The methods of histologic staging of primary Stage I melanoma and the relation to lymph node metastases and survival after surgery was evaluated in 151 patients with extremity melanoma only. Microstaging by depth of invasion showed a better prognostic correlation than by histologic typing (into superficial spreading, or nodular melanoma). A correlation existed between depth of invasion (Clark's levels) and incidence of nodal metastases at elective node dissection. This incidence was 5% at Level II, 4% at Level III, 25% at Level IV and 75% at Level V. The measured depth of invasion added prognostic insight to each Clark's level; the minimal invasion at which nodal metastases occurred was 0.6 mm for Level II, 0.9 mm for Level III, 1.5 mm for Level IV and over 4 mm for Level V. The 5 year disease-free survival after surgery was 100% for Clark Level II, 88% for Level III, 66% for Level VI and 15% for Level V. There was a direct relation between the measured depth of invasion and survival and mortality from disease at 5 years. Mortality from disease at 5 years could be directly equated with 10 times microinvasion in mm. Microstaging by direct measurement gave a better prognostic correlation than was found using Clark's levels for more deeply invading melanoma. At this time there is suggestive evidence that patients with certain higher risk lesions may do significantly better with wide excision and elective node dissection than with wide excision alone. These high risk lesions include Clark Level III to V, lesions measuring 0.9 mm or greater and all nodular melanomas.
Wanebo et al. (Mon,) studied this question.