Dubois RW, Swetter SM, Atkins M, McMasters K, Halbert R, Miller SJ, Shiell R, Kirkwood J. Developing Indications for the Use of Sentinel Lymph Node Biopsy and Adjuvant High-Dose Interferon Alfa-2b in Melanoma. Arch Dermatol. 2001;137(9):1217-1224. doi:10.1001/archderm.137.9.1217
DamianoAbeniMD, MPHMichaelBigbyMDPaoloPasquiniMD, MPHMoysesSzkloMD, MPH, DrPHHywelWilliamsPhD, FRCP
To convene a multidisciplinary panel of dermatologists, surgical oncologists, and medical oncologists to formally review available data on the sentinel lymph node (SLN) biopsy procedure and high-dose adjuvant interferon alfa-2b therapy for patients with melanoma and to rate the "appropriateness," "inappropriateness," or "uncertainty" of the procedure and therapy to guide clinical decision making in practice.
The panel comprised 13 specialists (4 dermatologists, 4 oncologists, and 5 surgeons) from geographically diverse areas who practiced in community-based settings (n = 8) and academic institutions (n = 5). Participants were chosen based on recommendations from the relevant specialty organizations.
A formal literature review was conducted by investigators at Protocare Sciences Inc, Santa Monica, Calif, on the risks and benefits of performing an SLN biopsy in patients with stage I or II melanoma and adjuvant interferon alfa-2b therapy in patients with stage II or III disease. The MEDLINE database was searched from 1966 through July 2000, and supplemental information was obtained from various cancer societies and cancer research groups. Panel participants were queried on additional sources of relevant information. Unpublished, presented data were included in abstract form on 1 recently closed clinical trial.
The RAND/UCLA Appropriateness Method was used to review and rate multiple clinical scenarios for the use of SLN biopsy and interferon alfa-2b therapy. The consensus method did not force agreement.
The panel rated 104 clinical scenarios and concluded that the SLN biopsy procedure was appropriate for primary melanomas deeper than 1.0 mm and for tumors 1 mm or less when histologic ulceration was present and/or classified as Clark level 4 or higher. The SLN biopsy was deemed inappropriate for nonulcerated Clark level 2 or 3 melanomas 0.75 mm or less in depth and uncertain in tumors 0.76 to 1.0 mm deep unless they were ulcerated or Clark level 4 or higher. Interferon alfa-2b therapy was deemed appropriate for patients with regional nodal and/or in-transit metastasis and for node-negative patients with primary melanomas deeper than 4 mm. The panel considered the use of interferon alfa-2b therapy uncertain in patients with ulcerated intermediate primary tumors (2.01-4.0 mm in depth) and inappropriate for node-negative patients with nonulcerated tumors less than 4.0 mm deep. Specialty-specific ratings were conducted as well.
IN AN IDEAL environment, physicians would have compelling high-quality evidence to drive clinical decision making when a new procedure or drug becomes available. Unfortunately, this rarely occurs. More typically, physicians have scanty or conflicting data. This latter situation currently confronts dermatologists, surgeons, and medical oncologists when they are faced with patients who have stage I or II malignant melanoma. For these patients, controversy abounds regarding the appropriate use of the recent sentinel lymph node (SLN) biopsy procedure and high-dose adjuvant interferon alfa-2b therapy.
Although the SLN biopsy procedure has become widely accepted as a means of staging the regional lymph nodes and is practiced as the standard of care in many major melanoma centers,1 some authors have argued that unless the use of SLN biopsy can be shown to improve overall survival, it should only be performed in the context of a clinical trial.2 Others argue that SLN biopsy is valuable principally as a staging procedure, and clinicians should not expect this diagnostic test to improve survival. Similarly, the role of adjuvant high-dose interferon alfa-2b for treatment of high-risk melanoma has been the subject of controversy because clinical trials have resulted in inconsistent data regarding overall survival. Regardless of one's viewpoint, these methods have become readily available: the US Food and Drug Administration approved interferon alfa-2b therapy for melanoma in 1998, and the American Medical Association has already acknowledged the widespread use of SLN biopsy by creating a common procedure terminology code for billing purposes.
