Surveillance and Follow-Up
Sarah A. Weiss and Mario Sznol
The incidence of melanoma has been rising over the past 20 years; however, the number of deaths attributed to melanoma has remained unchanged (1). Over 76,000 new melanoma cases are expected to be diagnosed in the United States in 2016, representing approximately 2% of the population. The vast majority of initial melanoma diagnoses are discovered in early stages, limited to a localized primary site and/or regional lymph nodes. While there are clear guidelines regarding surgical management of primary melanomas (2), postsurgical follow-up practices are not standardized and vary among countries, institutions, and individual clinicians (3).
The purpose of follow-up in melanoma patients is to identify treatable melanoma recurrences, ideally sooner than would be discovered without any active surveillance strategy in place, as well as to detect second primary melanomas for which these patients are at increased risk. However, the need for frequent follow-up is controversial and has been challenged (4,5). For example, up to three fourths of patients with a history of melanoma detect their own recurrences, either because the recurrence is visibly identifiable or because they are prompted to seek care due to new suspicious symptoms (6,7). Furthermore, no significant survival differences have been found for patients whose recurrence was discovered by the physician versus the patients themselves, although these studies were performed before highly effective systemic treatments were introduced for advanced melanoma (8). There is also increasingly more attention placed on limiting inefficient health care spending in an era with increasingly more cancer survivors who are undergoing follow-up. The follow-up care of patients with a history of melanoma is estimated to cost the U.S. health care system at least $500 million per year (9). These points emphasize the importance of establishing evidence-based follow-up strategies in early stage melanoma patients that ideally reflect the risk of recurrence and recurrence patterns of each stage.
RISK OF MELANOMA RECURRENCE
The American Joint Committee on Cancer’s (AJCC) 7th edition of the melanoma staging system incorporates independent prognostic features including primary melanoma characteristics (tumor thickness and presence of ulceration or mitoses), degree of lymph node involvement, and extent of distant metastases (10). Melanoma stage, therefore, is the principal tool used by clinicians to judge risk of recurrence and to accordingly guide follow-up practices. Patients with stage I melanoma have a 5% rate of recurrence. Five-year overall survival (OS) rates for stage IA and stage IB melanoma are 97% and 92%, respectively (7,11,12). Patients at higher risk of recurrence include patients with stage II and III melanoma. Stage II melanoma has a 30% recurrence rate, with 5-year OS rates of 81%, 70%, and 53% for stage IIA, IIB, and IIC, respectively (7). Finally, patients with stage III melanoma have a recurrence rate of at least 50%, with 5-year OS rates of 78%, 59%, and 40% for stage IIIA, IIIB, and IIIC, respectively (7). Despite these estimates, the staging system is imperfect and there is currently no reliable way to predict which individual patient will recur or when. Therefore, postsurgical follow-up practices have been instituted to monitor patients who remain at high risk for development of future melanoma recurrence, distant metastases, and second primary melanomas.
TIMING OF RECURRENCE DETECTION
Although many patients have historically discovered their recurrences outside of a scheduled follow-up visit, timing of recurrence detection may matter. Surveillance studies completed prior to the current treatment era showed that detection of limited melanoma recurrences, defined as lymph node (less than five affected nodes) or organ involvement measuring less than or equal to 2 cm and amenable to curative-intent surgical resection, are associated with longer survival compared with detection of extensive recurrences. For example, in a prospective German study of 1,969 stage I to III cutaneous melanoma patients undergoing surveillance every 3 months, limited compared with extensive recurrences had longer 3-year OS rates for both stage I to II (76% vs. 38%) and stage III (60% vs. 18%) melanoma (11). In contrast to previous retrospective data, over 70% of recurrences in this study were detected as a result of clinical and imaging follow-up examinations, and only 17% of recurrences were initially detected by the patient. This is likely related to the more frequent and intense clinical follow-up and use of sensitive imaging techniques in this study (6,7). Ten-year follow-up of this population reported OS rates of 43% for patients with limited recurrences compared with only 26% for extensive recurrences. The survival differences remained significant when the data were adjusted for both melanoma stage at diagnosis and potential lead time bias (13).
Despite these data, the longer OS survival rates reported in this surveillance study for patients with limited recurrences may simply reflect the natural history of melanoma. Patients who have limited compared with extensive disease recurrence would be expected to live longer, even if the recurrence is not resected. Furthermore, the majority of surveillance studies examining a benefit for limited versus extensive recurrence detection were conducted prior to the use of highly effective therapies such as CTLA-4 and PD-1 immune checkpoint inhibitors. Therefore, it is unknown whether the extent of recurrence or undergoing surgical resection makes as much of a difference in the current treatment era, and any reported survival benefit would need to be established in a randomized clinical trial.
