Beware: Discordance abounds among pathologists in the diagnosis of melanocytic neoplasms

By Jason B. Lee, MD
Dec. 4, 2019
Vol. 1, No. 38

Discordance in the histological diagnosis of melanocytic lesions has been an underemphasized and understudied problem, and, therefore, the scale and the impact of the problem is largely unknown. Over the years, there were small studies and observations suggesting that the problem of discordance is a significant one that needed more attention from all involved disciplines. (1,2) The magnitude of the problem was highlighted in the latest and largest iteration of study of its kind by Elmore JG et al., involving 240 cases and 187 U.S. pathologists. (3) Participating pathologists independently interpreted preselected cases of melanocytic lesions using their own diagnostic terms. Melanocytic lesions were categorized into 5 diagnostic classes based on Melanocytic Pathology Assessment Tool and Hierarchy for Diagnosis (MPATH-Dx), a recently developed classification scheme that takes the multitude of diagnostic terms that are utilized by pathologists into account (Table 1). The accuracy of the diagnosis was compared to a consensus reference diagnosis predetermined by a panel of three experienced dermatopathologists. The concordance rate for Class II lesions was 25%, for Class III lesions, 40%, and for Class IV lesions, 45%. A wide swath of melanocytic lesions spans these three classes — moderate to severe dysplastic nevi, Spitz nevi, atypical Spitz nevi, and melanomas both in situ and Stage 1A lesions. The authors of the study concluded, in their own words: “the diagnoses spanning moderately dysplastic nevi to early stage invasive melanoma were neither reproducible nor accurate.” This is an unsettling conclusion that questions the very gold standard that has long been accepted in diagnosing melanocytic neoplasms.

Table 1


MPATH-Dx* Class	Types of lesions	% Concordance with reference dx (95% CI)
I	Common melanocytic nevus; blue nevus; mildly dysplastic nevus	92 (90 to 94)
II	Moderately dysplastic nevus; Spitz nevus	25 (22 to 28)
III	Melanoma in-situ, severely dysplastic nevus; atypical Spitz tumor	40 (37 to 44)
IV	Melanoma pT1a (<0.8mm depth)	43 (39 to 46)
V	Melanoma pT1b or higher	72 (69 to 75)

*Melanocytic Pathology Assessment Tool and Hierarchy for Diagnosis
Adapted from Elmore, JG et al. BMJ 2017; 357: j2813. Web.




In the previous era, pathologists were presented with thick and bulky melanomas to interpret and discordance was not an issue as the diagnosis was predictive of biologic behavior. In the modern era, pathologists are presented with ever so small and thin melanocytic lesions to interpret, driven by the goal of capturing early or incipient melanomas by clinicians. They may be histologically interpreted as atypical or even early melanomas, but the diagnoses of these clinically subtle lesions are no longer tied to biologic behavior — our ability to diagnose these lesions with specificity is no better than flipping a coin according to this latest study.

Given the high discordance rate among pathologists in the diagnosis of melanocytic neoplasms, I have to wonder, how many of the countless “dysplastic” nevi that have been biopsied and subsequently excised in the last three decades were actually necessary procedures? How many of the melanoma in-situ lesions that have doubled in incidence in the last decade are authentic melanomas? These are the very lesions in which interobserver agreement was poor.

The process of interpreting sections on a slide is entirely subjective, and, as such, discordant interpretations are expected. The magnitude of discordance revealed in this study, however, calls into question the validity of the histologic interpretation melanocytic neoplasms by pathologists. Compared to diagnostic discordance in other disciplines, the authors claim that “the findings reported here are more pronounced than in other disciplines of medicine.”