In the absence of clear and convincing proof, what should clinicians do? Should physicians never use these interventions or use them for selected patients? If the latter is appropriate, in which patients do the potential benefits outweigh the risks? In response to these questions, we present the results of a multidisciplinary expert panel that used a well-validated consensus methodology. The panel examined the available scientific literature and quantitatively rated the appropriateness of the SLN biopsy procedure and adjuvant interferon alfa-2b therapy for 104 different clinical scenarios. Unlike other consensus statements that have had representation just from oncology, surgery, and pathology,3 this project included dermatologists as an equal participant throughout the process. These guidelines may assist but should not constrain clinicians in their decision making.
To develop appropriateness measures for the use of SLN biopsy and high-dose adjuvant interferon alfa-2b therapy for malignant melanoma, we (1) undertook a computer search of the literature to determine the indications for the use, risks, and benefits of these interventions; (2) developed a list of indications or clinical scenarios for which SLN biopsy or interferon alfa-2b therapy might be considered; and (3) convened a geographically diverse, multidisciplinary consensus panel (representing the specialties of dermatology, surgery, and medical oncology) to review and rate each indication for SNL biopsy and for interferon alfa-2b therapy. We used the validated RAND/UCLA Appropriateness Method for this process.4
The consensus panel received 2 reviews of the literature (1 for SNL biopsy and 1 for interferon alfa-2b therapy). Each review described the benefits and risks of these interventions. Articles were identified by searching the MEDLINE database from 1966 through July 2000 under the following keywords: melanoma, sentinel, adjuvant therapy, and interferon. In addition, articles from the Internet Web sites of the American Society of Clinical Oncology (ASCO) (http://www.asco.org), CancerNet (http://www.cancernet.nci.nih.gov), CancerTrials (http://cancertrials.nci.nih.gov), Sunbelt Melanoma Trial (http://www.sunbeltmelanoma.com), and the Eastern Cooperative Oncology Group (ECOG) (http://ecog.dfci.harvard.edu) were included. Finally, cross-referencing reference lists from the articles generated additional sources. We included clinical data from the ECOG 1694 trial5 in abstract form. We recognize that material in abstracts represents a lower level of peer review and is not the norm for literature reviews of this type. Nevertheless, some clinical trials are only available in abstract form, and these types of data are routinely used in this field.
We attempted to create a comprehensive list of all possible indications for the use of SLN biopsy and interferon alfa-2b therapy that might arise in clinical practice. The indications were divided into 2 general categories: those pertaining to the use of SLN biopsy (n = 48) and those pertaining to the use of interferon alfa-2b therapy (n = 56). Patients were categorized according to hypothetical situations or indications based on permutations of various clinical factors. Examples of indications are included in Table 1.
For SLN biopsy, factors include:
Location of lesion (head, trunk, or extremity)
Depth of lesion (≤0.75 mm, 0.76-1.0 mm, 1.01-4.00 mm, or >4.0 mm)
Clark level (2, 3, or ≥4)
Ulceration (present or absent)
For interferon alfa-2b therapy, factors include:
Depth of lesion (≤1.00 mm, 1.01-2.00 mm, 2.01-4.00 mm, or >4.0 mm)
Lymph node involvement (none, 1, 2, 3, or ≥4)
Presence of micrometastases vs macrometastases (for patients with positive nodes). Micrometastases are defined as lymph node metastases not detectable clinically, which are diagnosed after an elective or sentinel lymphadenectomy.
We convened a panel of 13 physicians representing a diversity of specialties (4 dermatologists, 4 oncologists, and 5 surgeons), practice settings (5 universities and 8 communities), and geographic sites. A pool of potential panelists was identified from a combination of specialty society recommendations (American Society for Clinical Oncology, American Academy of Dermatology, and the American College of Surgeons), authors of key clinical trials, and recognized community-based practitioners. The final selection balanced the specialty, setting, and geographic factors. The list of consensus panel members and their specialties is given in the acknowledgments.
The panelists received the literature review and an initial set of indications by mail, which they rated. During the meeting, the panelists reviewed the summarized first-round ratings, revised the indications structure, modified the definitions of key terms, discussed reasons for the degree of agreement or disagreement in ratings from the first round, and confidentially "rerated" all indications. Each indication was rated on a 9-point scale of appropriateness (9 indicated extremely appropriate; 5, uncertain; and 1, extremely inappropriate). Appropriateness was defined as the expected health benefits of the procedure or therapy exceeding its expected negative health consequences by a sufficiently wide margin to justify performing the procedure or giving the therapy. The final rating was the median score of the 13 panelists. We considered that indications were appropriate for median ratings between 7 and 9 (without disagreement), inappropriate for median ratings between 1 and 3 (without disagreement), and uncertain for median ratings between 4 and 6 or if panelists disagreed. The consensus method did not force agreement. We defined disagreement as occurring when at least 3 panelists rated the indication appropriate and at least 3 rated the indication inappropriate regardless of the median rating.
