Final Report Summary - NEVUS CLASSIFICATION (Classification and longitudinal follow-up of common melanocytic nevi with in vivo reflectance confocal microscopy)
Individuals with many melanocytic nevi (“moles”) are at increased risk of melanoma. Until recently, nevi were regarded as a single entity, implying uniform risk to patients with numerous nevi; however recent data suggests that nevi consist of distinct subsets which may be associated with different levels of melanoma risk. A new classification reflecting the diversity of nevi was clearly needed. Until recently, nevus classification has been based on histopathological analysis, and precise cataloging of nevi has required surgical excision to allow microscopic examination of the nevi. In the IRG grant, we proposed to harness Reflectance Confocal Microscopy (RCM), a cutting-edge technology which enables “optical skin biopsy” – non-invasive (i.e. without cutting or injuring) skin imaging at cellular-level resolution, for formulating a new nevus classification scheme and for monitoring the nevi for change over time. The working hypothesis was that RCM can identify distinct subsets of nevi. We also conjectured that RCM would allow follow-up of nevus evolution at the microscopic cellular level.
As the first step, we imaged with RCM, prior to surgical removal, 180 benign nevi excised for cosmetic reasons or patient request. We showed that five distinct patterns can be readily identified with RCM among nevi and that these RCM patterns have consistent histopathological correlates. These patterns have formed the basis for our RCM nevus classification.
Further, we have completed accrual of 116 participants, each contributing randomly selected nevi on the back and legs, for RCM imaging of 348 nevi. We selected for the study only nevi that were undoubtedly benign based on their clinical appearance. We analyzed the distribution of RCM patterns of these nevi and examined whether two independent observers can agree on the RCM classification of these nevi. The most prevalent RCM pattern of nevi was “Ringed” pattern seen in 43% of nevi, followed by "Clods" pattern seen in 27% of nevi. We found fair to good interobserver agreement on classification of nevi to RCM patterns, and were able to reach consensus classification in all nevi imaged with RCM.
In addition, we reimaged with RCM 201 nevi from 77 of the participants at one-year follow-up. We are currently analyzing the stability of pattern for the 201 nevi monitored for one year, and have thus far found substantial stability of pattern. This suggests that nevi that have an undoubtedly benign clinical appearance are highly stable at the cellular level.
Remarkably, our IRG study has led to a methodological breakthrough in RCM imaging: we were the first to demonstrate the feasibility of using RCM to track individual cellular structures in nevi over time (Figure 1, see legend below). We have described in detail the methodology of longitudinally tracking cellular structures in nevi and have shown that different observers can readily recognize these structures over time. This discovery has opened up new research avenues for using RCM to monitor skin lesions, at cellular level, for change versus stability.
There are several implications for the RCM-based nevus classification and longitudinal tracking. First, as we have systematically investigated RCM patterns of benign nevi, our study will advance the efficacy of RCM for differentiating benign nevi from malignant melanoma. One needs to recognize thoroughly the appearance of benign nevi to be able to identify deviations that indicate a malignancy. The publications of our work have already and will continue to introduce a glossary of images illustrating the RCM patterns of nevi, which will allow other European research groups to utilize RCM nevus classification for their studies. Second, RCM-based classification of nevi will be useful for case-control studies comparing pattern between subjects at high versus low-melanoma risk; this comparison will facilitate the use of RCM for identifying patients at higher melanoma risk based on the microscopic pattern of their nevi. Third, based on our newly described tracking technique, we would be able to monitor suspect lesions for changes. Prior to our research, RCM has only been used to assess the morphology of skin lesions at one point in time similar to the way histopathology assessment (i.e. microscopic examination of surgically removed tissue) is being done. However, differently that histopathology, RCM is a non-invasive cellular imaging technique and hence can be used to assess the microscopic changes of skin lesions over time. With this approach, stability at the microscopic level would indicate a benign lesion, while rapid change may identify early skin cancers. The monitoring of skin lesions will provide a unique opportunity to observe non-invasively how cells in changing lesions move, grow or proliferate within the intact live tissue and generate new knowledge that may lead to more effective skin cancer treatment. Finally, in vivo RCM devices are currently being developed and tested for cellular imaging in the fields of oral, gastrointestinal, pulmonary and genito-urinary medicine; the principle of dynamic monitoring of neoplasms for cellular changes is currently being tested in the skin, but is likely to become relevant for cancer diagnosis well beyond the skin.
