[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure.
Comparison of the Visual and Histopathologic Assessment (VAH) and the Pigmented Lesion Assay (PLA) Pathways
Comparison of the Visual and Histopathologic Assessment (VAH) and the Pigmented Lesion Assay (PLA) Pathways

The diagram provides an overview on modalities physicians can select to rule out melanoma in pigmented lesions suggestive of melanoma with 1 or more ABCDE (asymmetry, border, color, diameter, and evolution) risk factors when the VAH and the PLA are available. NPV indicates negative predictive value.

aNegative results are found for 87% of PLA tests, dramatically reducing the number of surgical biopsies and follow-up excisions. The high NPV reduces the probability of missing melanomas to less than 1% compared with 17% for the VAH.

Table 1.  
Real-World PLA Results at 2 US Dermatology Sitesa
Real-World PLA Results at 2 US Dermatology Sitesa
Table 2.  
Input Variables of Cost and Test Characteristics and Data Sources
Input Variables of Cost and Test Characteristics and Data Sources
Table 3.  
Base-Case Results
Base-Case Results
Table 4.  
One-Way Deterministic Sensitivity Analysis
One-Way Deterministic Sensitivity Analysis
1.
Friedman  RJ, Farber  MJ, Warycha  MA, Papathasis  N, Miller  MK, Heilman  ER.  The “dysplastic” nevus.  Clin Dermatol. 2009;27(1):103-115. doi:10.1016/j.clindermatol.2008.09.008PubMedGoogle ScholarCrossref
2.
Schäfer  T, Merkl  J, Klemm  E, Wichmann  HE, Ring  J; KORA Study Group.  The epidemiology of nevi and signs of skin aging in the adult general population: results of the KORA-survey 2000.  J Invest Dermatol. 2006;126(7):1490-1496. doi:10.1038/sj.jid.5700269PubMedGoogle ScholarCrossref
3.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, I: common and atypical naevi.  Eur J Cancer. 2005;41(1):28-44. doi:10.1016/j.ejca.2004.10.015PubMedGoogle ScholarCrossref
4.
Rigel  DS, Russak  J, Friedman  R.  The evolution of melanoma diagnosis: 25 years beyond the ABCDs.  CA Cancer J Clin. 2010;60(5):301-316. doi:10.3322/caac.20074PubMedGoogle ScholarCrossref
5.
Monheit  G, Cognetta  AB, Ferris  L,  et al.  The performance of MelaFind: a prospective multicenter study.  Arch Dermatol. 2011;147(2):188-194. doi:10.1001/archdermatol.2010.302PubMedGoogle ScholarCrossref
6.
Argenziano  G, Cerroni  L, Zalaudek  I,  et al.  Accuracy in melanoma detection: a 10-year multicenter survey.  J Am Acad Dermatol. 2012;67(1):54-59. doi:10.1016/j.jaad.2011.07.019PubMedGoogle ScholarCrossref
7.
Nault  A, Zhang  C, Kim  K, Saha  S, Bennett  DD, Xu  YG.  Biopsy use in skin cancer diagnosis: comparing dermatology physicians and advanced practice professionals.  JAMA Dermatol. 2015;151(8):899-902. doi:10.1001/jamadermatol.2015.0173PubMedGoogle ScholarCrossref
8.
Wilson  RL, Yentzer  BA, Isom  SP, Feldman  SR, Fleischer  AB  Jr.  How good are US dermatologists at discriminating skin cancers? a number-needed-to-treat analysis.  J Dermatolog Treat. 2012;23(1):65-69. doi:10.3109/09546634.2010.512951PubMedGoogle ScholarCrossref
9.
Hansen  C, Wilkinson  D, Hansen  M, Argenziano  G.  How good are skin cancer clinics at melanoma detection? number needed to treat variability across a national clinic group in Australia.  J Am Acad Dermatol. 2009;61(4):599-604. doi:10.1016/j.jaad.2009.04.021PubMedGoogle ScholarCrossref
10.
Wang  DM, Morgan  FC, Besaw  RJ, Schmults  CD.  An ecological study of skin biopsies and skin cancer treatment procedures in the United States Medicare population, 2000 to 2015.  J Am Acad Dermatol. 2018;78(1):47-53. doi:10.1016/j.jaad.2017.09.031PubMedGoogle ScholarCrossref
11.
Ferris  LK, Jansen  B, Ho  J,  et al.  Utility of a noninvasive 2-gene molecular assay for cutaneous melanoma and effect on the decision to biopsy.  JAMA Dermatol. 2017;153(7):675-680. doi:10.1001/jamadermatol.2017.0473PubMedGoogle ScholarCrossref
12.
Urso  C, Rongioletti  F, Innocenzi  D,  et al.  Histological features used in the diagnosis of melanoma are frequently found in benign melanocytic naevi.  J Clin Pathol. 2005;58(4):409-412. doi:10.1136/jcp.2004.020933PubMedGoogle ScholarCrossref
13.
Elmore  JG, Barnhill  RL, Elder  DE,  et al.  Pathologists’ diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study.  BMJ. 2017;357:j2813. doi:10.1136/bmj.j2813PubMedGoogle ScholarCrossref
14.
Malvehy  J, Hauschild  A, Curiel-Lewandrowski  C,  et al.  Clinical performance of the Nevisense system in cutaneous melanoma detection: an international, multicentre, prospective and blinded clinical trial on efficacy and safety.  Br J Dermatol. 2014;171(5):1099-1107. doi:10.1111/bjd.13121PubMedGoogle ScholarCrossref
15.
Lott  JP, Boudreau  DM, Barnhill  RL,  et al.  Population-based analysis of histologically confirmed melanocytic proliferations using natural language processing.  JAMA Dermatol. 2018;154(1):24-29. doi:10.1001/jamadermatol.2017.4060PubMedGoogle ScholarCrossref
16.
Hodis  E, Watson  IR, Kryukov  GV,  et al.  A landscape of driver mutations in melanoma.  Cell. 2012;150(2):251-263. doi:10.1016/j.cell.2012.06.024PubMedGoogle ScholarCrossref
17.
Hocker  TL, Alikhan  A, Comfere  NI, Peters  MS.  Favorable long-term outcomes in patients with histologically dysplastic nevi that approach a specimen border.  J Am Acad Dermatol. 2013;68(4):545-551. doi:10.1016/j.jaad.2012.09.031PubMedGoogle ScholarCrossref
18.
Reddy  KK, Farber  MJ, Bhawan  J, Geronemus  RG, Rogers  GS.  Atypical (dysplastic) nevi: outcomes of surgical excision and association with melanoma.  JAMA Dermatol. 2013;149(8):928-934. doi:10.1001/jamadermatol.2013.4440PubMedGoogle ScholarCrossref
19.
Duffy  KL, Mann  DJ, Petronic-Rosic  V, Shea  CR.  Clinical decision making based on histopathologic grading and margin status of dysplastic nevi.  Arch Dermatol. 2012;148(2):259-260. doi:10.1001/archdermatol.2011.2045PubMedGoogle ScholarCrossref
20.
Strazzula  L, Vedak  P, Hoang  MP, Sober  A, Tsao  H, Kroshinsky  D.  The utility of re-excising mildly and moderately dysplastic nevi: a retrospective analysis.  J Am Acad Dermatol. 2014;71(6):1071-1076. doi:10.1016/j.jaad.2014.08.025PubMedGoogle ScholarCrossref
