2252 Follow-Up Recommendations for Patients with American Joint Committee on Cancer Stages I–III Malignant Melanoma Wen-Jen Poo-Hwu, M.D., Ph.D. Stephan Ariyan, M.D., M.B.A. Lynne Lamb, R.N., M.S.N. Rose Papac, M.D. Daniel Zelterman, Ph.D. Grace L. Hu, Ph.D. Janis Brown, M.D. David Fischer, M.D. Jean Bolognia, M.D. Antonio C. Buzaid, M.D. Yale Melanoma Unit, Yale Comprehensive Cancer Center, Yale University School of Medicine, New Haven, Connecticut. Supported by National Institutes of Health research grant CA-16359 from the National Cancer Institute. The authors are grateful to Ms. Diane Mozdziesz for her special technical assistance in the preparation of this article. Dr. Antonio C. Buzaid is currently at Hospital Sirio Libanes, Sao Paulo, Brazil. Dr. Grace L. Hu is currently at Novartis Pharmaceuticals Corporation, Summit, New Jersey. Address for reprints: Dr. Wen-Jen Poo-Hwu, M.D., Ph.D., Section of Melanoma, Clinical Immunology Service, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. Received March 30, 1999; accepted April 27, 1999. © 1999 American Cancer Society BACKGROUND. Guidelines for follow-up of melanoma patients are not established. In 1987, a follow-up protocol was instituted at the Yale Melanoma Unit to improve upon the detection of disease recurrence in patients with American Joint Committee on Cancer Stage I–III cutaneous melanoma. The follow-up protocol consists of a patient education program and a surveillance schedule based on stage of disease. METHODS. The authors retrospectively reviewed the records of 373 patients who were seen and followed according to the surveillance protocol in the Yale Melanoma Unit between January 1988 and December 1994 to determine 1) the time interval between the initial visit and recurrence; 2) the most common method of detecting recurrences; 3) whether the surveillance schedule or the patient detects more recurrences, i.e., asymptomatic recurrences versus symptomatic recurrences; 4) whether there is any survival difference between asymptomatic and symptomatic recurrences. RESULTS. The 5-year overall survival rates for Stage I, II, and III patients were 95%, 72%, and 52%, respectively. Of the 78 recurrences, 44 (56%) were detected by physician-directed surveillance examinations and 34 (44%) by patients. Most recurrences were found within the first (47%) or second (32%) year of follow-up. The estimated 6-month hazard rates for death or recurrence were 0.0044, 0.0088, and 0.0278 for Stage I, II, and III patients, respectively. The group of asymptomatic patients with recurrence had a survival advantage over the symptomatic recurrence group. In addition, patients with locoregional recurrence had better survival than those with distant recurrence. CONCLUSIONS. Although many recurrences arise rapidly and are recognized early by patients, in this study more than half were found by surveillance examinations before symptoms were manifest. Based on the hazard ratio for recurrences, the authors recommend the following surveillance schedules in addition to the patient education program for detection of recurrences: 1) Stage I, annually; 2) Stage II, every 6 months for Years 1–2 and annually thereafter; 3) Stage III, every 3 months for Year 1, every 4 months for Year 2, and every 6 months for Years 3–5; 4) at Year 6 and beyond, all patients should have surveillance annually, due to the risk of late recurrence and/or metachronous multiple primaries. Cancer 1999;86:2252– 8. © 1999 American Cancer Society. KEYWORDS: medical oncology, cutaneous melanoma, recurrence detection, follow-up recommendation. T here will be 44,200 new cases of cutaneous melanoma diagnosed in the United States in 1999, and over 80% of these cases will be localized melanoma.1 The incidence of melanoma is increasing at an annual rate of 6%. This rate of increase is faster than for any other cancer in men and second to lung carcinoma in women.2 The death rate from melanoma continues to rise '2% annually. The recurrence Follow-Up of AJCC Stage I–III Melanoma/Poo-Hwu et al. rates and survival statistics of patients with localized melanoma vary with the stage of disease. For this study, tumors were staged according to the 1983 staging system of the American Joint Committee on Cancer (AJCC)3 (the 1997 definitions of Stages II and III4 are somewhat different from 1983). According to that staging system, Stage I tumors are defined as having a tumor thickness # 1.5 mm (Stage IA, #0.75 mm, Stage IB, 0.76 –1.50 mm). Stage II tumors have a thickness . 