—  SPECIALTY CONFERENCE  —

Dermatopathology

Cases 7&8 - Microstaging in Cutaneous Melanoma
D.J. Ruiter, A.Spatz, J.J. van den Oord and M.G. Cook

Dirk J. Ruiter
University Medical Center Nijmegen
Nijmegen, The Netherlands


Click on each slide thumbnail image for an enlarged view
Clinical Histories:
Case 7 : A 44 year old man with a blue-grey skin tumor of the left shoulder, 1.1 cm.


Case 7 - Figure A - A low power micrograph shows a confluent deforming dermal nodule with attenuation of the overlying epidermis. The tumor extends deeply into the reticular dermis.
Case 7 - Figure B - Intermediate power shows an intra-epidermal neoplastic component. The vertical growth phase nodules are larger than the overlying junctional nests or the nests seen in the adjacent epidermis in the microscopic field.
Case 7 - Figure C - (high power): The vertical growth phase nodule shows fully transformed malignant features in melanocytes showing a confluent growth pattern with frequent and atypical mitotic figures.

Case 8 : A 32 year old woman with an irregular black, partly grey-white slightly elevated mole of the right upper arm, 1.8 cm.


Case 8 - Figure A - A low power micrograph shows a sheet of melanocytes in the right half of the field with a confluent proliferation of intra-epidermal and intradermal neoplastic melanocytes in the left half of the field.
Case 8 - Figure B - An intermediate power micrograph shows regression with a confluent population of lymphocytes admixed with a band of melanophages in the absence of demonstrable neoplastic cells.
Case 8 - Figure C - A high power micrograph shows the intra-epidermal radial growth phase component manifesting nested upward migration in concert with a vertical growth phase dermal component in the dermis.

Optimal cancer diagnosis and treatment requires a specialized and multidisciplinary approach, involving a surgeon, radiologist, oncologist and pathologist. The pathologist's role is not only pivotal in the diagnosis, but also in staging of cancer patients and the identification of early recurrence. In cutaneous melanoma, a relatively aggressive tumour that may be diagnostically excised by a general practitioner, a dermatologist or surgeon, the pathologist is required to provide a detailed diagnosis and microscopical stage (microstage), including the melanoma type, the thickness according to Breslow (tumor thickness), the level of invasion according to Clark, the presence of ulceration, the relative number of mitotic figures and the presence of regression. Furthermore, the presence of microsatellites is included. In the past decade other microstaging findings were put forward such as presence of vertical growth phase and presence of tumour infiltrating lymphocytes. Inclusion of such detailed findings in the pathologist's report was motivated by their major prognostic implication. Recently a few important papers have addressed critical issues regarding microstaging of cutaneous melanoma that deserve discussion here. Furthermore, some recommendations based on extensive discussions within the EORTC Melanoma Group are given. These issues concern a) the new staging proposal by the Melanoma Staging Committee of the American Joint Committee on Cancer (AJCC), b) the presentation of new prognostic models that seem applicable in daily practice, and c) new immunohistochemical findings that demonstrate prognostic information independent of the conventional major factors.

Based on clinical experience with over 40,000 patients treated in major cancer centers worldwide Balch et al(1)  on behalf of the AJCC have proposed major revisions of the melanoma TNM and stage grouping criteria. The new staging system, shown in Table 1, intends to better reflect independent prognostic factors that are used in clinical trials and in reporting the outcomes of various melanoma treatment modalities. As the new AJCC staging system is largely based on histopathological features, practicing pathologists should be well aware of the revisions proposed and their clinical consequences. Major changes include: 1) melanoma thickness and ulceration, but not level of invasion, apart for tumour thickness of less than 1 mm to be used in the T classification; 2) the number of lymph nodes with metastases, rather than their gross dimensions, the delineation of microscopical versus macroscopical lymph node metastasis, and presence of ulceration of the primary melanoma to be used in the N classification; 3) the site of distant metastasis and the presence of elevated serum LDH , to be used in the M classification; 4) an upstaging of all patients with Stage I,II, and III disease when a primary tumour is ulcerated; 5) a merging of satellite metastases around the primary melanoma and in-transit metastases into a single staging entity that is grouped into Stage III disease; and 6) a new convention for defining clinical and pathological staging so as to take into account the new staging information gained from intraoperative lymphatic mapping and sentinel node biopsy. In our opinion these changes are improvements as they adequately tackle disadvantages of the previous TNM system. For instance, the deletion of the level of tumour invasion makes sense in view of its low reproducibility of assessment by pathologists and its moderate prognostic power as compared to tumour thickness and presence of ulceration. The rationale of the new proposal is sound and logical because it looks practical, reproducible and applicable. Even more important, it quite accurately reflects the clinico-biological behavior of the disease. Furthermore, it is –with exception of some factors- evidence-based and takes the dominant prognostic factors into account that were consistently identified in multivariate regression analyses. More details of these general comments can be found elsewhere(2).  Here we shall focus on specific comments that are worthwhile to discuss from a pathologist's perspective.