21.
Carrera  C, Marchetti  MA, Dusza  SW,  et al.  Validity and reliability of dermoscopic criteria used to differentiate nevi from melanoma: a web-based International Dermoscopy Society study.  JAMA Dermatol. 2016;152(7):798-806. doi:10.1001/jamadermatol.2016.0624PubMedGoogle ScholarCrossref
22.
Tsao  H, Olazagasti  JM, Cordoro  KM,  et al; American Academy of Dermatology Ad Hoc Task Force for the ABCDEs of Melanoma.  Early detection of melanoma: reviewing the ABCDEs.  J Am Acad Dermatol. 2015;72(4):717-723. doi:10.1016/j.jaad.2015.01.025PubMedGoogle ScholarCrossref
23.
Nufer  KL, Raphael  AP, Soyer  HP.  Dermoscopy and overdiagnosis of melanoma in situ.  JAMA Dermatol. 2018;154(4):398-399. doi:10.1001/jamadermatol.2017.6448PubMedGoogle ScholarCrossref
25.
Howlader  CK, Noone  AM, Krapcho  M,  et al. SEER cancer statistics review, 1975-2014. https://seer.cancer.gov/csr/1975_2014/sections.html. Updated April 2, 2018. Accessed July 21, 2017.
26.
Gerami  P, Yao  Z, Polsky  D,  et al.  Development and validation of a noninvasive 2-gene molecular assay for cutaneous melanoma.  J Am Acad Dermatol. 2017;76(1):114-120.e2. doi:10.1016/j.jaad.2016.07.038PubMedGoogle ScholarCrossref
27.
Yao  Z, Moy  R, Allen  T, Jansen  B.  An adhesive patch-based skin biopsy device for molecular diagnostics and skin microbiome studies.  J Drugs Dermatol. 2017;16(10):979-986.PubMedGoogle Scholar
28.
US Centers for Medicare & Medicaid Services. Physician fee schedule search. https://www.cms.gov/apps/physician-fee-schedule/overview.aspx. Accessed July 21, 2017.
29.
Caro  JJ, Briggs  AH, Siebert  U, Kuntz  KM; ISPOR-SMDM Modeling Good Research Practices Task Force.  Modeling good research practices—overview: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force, 1.  Value Health. 2012;15(6):796-803. doi:10.1016/j.jval.2012.06.012PubMedGoogle ScholarCrossref
30.
National Comprehensive Cancer Network. Melanoma NCCN evid. blocks (version 1.2017). https://www.nccn.org/professionals/physician_gls/default.aspx#melanoma. Accessed March 30, 2017.
31.
Egnatios  GL, Dueck  AC, Macdonald  JB,  et al.  The impact of biopsy technique on upstaging, residual disease, and outcome in cutaneous melanoma.  Am J Surg. 2011;202(6):771-777. doi:10.1016/j.amjsurg.2011.06.037PubMedGoogle ScholarCrossref
32.
Mills  JK, White  I, Diggs  B, Fortino  J, Vetto  JT.  Effect of biopsy type on outcomes in the treatment of primary cutaneous melanoma.  Am J Surg. 2013;205(5):585-590. doi:10.1016/j.amjsurg.2013.01.023PubMedGoogle ScholarCrossref
33.
Tong  LX, Wu  PA, Kim  CC.  Degree of clinical concern and dysplasia affect biopsy technique and management of dysplastic nevi with positive biopsy margins: results from a survey of New England dermatologists.  J Am Acad Dermatol. 2016;74(2):389-391.e2. doi:10.1016/j.jaad.2015.09.055PubMedGoogle ScholarCrossref
34.
Grelck  K, Sukal  S, Rosen  L, Suciu  GP.  Incidence of residual nonmelanoma skin cancer in excisions after shave biopsy.  Dermatol Surg. 2013;39(3, pt 1):374-380. doi:10.1111/dsu.12056PubMedGoogle ScholarCrossref
35.
Fleming  NH, Egbert  BM, Kim  J, Swetter  SM.  Reexamining the threshold for reexcision of histologically transected dysplastic nevi.  JAMA Dermatol. 2016;152(12):1327-1334. doi:10.1001/jamadermatol.2016.2869PubMedGoogle ScholarCrossref
36.
Winkelmann  RR, Rigel  DS.  Management of dysplastic nevi: a 14-year follow-up survey assessing practice trends among US dermatologists.  J Am Acad Dermatol. 2015;73(6):1056-1059. doi:10.1016/j.jaad.2015.06.061PubMedGoogle ScholarCrossref
37.
Losina  E, Walensky  RP, Geller  A,  et al.  Visual screening for malignant melanoma: a cost-effectiveness analysis.  Arch Dermatol. 2007;143(1):21-28. doi:10.1001/archderm.143.1.21PubMedGoogle ScholarCrossref
38.
Hieken  TJ, Hernández-Irizarry  R, Boll  JM, Jones Coleman  JE.  Accuracy of diagnostic biopsy for cutaneous melanoma: implications for surgical oncologists.  Int J Surg Oncol. 2013;2013:196493. doi:10.1155/2013/196493PubMedGoogle Scholar
39.
Styperek  A, Kimball  AB.  Malignant melanoma: the implications of cost for stakeholder innovation.  Am J Pharm Benefits. 2012;4(2):66-76.Google Scholar
40.
Alexandrescu  DT.  Melanoma costs: a dynamic model comparing estimated overall costs of various clinical stages.  Dermatol Online J. 2009;15(11):1.PubMedGoogle Scholar
41.
US Bureau of Labor Statistics. Table B-3: average hourly and weekly earnings of all employees on private nonfarm payrolls by industry sector, seasonally adjusted. https://www.bls.gov/news.release/empsit.t19.htm. Modified May 4, 2018. Accessed July 21, 2017.
42.
Tufts Medical Center. Cost-effectiveness analysis registry. http://healtheconomics.tuftsmedicalcenter.org/cear4/home.aspx. Accessed April 7, 2017.
43.
Seidler  AM, Bramlette  TB, Washington  CV, Szeto  H, Chen  SC.  Mohs versus traditional surgical excision for facial and auricular nonmelanoma skin cancer: an analysis of cost-effectiveness.  Dermatol Surg. 2009;35(11):1776-1787. doi:10.1111/j.1524-4725.2009.01291.xPubMedGoogle ScholarCrossref
44.
King  SMC, Bonaccorsi  P, Bendeck  S,  et al.  Melanoma quality of life: pilot study using utility measurements.  Arch Dermatol. 2011;147(3):353-354. doi:10.1001/archdermatol.2010.340PubMedGoogle ScholarCrossref
45.
American Cancer Society. Survival rates for melanoma skin cancer, by stage. https://www.cancer.org/cancer/melanoma-skin-cancer/detection-diagnosis-staging/survival-rates-for-melanoma-skin-cancer-by-stage.html. Accessed July 22, 2017.
46.
Rosendahl  C, Williams  G, Eley  D,  et al.  The impact of subspecialization and dermatoscopy use on accuracy of melanoma diagnosis among primary care doctors in Australia.  J Am Acad Dermatol. 2012;67(5):846-852. doi:10.1016/j.jaad.2011.12.030PubMedGoogle ScholarCrossref
47.
Conic  RZ, Cabrera  CI, Khorana  AA, Gastman  BR.  Determination of the impact of melanoma surgical timing on survival using the National Cancer Database.  J Am Acad Dermatol. 2018;78(1):40-46.e7. doi:10.1016/j.jaad.2017.08.039PubMedGoogle ScholarCrossref
Views 2,106
Citations 0
Original Investigation
July 11, 2018