1.50 mm (Stage IIA, 1.51– 4.00 mm, Stage IIB, .4.0 mm). Stage III tumors have spread to regional lymph nodes and/or in-transit metastases. Approximately 10 –30% of patients with AJCC Stage I and II disease and over 60% of patients with Stage III disease eventually develop metastatic disease after initial surgery.5 Up to 10% of melanoma patients will develop a second primary melanoma.6 At present, there is no general agreement on the proper frequency of clinical examinations, on the kind of diagnostic procedures, or on the follow-up duration for patients with AJCC Stage I–III melanoma after surgical resection. Model schedules for follow-up intervals for patients with Stage I melanoma were developed by Romero et al.7 based on a survey of eight expert physicians. Several groups suggested a follow-up schedule for patients with Stage I melanoma after analyzing the risk and pattern of recurrence.8,9 Reintgen et al.,10 by evaluating metastatic patterns, suggested frequent follow-up during the first 3 years after surgical excision. Other groups have reported both data and recommendations regarding the utility of follow-up tests for various stages of melanoma.11,12 Of particular note is a study that was performed with intermediate and high risk melanoma patients who were enrolled in an adjuvant trial and were followed with frequent blood analyses and chest X-rays.11 The majority of recurrences were discovered by review of systems and/or physical examination, despite the frequent use of other follow-up tests. The data suggest that diagnostic tests have limited value in the follow-up of patients with resected intermediate and high risk melanoma (Stage II and III melanoma). Guidelines regarding the follow-up of patients with AJCC Stage I–III cutaneous melanoma are needed because of the high frequency of disease recurrence and second primary melanomas. Follow-up visits after potentially curative surgery for melanoma serve a number of functions, including ensuring rehabilitation, evaluation of treatment, detection of new primary lesions, and early detection of recurrences that may be amenable to treatment, including new investigational therapies. 2253 TABLE 1 Protocol Follow-Up Schedule for Patients with Stage I–III Melanoma Month Stage I Stage II Stage III 3 4 6 8 9 12 15 16 18 20 21 24 27 28 30 32 33 36 42 48 54 60 Annually thereafter — — X — — X — — X — — X — — X — — X — X — X X — X — X — X — X — X — X — X — X — X X X X X X X — X — X X X — X — X X X — X — X X X X X X X MATERIALS AND METHODS Follow-Up Protocol At the Yale Melanoma Unit, patients with resected AJCC Stage I–III cutaneous melanoma are followed by a multidisciplinary team of dermatologists, medical oncologists, and plastic surgeons. In September 1987, a uniform follow-up protocol was adopted that combined frequent, comprehensive examinations with extensive patient education. The patient education was provided by the physicians and by clinical nurse specialists with direct discussion of clinical characteristics of melanoma, in-transit metastases, and lymph node drainage. During the first and/or second clinic visit, all patients received instructions in performing self-examination of the skin and a list of signs and symptoms of recurrence (i.e., pain, progressive fatigue, weight loss, nausea and emesis, headache, shortness of breath) that should alert them to contact their physicians. Pamphlets and videotape were used to educate patients and family members for photoprotection and melanoma prevention. Patients who were clinically free of disease after surgery were followed regularly according to the follow-up protocol shown in Table 1. Follow-up visits were scheduled based on the stage of disease according to the 1983 AJCC four-stage system.3 At each visit, a history, physical examination, complete blood count, and liver function tests (serum glutamic oxaloacetic transaminase, serum glutamic pyruvic 2254 CANCER December 1, 1999 / Volume 86 / Number 11 transaminase, alkaline phosphatase, and lactate dehydrogenase [LDH]) were performed. Chest X-rays were obtained annually for all Stage I and II patients and every 6 months for Stage III patients during the first 5 years of follow-up. All patients with Stage III disease had a baseline computed tomography (CT) scan for complete staging examination. Follow-up CT scans were obtained in 6 –12 months only if there were abnormal findings initially that were not clearly