melanocytic lesions such as blue nevi, Spitz nevi and congenital nevi can release cells into lymph nodes, so just because there’s a cell in the lymph node does not necessarily mean it is malignant,” Grichnik said. “There is no guarantee that a lymph node with a few melanocytic cells in it is going to end up being clinically palpable and lead to fully metastatic disease.”

However, mature data from MSLT-1 eased concerns about false-positive SLNB results, Sondak said.

PAGE BREAK

“Patients who have positive sentinel nodes are patients who, if nothing else is done, sooner or later will have a lymph node that shows up as a palpable node,” Sondak said. “In patients with thick melanoma on the trial, every single case on the sentinel node biopsy arm that was positive was matched with a patient on the observation arm who had a positive node develop clinically.”

Questions also persist about SLNB’s cost-effectiveness.

“There are certainly not insignificant costs associated with the procedure,” Karakousis said. “In this health care climate — in which costs become more important — the cost-to-benefit ratio is something that is going to have to be carefully examined.”

The fact that the procedure cannot be completed in-office adds to the cost.

“You have tremendous cost issues, because wide excision alone — which for most patients could have been completed in-office — now has to be performed in the operating room under general anesthesia to allow for sentinel node staging,” Brady said. “It’s a real commitment socially to pay for a prognostic tool in the absence of any level one data that using it will improve survival.”

However, SLNB adds minimal costs to wide excision, Andtbacka said.

“If you did not complete sentinel node biopsy, 21% of patients would recur based on MSLT-1 data,” Andtbacka said. “What is the cost and morbidity of treating patients at that time? A study on the full cost value has not been completed, but you have to assess this from multiple perspectives.”

Patients’ fears about whether their disease has spread to the lymph nodes is an indirect cost of forgoing SLNB, Andtbacka said. Karakousis expressed a similar sentiment.

“It is hard to put a price on prognostic information, or the ability to tell patients whether you know if they have a favorable tumor,” Karakousis said. “It is not something that can readily be quantified.”

Thin melanomas

A thorough evaluation of the risk–benefit ratio is especially pertinent for thin melanomas, or those less than 1 mm in Breslow thickness. An estimated 60% to 70% of patients with melanoma fall in this category. Although SLNB is not universally recommended for patients with thin melanomas, the MSLT-1 trial contained inadequate data to draw any conclusions in this area.

Guidelines for thin melanomas also are scant. The joint ASCO and Society of Surgical Oncology guidelines state “there is insufficient evidence to support routine SLNB” in this setting. National Comprehensive Cancer Network guidelines state: “Because patients with thin melanoma have a generally favorable prognosis, the role of SLNB in this cohort is unclear.”

Clinicians take several clinical and pathologic factors into account to determine a patient’s risk for positive nodes, Karakousis said. These factors help clinicians weigh the likelihood of finding nodal metastasis against the potential risks of the procedure. Common pathologic factors considered include the depth of the thin lesion, the mitotic index, lymphovascular invasion, Clark level, ulceration status and deep margin status.

“If you have a patient with a melanoma thicker than .75 mm that’s Clark level IV/V with mitoses, for instance, their risk for sentinel node positivity may be on the order of 8% to 10%, and that would be in the range of some patients whose melanoma is just over 1 mm in thickness for whom we routinely recommend sentinel node biopsy,” Karakousis said. “If we are going to offer it to that group, it seems justified that we would offer it to carefully selected patients with thin lesions who have characteristics associated with an increased risk of sentinel lymph node metastasis.”

A study published last year in Journal of Clinical Oncology suggests that SLNB may be indicated for patients with melanomas as thin as 0.75 mm. In the study, Han and colleagues found Breslow thickness ≥0.75 mm was significantly associated with sentinel node positivity (P=.03), and 6.3% of these patients had node-positive disease.

PAGE BREAK

However, Brady said she tends not to recommend SLNB for patients with thin melanomas.

“In most patients with thin melanoma, the risk of a positive node is 5% to 7%,” Brady said. “This means you are undertaking the risks of the surgical procedure in a setting where you have a 93% to 95% chance of having negative nodes.”

Because of this lack of clarity, the decision to use SLNB in thin melanomas often comes down to patient preferences.

“We need to avoid major blanket statements in this group, and take everything into consideration about the individual tumor and patient to fine-tune the decision,” Sondak said. “The decision-making regarding thin melanomas is one of the most personalized decisions we have to make.”

Pediatric melanomas

The role of SLNB in pediatric melanoma also is a subject of intense debate, and questions surrounding the use of the procedure in this patient population are particularly pertinent due to the increasing incidence of pediatric melanoma in the United States.

A study by Wong and colleagues, published in 2013 in Pediatrics, showed incidence of pediatric melanoma has increased an average of 2% (95% CI, 1.4-2.7) each year from 1973 to 2009.

