Dermatoscopic pitfalls
There are many pitfalls in dermoscopy but the main objective is a) not to miss a melanoma and b) to avoid unneccessary excision of benign lesions. Throughout this course the emphasis has been on developing safe approaches to managing pigmented and non pigmented lesions. You have learned to look carefully at the ugly duckling lesion, look carefully at a lesion that has changed, either on the patient's word or because you have monitored it, and carefully examine any lesion a patient brings to your attention as a concern. If a patient has previously had a melanoma excised you generally adopt a more aggressive biopsy policy with new or changing pigmented lesions. By this stage in the course the pitfals should be fewer in number than at the beginning but experience teaches you that the variation in dermatoscopic morphology of both benign and malignant lesions is considerable and it takes a long time to be visually acquainted with all these variations.
In this module I will look at some of the cases that have appeared in the Skin Cancer Society Blog which have caused the experts problems and hopefully you will learn to avoid them when you meet them again in the future.
The Pink Nodule
We see pink nodules all the time when we examine patients. Most are dermal nevi, neurofibromas or dermatofibromas if they have been present for a long time. A new pink nodule should be carefully examined but one that arises in association with a pigmented lesion or which has flecks of pigment either within it or around the circumference should definately merit our careful examination with a dermatoscope. We are looking at vessels and any pigment we can see and trying to decide if the lesion is a dermal nevus, BCC, dermatofibroma or amelanotic melanoma.
The Pink Nodule
When it erupts suddenly it merits our attention. The main concern is to avoid missing a diagnosis of amelanotic melanoma. If the vascular pattern is polymorphic with both dot vessels and corkscrew vessels and linear irregular vessels then consider an amelanotic melanoma. The vessels in a dermal nevus will be mainly comma shaped but if there is a mixture of both thick and thin comma vessels then again consider an amelanotic melanoma. The pink basal cell skin cancer will usually have arborising vessels. The pink dermatofibroma may not have much in the way of vessels at all. Unusual skin tumours such as Merkel cell, atypical fibroxanthoma and rapidly growing squamous cell carcinoma may again have a variety of vessels. When lesions grow rapidly looped vessels will often occur and are known as hairpins. If they have a white halo around them made up of keratinocytes then it is likely to be a keratinocytic lesion rather than a rapidly growing nodular melanoma.
In practical terms any rapidly growing pink nodule gets biopsied because usually the vascular patterns are not specific enough to allow you to make a definitive diagnosis.
Consider the following cases.
Image 1 This is Harald's analysis "2. a lesion with 2 patterns (lines reticular and structureless), combined asymmetrically, more than one color (light brown and skin colored), needs just one clue to make the diagnosis of a melanoma: the clue here are linear, serpentine vessels that are coiled like a corksrew.
Pathologic diagnosis: melanoma arising in a congenital nevus."
Image 2 from Greg Canning is of a dermatofibroma on the dorsum of the foot showing both hairpin and dot vessels but this is very unusual site and morphology.
Image 3 from Cliff Rosendahl has many lessons. It is a merkel cell tumour that was badly mismanaged elsewhere. The lesson is never to physically obliterate an elevated pink lesion if you do not know the pathology but you really should read this case so that you avoid this pitfall if you ever come across it.
Image 1 This lesion shows obvious regression clinically with the pigmented areas separated by pale whiter areas. Harald Kittlers analysis -"Patterns: > 1 (clods, dots, structureless zone)
Color: >1
Clues: gray dots, white structureless area
Diagnosis: Melanoma"
Image 2 is a fantastic example of regression. There are no real pitfalls here although with all that regression and the peripheral pigment you might just think it was a pigmented BCC but only for a millisecond ! This is Cliff Rosendahl's description using Harald's algorithm. "Pattern - Structuress zones, clods,scrap of lines reticular(extreme NW), peripheral lines radial
Colours - A Kaleidoscope!
Clues
1. Peripheral focal lines radial (radial streaming)
2. Eccentric structureless blue, white,pink
3. Peripheral black clods
4. Patch of white lines reticular (negative network in the north)
5. Vessels dots+curved+one helical(corkscrew)dead centre last image
6.Pseudopods (NE)
7. Focal grey dots (due north last image)
and finally the "rising hair on the back of the neck" sign!"