indicative of metastatic disease. Patients who developed multiple primary melanomas were continued on the follow-up schedule according to the highest stage of the invasive melanoma. Patient Selection We retrospectively evaluated the records of 419 patients with Stage I–III melanoma who were seen and followed in the Yale Melanoma Unit from January 1988 to December 1994. To be eligible for the study, patients were required to have pathologically confirmed cutaneous melanoma and no clinical evidence of disease after surgery. Patients who presented with AJCC Stage IV disease (tumor metastatic to lymph nodes beyond the regional lymph nodes or to distant sites) or noncutaneous melanoma were excluded. Patients who had inadequate medical records or inadequate follow-up at the Yale Melanoma Unit also were excluded from analysis. Forty-six patients were excluded from this study: 4 patients had noncutaneous melanoma (mucosal melanoma), 39 patients had inadequate follow-up at Yale (#2 visits), 2 patients had nonresectable Stage III disease, and 1 patient had 2 in situ lesions only. In all, 373 patients were surveyed. The following issues were evaluated by a review of the records for all patients with recurrent melanomas: 1) What was the time interval between the initial visit and diagnosis of recurrence? 2) Was the recurrence detected during a scheduled visit by a physician, or was recurrence recognized by the patient between scheduled visits by self examination or symptoms? 3) Which procedure(s) identified recurrence(s) in the asymptomatic patient (i.e., review of systems and physical examination, complete blood count, liver function tests, chest X-ray, or others)? 4) Where was (were) the site(s) of disease recurrence, locoregional metastases, and/or systemic metastases? 5) What was the survival after recurrence? 6) Did the patient develop another primary melanoma during the follow-up period? Statistical Analysis For each patient who experienced disease recurrence, a data abstract sheet summarized the demographic information, the date recurrence was diagnosed, and symptoms or procedures that led to the diagnosis of recurrence. It also recorded the various treatments the patient received for recurrent disease, such as surgery, neoadjuvant or adjuvant therapy, and the treatment outcome. All analyses were descriptive, intended to extract as much information as possible from existing data, to help formulate better strategies for monitoring patients with early stage melanoma. The results from procedures (review of systems, physical examination, complete blood count, liver function tests, and chest X-ray) that were abnormal at the time of disease recurrence were tabulated. The risk of recurrence for every 6-month period after surgery and initial visit was estimated by the hazard function. For every 6-month period after the remission of disease, the hazard function is the ratio of the number of patients who experience disease recurrence during the period divided by the number of disease free patients at the start of that period. The cumulative disease free survival curves were calculated according to the Kaplan–Meier product-limit method. The population was evaluated and stratified by stage of disease. Disease free interval and survival curves were constructed separately for the group of patients with recurrences and for subsets of patients defined by the methods of detection (symptomatic vs. asymptomatic) and by the site of recurrence (locoregional vs. distant). RESULTS Patient Population The analysis in the current study was based on 373 patients who were followed according to the protocol schedule. Of these patients, 193 patients (52%) had AJCC Stage I disease (Stage IA, n 5 84 patients; Stage IB, n 5 109 patients), 117 patients (31%) had Stage II disease (Stage IIA, n 5 85 patients; Stage IIB, n 5 32 patients), and 63 patients (17%) had Stage III disease. All patients had more than 2 routine clinic visits. The clinical characteristics of these patients are shown in Table 2. The cumulative disease free survival was analyzed for each disease stage, as shown in Figure 1. The minimum follow-up for all patients is 2 years. The 5-year overall survival rates for patients with Stage I, II, and III disease were 95%, 72%, and 52%, respectively. Of the 78 patients who experienced disease recurrence, 34 patients (44%) developed symptoms that indicated recurrence, and the other 44 patients (56%) were diagnosed by procedure(s) performed