Clark's Level Classification for Skin Tumors
Clark's level is an anatomic classification system that describes the depth of melanoma invasion through the layers of the skin, ranging from Level I (confined to epidermis) to Level V (invasion into subcutaneous fat), but it is now considered optional in pathology reporting and has been largely superseded by Breslow thickness measurement for prognostic purposes. 1
The Classification System
Clark's level divides melanoma invasion into five anatomic levels:
- Level I: Melanoma confined to the epidermis (melanoma in situ)
- Level II: Invasion into the papillary dermis
- Level III: Filling/expansion of the papillary dermis
- Level IV: Invasion into the reticular dermis
- Level V: Invasion into subcutaneous fat
Current Clinical Role and Limitations
Diminished Prognostic Importance
The most recent American Academy of Dermatology guidelines (2019) clearly state that Clark level is now optional for all melanoma stages 1. This represents a significant shift from earlier staging systems where it was a reportable feature, particularly for thin melanomas (≤1 mm) when mitotic rate could not be assessed.
When Clark Level May Still Be Useful
Despite its optional status, Clark level retains some clinical utility in specific scenarios:
- When mitotic rate information is unavailable - Several studies demonstrate relevance in management decisions under these circumstances 1
- Thin melanomas (0.76-1.50 mm) - Research shows that within this thickness range, deeper invasion correlates with significantly worse prognosis (p=0.0003), suggesting Clark level adds prognostic value when Breslow thickness alone is insufficient 2
- Supplementary prognostic information - For melanomas measuring 0.76-1.50 mm thick, additional classification by depth of invasion significantly improves prognostic accuracy 2
Practical Challenges with Clark Level
Reproducibility Issues
Multiple studies have documented significant inter-observer variability in applying Clark's classification:
- Initial agreement among pathologists was poor, with consensus reached in only 6/11 cases without discussion 3
- Even after discussion and slide review, agreement improved to only 11/13 cases 3
- Distinguishing Level II from III, or Level III from IV is particularly difficult in many cases 4
Technical Requirements
For accurate Clark level assessment, the pathology specimen must meet specific criteria:
- Tumor must be sectioned through its largest extension
- Bordering epidermis must be examinable
- Specimen must include subcutaneous fat
- Van Gieson Elastica stain should be available in addition to routine H&E staining to accurately determine invasion level 4
Breslow Thickness: The Superior Alternative
Breslow thickness (measured in millimeters) has superior prognostic value compared to Clark level 2. This continuous measurement is:
- More reproducible between pathologists
- Better correlated with survival outcomes
- The primary determinant for treatment decisions in current AJCC staging
The 2001 consensus guidelines required Clark level as a mandatory element in histopathology reports 5, but this has evolved to optional status in contemporary practice 1.
Reporting Recommendations
Current pathology reports may include Clark level as an optional feature that could assist in clinical management, particularly when:
- Mitotic rate cannot be assessed
- The melanoma falls in the intermediate thickness range (0.76-1.50 mm)
- Additional prognostic stratification is desired
However, Breslow thickness remains the essential measurement that must be reported for all invasive melanomas and forms the basis for treatment planning and staging.
Common Pitfalls to Avoid
- Do not rely solely on Clark level for treatment decisions - Breslow thickness should guide surgical margins and sentinel lymph node biopsy decisions
- Ensure adequate specimen preparation - Inadequate depth of excision or improper sectioning makes accurate Clark level assessment impossible
- Recognize the limitations - The discontinuous nature of Clark's classification (discrete levels vs. continuous measurement) inherently creates classification challenges at boundaries between levels 3