Economic Analysis of a Noninvasive Molecular Pathologic Assay for Pigmented Skin Lesions

Author Affiliations
  • 1Department of Internal Medicine, Stanford University, Stanford, California
  • 2Cedar Associates, Menlo Park, California
  • 3Department of Dermatology, State University of New York Downstate Medical Center, Brooklyn
  • 4Department of Dermatology, Brooklyn Veterans Administration Medical Center, Brooklyn, New York
JAMA Dermatol. Published online July 11, 2018. doi:10.1001/jamadermatol.2018.1764
Key Points

Question  What are the economic implications of using a noninvasive gene expression test (the pigmented lesion assay) to rule out melanoma and the need for surgical biopsy of atypical pigmented lesions suggestive of melanoma?

Findings  In this cost-savings analysis of the pigmented lesion assay using model inputs for patients with pigmented lesions suggestive of melanoma, a $447 (47%) cost reduction per assessed pigmented lesion vs the current histopathologic standard of care was achievable if the assay was priced at $500. Savings were driven by reduction in initial biopsies and excisions and reduced stage-related treatment costs from missing fewer melanomas.

Meaning  These findings suggest that the pigmented lesion assay reduces cost and may improve the care of patients with primary pigmented skin lesions suggestive of melanoma.

Abstract

Importance  A recently described noninvasive gene expression test (the pigmented lesion assay [PLA]) with adhesive patch–based sampling has the potential to rule out melanoma and the need for surgical biopsy of pigmented lesions suggestive of melanoma with a negative predictive value of 99% compared with 83% for the histopathologic standard of care. The cost implications of using this molecular test vs visual assessment followed by biopsy and histopathologic assessment (VAH) have not been evaluated.

Objective  To determine potential cost savings of PLA use vs the VAH pathway.

Design, Setting, and Participants  This health economic analysis performed from a US payer perspective was based on consensus treatment guidelines and fee schedules from the Centers for Medicare & Medicaid Services. Data for model input were derived from routine use of the test in US dermatology practices and literature. Participants included patients with primary cutaneous pigmented lesions suggestive of melanoma. Data were analyzed from February 8 to December 1, 2017.

Main Outcomes and Measures  The primary analysis consisted of the relative reduction in costs of diagnostic surgical procedures for PLA vs VAH management. Additional analyses included stage-related treatment costs associated with delays in diagnosis.

Results  In the cost analysis for this economic model, the relative reduction in surgical procedure costs (biopsy and subsequent excision), assuming $0 for the PLA to facilitate multiple comparison scenarios, was −$395 compared with VAH. The relative reduction in stage-related treatment costs associated with the PLA was −$433 compared with VAH, primarily associated with avoidance of delays due to false-negative diagnoses. Surveillance costs were reduced by −$119 with the PLA. The total cost of fully adjudicating a lesion suggestive of melanoma by VAH was $947. At a mean selling price reference point for PLA of $500, cost savings of $447 (47%) per lesion tested could be realized.

Conclusions and Relevance  The results of this analysis suggest that the PLA reduces cost and may improve the care of patients with primary pigmented skin lesions suggestive of melanoma.