during a scheduled visit. The detection of recurrences by patients or at scheduled follow-up visits are compared in Table 3. There were 9 recurrences in patients with Stage I disease (5%), with a median time interval between initial visit and diagnosis of recurrence of 22.0 months (range, 2.0 – 60.5 months); 35 recurrences in patients with Stage II disease (30%), with a median time inter- Follow-Up of AJCC Stage I–III Melanoma/Poo-Hwu et al. TABLE 2 Clinical Characteristics of 373 Patients with American Joint Committee on Cancer Stage I–III Cutaneous Melanoma Characteristic Ages (yrs) Mean Range Stage I II III Site of primary tumor Head/neck Trunk Upper limb Lower limb Unknown 2255 TABLE 3 Distribution of 78 Recurrences in Patients with Stage I–III Cutaneous Melanoma Males (n 5 210) Females (n 5 163) Total (n 5 373) 52.3 16–81 47.2 12–81 49.8 12–81 91 74 45 102 43 18 193 117 63 47 106 26 27 4 8 50 31 73 1 55 156 57 100 5 Characteristic Gender Male Female Total (%) Stage I II III Location of recurrences Distant metastases (%) Locoregional (%) Local relapses Regional lymph nodes Patient detected Physician detected Two-sided exact P valuea 24 10 34 (44) 28 16 44 (56) — — 0.63 2 16 16 7 19 18 — 0.47 — 14 of 34 (41) 20 of 34 (59) 12 8 25 of 44 (57) 19 of 44 (43) 7 12 — — 0.25 a P values were determined by using the Fisher exact two-sided test of equality of symptomatic and asymptomatic rates. TABLE 4 Distribution of Physician-Detected Recurrences by Detection Methods FIGURE 1. Graph showing overall survival for 373 patients with American Joint Committee on Cancer (AJCC) Stage I–III cutaneous melanoma: 193 patients with Stage I disease, 117 patients with Stage II disease, and 63 patients with Stage III disease. val between initial visit and diagnosis of recurrence of 13.2 months (range, 2.4 –71.0 months); and 34 recurrences in patients with Stage III disease (54%), with a median time interval between initial visit and diagnosis of recurrence of 10.6 months (range, 2.3–53.8 months). Half of the locoregional recurrences (20 of 39 recurrences) were detected by patients, whereas more than half of the distant metastases were detected by physicians (25 of 39 recurrences). It is illustrated in Table 4 that 25 of the 44 physician-detected recurrences (57%) were identified by either history or physical examination. There were 8 recurrences identified by abnormal chest X-ray and 10 by other imaging examinations (CT scan or magnetic resonance imaging [MRI]). CT scans and MRIs were obtained either because there were abnormal findings on the baseline scan that were not clearly indicative of metastatic Method No. of recurrences Physician-detected recurrences (%) All recurrences (%) H&P CBC/LFTs Chest X-ray Othersa 25 1 8 10 57 2 18 23 32 1 10 13 H&P: history and physical examination; CBC: complete blood cell count; LFTs: liver function tests (serum glutamic oxaloacetic transaminase, serum glutamic pyruvic transaminase, alkaline phosphatase, and lactate dehydrogenase). a Computed tomography or magnetic resonance imaging scans. disease or there were suspicious findings on physical examination. Although the results of laboratory studies were abnormal at the time of detection of recurrence in 38% of patients, there was only 1 patient in whom an abnormal laboratory test result (elevated LDH) was the sole indicator of recurrent disease. Most patients developed recurrences within either the first year (47%) or the second year (32%) of followup, and these were divided equally between locoregional sites and distant sites. Five patients (6%) developed recurrences after 5 years of follow-up; 1 patient developed symptoms, 2 were detected by physical examination, and 2 were detected by annual chest X-ray. The group with locoregional recurrences, as expected, had a better survival rate compared with the patients with distant recurrences. The median survivals were 34 months and 13 months for locoregional and distant recurrences, respectively (log-rank test; P 5 0.03) (Fig. 3). Patients with regional lymph node metastases were 2256 CANCER December 1, 1999 / Volume 86 / Number 11 FIGURE 2. Graph showing the survival of patients after disease recurrence comparing locoregional recurrences with distant recurrences. The median survivals were 34 months (39 patients) and 13 months (39 patients) for locoregional recurrences and distant recurrences, respectively (log-rank test; P 5 0.03). enrolled in adjuvant immunotherapy protocols after