Small melanomas
Small diameter melanomas are usually defined as those that are 5mm or less in diameter. You might think that most of these would be thin lesions because it is relatively early in the development of a melanoma. Studies have shown that a third may be in situ and two thirds invasive. With small melanomas it is difficult to use concepts of symmetry and the nature of the border in your clinical assessment and often it is only the dermatoscopic view that alerts you to melanoma features such as blue/grey veils and peripheral dots and globules or streaking. It has been noted that small melanomas tend to occur in younger individuals but there is no difference in the normally found sex ratio.
The small melanoma may simply be noted to be an ugly duckling in that it is often much darker than other similarly sized nevi. If a suspected small melanoma is raised it is prudent to just excise it using a small 4mm to 6mm punch rather than following the lesion. Any papular or raised pigmented lesion really should not be monitored. The problem of monitoring is that small nodular melanomas can grow at a rate of 0.5 of a millimetre a month.
A small
black nevus can be a diagnostic pitfall. The black pigment is so intense centrally that it obliterates the underlying reticular network. This is because of pigmentation in dead keratinocytes in the stratum corneum. These can be removed by tape stripping and the underlying benign network revealed. Some of these lesions lack the surrounding network as a peripheral halo and these are the lesions that can be mistaken for a melanoma and over aggressively excised. Remember the tape stripping!
Pitfalls presented by the Amelanotic melanoma
We have talked about the amelanotic melanoma above in the context of a pink nodule. However some amelanotic melanomas are macular lesions that present as a pink patch. Because they are thin lesions the predominant vessel pattern is a dot pattern. These dot vessels represent the vertical vessels coming from the deeper dermal plexus into the papillary dermis. They are coming up vertically, so looking down on them we see them as dots. They are mainly seen in thin lesions. Obviously dot vessels can be seen in other pink flat lesions, including psoriasis and dot vessels are common on the lower legs as part of venous stasis. Most amelanotic melanomas are better described as being hypo melanotic melanomas in that they usually have some remnant pigmented network, particularly at the peripheries. In these particular cases we are not dealing with regression or at least we are assuming the lesions do not have regressive features such as grey peppery pigmentation dots in the dermis.
Sometimes an amelanotic melanoma is incorrectly diagnosed clinically as a superficial BCC and it may be treated with cryotherapy or curette and cautery without histology being obtained. Unless you can diagnose the superficial BCC quite confidently under your dermatoscope it is best to biopsy these lesions and to avoid cryotherapy.
Amelanotic melanomas or hypomelanotic melanomas may become that way because they intrinsically lack melanin and melanocytes or because of regression where the melanocytes have been destroyed. If the latter has occurred you would expect to see a grey dot peppering and whitish areas that are whiter than the surrounding skin. All the criteria normally used to diagnose melanoma become irrelevant in a truly amelanotic melanoma but with the loss of pigment
the vascular structures become more easily visible and these patterns are what must be learned and recognised.
Combinations of amelanotic melanoma with another lesion, for example a basal cell skin cancer, are very unusual and as you can imagine extremely difficult to diagnose.
The big problem in examining amelanotic melanomas is in using a dermatoscope with a contact plate and exerting too great a pressure, thereby compressing the vascular structures. A gel should be used if you want to examine vessels and should be placed carefully over the lesion to try and avoid any bubbles and then the lesion examined without lifting the dermatoscope and introducing bubbles into the medium. Gel allows you to look at the vessels without putting pressure on them. Polarised light dermatoscopes may give an even better view.
Vessels in an amelanotic melanoma do not have keratin surrounding them. This should help you in differentiating an amelanotic melanoma from an SCC or other keratinising tumour.
The vascular patterns in amelanotic melanoma have been given a variety of names ranging from pin point or point like, through dotted vessels and hairpin like or linear irregular vessels. But point or looped vessels are just part of the normal vascular structure of the dermal plexus sending vessels into a growing tumour. The thicker the tumour the more the vessels are distorted and looped rather than pinpoint. Dot vessels are a feature of thin lesions. The thicker the lesion the more likely you are to get varying vessels because the loops are distorted more and you may see twisting or corkscrewing and even glomerular like structures.