Introduction

Management of atypical pigmented lesions involves ruling out melanoma via visual assessment, followed by surgical biopsy and histopathologic assessment (VAH).1-3 The goal of this assessment is to identify melanomas at their earliest stages (in situ or stage I) when a high cure rate is possible by wide excision.4 Although the purpose of the VAH pathway is to rule out melanoma, the poor performance metrics of this diagnostic pathway lead to a low negative predictive value (NPV) for early-stage disease. The low specificity of visual assessment (3.7%-32.0%) results in a high number of lesions with false-positive biopsy results.5-11 Therefore, the primary and difficult role of histopathologic assessment in this setting is to identify the small number of true-positive lesions from a large pool, including a large number of false-positive lesions. However, significant overlap in the histopathologic diagnostic criteria exists between atypical nevi and early-stage melanoma, invariably leading to false-negative diagnoses and a relatively low sensitivity of histopathologic assessment (81%-84%).12-14 With the prevalence of early-stage melanoma in biopsy specimens at approximately 6% and ranging from 2% to 10% in many settings,1,2,14-16 the NPV of the surgical biopsy plus histopathologic diagnostic paradigm is unexpectedly low in most settings. In a study by Malvehy et al,14 206 cases of melanoma in situ and stage IA invasive melanoma (thickness <0.75 mm) were diagnosed with a sensitivity of 81%, a specificity of 10%, and an NPV of 83%.

This low NPV for the current standard of care pathway is accompanied by a high number of unnecessary surgical procedures driven by the poor specificity of visual assessment.8 The mean number of surgical biopsies needed to identify 1 melanoma (number needed to biopsy [NNB]) is approximately 20 and ranges from 8 to 30 depending on the setting.9-14 Conservative management of biopsied lesions with atypia and positive margins leads to a high number of subsequent unnecessary excisions with margins.13-17 Approximately 5.2 excisions with margins are performed per melanoma identified.13,17-22 Less than 1.0% of lesions with atypia and positive margins that undergo excision are diagnostically upgraded to melanoma.12,13 This notion that the current pathway has a significant number of unnecessary surgical procedures is also supported by other investigators,13-17,21,23 who recently found that more than 90% of skin biopsies to rule out melanoma were attributed to benign and low-risk lesions. Approximately 3.0 million surgical biopsies and 780 000 excisions are performed in the United States annually to find approximately 150 000 in situ and invasive melanomas as part of the current diagnostic pathway for atypical pigmented lesions.8,9,23-25 These findings stress the need for cost-effective tools that may improve management of pigmented lesions.

One such tool, the pigmented lesion assay (PLA), is a gene expression test that helps rule out melanoma and the need for a surgical biopsy of atypical pigmented lesions. The PLA is based on a new platform technology for noninvasive genomic testing of the skin, which allows the analysis of samples collected from adhesive patches.11,22,24-27 In contrast to the current VAH pathway, the PLA has a very high NPV (>99%) that reduces the probability of missing melanomas from about 17% using the current standard of care (NPV of 83%) to less than 1%.26,27 In addition, the noninvasive sampling leads to dramatic reduction in surgical biopsies and subsequent excisions. The aim of this study was to demonstrate the economic implications of the noninvasive PLA compared with the current standard of care, the VAH pathway.

Methods

The Figure depicts the decision paradigm used in the model. The PLA and the VAH rely on visual assessment to identify lesions suggestive of melanoma with 1 or more ABCDE (asymmetry, border, color, diameter, and evolution) criteria. The PLA provides a dichotomous result (positive or negative), with positive test results entering the surgical biopsy pathway. For the VAH, all (100%) lesions suggestive of melanoma enter the biopsy pathway. Management after biopsy is based on histopathologic diagnosis. All studies used for the economic analysis were approved by the Western-Copernicus Group independent review board and were conducted in accordance with the Declaration of Helsinki principles. Informed consent was obtained for all the underlying studies required to conduct this analysis.

Data were analyzed from February 8 to December 1, 2017. An economic model was constructed from the US payer perspective28 aimed at assessing the initial surgical costs associated with the VAH. For this purpose, data were presented with (1) a theoretical zero cost ($0) relative reference point for the PLA to facilitate assessment of various cost-saving scenarios and (2) pricing of the PLA at $500, close to the expected mean selling price. Direct medical costs for initial surgical management included office visits, biopsy procedures, histopathologic evaluation and special stains, and management of repairs and complications. Unit costs were based on the Centers for Medicare & Medicaid Services reimbursement rates for the corresponding Current Procedural Terminology codes.28 All costs were adjusted to 2017 US dollars. Primary cost outcomes included were initial surgical procedures (biopsy and excisions), stage-related melanoma treatment costs, indirect costs, and costs related to patient experience (quality of life).20,29-45

Key diagnostic test input variables to compare costs for initial surgical procedures include the sensitivity and specificity of the modalities, the type of initial biopsy procedure, and the excision rate for histopathologically atypical and/or ambiguous lesions with positive margins. For the PLA, the sensitivity and specificity were derived from a retrospective medical record review of real-world use of 280 PLA-negative and 39 PLA-positive cases at 2 dermatology sites to maximize clinical relevance of the data obtained (Table 1). For the VAH, a specificity of 32% was used based on a utility study by Ferris et al11 that compared the PLA with the VAH in 60 pigmented lesions. The sensitivity of the VAH is well published in the literature.1-22,46 However, to minimize bias and to facilitate comparisons, the concordance of histopathologic diagnoses at the primary site against consensus diagnoses obtained by a panel of 3 dermatopathologists in 128 cases of in situ and stage I invasive melanoma of our own archival cases was evaluated in this study. This review demonstrated a sensitivity of 84% for the VAH for representative early-stage lesions. A tangential biopsy (shave or scoop), as performed most often in the United States, was assumed at a rate of 70%, with the remaining 30% assumed to be excisional or punch procedures.31 The analysis includes the incidence of procedure-related complications, such as infection and bleeding.31-33 A literature review indicated that a rate of excision after initial biopsy of atypia with positive tissue margins of 20% was suitable.20,34-36