Our review identified 2 randomized controlled6,7 and 10 nonrandomized studies with multivariate analyses.1,8- 16 Sentinel lymph node status is the most important predictor of survival for patients with melanoma. In the report by Gershenwald et al,12 the hazard ratio for survival associated with a negative sentinel node was 6.43, much greater than that for any other prognostic factor. For the purposes of this review, a positive SLN biopsy result is defined as the presence of identifiable melanoma cells on routine hematoxylin-eosin stains and/or by immunohistochemical stains, including HMB-45 and S100. Despite its prognostic importance, it has not been proven that SLN biopsy confers a direct therapeutic or survival benefit to the patient (Table 2).
Preliminary results of the Multicenter Selective Lymphadenectomy Trial (MSLT) were reported at the American Society of Clinical Oncology 1999 Annual Convention in abstract form.7 After a 3-year follow-up period, the SLN-positive group had a 3-year overall survival of 72% and a 3-year disease-free survival of 60%, compared with a 3-year overall survival of 91% and 3-year disease-free survival of 85% for the SLN-negative group. This ongoing study randomizing patients to SLN biopsy vs nodal observation will definitively determine whether SLN biopsy is associated with improved survival. However, as discussed previously, many believe that SLN biopsy is valuable mainly as a staging procedure, which helps define patient populations who may benefit from early therapeutic lymph node dissection and adjuvant therapy. Accordingly, SLN biopsy is likely to continue as a diagnostic procedure to determine regional nodal stages regardless of the results of the MSLT study.
In a retrospective review of the combined experience (n = 580) of the M. D. Anderson Cancer Center, University of Texas, Houston, and the H. Lee Moffitt Cancer Center, University of South Florida, Tampa, Gershenwald et al12 found a 3-year disease-free survival difference between SLN-positive and SLN-negative patients (55.8% vs 88.5%; P<.001). An SLN-negative biopsy result was associated with a 58.6% increase in disease-free survival in their analysis.
In a prospective cohort study of 200 cases, Jansen et al14 also found statistically significant differences in 3-year overall survival (93% vs 67%; P<.001) and disease-free survival (88% vs 58%; P<.001) between SLN-negative and SLN-positive patients. Examining survival patterns in patients with thick melanomas (>4 mm), Gershenwald et al11 found statistically significant differences in 3-year disease-free survival of SLN-negative and SLN-positive patients (82.4% vs 58%, respectively; P<.03) and overall survival (89.8% vs 64.4%, respectively; P = .006) in sentinel node-negative and node-positive patients. Of the node-positive patients, 38% received adjuvant interferon alfa-2b treatment.11
In a retrospective matched-pair analysis of 534 stage I lesions, Essner et al9 did not find any differences in disease-free and overall survival between the group undergoing lymphatic mapping and SLN biopsy and the group undergoing elective lymph node dissection. They did note a trend toward better survival in node-positive patients in the SLN biopsy group (P = .08), but this finding could be related to the shorter follow-up period of this group. Collectively, the above studies provide compelling evidence of the prognostic value of SLN biopsy for patients with melanoma.
Several clinical factors correlate with the likelihood of lymph node involvement and thus influence the decision regarding SLN biopsy. Breslow thickness (tumor thickness in millimeters from the top of the granular layer or base of superficial ulceration to the deepest point of tumor penetration at the primary melanoma site) has an important impact on lymph node status. In general, for lesions less than 1.0-mm thick, the SLN positivity rate varies from 0% to 6%. For lesions between 1.0 and 1.49 mm, the SLN positivity rate varies from 0% to 14%; for lesions between 1.5 and 4.0 mm, the rate varies between 11% and 32%; and for lesions thicker than 4.01 mm, the rate varies between 20% and 67%.