resection. Patients with unresectable locoregional recurrences were treated on neoadjuvant chemotherapy protocols followed by resection of residual disease. Twelve patients with locoregional recurrences (30%) remained disease free after 5 years. Of the 39 patients with systemic metastases, all were enrolled in the ongoing clinical trials of systemic therapy. As a group, these patients had a shorter survival compared with the patients who had locoregional recurrences (Fig. 2). However, 4 patients (10%) responded to systemic therapy, underwent resection of residual visceral disease, and remained disease free at 5 years. Comparison of survival after disease recurrence detected either at routine examination (asymptomatic) or by the patients (symptomatic) is shown in Figure 3. The median survival of asymptomatic and symptomatic recurrences were 27 months and 14.5 months, respectively. This was statistically significant (P 5 0.02) in a multivariate analysis (Cox regression) controlling for stage, symptomatic versus asymptomatic, and local versus distant recurrences. The estimated 6-month hazard rates for death or recurrence after the date of the first Yale Melanoma Unit visit were 0.0044, 0.0088, and 0.0278, respectively, for patients with Stage I, II, and III disease or roughly a ratio of 1:2:6.3 (Fig. 4). All 19 patients with multiple primary melanomas were identified by physician-performed physical examination. Three patients had synchronous lesions at the initial diagnosis. Seventeen patients had 2 primary melanomas, 1 patient had 3 primary melanomas, and 1 patient had 12 primary melanomas. Seven of the patients with metachronous melanomas were diag- FIGURE 3. Graph showing the survival of patients after recurrence comparing asymptomatic recurrences with symptomatic recurrences. The median survivals of asymptomatic patients (n 5 44 patients) and symptomatic patients (34 5 patients) patients were 27 months and 14.5 months, respectively (Cox regression analysis; P 5 0.02). FIGURE 4. Smoothed hazard function by stage of disease. nosed in the first year of follow-up, 4 patients were diagnosed in the second year, and 5 patients were diagnosed thereafter (range, 32–125 months). The thickness of the initial melanoma ranged from 0.3 mm to 6.0 mm, and the subsequent melanoma ranged in thickness from melanoma in situ to 3.0 mm. In general, the subsequent melanomas were thinner than the first primary lesions, as reported previously by our group.6 DISCUSSION There are no established surveillance programs for patients with localized melanoma. Because of the substantial risk for recurrent disease and the tendency for late recurrence,13,14 extended postoperative follow-up visits at frequent intervals have become a common strategy for all melanoma centers. Although several centers have examined the outcome of surveillance programs, there is Follow-Up of AJCC Stage I–III Melanoma/Poo-Hwu et al. no published evidence documenting a beneficial effect of these procedures.9,15–17 Nevertheless, such programs are useful in identifying new primary lesions, in providing patient reassurance, and in sometimes detecting a cohort of patients with resectable lesions with a potential for long term disease control. Locoregional recurrence possibly is curable in some instances, and even solitary visceral sites, when they are resectable, are known to be associated with an improved prognosis.9,18 In 1987, we developed a patient education program and a follow-up schedule to provide a framework for early detection of recurrence by both patients and physicians. This protocol assumed that early recurrences may be more amenable to treatment. Although current chemotherapeutic regimens have had only marginal success in treating metastatic melanoma, numerous clinical trials are investigating new strategies and drugs for this disease. At centers where such trials are offered, early detection of recurrence offers potential value in the development of effective therapy. In 1994, our preliminary evaluation concluded that '90% of the recurrences were detected by patients.19 In 1997, 10 years after the commencement of these educational and surveillance programs, we reviewed the outcome of an expanded series with additional follow-up, which led to a more accurate construction of guidelines for the follow-up of patients with Stage I–III malignant melanoma. The modified conclusion was reported briefly in a letter to the editor of The Journal of the American