To assess stage-related treatment costs, the false-negative rate of the PLA and VAH were used to establish the proportion of patients with delayed diagnosis. A delay of at least 3 months in diagnosing melanoma was assumed based on guidelines for interval surveillance recommended by the National Comprehensive Cancer Network.30 The effects on clinical stage of delayed diagnoses were derived from Losina et al37 in a model that suggests that the rate of progression of an undiagnosed lesion to a more advanced stage is 10% per year.47 Distribution and overall survival of melanoma by stage is based on population statistics reported by the American Cancer Society.45

Stage-related costs attributable to melanoma after initial surgical management varied from a low of $6041 for management of stage 0 disease (primarily surveillance of new lesions) to more than $200 000 for management of stage IV disease (systemic therapy, including recently approved options and end-of-life care).39,40 Indirect costs were estimated based on assumptions made on the time to attend a clinic visit, including commute time (0.75-2.00 hours), with a unit cost of $26 per hour.40 King et al44 reported disutilities for tangential biopsy procedures at −0.001, excision procedures at −0.004, and subsequent repair procedures at −0.021. The model multiplied mean durations of overall survival by the utility for each stage, weighted by the proportion of patients diagnosed in each stage to derive mean melanoma-related quality-adjusted life-years. A univariate sensitivity analysis was performed on key variables based on published estimates with 95% CIs of −33% to 33%.28

Results
Economic Model Base Case

The Figure compares how use of the PLA and the VAH compare at the decision diagram level. Key economic model input variables and base-case results are summarized in Table 2 and Table 3. For the base-case model, we assumed the pretest probability of melanoma was 6% (range, 2%-10%) and that 56% of lesions evaluated would be considered clinically suggestive enough to recommend biopsy.1,2,13,22

The proportion of patients undergoing initial surgical biopsy of any type was 69.0% with the VAH vs 13.3% with the PLA. The improved accuracy of the PLA reduces the number of patients undergoing subsequent excision for melanoma from 18.8% with the VAH to 7.5% with the PLA. The NNB dropped from 15.7 with VAH to 2.7 with the PLA. This result aligns well with the NNB of 2.8 found at the 2 dermatology clinic sites assessed. With the VAH, 68.0% of patients with melanoma had an immediate and accurate diagnosis of melanoma compared with 75.5% of patients undergoing PLA assessment, such that 7.5% of patients would avoid a delayed diagnosis and progress to a more advanced stage.

Cost Differential PLA vs VAH

Assuming a theoretical $0 PLA cost reference point to facilitate comparisons of multiple scenarios, costs associated with initial surgical biopsies were lower for the PLA ($76) compared with the VAH ($396) (relative difference, −$320) (Table 3). The cost of excisional procedures was also lower with the PLA ($50) vs the VAH ($125), demonstrating that the clinical performance of the PLA reduced overall costs of surgical procedures associated with the assessment of pigmented lesions (PLA vs VAH, −$395). No difference existed in patient visits for initial diagnostic evaluation, but the PLA had a lower cost for subsequent office visits for surveillance (PLA vs VAH, −$119). Because of the high treatment costs for stage-related management of melanoma, particularly with newer targeted therapies, avoidance of diagnostic delays significantly affects treatment costs. Lower false-negative diagnoses with the PLA reduces these treatment costs by $433 (PLA, $1616 per melanoma; VAH, $2049 per melanoma).39 Indirect costs from lost employee wages and productivity were $101 for VAH and $54 for the PLA. Applying the utility estimates by melanoma stage from King et al,44 the difference in quality-adjusted life-years was 0.016 (VAH, 16.707; PLA, 16.743). Use of the PLA at an assumed mean selling price reference point of $500 leads to $447 in direct cost reductions relative to the VAH for each evaluated primary pigmented skin lesion clinically suggestive of melanoma.

Sensitivity Analysis

The PLA pathway was dominant compared with the VAH pathway in all sensitivity analyses (Table 4), meaning it reduced total costs and improved the patient experience. The most influential variables on costs were the estimate of increased specificity for PLA relative to VAH (range, −$1014 to −$457) and melanoma treatment costs. The probabilistic sensitivity analysis showed that the incremental costs and quality-adjusted survival points always fell in the bottom right quadrant of the cost-effectiveness plane. The PLA was a less costly and more effective strategy in all simulations regardless of which utilities for melanoma stage were applied.

Discussion

The PLA is used as a rule-out test to aid the decision by physicians to perform a surgical biopsy and not as a screening test on pigmented lesions without clinical risk factors. Positive PLA test results are followed up with a surgical biopsy and histopathologic assessment, whereas the negative test results (a much larger group) are followed up with surveillance per standard of care. This economic model demonstrates that incorporating the noninvasive PLA into the diagnostic paradigm for pigmented skin lesions suggestive of melanoma can yield substantial cost savings relative to the VAH standard of care.

The economic benefits of the PLA are driven by high specificity that substantially reduces the NNB to find melanoma by 5-fold (from 15.7 to approximately 2.7). Most surgical biopsies have negative findings and can be considered unnecessary. In a study by Lott et al15 of 80 368 surgical biopsies, 83% were considered class I (benign or mildly atypical) and 8% were considered class II (moderately to severely atypical). In a review of more than 15 000 PLA test results, 88% of tests had negative results, eliminating unnecessary surgical procedures (Laura Ferris, MD, PhD; written communication; February 1, 2018).

As many as 30% of biopsied lesions have cellular atypia on histopathologic assessment with positive tissue margins.20 Conservative management of these results leads to a high number of subsequent excisions with margins. Only 0.2% to less than 1.0% of lesions with atypia and positive margins that undergo excision are diagnostically upgraded to melanoma in situ, suggesting that most of these excisions are similarly unnecessary.14,16 Thus, as the model demonstrates, reducing the number of unnecessary surgical biopsies will also reduce unnecessary excisional procedures, thereby driving even more savings.

In this study, a relatively generous specificity of 32% was given to the VAH pathway. Work by Monheit et al5 and Argenziano et al6 describe visual assessment specificity of 3.7% to 6.0%, and the VAH specificity is likely less than 10%. In the real world, this lower specificity would be associated with even more surgical procedures, and an even larger economic benefit of the noninvasive PLA may be realized.