Specifically, Belli et al17 reported SLN positivity rates of 16% in a series of trunk and limb lesions thicker than 1.0 mm. There was a 7% SLN positivity rate in lesions between 1.00 and 1.9 mm, an 11% positivity rate for lesions between 2.0 and 2.9 mm, and a 20% rate for those thicker than 3.0 mm. Reporting on the World Health Organization (WHO) Melanoma Program experience, Cascinelli et al1 reported positivity rates of 2% (<1.0 mm), 7% (1.0-1.99 mm), 13% (2.0-2.99 mm), and 31% (≥3.0 mm) among 829 patients. Reporting their findings in a retrospective review of 540 cases, Lenisa et al18 found an overall SLN positivity rate of 15%, with 1% of the lesions less than 1-mm thick being node positive, as were 5% of the lesions between 1.0 and 1.99 mm, 18% of the lesions between 2.0 and 2.99 mm, and 27% of the lesions thicker than 3.0 mm. Using a similar scale, Mraz-Gernhard et al19 found that 16% of the lesions between 1.0- and 1.99-mm thick were associated with a positive SLN, with 50% of the lesions between 3.0 and 3.99 mm associated with a positive SLN.
Clark level describes the anatomic extent of skin invasion. Level 1 is intraepidermal (in situ); level 2 penetrates the papillary dermal collagen; level 3 expands and fills the papillary dermis; level 4 penetrates the reticular dermal collagen; and level 5 penetrates into the subcutaneous fat or deeper. Few studies present rates of SLN positivity in terms of Clark levels. Bongers et al20 found no node-positive patients whose primary lesions were Clark level 2, while 17% of Clark level 3 lesions were SLN positive, as were 16% of Clark level 4 and 25% of Clark level 5 lesions. Data from the MSLT showed a 14% SLN positivity rate for Clark level 3 lesions, rising to 20.6% for Clark level 4 and 31.2% for Clark level 5 lesions.6 In an analysis of patients with thick lesions (>4.0 mm), Gershenwald et al11,12 did not find a Clark level higher than 3 to be a significant predictor of SLN positivity.
Ulceration is defined by the American Joint Committee on Cancer (AJCC) as the absence of an intact epidermis overlying a portion of the primary melanoma based on pathologic microscopic observation of the histologic sections.3 There is a comparative dearth of studies linking tumor ulceration to SLN status. Gershenwald et al12 found a statistically significant difference between SLN-negative and SLN-positive patients according to the ulceration status; they found that 43.5% of the SLN-positive patients had ulceration, while only 20.2% of the SLN-negative patients had tumor ulceration (P<.001). In an analysis of thick lesions (>4.0 mm), Gershenwald et al11 found that ulcerated lesions were more likely to have a positive SLN (P<.01). Other studies have used multivariate analyses to demonstrate that ulceration is associated with SLN positivity.1,9,19
There are no published studies that explicitly investigate the risks associated with the SLN biopsy procedure. However, one may assume that SLN biopsy comes with the general risks associated with any minor surgical procedure (eg, wound infection, dehiscence, and seroma). In their series describing the experience of the WHO Melanoma Program, Cascinelli et al1 observed no major complications associated with the procedure in 892 procedures performed in 829 patients. Wrone et al21 described that lymphedema can occur in a small percentage of cases.
False-negative results also may be considered a risk in this setting. False-negative results represent a risk because patients inaccurately labeled as having no nodal involvement might not be offered potentially effective therapy. Most studies report a 4% to 5% false-negative rate; however, there are 2 significant outliers. Carlson et al22 reported a false-negative rate of 21.3% from their series of 58 patients with stage I and II head and neck melanomas. The authors postulate that 24% of their cases had drainage to the parotid region, "where high radioactive counts made localizing focal activity difficult."22 Jansen et al13 also reported a 20% false-negative rate. In their relatively small series (n = 30) of head and neck lesions thicker than 1 mm, the authors could not find a sentinel node in 10% of their patients, and 15% of the identified sentinel nodes could not be retrieved. These studies are obviously qualified by their small sample size.