Medical Association.20 The outcome as a result of educating patients during repetitive clinic visits to carry out self-examination and to recognize the signs and symptoms of recurrent melanoma revealed that 44% of recurrences were detected by the patients (Table 3). Although recurrence may develop rapidly and may be recognizable by patients between scheduled visits, 56% of the recurrences were detected only by thorough examination by physicians, and the majority were found as a result of history and physical examination. In one type of recurrence, namely, recurrence at a locoregional site, detection by patients was more frequent than detection by physicians (59% by patients). This emphasizes the equal importance of patient education and surveillance examination. In our study, the 5-year overall survival rates for patients with Stage I and II disease were 95% and 72%, respectively. These figures are very close to the rates reported by Balch et al.5 The 5-year survival rate for patients with Stage III disease was 55%, which is higher than the reported rate of 40%.5 Because of the relatively small number of patients in our study (63 patients vs. 520 patients), this difference is not statistically significant. In our study, the survival advantage was seen in the asymptomatic recurrence over the 2257 symptomatic recurrence. The same advantage also was seen in patients with locoregional recurrence over distant recurrence. This may be accounted for in part by aggressive combined surgery and neoadjuvant and adjuvant therapy, as reported previously by us.21 The majority of melanoma recurrences in our study occurred by the end of the third year of follow-up. Ideally, follow-up visits would be frequent in the first 3 years to facilitate early detection of recurrence. However, because the numbers of patients who require follow-up are increasing, it is important to find a compromise between the need for early detection of recurrences and the resources available to detect them. Accordingly, follow-up based on the risk of recurrence for individual patients, as estimated by their stage specific hazard functions, is discernible. We estimate that these are roughly at a ratio of 1:2:6 for patients with Stage I, II, and III disease, respectively. Over the course of this series, the hazard rate for Stage I patients did not change substantially; therefore, we recommend only annual follow-up for such patients. Stage II patients have twice the risk of recurrence for the first 2 years; therefore, twice yearly follow-up visits for the first 2 years and annual follow-up after that is recommended. Stage III patients have an initial hazard rate 4 – 6 times greater than that of Stage I patients, and we recommend examination of these patients at 3-month intervals for the first year. The estimated hazard rate for Stage III patients then decreases quickly, and our recommendation for their subsequent follow-up is every 4 months for the second year and every 6 months thereafter up to 5 years. Each scheduled visit should include a complete review of systems, physical examination, complete blood count, and liver function tests. Chest X-rays should be performed annually. These guidelines for testing are relatively simple and inexpensive. Chest X-rays will detect the commonest site of visceral disease, and the LDH level is often a sensitive marker for tumor burden. Abnormalities in the complete blood count and liver function tests indicate a need for more sophisticated testing. A similar strategy was recommended independently in a recent review.22 Patients with second primary melanomas should be continued on the surveillance schedule in accordance with their highest stage of invasive melanoma. Because of the extremely high risk of recurrence, all patients with Stage III disease should be encouraged to participate in adjuvant therapy trials. All adjuvant trials have a requirement of baseline body CT scans for complete staging examination. Follow-up CT scans often are performed in 3– 6 months if there are abnormal findings. 2258 CANCER December 1, 1999 / Volume 86 / Number 11 TABLE 5 Recommended Follow-Up Schedule for Patients with Stage I–III Melanoma 6. 7. Month Stage I Stage II Stage III 3 6 9 12 16 18 20 24 30 36 42 48 54 60 Annually thereafter — — — X — — — X — X — X — X X — X — X — X — X — X — X — X X X X X X X — X X X X X X X X X 8. 9. 10. 11. 