In contrast to the current pathway, the PLA also has a very high NPV (>99%),26 which reduces the likelihood that a melanoma will be missed and subject to a delayed diagnosis. Melanomas with a delayed diagnosis have higher stage-related treatment costs. This finding may be more relevant for early-stage disease. Conic and colleagues47 recently described a delay of surgery of more than 29 days as having a negative effect on overall survival for stage I melanoma while not providing information on melanoma-specific survival. Their study leaves unclear whether delays in treatment resulted in patients dying of melanoma, but the data can be seen as supporting the notion that minimizing delays in treatment may benefit patients with melanoma. The PLA test has a 72-hour turnaround time; although patients with positive test results must return for a biopsy, that return visit is unlikely to lead to a significant delay.

Limitations

Limitations of the PLA test include that it does not work on palms of hands and soles of feet owing to insufficient RNA in tissue samples obtained from these locations via adhesive patches. Adhesive patches are also not suited to harvest of tissue from mucous membranes. The potentially biggest limitation of this economic model stems from the assumption that small delays in diagnosis and stage progression owing to a missed melanoma combined with the significant cost of late-stage therapy create meaningful opportunities for cost savings. We attempted to base this notion on evidence from the literature most prominently supported by the work by Conic et al,47 who found that overall survival decreased in patients waiting longer than 90 days for surgical treatment regardless of stage and that delay of surgery beyond the first 29 days had a negative effect on overall survival of stage I melanoma. Other limitations are based on assumptions that attempted to generalize direct medical costs for office visits (which vary by state and are influenced by a patient’s insurance type and coverage), for biopsy procedures and practices (which also vary among health care professionals and dermatology offices and centers), for histopathologic practices (including the frequency of use of special stains), and for the differential management of repairs and complications. Other variables we attempted to capture within the confines of assumptions include conditions for reexcisions, margin selection, and stage-related treatment costs. With generalized assumptions, we were also unable to fully capture the rapidly changing cost structure and landscape of checkpoint inhibitors, BRAF antagonists, and emerging combination strategies for late-stage melanoma and the future use of these high-cost treatment options and adjuvant and neoadjuvant settings.

Conclusions

The PLA provides clinical utility in the assessment of pigmented lesions suggestive of melanoma by its ability to transform the current clinical pathway from one that is subjective (hence variable in implementation), invasive, and with a relatively low NPV to one that is objective (hence more predictable), noninvasive, and with a high NPV.5,6,13,14,20,26 Pricing of the PLA at $500 would lead to $447 or 47% in health care cost savings vs the VAH standard of care pathway per assessed primary cutaneous lesion suggestive of melanoma. Use of the PLA reduces surgical procedures, missed melanomas, and cost.

Back to top
Article Information

Accepted for Publication: April 30, 2018.

Corresponding Author: Daniel M. Siegel, MD, MS, CPCD, Department of Dermatology, State University of New York Downstate Medical Center, 450 Clarkson Ave, Basic Science Building, PO Box 46, Brooklyn, NY 11203 (cyberderm@dermsurg.org).

Published Online: July 11, 2018. doi:10.1001/jamadermatol.2018.1764

Open Access: This article is published under the JN-OA license and is free to read on the day of publication.

Author Contributions: Drs Hornberger and Siegel had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Both authors.

Acquisition, analysis, or interpretation of data: Both authors.

Critical revision of the manuscript for important intellectual content: Both authors.

Statistical analysis: Hornberger.

Administrative, technical, or material support: Siegel.

Supervision: Siegel.

Conflict of Interest Disclosures: Dr Hornberger reported serving as a consultant to DermTech, Inc, and being an employee of Cedar Associates, which received funding from the sponsor to conduct the research. Dr Siegel reported serving as a consultant and a Scientific Advisory Board member for DermTech, Inc, and receiving stock ownership or options in DermTech, Inc. No other disclosures were reported.

Funding/Support: This study was supported in part by DermTech, Inc.

Role of the Funder/Sponsor: The funder was not involved in the design and conduct of the study or the collection, management, analysis, and interpretation of the data. DermTech, Inc, provided assistance with the preparation and review, but not the approval of the manuscript or the decision to submit the manuscript for publication

Additional Contributions: Laura Ferris, MD, PhD, University of Pittsburgh, Pittsburgh, Pennsylvania, and Darrell Rigel, MD, and Ryan Svoboda, MD, New York University Medical Center, New York, New York, provided critical review of the manuscript. No compensation for the review was given.