Our review identified and evaluated 9 randomized controlled trials of adjuvant interferon alfa-2b therapy for cutaneous melanoma. Three of these trials used high-dose interferon alfa-2b therapy, 5 studies used low-dose interferon alfa-2b therapy, and 1 study compared both low-dose and high-dose interferon alfa-2b therapies. Since the strongest evidence favoring interferon alfa-2b therapy has been obtained using the high-dose approach, we will limit our discussion to the high-dose interferon alfa-2b therapy. Low-dose interferon alfa-2b therapy has not been shown to improve relapse-free or overall survival in patients with melanoma.23- 25 We defined high-dose therapy as the intravenous administration of 20 million IU/m2 5 days per week for 4 weeks (induction) followed by the subcutaneous injection of 10 million IU/m2 3 days per week for 11 months (maintenance), which was reviewed and approved by the Food and Drug Administration in 1995. Studies of high-dose interferon alfa-2b therapy5,26- 28 are summarized in Table 3.
In the North Central Cancer Treatment Group (NCCTG) Trial reported by Creagan et al26 in 1995, 262 patients with deep primary lesions (>1.69 mm) or regional lymph node metastases were randomized to receive high-dose interferon alfa-2a therapy or to be observed. Lymph node dissection was only performed on patients with clinically positive nodes. Five-year relapse-free survival was 43% in the interferon alfa-2b therapy group and 36% in the observation group (P = .24). Five-year overall survival was 54% in the interferon alfa-2b therapy group and 48% in controls (P = .53). Subgroup analysis suggested a statistically significant improvement in relapse-free survival among patients with lymph node involvement.
In the ECOG 1684 study,27 280 patients were prospectively randomized (observation group [n = 137]; treatment group [n = 143]) with thick primary lesions (Breslow thickness, >4 mm; T4) without nodal metastasis or patients with primary tumors of any thickness with regional lymph node metastases (N1 tumors). All patients were required to undergo elective regional lymph node dissection prior to receiving adjuvant therapy. After a median follow-up of 6.9 years, both the relapse-free survival and overall survival rates for the interferon alfa-2b treatment group improved. The median relapse-free survival was 1.72 years in the treatment group (95% confidence interval [CI], 1.07-2.88) compared with 0.98 years in the observation group (95% CI, 0.50-1.65). The median overall survival increased from 2.78 years (95% CI, 1.83-4.03) to 3.82 years for the treatment group (95% CI, 2.34-7.08). The estimated 5-year relapse-free survival was 37% in the treatment group and 26% in the observation group (P = .002). The estimated 5-year overall survival was 46% for the treatment group and 37% for the observation group (P = .02).
The ECOG 1690 study28 was designed to compare the high-dose interferon alfa-2b regimen used in ECOG 1684 study27 with the low-dose regimen used in Europe for patients with deep primary lesions (>4.0 mm) or regional lymph node metastases. Compared with the ECOG 1684 study, a greater number of patients with clinically negative lymph nodes were enrolled to examine whether node-negative patients with thick primary lesions would benefit from adjuvant therapy. For both high-dose and low-dose groups, interferon alfa-2b treatment had no apparent impact on overall survival compared with the observation group. The high-dose therapy group had an estimated 5-year relapse-free survival rate of 44% vs 35% in the observation group (P = .05, 2-tailed). In a Cox proportional hazards regression model analysis that adjusted for stage and number of positive nodes, a significant factor for relapse-free survival was reached for the high-dose therapy group but not for the low-dose therapy group.
The final study, ECOG 1694,5 has been presented recently to the European Society of Medical Oncology and published to date only in abstract form. In this study, of 880 high-risk patients with melanoma (deep primary lesions >4.00 mm or regional lymph node metastases), 774 eligible patients were randomized to receive high-dose interferon alfa-2b therapy vs a GM2 ganglioside vaccine. The study was closed early by the data monitoring committee because of the clear superiority of interferon alfa-2b therapy in terms of both disease-free (hazard ratio, 1.47; P = .003, 1-tailed) and overall (hazard ratio, 1.52; P = .018, 1-tailed) survival compared with the vaccine arm.
Almost all patients experience adverse effects associated with interferon therapy.26- 30 The most frequent effects include fatigue, neutropenia, myalgia, headache, fever, chills, and an increased aspartate aminotransferase level. Other frequently occurring adverse effects are nausea, vomiting, depression, alopecia, diarrhea, thrombocytopenia, altered taste sensation, and dizziness and/or vertigo. As with all therapies with interferons, very infrequent reports of suicidal behavior (ideation, attempts, and completed suicides) have been associated with treatment. Table 4 summarizes the data describing adverse effects of interferon alfa-2b in the treatment of primary melanoma.