Most melanoma recurrences are detected in the first or second year after diagnosis and are most reliably predicted by AJCC stage over any other factor (9,11). About half of the identified recurrences occur within the first year after diagnosis, and another 30% in the second year (7,14). A retrospective review of patients followed over a 20-year period at the Sydney Melanoma Unit in Australia reported that 52% of all first-time recurrences were to regional nodes, 17% were local recurrences, and 8% were in-transit metastases (Figure 14.1); there were no differences between local recurrences and in-transit recurrences either in timing of diagnosis or overall prognosis (14). Of note, the majority of patients in this study had thick (>1 mm) primary melanomas, but only 21% underwent elective lymph node dissection. Therefore, the data demonstrating regional lymph nodes as the first site of recurrence is unsurprising given that the regional nodes were not uniformly treated at the time of the initial primary surgery.
Similarly, a retrospective study of recurrence patterns in 11,615 stage I to III melanoma patients at Duke University found that the risk of first melanoma recurrence peaked at 1 year. When recurrences were stratified by location, loco-regional sites including skin and lymph node recurrences peaked at 8 months while more distant site involvement peaked later, at approximately 2 years (15). Melanoma recurrences arising 5 or more years after initial diagnosis are less common and occur at a rate of approximately 6% (7); however, even with time, the risk of recurrence never falls to zero (15). Late recurrences 10 years or more from the time of initial melanoma diagnosis have been reported in the literature, even in cases of originally thin melanomas located on the extremities that typically portend a favorable prognosis (16–18). A retrospective analysis of 7,104 melanoma patients at Duke University cites only a 2.4% rate of melanoma recurrence 10 years or more after initial diagnosis (19). Therefore, postoperative follow-up on a regular basis for the first 1 to 5 years after melanoma diagnosis is a generally accepted practice, followed by annual follow-up from years 6 to 10. These patients’ medical histories should always include their melanoma diagnosis. Any development of new skin lesions or suspicious symptoms should prompt a workup to rule out recurrent melanoma, even beyond the 10-year follow-up time frame.
METHOD OF RECURRENCE DETECTION
The majority of identified melanoma recurrences are typically found during follow-up through either patient history or by physical examination, less so from routine radiographic imaging in asymptomatic patients, and rarely from laboratory tests (6,20). With the advent of new immunotherapies and targeted therapies for advanced melanoma, patients treated for recurrences and/or distant metastases have the potential for extended survival or long-term durable responses, which were not possible even 5 to 10 years ago. Thus, meaningful interventions can be employed if a patient is found to have a recurrence or metastatic disease and this argues for the importance of well-timed follow-up.
Much of the literature reports that at least 50% of recurrences are detected by the patients themselves. For example, in a prospective study of 1,062 stage I and II melanoma patients at Memorial Sloan Kettering Cancer Center in which 19% of patients developed a recurrence during their follow-up, recurrences were detected more by the patient (55%) than by the physician (45%). Patients tended to detect slightly more of the local, in-transit, and nodal recurrences on their own physical inspection, whereas physicians detected recurrences more by radiologic examinations (20). In another prospective study of 290 high-risk stage IIB to III melanoma patients at the University of Barcelona, an intensive follow-up protocol was conducted, consisting of physical examination and laboratory tests every 3 months and total body CT and brain MRI every 6 months for the first 1 to 2 years, followed by each of those examinations every 6 months in years 3 to 5, and then physical examinations and laboratory tests annually in years 5 to 10. The majority of recurrences were detected by follow-up imaging (57%), and then by clinical examination (24%), patient self-examination (18%), and rarely by laboratory tests such as S100B protein and/or melanoma-inhibitory activity. Half of the initial recurrences were detected in either the skin or in lymph nodes. Almost all of the skin recurrences were detected on physical examination by physician or patient, while the lymph node recurrences were detected more by imaging (55%) than by physical examination. Eighty-four percent of distant metastases were detected by imaging, but patients had corresponding clinical symptoms in only 8% of cases (21). The data therefore emphasize that while patient self-awareness is critical, clinical and radiographic follow-up also detects a substantial amount of recurrences not identified by an asymptomatic patient.
For physician-detected recurrences, patient history and physical examination remain central components of follow-up. A medical history at each follow-up visit should focus on any positives from a complete review of systems that could be an indication of recurrence. Physical examination should concentrate on identifying new skin lesions, subcutaneous nodules, or palpable lymph nodes that may be concerning for recurrence. Additionally, the physician should pay close attention to the primary melanoma surgical site including the scar, the surrounding skin, and the regional area, including the draining lymph node basins. Suspicious lesions should be biopsied and new systemic complaints should prompt imaging, when appropriate (22).
ROLE OF IMAGING