A specific shortcoming of the new proposal is the unsatisfactory definition of ulceration, one of the most prominent new features. According to the AJCC proposal ulceration is defined as "the absence of an intact epidermis overlying a portion of the primary melanoma based on pathologic microscopic observation of the histologic sections". This definition lacks precision, as it would also include superficial interruption of the epidermis (erosion), that may be caused by minimal trauma such as scratching or previous shaving, or by technical artifacts. In contrast to erosion, in tumoural ulceration fibrin and neutrophils cover the underlying melanoma cells. Moreover, the epidermal edges of the ulcer often show pagetoid growth of melanoma cells as single cells or nests, suggesting tumoural destruction. Obviously, a distinction between ulceration and traumatic/artificial erosion is crucial as overcalling the latter as the former would lead to overstaging of the patient. In our opinion ulceration should be defined as "the interruption of the whole thickness of the epidermis above a portion of the primary melanoma, based on histopathological examination". Another point that needs to be addressed here is the width of ulceration, that may tremendously vary among cases and has been shown earlier to be a more accurate prognostic parameter than ulceration per se(3,4).  This is not mentioned in the new proposal at all. Therefore, we would like to recommend the use of our refined definition and the inclusion of ulceration width more than 0.1 mm. Only the use of a consistent definition can provide an evidence-based feature in a staging proposal. This could be checked in the databases used. Finally, one may wonder why ulceration in a primary melanoma (and probably also in other tumour types) harbors an unfavorable prognosis and is such a dominant and independent prognostic parameter. Ulceration may be a reflection of rapid tumour growth leading to destruction of the overlying epidermis and invasion of intradermal lymph and/or blood vessels that is an essential event in the pathomechanism of metastasis. A few studies indeed have shown a correlation between ulceration and number of mitotic figures in primary melanoma(5).  Another possible explanation for a more aggressive tumour growth and increased metastatic potential may be stimulation of ulcer surrounding melanoma cells related to the inflammatory and reparative host response inflicted by the tissue damage in the ulcer area. As ulceration results in a defect of the tumour surface it leads to a decrease of tumour thickness, which would give an incorrect estimation of the prognosis of the patient. However, when corrected for thickness ulceration still remains an independent prognostic factor. Further research on the molecular and cellular mechanisms is needed to shed more light on these intriguing events that cause or accompany ulceration and alleviate the capacity of the melanoma cells to metastasize.

Somewhat surprisingly, in the new AJCC proposal in the T1 group (i.e. tumour thickness less than 1.0 mm) levels IV and V are incorporated and indicated as T1b. However, based on its limited consistency it does not seem appropriate to still include levels of invasion. As an alternative, the presence of vertical growth melanoma (VGP) as put forward by Clark and Elder(6)  could be used as it indicates a potential to metastasize. It would include a minority of level II and all levels III, IV, and V cases. In addition, the presence of VGP can be reliably assessed by general histopathologists as demonstrated by Cook et al(7).  By the way, T1 in our opinion should be depicted for tumour thickness up to 0.99 mm as most clinical protocols indicate a sentinel node biopsy on patients with a thickness more than 1.0 mm.