Figure 1 Legend: Precise follow-up of micro-anatomic structures of nevi with RCM.
We imaged a benign nevus with RCM at baseline (A) and at 1-year follow-up (B). We were able to identify with RCM precisely the same subsurface micro-anatomic tissue structures (annotated by corresponding circles, triangles and asterisks) at both time points. Remarkably, the same individual cells, seen as distinct bright dots (annotated by corresponding arrows), can also be identified at both times points.
As the first step, we imaged with RCM, prior to surgical removal, 180 benign nevi excised for cosmetic reasons or patient request. We showed that five distinct patterns can be readily identified with RCM among nevi and that these RCM patterns have consistent histopathological correlates. These patterns have formed the basis for our RCM nevus classification.
Further, we have completed accrual of 116 participants, each contributing randomly selected nevi on the back and legs, for RCM imaging of 348 nevi. We selected for the study only nevi that were undoubtedly benign based on their clinical appearance. We analyzed the distribution of RCM patterns of these nevi and examined whether two independent observers can agree on the RCM classification of these nevi. The most prevalent RCM pattern of nevi was “Ringed” pattern seen in 43% of nevi, followed by "Clods" pattern seen in 27% of nevi. We found fair to good interobserver agreement on classification of nevi to RCM patterns, and were able to reach consensus classification in all nevi imaged with RCM.
In addition, we reimaged with RCM 201 nevi from 77 of the participants at one-year follow-up. We are currently analyzing the stability of pattern for the 201 nevi monitored for one year, and have thus far found substantial stability of pattern. This suggests that nevi that have an undoubtedly benign clinical appearance are highly stable at the cellular level.
Remarkably, our IRG study has led to a methodological breakthrough in RCM imaging: we were the first to demonstrate the feasibility of using RCM to track individual cellular structures in nevi over time (Figure 1, see legend below). We have described in detail the methodology of longitudinally tracking cellular structures in nevi and have shown that different observers can readily recognize these structures over time. This discovery has opened up new research avenues for using RCM to monitor skin lesions, at cellular level, for change versus stability.
There are several implications for the RCM-based nevus classification and longitudinal tracking. First, as we have systematically investigated RCM patterns of benign nevi, our study will advance the efficacy of RCM for differentiating benign nevi from malignant melanoma. One needs to recognize thoroughly the appearance of benign nevi to be able to identify deviations that indicate a malignancy. The publications of our work have already and will continue to introduce a glossary of images illustrating the RCM patterns of nevi, which will allow other European research groups to utilize RCM nevus classification for their studies. Second, RCM-based classification of nevi will be useful for case-control studies comparing pattern between subjects at high versus low-melanoma risk; this comparison will facilitate the use of RCM for identifying patients at higher melanoma risk based on the microscopic pattern of their nevi. Third, based on our newly described tracking technique, we would be able to monitor suspect lesions for changes. Prior to our research, RCM has only been used to assess the morphology of skin lesions at one point in time similar to the way histopathology assessment (i.e. microscopic examination of surgically removed tissue) is being done. However, differently that histopathology, RCM is a non-invasive cellular imaging technique and hence can be used to assess the microscopic changes of skin lesions over time. With this approach, stability at the microscopic level would indicate a benign lesion, while rapid change may identify early skin cancers. The monitoring of skin lesions will provide a unique opportunity to observe non-invasively how cells in changing lesions move, grow or proliferate within the intact live tissue and generate new knowledge that may lead to more effective skin cancer treatment. Finally, in vivo RCM devices are currently being developed and tested for cellular imaging in the fields of oral, gastrointestinal, pulmonary and genito-urinary medicine; the principle of dynamic monitoring of neoplasms for cellular changes is currently being tested in the skin, but is likely to become relevant for cancer diagnosis well beyond the skin.
Figure 1 Legend: Precise follow-up of micro-anatomic structures of nevi with RCM.
We imaged a benign nevus with RCM at baseline (A) and at 1-year follow-up (B). We were able to identify with RCM precisely the same subsurface micro-anatomic tissue structures (annotated by corresponding circles, triangles and asterisks) at both time points. Remarkably, the same individual cells, seen as distinct bright dots (annotated by corresponding arrows), can also be identified at both times points.