The authors provide several recommendations to address the problem. First, a simplified management-oriented classification of melanocytic neoplasm is recommended, citing the MPATH-Dx classification as an example. The authors plead for a standardized diagnostic terminology by detailing a case in which 18 different diagnostic terms were attributed to it. The second recommendation is the acknowledgment of limitation in the ability to classify melanocytic lesions accurately. Lastly, they recommend the development of validated adjunctive molecular tests to support the diagnostic assessment of the pathologist. Molecular diagnostics offer the promise of a more precise and accurate method of diagnosing melanocytic lesions, but these tests have their own limitations as consensus diagnosis is used for reference. (4)

For cases in the middle of the histopathological spectrum, such as flat thin and Spitzoid melanocytic neoplasms, more outcome studies are needed to determine their biological behavior. Armed with better outcomes data, clinicians will have more confidence in the management of problematic middle-of-the-histopathological-spectrum lesions. For example, multiple outcomes studies thus far have shown that the association of melanoma and dysplastic nevi histologically graded as moderate or severe is so low, they can be monitored clinically rather than further procedural intervention. (5,6) To take it one step further, the practice of histological grading and providing comments on margins of dysplastic nevi could be discouraged since they are no longer considered precursor lesions. (7) Any histologic grading or margin comments on pathology reports exerts some degree of pressure on clinicians to intervene rather than recommend no further treatment. Pathologists, instead, could provide explicit excision recommendation for those lesions is which melanoma cannot be confidently excluded.

The major limitation of the study is the use of preselected cases, that is, cases that have been managed already. As the diagnostic assessment of preselected cases has no consequences for the participating pathologists, their diagnostic threshold may be different than in real-world practice. Designing a study that reflects the real-world practice of pathology, however, may be logistically very difficult to overcome and may not be feasible.

The authors of this large accuracy and reproducibility study confirm previous observations of discordance among pathologists in the diagnosis of melanocytic neoplasms, particularly the middle histopathological spectrum. Stakeholders from all disciplines need to heed the discordance problem and work together to methodically address the large and complex problem. The first step is to acknowledge the problem, done laudably by the authors of the study.

Point to Remember: Discordance in interpreting melanocytic lesions by dermatopathologists is common. Rectifying this problem is essential in providing optimal patient care, especially for those patients with borderline atypical lesions.

---

Our Editor's Viewpoint

Warren R. Heymann, MD

Dr. Lee offers an outstanding perspective on the discordance between dermatopathologists in their interpretation of melanocytic lesions. Until molecular studies are available that precisely reveal a lesion’s genetic fingerprint allowing for accurate prognostication, we will need to rely on the subjective histologic assessment of the dermatopathologist. Unquestionably, one dermatopathologist’s report of a severely atypical compound nevus is another dermatopathologist’s melanoma. The key is communicating with your dermatopathologist so you understand precisely what is meant in the report. While I still comment on margins (patients often inquire — “Doctor, was the lesion removed?”), I only recommend a re-excision of the site when the lesion is severely atypical. I am confident that in short order molecular techniques with precise prognostication will render this discussion moot.

1. Farmer ER, Gonin R, Hanna MP. Discordance in the histopathologic diagnosis of melanoma and melanocytic nevi between expert pathologists. Hum Pathol 1996;27:528-31.

2. Corona R, Mele A, Amini M, et al. Interobserver variability on the histopathologic diagnosis of cutaneous melanoma and other pigmented skin lesions. J Clin Oncol 1996;14:1218-23. doi:10.1200/

3. Elmore, JG et al. Pathologists’ diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility Study.” BMJ. 2017; 357: j2813. Web.

4. Bitterman, A. Molecular testing to differentiate melanoma from benign nevi: the gold standard limitation. J Amer Acad Dermatol. 2016; 75: 849–850.

5. Reddy KK, Farber MJ, Bhawan J, Geronemus RG, Rogers GS. Atypical (dysplastic) nevi: Outcomes of surgical excision and association with melanoma. JAMA Dermatol. 2013;149:928-34.

6. Engeln, K et al. “Dysplastic Nevi with Severe Atypia: Long-Term Outcomes in Patients with and Without Re-Excision.” J Amer Acad Dermatol. 2017; 76: 244–249.

7. Duffy K, Grossman D. The dysplastic nevus: From historical perspective to management in the modern era: Part I. historical, histologic, and clinical aspects. J Am Acad Dermatol. 2012;67:1.e1-18.



---