The challenges clinicians face during treatment of adults with melanoma are magnified in the pediatric setting. For example, a paper by Niebling and colleagues published in July in Annals of Surgical Oncology showed 5.1% of melanoma diagnoses changed after pathologic review. The rate for discrepancy often is much higher in pediatric patients, Sondak said.

“Sometimes the pathologist has a high degree of uncertainty about the exact pathology that adds to the problematic nature of having a young child with a cancer diagnosis,” Sondak said. “We have found that SLNB is helpful in dealing even with those cases in which the malignant diagnosis is uncertain because it can eliminate remaining uncertainty.”

Still, there is an increased risk for false-positive SLNB in pediatric patients given that benign nevi also can seed nodes.

“As patients get older, the number of real melanomas goes up, and the number of growing moles goes down,” Grichnik said. “But in children and young adults, the number of growing moles is huge and the number of melanomas is rare.”

The incidence of lymph node metastases is more prevalent in younger patients, even though survival often is favorable. A study by Averbrook and colleagues, published in 2012 in Cancer, showed 30% of pediatric patients who underwent SLNB had node-positive disease. However, 10-year OS ranged from 100% of those aged birth to 10 years, to 79.5% of patients aged older than 15 years.

“It is a little counterintuitive, in that younger patients appear to have a higher incidence of nodal metastases but a better OS compared to older patients with similar clinical stage tumors,” Karakousis said. “Moreover, the value of SLN biopsy in pediatric patients with melanocytic tumors of uncertain malignant potential — in which a pathologic diagnosis of melanoma cannot be firmly established — is less clear. These patients, when atypical cells are identified in their lymph nodes, tend to have a very favorable prognosis making it challenging to interpret the clinical significance of their SLN findings.”

Yet, SLNB can help provide some clarity, Sondak said.

“This is one of the real gray areas of modern medicine, and we are using SLNB as one piece of evidence among many that influences how we treat our children,” Sondak said.

The future of SLNB

Anti-CTLA-4 and PD-1 antibodies, along with BRAF- and MEK-targeted inhibitors, are poised to change treatment standards in melanoma.

SLNB can help researchers identify high-risk patients eligible to participate in clinical trials designed to assess the efficacy of these novel agents.

“There are many newer treatments that are being evaluated, and knowing the exact tumor burden and whether there is presence of lymph node metastases is extremely important,” Andtbacka said. “The role of a sentinel lymph node biopsy in that setting has become absolutely essential so that we can have better-defined patient populations in adjuvant studies.”

PAGE BREAK

Although the therapeutic value of these targeted agents in the adjuvant setting is yet unknown, these therapies have the potential to improve OS after SLNB, Karakousis said.

The identification of biologic markers to predict recurrence — particularly among patients who have a negative sentinel node — also could help improve the prognostic utility of SLNB, Andtbacka said.

“It is critical that we have more data and better markers to determine the patients who will and will not benefit from adjuvant therapy,” Andtbacka said. “We are getting a much better understanding of the biology of the disease, and I am very hopeful that within the next 5 years, the way we look at melanoma from risk-of-recurrence and staging perspectives will change and incorporate more biological prognostic factors.”

Molecular diagnostic tools or circulating tumor cell assays may capture some of these biological risk factors, Grichnik said.

Data presented at ASCO by Lawson and colleagues showed that a gene-expression profile test and SLNB demonstrated comparable positive predictive values for distant metastases in node-positive patients (50% vs. 55%), yet the negative predictive value was greater with the gene assay (82% vs. 67%).

Further, patients who had a low-risk gene test result and negative SLNB demonstrated increased 5-year distant metastasis-free survival than patients with a negative SLNB alone (86% vs. 64%).

Some researchers say identification of high-risk markers in the primary tumor may eliminate the need for SLNB.

“I foresee a future in which we can remove a lesion with pathologic characteristics suggestive of melanoma and, through molecular analyses, determine the lethal (or non-lethal) potential of that lesion and best therapies for that patient,” Grichnik said. “It may very well be that if they have a high-risk lesion, we should immediately start them on an adjuvant protocol without having to check the sentinel node.”

Any attempt to eliminate SLNB from the prognostic paradigm would be met with considerable resistance.

“There is a lot of surgical history in removing lymph nodes from melanoma, and so a lot of people are holding on to concepts that in the past they have felt very strongly about,” Brady said. “It is very difficult to get people to rethink whether this is a valuable procedure.”

Yet, frequent re-evaluation is necessary to ensure progress, Grichnik said.

“SLNB currently has a role, and it clearly has prognostic and staging value,” Grichnik said. “We just have to be careful deciding something is the gold standard, and then never challenging it.”