Of the indications for interferon alfa-2b therapy, 89% were rated as appropriate; 2%, uncertain; and 9%, inappropriate. For SLN biopsy, 88% were rated appropriate; 6%, uncertain; and 6%, inappropriate. According to our definition, the panel disagreed about 2% of the interferon alfa-2b therapy indications and 6% of the SLN biopsy indications in the final ratings, falling from 14% and 8%, respectively, in the first-round ratings.
Despite the complexity of the rating structure and its permutation of multiple clinical factors, the final ratings were readily grouped for simpler presentation. We collapsed the separate indications in which the categorization of appropriateness did not differ based on clinical factors (eg, all were appropriate, uncertain, or inappropriate). For SLN biopsy, the 48 indications could be simplified into 9 categories (Table 5). Similarly, for interferon alfa-2b therapy, the 56 indications could be simplified into 9 categories (Table 6).
We analyzed the ratings separately for each clinical specialty. For SLN biopsy, the mean rating for oncologists was 7.7; for dermatologists, 7.1; and for surgeons, 8.0. For interferon alfa-2b therapy, the mean rating for oncologists was 7.9; for dermatologists, 7.0; and for surgeons, 7.8. We did not perform statistical testing because of the descriptive nature of this analysis and the small sample size. We also analyzed the ratings by specialty for "break-point" indications or indications for which the panel disagreed and for which small changes in the clinical factors were associated with large changes in the ratings. The break-point indication for SLN biopsy was patients with 0.76- to 1.0-mm thick trunk lesions that were Clark level 2 or 3 without histologic ulceration. On average, oncologists rated this indication 3.5; dermatologists, 4.0; and surgeons, 6.6. The break-point indication for interferon alfa-2b therapy was node-negative patients with 2-mm deep ulcerative primary melanomas (classified T3b N0M0 by the revised AJCC melanoma staging system3). On average, oncologists rated this indication 6.0; dermatologists, 5.3; and surgeons, 4.4 (Table 7).
We used an evidence-based, validated, consensus method to develop indications for SLN biopsy in patients with stage I or II melanoma and high-dose adjuvant interferon alfa-2b therapy in patients with stage II or III melanoma. These guidelines can assist clinicians in decision making when definitive data do not exist.
The panel rated SLN biopsy in 48 real-world clinical situations and found it appropriate for patients with lesions thicker than 1.0 mm (T2) or with lesions 1.0-mm thick or less when ulceration or reticular dermal invasion (Clark level ≥4) was present. They rated patients with lesions 0.75-mm thick or less as inappropriate because of the rare likelihood of nodal involvement (unless they had ulceration or Clark level ≥4 lesions) and rated patients with lesions between 0.76- and 1.0-mm thick as uncertain because of the low (but not zero) risk of nodal involvement. In arriving at these ratings, the panel balanced the potential benefits of the procedure (eg, prognostic information and therapeutic impact) with the risks of that procedure (eg, minor morbidity and false-negative possibility). As the results of the literature review show, no trial has definitively demonstrated a survival benefit in the use of SLN biopsy. Despite this lack of evidence, the panel clearly delineated patients for whom the procedure was appropriate, inappropriate, and uncertain using a risk-benefit metric.
In a similar fashion, the panel rated 56 clinical scenarios for the use of high-dose adjuvant interferon alfa-2b therapy. Again, balancing the evidence of benefit against both the risk of the specific therapy and the risk of no treatment, they found the use of interferon alfa-2b therapy appropriate for patients with either thick lesions (>4.0 mm; T4 N0 lesion) or with positive lymph nodes (N1-N3 lesion). For node-negative patients, they rated the use of interferon alfa-2b therapy inappropriate for lesions less than 2.0 mm in thickness (T2 lesion) or between 2.1 and 4.0 mm if no ulceration was present histologically (T3a lesion). They rated the use of interferon alfa-2b therapy as uncertain for patients with lesions between 2.1- and 4.0-mm thick with ulceration (T3b lesion). Thus, the panel provided specific guidance for the use of adjuvant interferon alfa-2b therapy in the currently unclear setting of the node-negative patient.
