Jump to content

Dr. Mark Hurt's Blog

  • entries
    37
  • comments
    96
  • views
    4,934

Is metastatic melanoma really metastatic melanoma?


Mark A. Hurt MD

1,710 views

 Share

I recently encountered a biopsy from a man in his forties. The lesion was composed of melanocytes in the epidermis and dermis. It's was relatively symmetrical, and it had a minor degree of maturation of melanocytes on progressive descent. There was greater than 1 mitotic figure per square millimeter, and, with Sox-10, there was a pagetoid pattern of melanocytes in the epidermis. Additionally, with Melan-A, the staining of melanocytes was not uniform. The depth was 0.95 mm.

Based on these parameters, my diagnosis was melanoma, T1b (AJCC criteria).

The patient had a sentinel lymph node biopsy, which contained a focus of melanocytes in the nodal parenchyma (so I am told; I have not had the opportunity to review the node). The melanocytes in the node, so I am told, was similar to those in the skin biopsy. What has followed is the request to send the biopsy for genetic analysis, which I was glad to do. I do not yet know the result.

My take on this case is that the patient has metastatic melanoma regardless of the outcome of the genetic analysis, because there is a lymph node metastasis, and the sensitivity of the aCGH is only about 80%. This means that 20% of melanomas are not detected by aCGH.

Am I wrong in my reasoning? If so, why? Is there really such a thing as a benign metastasis? If so, why?

[b]Follow-Up (8/5/13): [/b]I learned later on that the lesion in the skin had genetic alterations commensurate with melanoma.
 Share

10 Comments


Recommended Comments

Dr. Phillip McKee

Posted

I couldn't agree with you more. The dermatopathology population in general have lost their minds over the issue of positive sentinel nodes. This obviously started with Spitzoid melanoma where so much rubbish has been published. Because the majority of Spitzoid melanomas seem to do well, many pathologists insist on calling them atypical Spitz nevus with positive sentinel node biopsy. Indeed there is one paper discussing this issue and includes 10 cases of atypical Spitz nevus all with positive sentinel nodes and including a patient with distant metastases! Perhaps I am old fashioned but tumor in a lymph node is a metastasis pure and simple. It does not however necessarily predict the outcome of the patient.
Link to comment
Mark A. Hurt MD

Posted

Amen to that. I have even see articles in which the patient died, and they were still considered as having benign disease. It really makes no sense to me.
Link to comment
I agree that melanoma metastasis is a likely diagnosis, but there is room for doubt. Sentinel nodes of melanoma patients may contain melanoma metastases or nevus cell aggregates. The latter appear to be more common in melanoma patients than in nonmelanoma patients (Ridolfi et al, 1977).
I, too, am unsure about CGH-based diagnosis in this context. I believe scrutiny of the morphology, (architecture; cytologic features nodal melanocytes) offers the best chance for a correct diagnosis.
To my mind, spread to a regional lymph node is not sufficient evidence of malignancy of a melanocytic neoplasm, since there is much evidence indicating that benign melanocytic nevi occasionally spread to lymph nodes as well. Nevus cell nests frequently bulge into cutaneous lymphatics. Nevus cells groups are encountered in lymph nodes draining the skin (Ridolfi et al. 1977; Carson et al 1996) and within the lymph node basin involve a sentinel node more often than the more ‘distal’ nodes of the same basin (Carson et al., 1996; Holt et al., 2004). The walls of afferent lymph vessels near lymph nodes containing nevus cells, occasionally contain nevus cells as well (Subramony & Lewis, 1985). Nevus cell groups are not encountered in lymph nodes that do not drain the skin (McCarthy et al., 1974). Taken together, these findings constitute strong – though, I must admit, perhaps not absolutely conclusive – evidence that benign nevi can spread to regional lymph nodes. The alternative hypothesis that precursors of these melanocytes arrive at lymph nodes by some presumed process of ‘aberrant migration’ has always stuck me as far-fetched (why would such aberrant migration of melanoblasts result in their arrival in sentinel nodes draining the skin only, rather than other nearby nodes? Why indeed in lymph nodes?) and to my knowledge there is no evidence to support that obsolete idea. In line with this (though the argument is more complex here), a ‘positive’ sentinel node in case of a melanocytic tumor of uncertain malignant potential does not carry the same prognostic information as a positive sentinel node in unequivocal melanoma (Scolyer et al., 2010).
Spread of benign tumour cells or even nonneoplastic cells to regional lymph nodes is not unique to melanocytes. Ity has been identified in endometrium, fallopian tube, decidua, thyroid, breast, salivary gland, mesothelium.
My concern is quite the reverse: some workers conclude, on the basis of what I would regard as insufficient evidence, that a melanocytic neoplasm is a melanoma. No doubt there is a rise in incidence of thin melanomas, but a more precise way of stating this would be that there is a rise in [i]diagnoses [/i]of thin melanoma. There is a proliferation of proposed new melanoma variants, even when distant metastasis and tumor-related death has not been identified in the seminal articles. I believe this is a problem.
Of course if a patient dies of distant metastasis, there is no room for discussion. But I seem to disagree with Ackerman and you both, that spread to a regional lymph nodes equals cancer metastasis and can be regarded as formal and unequivocal proof of malignancy. I truly believe the issue is more complex than that.
Link to comment
Mark A. Hurt MD