In the new AJCC proposal a microscopical satellite (microsatellite), another important prognostic factor, is defined as a nest of tumour cells measuring 0.05 mm or greater that is present in the section in which the maximum thickness has been made and is distinctly separate from the main tumour mass. A minimal size of 0.05 mm means a group of about 100 tumour cells that have detached from the primary tumour and could therefore be interpreted as micrometastasis. Based on this clinico-biological behavior, patients with microsatellites, in the new proposal, were moved to the N classification. The same was done for in-transit metastases between the primary melanoma and the regional lymph nodes, as they represent "intralymphatic metastases" and carry a very poor prognosis. The concept of intralymphatic or intravascular metastasis implies the presence of tumour tissue, i.e. both neoplastic and non-neoplastic cells within the intravascular compartment. This would mean the presence of an appropriate tumour microenvironment within the vascular compartment, resulting in a markedly increased capacity to metastasize as compared to individual intravascular tumour cells or microaggregates thereof(8).  Further research on the challenging concept of intravascular metastasis is needed to establish its biological and clinicopathological context and implications. This also accounts for the presence of tumour cells in a sentinel node. The definition of micrometastasis in a sentinel node given by the AJCC proposal can be considered only provisional, as the extent of microscopic involvement may vary tremendously between a sporadic melanoma cell, a small focus or a substantial growth within the contour of the node. It seems plausible that the degree of involvement represents a spectrum that is reflected by a different clinical course. Therefore, we recommend systematic registration of the extent of tumour involvement in a sentinel node in order to make a future subdivision of the "micrometastasis group" possible. As a final practical comment on microsatellites, in our opinion it follows from the AJCC proposal that they no longer should be included into the tumour thickness measurement, as they were shifted to the N classification. This point however needs clarification. Another practical point of concern is the number of lymph nodes in a "radical" dissection specimen reported, because a limited number of nodes reported may be due to inadequate surgery, but also may be due to a limited evaluation of the surgical specimen by the pathologist. Therefore, it would be worthwhile to give recommendations on the number of nodes to be reported for each dissection basin (neck, axilla, groin and/or pelvis) to be reported, based on the database from 40,000 patients reviewed by the AJCC.

Next to TNM staging other approaches have been chosen, in which a varying number of clinical and histopathological parameters were combined in a prognostic model. The Pigmented Lesion Group from Pennsylvania University has developed two multivariate logistic regression models to predict survival in patients who have primary cutaneous melanoma. The first model(9,10)  assigns the 488 patients studied to two groups based on radial growth phase (RGP) or vertical growth phase (VGP). The probability of survival for those patients with VGP melanomas was further determined based on a model using six parameters (mitotic rate, tumor infiltrating lymphocytes, tumor thickness, anatomic site of the primary tumor, gender, and histological regression) that have the greatest strength as independent predictors of survival. According to the authors, this model is 89% accurate for predicting survival in patients with VGP melanomas. In addition, a second model(9,10)  has been developed that uses readily available clinical parameters to predict survival. Four of these parameters (tumour thickness, anatomic site, age and gender) proved to be powerful independent predictors. Clinical algorithms for assessing risk of individual patients were provided. Such an approach was also chosen by Cochran et al(3)  based on a histology database of 1,042 sequential melanoma patients. A step-up multivariate analysis showed five variables to be linked to survival, i.e. gender, anatomical site of primary tumour, age relative to 60 years, tumour thickness, and presence and width of ulceration. Probability of survival was calculated using a two-step approach. The survival-linked variables are multiplied to give an individualized risk score. This is converted into probability of survival by the formula .987 (risk score) for 3-year survival, .975 (risk score) for 5-year survival, and .960 (risk score) for 10-year survival. Similar techniques were used to develop individualized risk scores for likelihood of recurrence. This relatively simple approach to prognostication uses readily available demographic information and is more likely to be more accurate than single-factor analysis. It may be useful to select patients for adjuvant treatment.