References
1.
Friedman  RJ, Farber  MJ, Warycha  MA, Papathasis  N, Miller  MK, Heilman  ER.  The “dysplastic” nevus.  Clin Dermatol. 2009;27(1):103-115. doi:10.1016/j.clindermatol.2008.09.008PubMedGoogle ScholarCrossref
2.
Schäfer  T, Merkl  J, Klemm  E, Wichmann  HE, Ring  J; KORA Study Group.  The epidemiology of nevi and signs of skin aging in the adult general population: results of the KORA-survey 2000.  J Invest Dermatol. 2006;126(7):1490-1496. doi:10.1038/sj.jid.5700269PubMedGoogle ScholarCrossref
3.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, I: common and atypical naevi.  Eur J Cancer. 2005;41(1):28-44. doi:10.1016/j.ejca.2004.10.015PubMedGoogle ScholarCrossref
4.
Rigel  DS, Russak  J, Friedman  R.  The evolution of melanoma diagnosis: 25 years beyond the ABCDs.  CA Cancer J Clin. 2010;60(5):301-316. doi:10.3322/caac.20074PubMedGoogle ScholarCrossref
5.
Monheit  G, Cognetta  AB, Ferris  L,  et al.  The performance of MelaFind: a prospective multicenter study.  Arch Dermatol. 2011;147(2):188-194. doi:10.1001/archdermatol.2010.302PubMedGoogle ScholarCrossref
6.
Argenziano  G, Cerroni  L, Zalaudek  I,  et al.  Accuracy in melanoma detection: a 10-year multicenter survey.  J Am Acad Dermatol. 2012;67(1):54-59. doi:10.1016/j.jaad.2011.07.019PubMedGoogle ScholarCrossref
7.
Nault  A, Zhang  C, Kim  K, Saha  S, Bennett  DD, Xu  YG.  Biopsy use in skin cancer diagnosis: comparing dermatology physicians and advanced practice professionals.  JAMA Dermatol. 2015;151(8):899-902. doi:10.1001/jamadermatol.2015.0173PubMedGoogle ScholarCrossref
8.
Wilson  RL, Yentzer  BA, Isom  SP, Feldman  SR, Fleischer  AB  Jr.  How good are US dermatologists at discriminating skin cancers? a number-needed-to-treat analysis.  J Dermatolog Treat. 2012;23(1):65-69. doi:10.3109/09546634.2010.512951PubMedGoogle ScholarCrossref
9.
Hansen  C, Wilkinson  D, Hansen  M, Argenziano  G.  How good are skin cancer clinics at melanoma detection? number needed to treat variability across a national clinic group in Australia.  J Am Acad Dermatol. 2009;61(4):599-604. doi:10.1016/j.jaad.2009.04.021PubMedGoogle ScholarCrossref
10.
Wang  DM, Morgan  FC, Besaw  RJ, Schmults  CD.  An ecological study of skin biopsies and skin cancer treatment procedures in the United States Medicare population, 2000 to 2015.  J Am Acad Dermatol. 2018;78(1):47-53. doi:10.1016/j.jaad.2017.09.031PubMedGoogle ScholarCrossref
11.
Ferris  LK, Jansen  B, Ho  J,  et al.  Utility of a noninvasive 2-gene molecular assay for cutaneous melanoma and effect on the decision to biopsy.  JAMA Dermatol. 2017;153(7):675-680. doi:10.1001/jamadermatol.2017.0473PubMedGoogle ScholarCrossref
12.
Urso  C, Rongioletti  F, Innocenzi  D,  et al.  Histological features used in the diagnosis of melanoma are frequently found in benign melanocytic naevi.  J Clin Pathol. 2005;58(4):409-412. doi:10.1136/jcp.2004.020933PubMedGoogle ScholarCrossref
13.
Elmore  JG, Barnhill  RL, Elder  DE,  et al.  Pathologists’ diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study.  BMJ. 2017;357:j2813. doi:10.1136/bmj.j2813PubMedGoogle ScholarCrossref
14.
Malvehy  J, Hauschild  A, Curiel-Lewandrowski  C,  et al.  Clinical performance of the Nevisense system in cutaneous melanoma detection: an international, multicentre, prospective and blinded clinical trial on efficacy and safety.  Br J Dermatol. 2014;171(5):1099-1107. doi:10.1111/bjd.13121PubMedGoogle ScholarCrossref
15.
Lott  JP, Boudreau  DM, Barnhill  RL,  et al.  Population-based analysis of histologically confirmed melanocytic proliferations using natural language processing.  JAMA Dermatol. 2018;154(1):24-29. doi:10.1001/jamadermatol.2017.4060PubMedGoogle ScholarCrossref
16.
Hodis  E, Watson  IR, Kryukov  GV,  et al.  A landscape of driver mutations in melanoma.  Cell. 2012;150(2):251-263. doi:10.1016/j.cell.2012.06.024PubMedGoogle ScholarCrossref
17.
Hocker  TL, Alikhan  A, Comfere  NI, Peters  MS.  Favorable long-term outcomes in patients with histologically dysplastic nevi that approach a specimen border.  J Am Acad Dermatol. 2013;68(4):545-551. doi:10.1016/j.jaad.2012.09.031PubMedGoogle ScholarCrossref
18.
Reddy  KK, Farber  MJ, Bhawan  J, Geronemus  RG, Rogers  GS.  Atypical (dysplastic) nevi: outcomes of surgical excision and association with melanoma.  JAMA Dermatol. 2013;149(8):928-934. doi:10.1001/jamadermatol.2013.4440PubMedGoogle ScholarCrossref
19.
Duffy  KL, Mann  DJ, Petronic-Rosic  V, Shea  CR.  Clinical decision making based on histopathologic grading and margin status of dysplastic nevi.  Arch Dermatol. 2012;148(2):259-260. doi:10.1001/archdermatol.2011.2045PubMedGoogle ScholarCrossref
20.
Strazzula  L, Vedak  P, Hoang  MP, Sober  A, Tsao  H, Kroshinsky  D.  The utility of re-excising mildly and moderately dysplastic nevi: a retrospective analysis.  J Am Acad Dermatol. 2014;71(6):1071-1076. doi:10.1016/j.jaad.2014.08.025PubMedGoogle ScholarCrossref
21.
Carrera  C, Marchetti  MA, Dusza  SW,  et al.  Validity and reliability of dermoscopic criteria used to differentiate nevi from melanoma: a web-based International Dermoscopy Society study.  JAMA Dermatol. 2016;152(7):798-806. doi:10.1001/jamadermatol.2016.0624PubMedGoogle ScholarCrossref
22.
Tsao  H, Olazagasti  JM, Cordoro  KM,  et al; American Academy of Dermatology Ad Hoc Task Force for the ABCDEs of Melanoma.  Early detection of melanoma: reviewing the ABCDEs.  J Am Acad Dermatol. 2015;72(4):717-723. doi:10.1016/j.jaad.2015.01.025PubMedGoogle ScholarCrossref