Posted

Dr. Mooi,

Thank you for weighing in. I don't really have an issue with melanocytic nevus cell nests in nodes, especially in the capsules, where they are usually encountered. I suspect that these are not part of any melanoma, and who knows where they come from?

I am curious to know whether one can identify whether a metastasis of a melanoma has the same genetic signature as the primary. Is that possible? I think it might help resolve some of the issues you have raised.

Still, it seems counterintuitive to me that one would consider a metastasis of a melanoma as benign disease. I can understand that some patients with melanoma might survive much better than others, just as some people have melanomas that look identical morphologically but some die quickly and others live for a long, long time, perhaps never dying of melanoma but some other cause. There is a famous story in St. Louis about a patient who lived for 40 years with his melanoma, metastases and all, confirmed by a number a reputable pathologists in the city. To my mind, this kind of phenomenon shows that the diagnosis of melanoma is established by morphological criteria; the prognosis is a statistical conclusion that was never meant to apply to a single case, despite the fact that it is often used that way.

For me, there is no melanocytic tumor of uncertain malignant potential. Rather, I frame it this way: there are melanocytic proliferations of uncertain diagnosis and, as a consequence, the prognosis is also uncertain [i]because[/i] the diagnosis is uncertain.

Comments?
Link to comment
Dear Dr Hurt,
many thanks for your thoughtful comments. I believe we agree on most issues.
I'm not sure that comparison of genetic signatures is as helpful in melanoma as it is in, say, head and neck cancer or lung cancer. Time (further work) will tell.
I agree that, when a melanoma spreads to another site, distant or regional, that is metastasis; I don't think there is a logical argument to call it anyting else. The point I tried to make is that [i]the mere presence of a melanocytic lesion in a node is insufficient to call the lesion a melanoma[/i]. If the lesion in a node (usually but not always in the capsule) looks like a naevus, we all know that it is of no consequence. This means that with respect to melanocytic lesions, presence in a lymph node is insufficient basis for a diagnosis of (metastatic) melanoma. The diagnosis of melanoma needs to be supported by histological evidence of malignancy of the lesion.
I do not regard 'melanocytic tumour of uncertain malignant potential' (MELTUMP) as a [i]diagnosis[/i]. Rather, I see it as a summary of a [i]situation[/i], the situation being that the pathologist is unable to say whether the lesion is a naevus or a melanoma. For instance, a verdict of MELTUMP can result from biopsy crush artifact or from incisional (rather than excisional) biopsy. To my mind, MELTUMP is not a diagnosis, not an entity. Perhaps I should have said that more clearly. Of course I - like anyone else - try to minimize that group of MELTUMPS as much as I can. A verdict of MELTUMP means that I have failed to establish a firm diagnosis of naevus or melanoma.
On a final note, let me say how much I enjoy your excellent website, that is full of life and contains so many treasures! I am very pleased that you have agreed to take over this work from our mutual good friend Phillip McKee. Keep up the good work! Kind regards, Wolter Mooi
Link to comment
Mark A. Hurt MD

Posted

Dr. Mooi,

It appears that we might be very close to each other in how we see the situation.

I agree with you that any concern about nests of melanocytes in nodes is [i]not [/i]that all are malignant; melanocytes are necessary for malignancy, but not sufficient for malignancy simply by being in the node capsule or parenchyma (although involvement of the parenchyma poses a special problem, in my opinion).

My experience is that the nodal melanocytes tend to form patterns similar to those in the dermis. I suspect that a number of examples of melanocytes found in nodes are true-true but, perhaps, unrelated to to the skin lesions in the vicinity. That issue is a cause for investigation, and I simply don't have the answer to any putative association.