The validity and utility of guidelines depend on the method used. We used the RAND/UCLA Appropriateness Method,4 a well-studied approach that blends evidence from the literature with an expert consensus process. No panel can divine the results of a definitive randomized controlled trial before it has been conducted. It can, however, assess what is known today in as quantitative and unbiased fashion as possible. The panel received an extensive review of the literature (a small component of which is summarized in the "Literature Review Results" section). The systematic review included data on the potential benefits and risks of SLN biopsy and interferon alfa-2b therapy. Prior published reviews of the interferon alfa-2b literature included the 2 key high-dose randomized trials (the ECOG 168427 and ECOG 169028 trials), which had conflicting results regarding overall survival benefit. At the time of the panel review, results from the ECOG 1694 study5 had just become available owing to the closure and unblinding of the data by the data monitoring committee after a clear advantage for the interferon alfa-2b–treated patients was demonstrated. Thus, the panel had an opportunity to consider these most recent data in detail.
Consistent with the RAND/UCLA Appropriateness Method,4 the panel had a multidisciplinary composition. It comprised clinicians from the key specialties of dermatology, oncology, and surgery, which is unlike other projects in which dermatologists were not included.1,3 The panel included representatives from academic as well as community-based practice. Finally, the panelists represented all major geographic regions of the country. This diverse panel reflects multiple viewpoints and mitigates potential bias.
The development of the guidelines used a quantitative process. Each panelist rated 104 clinical scenarios on a 9-point risk-benefit scale from least appropriate (1) to most appropriate (9). These scenarios reflect real-world combinations of clinical factors. The final ratings were calculated as the median scores among the 13 panelists, and when statistical disagreement occurred (ie, ≥3 panelists rated a scenario appropriate while ≥3 panelists rated that same scenario inappropriate), that scenario was given a rating of uncertain.
We examined the ratings by specialty to determine whether systematic differences existed among the panelists. On average, surgeons had the highest ratings for SLN biopsy and medical oncologists had the highest ratings for interferon alfa-2b therapy. These differences became most marked for break-point indications. We identified a break-point indication when the panel rated that scenario as uncertain and only a slightly different combination of clinical factors was rated appropriate or inappropriate. For example, the break-point indication for the use of SLN biopsy was a patient with a 0.76- to 1.0-mm thick trunk lesion that was classified as Clark level 2 or 3 and without ulceration. The panel rated this scenario as uncertain. If the lesion was 0.75-mm thick or less, the use of SLN biopsy was rated as inappropriate; if the tumor was thicker than 1.0 mm, Clark level 4 or higher, or with ulceration, the procedure was rated as appropriate. In this break-point scenario, oncologists rated the appropriateness of the use of SLN biopsy the lowest (3.5 on the 9-point scale); surgeons, the highest (6.6); and dermatologists were in the middle (4.0).
We observed the opposite tendency for the appropriateness of interferon alfa-2b therapy for which oncologists rated the highest (6.0); surgeons, the lowest (4.4); and dermatologists were again in the middle (5.3). These findings argue strongly for inclusion of all relevant clinical disciplines in any guideline process. Too often consensus panels reflect the opinion of just selected specialties.
Clinical decision making is almost never black and white: it requires application of the existing evidence and the judgment of experienced clinicians. This study provides an objective method for integrating the existing data to a variety of patient care scenarios. While treatment decisions need individualization in many cases and the conclusions of this panel should not be interpreted as a standard of care, we believe that these guidelines can aid the decision-making process for clinicians regarding patients with melanoma.
A cooperative effort of the Clinical Epidemiology Unit of the Istituto Dermopatico dell' Immacolata–Istituto di Ricovero e Cura a Carattere Scientifico (IDI-IRCCS) and the Archives of Dermatology.
Accepted for publication April 5, 2001.
This study was partially funded by Schering Plough Pharmaceuticals, Madison, NJ.
Consensus panel members include Michael Atkins, MD (oncology), Rodolfo Bordoni, MD (oncology), Larry Dillon, MD (surgery), Allan Halpern, MD (dermatology), Ryan Holbrook, MD (surgery), Micheline Hyacinth, MD (surgery), John Kirkwood, MD (oncology), Kelly McMasters, MD (surgery), Merrick Ross, MD (surgery), Susan Swetter, MD (dermatology), Thomas Sweeney, MD (oncology), Ronald Shiell, MD (dermatology), and Stanley Miller, MD (dermatology).
Corresponding author and reprints: Robert W. Dubois, MD, PhD, Zynx Health, Inc, 9100 Wilshire Blvd, Beverly Hills, CA 90211 (e-mail: email@example.com).