23.
Nufer  KL, Raphael  AP, Soyer  HP.  Dermoscopy and overdiagnosis of melanoma in situ.  JAMA Dermatol. 2018;154(4):398-399. doi:10.1001/jamadermatol.2017.6448PubMedGoogle ScholarCrossref
25.
Howlader  CK, Noone  AM, Krapcho  M,  et al. SEER cancer statistics review, 1975-2014. https://seer.cancer.gov/csr/1975_2014/sections.html. Updated April 2, 2018. Accessed July 21, 2017.
26.
Gerami  P, Yao  Z, Polsky  D,  et al.  Development and validation of a noninvasive 2-gene molecular assay for cutaneous melanoma.  J Am Acad Dermatol. 2017;76(1):114-120.e2. doi:10.1016/j.jaad.2016.07.038PubMedGoogle ScholarCrossref
27.
Yao  Z, Moy  R, Allen  T, Jansen  B.  An adhesive patch-based skin biopsy device for molecular diagnostics and skin microbiome studies.  J Drugs Dermatol. 2017;16(10):979-986.PubMedGoogle Scholar
28.
US Centers for Medicare & Medicaid Services. Physician fee schedule search. https://www.cms.gov/apps/physician-fee-schedule/overview.aspx. Accessed July 21, 2017.
29.
Caro  JJ, Briggs  AH, Siebert  U, Kuntz  KM; ISPOR-SMDM Modeling Good Research Practices Task Force.  Modeling good research practices—overview: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force, 1.  Value Health. 2012;15(6):796-803. doi:10.1016/j.jval.2012.06.012PubMedGoogle ScholarCrossref
30.
National Comprehensive Cancer Network. Melanoma NCCN evid. blocks (version 1.2017). https://www.nccn.org/professionals/physician_gls/default.aspx#melanoma. Accessed March 30, 2017.
31.
Egnatios  GL, Dueck  AC, Macdonald  JB,  et al.  The impact of biopsy technique on upstaging, residual disease, and outcome in cutaneous melanoma.  Am J Surg. 2011;202(6):771-777. doi:10.1016/j.amjsurg.2011.06.037PubMedGoogle ScholarCrossref
32.
Mills  JK, White  I, Diggs  B, Fortino  J, Vetto  JT.  Effect of biopsy type on outcomes in the treatment of primary cutaneous melanoma.  Am J Surg. 2013;205(5):585-590. doi:10.1016/j.amjsurg.2013.01.023PubMedGoogle ScholarCrossref
33.
Tong  LX, Wu  PA, Kim  CC.  Degree of clinical concern and dysplasia affect biopsy technique and management of dysplastic nevi with positive biopsy margins: results from a survey of New England dermatologists.  J Am Acad Dermatol. 2016;74(2):389-391.e2. doi:10.1016/j.jaad.2015.09.055PubMedGoogle ScholarCrossref
34.
Grelck  K, Sukal  S, Rosen  L, Suciu  GP.  Incidence of residual nonmelanoma skin cancer in excisions after shave biopsy.  Dermatol Surg. 2013;39(3, pt 1):374-380. doi:10.1111/dsu.12056PubMedGoogle ScholarCrossref
35.
Fleming  NH, Egbert  BM, Kim  J, Swetter  SM.  Reexamining the threshold for reexcision of histologically transected dysplastic nevi.  JAMA Dermatol. 2016;152(12):1327-1334. doi:10.1001/jamadermatol.2016.2869PubMedGoogle ScholarCrossref
36.
Winkelmann  RR, Rigel  DS.  Management of dysplastic nevi: a 14-year follow-up survey assessing practice trends among US dermatologists.  J Am Acad Dermatol. 2015;73(6):1056-1059. doi:10.1016/j.jaad.2015.06.061PubMedGoogle ScholarCrossref
37.
Losina  E, Walensky  RP, Geller  A,  et al.  Visual screening for malignant melanoma: a cost-effectiveness analysis.  Arch Dermatol. 2007;143(1):21-28. doi:10.1001/archderm.143.1.21PubMedGoogle ScholarCrossref
38.
Hieken  TJ, Hernández-Irizarry  R, Boll  JM, Jones Coleman  JE.  Accuracy of diagnostic biopsy for cutaneous melanoma: implications for surgical oncologists.  Int J Surg Oncol. 2013;2013:196493. doi:10.1155/2013/196493PubMedGoogle Scholar
39.
Styperek  A, Kimball  AB.  Malignant melanoma: the implications of cost for stakeholder innovation.  Am J Pharm Benefits. 2012;4(2):66-76.Google Scholar
40.
Alexandrescu  DT.  Melanoma costs: a dynamic model comparing estimated overall costs of various clinical stages.  Dermatol Online J. 2009;15(11):1.PubMedGoogle Scholar
41.
US Bureau of Labor Statistics. Table B-3: average hourly and weekly earnings of all employees on private nonfarm payrolls by industry sector, seasonally adjusted. https://www.bls.gov/news.release/empsit.t19.htm. Modified May 4, 2018. Accessed July 21, 2017.
42.
Tufts Medical Center. Cost-effectiveness analysis registry. http://healtheconomics.tuftsmedicalcenter.org/cear4/home.aspx. Accessed April 7, 2017.
43.
Seidler  AM, Bramlette  TB, Washington  CV, Szeto  H, Chen  SC.  Mohs versus traditional surgical excision for facial and auricular nonmelanoma skin cancer: an analysis of cost-effectiveness.  Dermatol Surg. 2009;35(11):1776-1787. doi:10.1111/j.1524-4725.2009.01291.xPubMedGoogle ScholarCrossref
44.
King  SMC, Bonaccorsi  P, Bendeck  S,  et al.  Melanoma quality of life: pilot study using utility measurements.  Arch Dermatol. 2011;147(3):353-354. doi:10.1001/archdermatol.2010.340PubMedGoogle ScholarCrossref
45.
American Cancer Society. Survival rates for melanoma skin cancer, by stage. https://www.cancer.org/cancer/melanoma-skin-cancer/detection-diagnosis-staging/survival-rates-for-melanoma-skin-cancer-by-stage.html. Accessed July 22, 2017.
46.
Rosendahl  C, Williams  G, Eley  D,  et al.  The impact of subspecialization and dermatoscopy use on accuracy of melanoma diagnosis among primary care doctors in Australia.  J Am Acad Dermatol. 2012;67(5):846-852. doi:10.1016/j.jaad.2011.12.030PubMedGoogle ScholarCrossref
47.
Conic  RZ, Cabrera  CI, Khorana  AA, Gastman  BR.  Determination of the impact of melanoma surgical timing on survival using the National Cancer Database.  J Am Acad Dermatol. 2018;78(1):40-46.e7. doi:10.1016/j.jaad.2017.08.039PubMedGoogle ScholarCrossref
×