In my practice, one of the biggest problems I encounter is the miscommunication about uncertainties in the diagnosis of melanocytic lesions. This is why I am so adamant about avoiding terms such as "atypia," "dysplasia," and "malignant potential." As a rule, anytime these words are raised, the lesion is considered malignant, by the clinician, when it might not be so. Unfortunately, because of the stakes involved, the lesions for which one is uncertain will inevitably be excised, and should be. This does not mean, however, that they are malignant. My observation is that there seems to be an epistemological "inversion" about such lesions. By this I mean that there is a logical fallacy of assigning "uncertainty" to the [u][i]lesion [/i][/u]in question, when uncertainty [u][i]really [/i][/u]belongs to the observer.

In my way of thinking, every lesion I am uncertain about always raises the question of not being able to exclude melanoma, [i]despite [/i]being uncertain. This always results in excision. What I find more interesting, epistemologically, is when dermatopathologists use the term "dysplasia" or "atypia" and "nevus" together in a diagnostic line. I see it as a contradiction. Do you?

Mark
Link to comment
What an interesting phrase, epistomological inversion! I agree with the basis of your thoughts, but do not have quite the same objection to MELTUMP. If we say that Handel's flute sonata HWV378 is of uncertain authenticity, we mean of course that [i]we [/i]are unsure wether it is authentic Handel or not. The 'uncertain authenticity' does not refer to the work itself (it is either authentic or it is not) but it refers to our own uncertainty. I believe we do this all the time, in daily life and in science. If we say that the intentions of mr Bloggs are unclear, the intentions themselves are probably perfectly clear to mr Bloggs, but he hasn't conveyed them to us. Again, we refer to our own uncertainty.
I do not use the term 'atypia' in the classification of a naevus, since a diagnosis of naevus implies that the lesion is benign. If the atypia makes us wonder whether the lesion could not be a naevoid melanoma after all, I believe we should not call it an 'atypical naevus'. Nor should we call it a melanoma. So, I use MELTUMP, STUMP, SAMPUS, in such situations. of course the uncertainty is mine, it is not an intrinsic element of the biology of the lesion. I always feel bad having to resort to those categories, and always try to establish an unequivocal diagnosis.
Regarding dysplastic naevus: I do use that term, since to my mind that is a distinctive subtype of naevus (ergo, benign!), and do not see why a benign lesion should not be called dysplastic - there are quite a number of examples of dysplastic benign lesions outside the skin. I believe its recognition aids in the distinction between naevus and melanoma; that's perhaps the main reason for retaining the entity.
I'm not sure you agree with all of this, but iI have certainly enjoyed thinking and talking about these issues.
Link to comment
Mark A. Hurt MD

Posted

Dr. Mooi,

Actually the full phrase is the "inversion of epistemology with metaphysics." This means, for those who believe tihis, that they have inverted the concept of the thing with the thing itself. Instead of a concept of "malignant-benign-I don't know" for the observer, the inversion is that the tumor, as such, doesn't know what it is or has to bear the intellectual responsibility for what it is, which is, of course, impossible. Instead of thinking, the humans are passive recorders, somehow looking at the lesions for some sort of divine revelation about the nature of the lesion -- as opposed to doing the hard thinking required to establish a diagnosis.

As for MELTUMP, STUMP, and SAMPUS, unfortunately, those acronyms make formal that uncertainty is treated as though it were certainty. In a sense this this true, as one is certain he doesn't know the diagnosis, and he uses one of these acronyms as a substitute for his uncertainty. But it does not solve the problem of uncertainty.

Yet, I think it is problematic that one do this. One should make the dermatologist as uncomfortable as possible in the uncertainty. For example, in St. Louis, my approach is to use the phrase "melanocytic proliferation." This is followed by a differential diagnosis and an explanation about why I am not able to be more definitive. My goal is to result in a excision of the lesion.

As for the term "dysplastic", I challenge you to revisit Clark's literature. I think if you do, you will find it has no basis for describing any kind of lesion under that terminology. I am am happy to continue this discussion, as I think it is extremely important. Here is a link to my position statement on the issue.

http://markahurt.com/files/documents/20050457hurt_ClarkNevusLMN_DysNev.pdf

Very best from

Mark
Link to comment
Dear Dr Hurt, dear Mark,
thanks for your interesting answer and questions. I do not believe we can settle this to mutual satisfaction in this way. My own views are laid down in the monograph I wrote with Thomas Krausz (2nd edition 2007) and a number of papers on the subject. If we feel uncertain about the DD naevus versus melanoma, it is mandatory that we communicate this in an unequivocal way to the clinician, I'm sure we agree on that point, and that is the most imorpatnt one! Kind regards, Wolter
Link to comment
×
×
  • Create New...