12. The importance of life-time follow-up for patients with melanoma is a recognized practice in many melanoma centers.9,22,23 All patients should continue to be examined by a physician annually, either a primary care physician or a dermatologist, after the fifth year. Our recommended surveillance schedule is shown in Table 5. Based on the estimated hazard rates, less frequent surveillance visits are needed. The results of the current study strongly suggest that these surveillance schedules and patient education programs complement each other; both have a place in the early detection of disease recurrence in patients with malignant melanoma. The effectiveness and benefit of these surveillance guidelines in the reduction of overall disease-related mortality can be evaluated only in large, prospective, randomized studies. REFERENCES 1. 2. 3. 4. 5. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J Clin 1999;49:8 –31. Rigel DS. Malignant melanoma: perspectives on incidence and its effects on awareness, diagnosis, and treatment. Cancer 1996;46:195– 8. American Joint Committee on Cancer. Manual for staging of cancer. 2nd ed. Philadelphia: JB Lippincott, 1983:117–22. American Joint Committee on Cancer. AJCC cancer staging manual, 5th ed. Philadelphia: JB Lippincott-Raven, 1997: 163–7. Balch CM, Cascinelli N, Drzewiecki KT, Eldh J, MacKie RM, McCarthy WH, et al. A comparison of prognostic factors world wide. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S-J, editors. Cutaneous melanoma. Philadelphia: JB Lippincott, 1992:188 –92. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Ariyan S, Poo WJ, Bolognia JL, Buzaid AC, Ariyan T. Multiple primary melanomas: data and significance. Plast Reconstr Surg 1995;96:1384 –9. Romero JB, Stefanato CM, Kopf AW, Bart RS. Follow-up recommendations for patients with Stage I malignant melanoma. J Dermatol Surg Oncol 1994;20:175– 8. Martini L, Brandani P, Chiarugi C, Reali UM. First recurrence analysis of 840 cutaneous melanomas: a proposal for a follow-up schedule. Tumori 1994;80:188 –97. McCarthy WH, Shaw HM, Thompson JF, Milton GW. Time and frequency of recurrence of cutaneous Stage I malignant melanoma with guidelines for follow-up study. Surg Gynecol Obstet 1988;166:497–502. Reintgen DS, Cox C, Slingluff CL, Seigler HF. Recurrent malignant melanoma: the identification of prognostic factors to predict survival. Ann Plast Surg 1992;28:45–9. Weiss M, Loprinzi CL, Creagan ET, Dalton RJ, Novotny P, O’Fallon JR. Utility of follow-up tests for detecting recurrent disease in patients with malignant melanomas. JAMA 1995; 274:1703–5. Buzaid AC, Sandler AB, Mani S, Curtis A, Poo WJ, Bolognia JL, et al. Role of computed tomography in the staging of primary melanoma. J Clin Oncol 1993;11:638 – 43. McEwan L, Smith JG, Matthews JP. Late recurrence of localized cutaneous melanoma: its influence on follow-up policy. Plast Reconstr Surg 1990;86:527–34. Pearlman NW, Takach TJ, Robinson WA, Ferguson J, Cohen AL. A case-control study of late recurrence of malignant melanoma. Am J Surg 1992;164:458 – 61. Shumate CR, Urist MM, Maddox WA. Melanoma recurrence surveillance: patient or physician based? Ann Surg 1995;221: 566 –71. Kersey PA, Iscoe NA, Gapski JAP, Osoba D, From L, DeBoer G, et al. The value of staging and serial follow-up investigations in patients with completely resected primary, cutaneous malignant melanoma. Br J Surg 1985;72:614 –7. Kelly JW, Blois MS, Sagebiel RW. Frequency and duration of patient follow-up after treatment of a primary malignant melanoma. J Am Acad Dermatol 1985;13:756 – 60. Soong S-J, Harrison RA, McCarthy WH, Urist MM, Balch CM. Factors affecting survival following local, regional, or distant recurrence from localized melanoma. J Surg Oncol 1998;67:228 –33. Jillella A, Mani S, Nair B, Poo WJ, Bolognia J, Ariyan S, et al. The role for close follow-up of AJCC Stage I–III melanoma patients who remain clinically free of disease: preliminary analysis of the Yale experience. Proc ASCO 1995;14:413. Poo WJ, Ariyan S. Patient vs. physician follow-up for melanoma: a clarification. JAMA 1996;276:450 –1. Sasson HN, Poo WJ, Henderson-Bakas M, Ariyan S. Prolonged survival in patients with advanced melanoma treated with neoadjuvant chemotherapy followed by resection. Ann Plast Surg 1996;37:286 –92. Ross MI. Staging evaluation and surveillance for melanoma patients in a fiscally restrictive medical environment. Surg Clin North Am 1996;76:1423–32. Tsao H, Cosimi AB, Sober AJ. Ultra-late recurrence (15 years or longer) of cutaneous melanoma. Cancer 79;2361–70.