Sondak agreed that the knowledge gained through molecular analyses can improve upon what researchers have learned with SLNB.

“If somebody comes along tomorrow with a new test for prognosis and prediction of outcome in melanoma, they need to show us how it adds to what we already know from sentinel node biopsy,” Sondak said. “It was a good platform, and just as Clark level and Breslow thickness got us to one place, sentinel node biopsy got us to the next. We certainly hope the future will show molecular analyses can get us to a new place that is even further along for our patients.” — by Alexandra Todak

References:

Averbrook BJ. Cancer. 2014;doi:10.1002/cncr.28289.

Coit DG. J Natl Compr Canc Netw. 2013;11:395-407.

Han D. J Clin Oncol. 2013;31:4387-4393.

Lawson DH. Abstract #9022. Presented at: ASCO Annual Meeting; May 30-June 3, 2014; Chicago.

Morton DL. New Engl J Med. 2014;370:599-609.

Niebling MG. Ann Surg Oncol. 2014;21:2245-2251.

Wong JR. Pediatrics. 2013;131:846-854.

Wong SL. J Clin Oncol. 2012;30:2912-2918.

Wrightson WR. Ann Surg Oncol. 2003;10:676-680.

For more information:

Robert H.I. Andtbacka, MD, CM, can be reached at Huntsman Cancer Institute, Patient Care Center, 2000 Circle of Hope, Salt Lake City, UT 84112; email: robert.andtbacka@hci.utah.edu.

Mary S. Brady, MD, FACS, can be reached at Memorial Sloan Kettering Cancer Center, 1275 York Ave., New York, NY 10065; email: bradym@mskcc.org.

James Grichnik, MD, PhD, can be reached at Sylvester Comprehensive Cancer Center, 1475 NW 12th Ave., Outpatient Clinic, Miami, FL 33136; email: jgrichnik@med.miami.edu.

Giorgos C. Karakousis, MD, can be reached at Hospital of the University of Pennsylvania, Department of Surgery, 3400 Spruce St., 4 Silverstein, Philadelphia, PA 19104; email: giorgos.karakousis@uphs.upenn.edu.

Vernon K. Sondak, MD, can be reached at Moffitt Cancer Center, 12902 Magnolia Drive, Tampa, FL 33612; email: vernon.sondak@moffitt.org.

Disclosure: Grichnik reports consultant roles with Amgen, Castle Biosciences and Novartis. Karakousis reports consultant fees from Amgen. Sondak reports personal fees from and consultant/advisory roles with Amgen, Bristol-Myers Squibb, GlaxoSmithKline, OncoSec, MabVax, Merck, Novartis and Polynoma. Andtbacka and Brady report no relevant financial disclosures. 

 

 POINTCOUNTER

Should all patients with microscopic positive nodes undergo completion lymph node dissection?

POINT

Additional surgery is the standard for sentinel node-positive melanoma.

Mark B. Faries, MD, FACS

Mark B. Faries

Most stage III melanoma is now diagnosed early, through sentinel lymph node biopsy (SLNB). When it was originally conceived, SLNB was intended only to determine the status of basins at risk, and not to render any benefit beyond staging. However, it has become evident that most patients with sentinel node metastases have all of their nodal disease removed with the SLNB, which has led to the question of whether the larger, more morbid completion dissection is warranted. The equipoise on this issue provided justification for the second Multicenter Selective Lymphadenectomy Trial (MSLT-II), in which patients will be randomly assigned to completion dissection or to clinical follow-up with nodal ultrasound. That trial has completed accrual and will provide the answer to this question.

PAGE BREAK

In the meantime, completion lymph node dissection (CLND) remains the “standard” option, though as is pointed out in the accompanying piece, there are reasons to question that standard. There are clear reasons, though, for continuing this practice, at least until we know the result of MSLT-II. The first is completeness of staging. The status of non-sentinel lymph nodes provides important prognostic information beyond that of the sentinel node, which has been demonstrated in several retrospective series. Patients with non-SLN metastases have a significantly worse long-term outlook, which appears similar to patients with clinically detectable disease (stage IIIB/IIIC). This added information may be helpful as patients weigh options for currently available toxic adjuvant therapies with modest clinical activity.

Approximately 10% to 20% of patients with sentinel node metastases will have additional non-SLN involvement, which is likely to result in regional nodal recurrences over time. These recurrences, and the distress they induce, can be substantially reduced by immediate completion dissection. Although nodal disease can frequently be removed at the time of nodal recurrence, data from MSLT-I suggest that the morbidity of those procedures, specifically lymphedema, may be increased if the operation is delayed.

Finally, at present, eligibility for adjuvant clinical trials generally requires CLND. Although one may question this requirement, it remains common.