Do immunohistochemical markers, in particular progression markers, have a place in the assessment of prognosis in primary cutaneous melanoma? As stated before, most progression markers do not appear to have a prognostic value, and only few are valuable in that respect(11).  However, the question is whether these markers have an independent prognostic power in a multivariate statistical analysis including established histopathological parameters. Ostmeier(5)  et al in a very large series of frozen sections derived from 691 melanoma patients found that the following 15 parameters were related to disease-free survival in univariate regression analysis: tumour thickness, ulceration, anatomical site of the primary tumour, gender, age, number of mitoses, and the immunohistochemical markers VLA-2, HLA-A,B,C, HLA-DR, gp100, Mel 14, ICAM-1, K-1-2, G-7-E2, and H-2-4-7. In multivariate analysis, only tumour thickness, ulceration and anatomical site exhibited independent significance. If the number of mitoses was replaced by ulceration, then the model performed slightly better, although ulceration was not significant in the presence of mitoses. However, the model could not be improved by any of the immunohistochemical markers(5).  In contrast a few recent immunohistochemical studies using paraplast embedded primary melanoma lesions in fairly large numbers of cases have shown independent prognostic value of certain markers. As indicated in Table 2, these include molecules involved in cell proliferation, matrix degradation, adhesion, transcription and cell differentiation(12-20).  Decreased survival was found for elevated expression of Ki 67, PCNA, cyclin A, MMP-2 and osteonectin, and decreased expression of p16 and gp100. Increased survival was found for elevated expression of tPA, CD44 and hyaluronic acid, and microphtalmia transcription factor. As exemplified by the paper of Florenes et al(14)  in this issue the prognostic significance accounts for a subgroup of melanoma patients only, i.e. those with superficial spreading melanoma. Comparable studies have referred to other subgroups such as thin (i.e. 1.0 mm thickness or less) melanomas(19),  and melanomas of the extremities(18).  This suggests that immunohistochemical markers still may yield independent prognostic contribution in subgroups of melanoma patients. Before this could lead to clinical implications the studies would have to be confirmed by other independent investigators. In addition, already existing immunohistochemical data could be pooled and statistically evaluated in order to provide confirmative data. Further, it would be highly interesting to compare immunohistochemical expression with that of gene expression obtained by cDNA microarray technology , as this with subsequent mathematical analysis of gene expression may make it possible to discern a subset of aggressive melanoma cases(21). 

