MOLEM (Management of Lesion to Exclude Melanoma) schema seems superior to Melanocytic Pathology Assessment Tool and Hierarchy for Diagnosis (MATH-Dx) in clinical relevancy for the management classification of melanocytic and nonmelanocytic lesions excised to exclude melanoma. A more standardized way of establishing diagnostic criteria is important for instructing artificial intelligence (AI) algorithms.

A study examined pathologists’ variability, reliability, and confidence in reporting melanocytic and nonmelanocytic lesions excised to exclude melanoma using the MOLEM schema in higher-risk patients.

Higher-risk patients were included in the study who gave Baseline demographic characteristics and related clinical details (eg, history of melanoma). Patients suspicious for melanoma, as reported by the primary care physician, were further subjected to clinical evaluation, dermoscopy imaging, and subsequent excision biopsy of the suspected lesion(s). 

217 lesions were removed and conventional histologic slides were prepared and reviewed by up to 9 independent pathologists (who were blinded for original histopathological diagnosis) for diagnosis using the MOLEM reporting schema.

The observations were then assessed further.

The following results were obtained-

• 197 patients were enrolled for the study of which 51.8% were male; while 48.2% were female.
• Mean (SD) age of the subjects was 64.2 (15.8) years, ranging between 24-93 years. 
• A total of 217 index lesions were assessed with a total of 1516 histological diagnoses. 
• Among 1516 diagnoses-

o 44.7% were classified as MOLEM class I

o 7.9% as MOLEM class II

o 37.2% as MOLEM class III

o 7.5% as MOLEM class IV

o 3.6% as MOLEM class V. 

• Concordance rates were- 

o 88.6% for class I

o 50.8% for class II

o 76.2% for class III

o 77.2% for class IV

o 74.2% for class V

• The quadratic weighted interrater agreement was found to be 91.3%, with a Gwet AC1 coefficient of 0.76. 
• The quadratic weighted agreement between pathologists’ ratings and majority MOLEM was found to be 94.7%, with a Gwet AC1 coefficient of 0.86. 
• The confidence in diagnosis data demonstrated-

o A relatively high level of confidence when diagnosing classes I and V  

o A lower level of pathologist confidence when diagnosing classes II and III. 

• The quadratic weighted interrater confidence rating agreement was found to be 95.2%, with a Gwet AC1 coefficient of 0.92. 
• The confidence agreement were- 

o 95.0% for class I

o 93.5% for class II

o 95.3% for class III

o 96.5% for class IV

o 97.5% for class V

Thus it was concluded that the proposed MOLEM schema is superior to the MPATH-Dx schema in lesions excised to exclude melanoma by combining diagnoses with similar prognostic outcomes for melanocytic and nonmelanocytic lesions into standardized classification categories. 

Pathologists’ level of confidence seemed to obey the MOLEM schema diagnostic concordance trend, ie, atypical naevi and melanoma in situ diagnoses were most conflicting and difficult for pathologists to confidently diagnose.

Source- JAMA Netw Open. 2021;4(12):e2134614. doi:10.1001/jamanetworkopen.2021.34614