Overall, completion dissection remains the standard, though an informed, reasonable patient may elect to choose a different course. Discussion of options, though, should include clear enumeration of the established benefits of dissection on staging, regional disease control and clinical trial eligibility.

Mark B. Faries, MD, FACS, is director of the Donald L. Morton Melanoma Research Program at John Wayne Cancer Institute.  He can be reached at mark.faries@jwci.org.  Disclosure: Faries reports consultant/advisory board roles with Amgen, Astellas and Genentech.

References:

Bamboat ZM. Ann Surg Oncol. 2014;21:3117-3123.

Faries MB. Ann Surg Oncol. 2010;17:3324-3329.

Leung AM. JAMA Surg. 2013;148:879-884.

Reintgen M. Ann Surg Oncol. 2013;20:668-674.

COUNTER

It may be very reasonable to offer observation as an alternative to complete lymph node dissection in selected patients with microscopic positive sentinel lymph nodes.

Michael S. Sabel, MD

Michael S. Sabel

Our tradition in oncology, particularly surgical oncology, is often to offer the most aggressive therapy as “standard of care,” and then require prospective, randomized data to pull back from that. The logic behind this is clear, but the cost remains the exposure of a large group of patients to the morbidity of therapy to benefit what might be a small minority of patients.

Despite a growing desire to create narrow guidelines, well-defined standards and strict quality measures, the true role of the surgeon is to educate patients about the risks and benefits of a procedure so they can choose what is appropriate for them. It cannot be considered inappropriate for a patient to consider a small — and perhaps unknown — benefit against a considerable impact on quality of life, and conclude that the treatment is not for them.

PAGE BREAK

Although it is certainly possible that the survival benefits of CLND for a positive SLN outweigh the potential risks, it is unclear to me how we can be so dogmatic in that belief given our difficulty in clearly demonstrating a survival benefit to either elective node dissection or SLNB as compared with observation alone. Although we actively await the results of a randomized trial, prompted by the strong possibility that completion node dissection offers no significant benefit to survival, we have only retrospective data to consider. Those few studies demonstrate no difference in survival, and perhaps regional recurrence, between those patients who undergo CLND and those who do not. And, although the criticism of these studies is the selective bias inherent to retrospective review, this does not invalidate the results, but rather supports the idea that with proper selection, we can avoid surgery in a subset of patients.

The remaining questions are: How large is that subset and how is it defined? By age, co-morbidities, risk of distant disease, tumor burden within the SLN or a combination of factors? Stratifying the potential benefit is critical, including not just OS but also regional control and how the information may influence adjuvant therapy decisions. These must then be weighed against the varying risks, which — like the benefits — are different from patient to patient, based on their co-morbidities and the site of the node dissection. The conversation must therefore be different, for example, between a 70-year-old obese patient facing an inguinal lymph node dissection for a few immunohistochemistry-positive cells and a 40-year-old patient considering axillary lymph node dissection for a SLN with 50% replacement.

It is disingenuous to have the equipoise to offer no CLND on a clinical trial but then not consider it a viable option off of a clinical trial, or to consider it a violation of quality care. Conversely, given the high number of patients not undergoing CLND for a positive SLN, it also is wrong to inform SLN-positive patients they don’t need a CLND, rather than presenting the potential risks and benefits and letting them choose. The art of oncology lies in a balanced presentation based on a thorough understanding of the available data, and letting patients maintain autonomy. Just as it is not unreasonable for a patient to either decline or desire adjuvant systemic therapy based on a risk-to-benefit analysis, it is not unreasonable for any individual patient to conclude that the present level of evidence does not support CLND for a positive SLN in their particular situation. As researchers, it becomes even more incumbent upon us to continue to generate data, both prospective and retrospective, to improve our ability to target our surgical therapies where they will do the most good.

Michael S. Sabel, MD, is associate professor of surgery at the University of Michigan. He can be reached at University of Michigan Health Systems, 1500 E. Medical Center Drive, 3303 Cancer Geriatric Center SPC 5932, Ann Arbor, MI 48109; email: msabel@umich.edu. Disclosure:  Sabel reports an advisory board role with Merck.

References:

Balch CM. Ann Surg. 1996;224:255-266.

Bilimoria KY. Ann Surg Oncol. 2008;15:683-690.

Cascinelli N. Lancet. 1998;351:793-796.

Frankel TL. Ann Surg Oncol. 2008;15:2403-2411.

Gershenwald JE. J Clin Oncol. 2008;26:4296-4303.

Kingham TP. Ann Surg Oncol. 2010;17:514-520.

Morton DL. N Engl J Med. 2014;370:599-609.

van Akkooi AC. Ann Surg. 2008; 248: 949-955.

Wong SL. Ann Surg Oncol. 2006;13:809-816.