References

  1. Balch CM, Buzaid AC, Atkins MB et al. A new American Joint Committee on Cancer staging system for cutaneous melanoma. Cancer 2000; 88: 1484-91.
  2. Ruiter DJ, Testori A, Eggermont AMM et al. The AJCC staging proposal for cutaneous melanoma: Comments by the EORTC Melanoma Group. Editorial. Ann Oncol 2001; 12: 1-3.
  3. Cochran AJ, Elashoff D, Morton DL et al. Individualized prognosis for melanoma patients. Hum Pathol 2000; 31: 327-31.
  4. Day CL Jr, Lew RA, Harrist TJ. Malignant melanoma prognostic factors 4: ulceration width. J Dermatol Surg Oncol 1984; 10: 23-4.
  5. Ostmeier H, Fuchs B, Otte F et al. Can immunohistochemical markers and mitotic rate improve prognostic precision in patients with primary melanoma? Cancer 1999; 85: 2391-9.
  6. Elder DE, Van Belle P, Elenitsas R et al. Neoplastic progression and prognosis in melanoma. Semin Cutan Med Surg 1996; 15: 336-48.
  7. Cook MG, Clarke TJ, Humphreys S et al. The evaluation of diagnostic and prognostic criteria and the terminology of thin cutaneous malignant melanoma by the CRC Melanoma Pathology Panel. Histopathology 1996; 28: 497-512.
  8. Ruiter DJ, Van Krieken JHJM, Van Muijen GNP et al. Tumour metastasis: is tissue an issue? Lancet Oncol 2001; 2: 109-12.
  9. Schuchter L, Schultz DJ, Synnestvedt M et al. A prognostic model for predicting 10-year survival in patients with primary melanoma. Ann Intern Med 1996; 125: 369-75.
  10. Halpern AC, Schuchter LM. Prognostic models in melanoma. Semin Oncol 1997; 24:S2-7.
  11. Ruiter DJ, Van Muijen GN. Markers of melanocytic tumour progression. Editorial. J. Pathol. 1998; 186: 340-2.
  12. Straume O, Sviland L, Akselen LA. Loss of nuclear p16 protein expression correlates with increased tumor cell proliferation (Ki-67) and poor prognosis in patients with vertical growth phase melanoma. Clin Cancer Res 2000; 6: 1845-53.
  13. Niezabitowski A, Czajecki K, Rys J et al. Prognostic evaluation of cutaneous malignant melanoma: a clinicopathologic and immunohistochemical study. J Surg Oncol 1999; 70: 150-60.
  14. Florenes VA, Maelandsmo GM, Faye R et al. Cyclin A expression in superficial spreading melanoma correlates with clinical outcome. J Pathol, 2001, this issue.
  15. Salti GI, Manougian T, Farolan M et al. Micropthalmia transcription factor: a new prognostic marker in intermediate-thickness cutaneous malignant melanoma. Cancer Res 2000; 60: 5012-6.
  16. Karjalainen JM, Tammi RH, Tammi MI et al. Reduced level of CD44 and hyaluronan associated with unfavorable prognosis in clincal stage I cutaneous melanoma. Am J Pathol 2000; 157: 957-65.
  17. Vaisanen A, Kallioinen M, Taskinen PJ et al. Prognostic value of MMP-2 immunoreactive protein (72kD type IV collagenase) in primary skin melanoma. J. Pathol 1998; 186: 51-8.
  18. Ferrier CM, Suciu S, Van Geloof WL et al. High tPA expression in primary melanoma of the limb correlates with good prognosis. Br J Cancer 2000; 83: 1351-9.
  19. Massi D, Franchi A, Borgognoni L et al. Osteonectin expression correlates with clinical outcome in thin cutaneous malignant melanomas. Hum Pathol 1999; 30: 339-44.
  20. Straume O, Akslen LA. Alterations and prognostic significance of p16 and p53 protein expression in subgroups of cutaneous melanoma. Int J Cancer 1997; 74: 535-9.
  21. Bittner M, Meltzer P, Chen Y et al. Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature 2000; 406: 536-40.
Table 1. Proposed AJCC melanoma TNM classification
T Classification Thickness Ulceration status
T1 ≤ 1.0 mm a: w/o ulceration
b: with ulceration or level IV or V
T2 1.01-2.00 mm a: w/o ulceration
b: with ulceration
T3 2.01-4.0 mm a: w/o ulceration
b: with ulceration
T4 >4.0 mm a: w/o ulceration
b: with ulceration
N Classification Number of Metastatic Nodes Nodal Metastatic Mass
N1 1 node a: micrometastasisa
b: macrometastasisb
N2 2-3 nodes a: micrometastasisa
b: macrometastasisb
c: in transit met(s)/satellite(s) without metastatic nodes
N3 4 or more metastatic nodes, or matted nodes, or in transit met(s)/satellite(s) and metastatic node(s)  
M Classification Site Serum LDH
M1a Distant skin, SQ or nodal mets Normal
M1bLung metastases Normal Normal
M1c All other visceral metastases or any distant metastasis Elevated
aMicrometastasis are diagnosed after elective or sentinel lymphadenectomy.
bMacrometastases are defined as clinically detectable nodal metastases confirmed by therapeutic lymphadenectomy or when nodal metastasis exhibits gross extracapsular extension.
Reproduced with permission of the AJCC.

Table 2. Prognostic markers in primary cutaneous melanoma independent from tumour thickness
Marker Number of patients Subgroup Survival References
Proliferation
Ki 67 (MIB-1) 202 VGP MOS (12)
PCNA 93 no DFS (13)
cyclin A 110 SSM DFS (14)
Transcription
Micropthalmia transcription factor 63 1.0-4.0 mm DFS MOS (15)
Adhesion
CD44 251 no DFS (16)
Matrix degradation
MMP-2 50 no MOS (17)
tPA 220 1.5-4.0 mm extremitiesDFS MOS (18)
Differentiation
gp100 (HMB45) 93 no DFS (13)
Other
Osteonectin 188 ≤ 0.75 mm DFS MOS (19)
p16 202, 102 NM DFS (12, 20)
p53 202 VGP MOS (12)
VGP = vertical growth phase
DFS = disease free survival
MOS = mean overall survival
SSM = superficial spreading melanoma
NM = nodular melanoma