In keratinocytic lesions you tend to find the vascular loops most prominent at the periphery of the lesion but in an amelanotic melanoma they are distributed fairly evenly across the surface, more so in early lesions than late ones. Remember again that the older and thicker an amelanotic melanoma is then the more variable will be the vascular morphology.
To summarise thin melanomas less than 0.5 of a millimetre will tend to have dot like vessels distributed over the surface quite regularly. Tumours 0.5mm to 2mm in thickness as well as having dot vessels will also have some hairpin vessels. Tumours greater than 2mm in thickness will have more distorted vessels. Once a tumour is more than 3mm thick it cannot get its vascular supply from below and it tends to develop it from the side of the lesion and so vessels grow over the surface and may come sharply into focus and resemble the arborising vessels of a basal cell skin cancer but even then the arborising is not as marked as you see in a BCC.
As an aside a logical analysis of vessels was published first in the late 90s by Kreusch in Clinical Dermatology 20; 248-254 titled Vascular Patterns and Skin Tumours. His sequence of questions was basically the following.
You ask yourself "Are there arborising vessels in which case it is BCC, Are there crown vessels in which case it is sebaceous hyperplasia, Are there comma vessels which are mainly found in a dermal nevus. Are there nonspecific vessels, which can indicate Kaposi sarcoma or regressive lesions? Are there pinpoint or hairpin vessels? If there is an absence of white halos the lesion may be melanocytic. If they are present it is probably a keratinocytic tumour. Are there any traces of melanin. If we are dealing with just a pink lesion traces of melanin make a melanocytic lesion more likely."
Picking up amelanotic melanomas is obviously difficult. You have to have almost a sixth sense as to which pink lesions need to be examined. Certainly any that the patient brings to your attention as being new or any that your eye picks up as being different from any background pink lesions, in any patient who has previously had a melanoma it becomes mandatory to examine any of the pink lesions whether the initial melanoma was amelanotic or not. Also in patients who have dysplastic nevi, any pink lesions need to be carefully examined. If we select patients who are at increased risk and learn to look at their pink non-melanocytic lesions as well as the darker melanocytic ones, then it is likely we are going to see a higher proportion of amelanotic melanomas. Also in any patient who presents with metastatic melanoma without an obvious primary it is important to examine any pink lesions that are still existing on the skin surface.
In summary if you look at a pink lesion and it seems to have no features at all then back off, put a bit of gel on and look again without pressure because you are probably obliterating the underlying vascular structures. Only then can you confidently exclude an amelanotic melanoma.
This is an interesting article from Medscape discussing the use of dermoscopy in diagnosing
amelanotic or hypomelanotic melanomas
There is also a recent article in the Archives of Dermatology by Scott Menzies on the dermoscopic diagnosis of hypomelanotic and amelanotic melanomas.
Download it here
Image 1 is an in situ amelanotic melanoma. It has one milium cyst . If you have read the Scott Menzies article you would know that this is not significant. In his study more than 3 milia was an excellent negative predictor of melanoma.
Image 2 is a lesion that clinically looks like an amelanotic melanoma arising from a pigmented lesion but the dermoscopy is of a Firoepithelioma of Pinkus (variant of BCC) arising next to a benign melanocytic nevus!
Image 3 is of an extensive melanoma in situ on the neck looking like an area of Bowen's disease. However these are dot vessels rather than the glomerular or coiled vessels more typical of Bowen's. Interestingly the inferior area of redness she has on her neck was an area of Bowen's!
The Benign Parallel Ridge Pattern of a Small Acral Nevus
This is always a potential pitfall. In fact small acral nevi can even be histologically misdiagnosed as melanoma when in fact they are benign. This is one of the areas of the body where histopathologists are particularly wary of using criteria such as pagetoid spread of melanocytic cells into the epidermis as 100% indicating melanoma. There have been cases in the literature of small acral nevi misdiagnosed as melanoma on the basis of a parallel ridge pattern and some mild pagetoid spread. With small acral nevi always have a look at the edges of the lesion because these are more likely to give you an accurate dermatoscopic picture and often will in fact show a parallel furrow pattern.
If a small lesion on the foot or hand is very dark and appears to show a parallel ridge pattern, look at it very carefully and make sure it is not a haematoma with subcorneal haemorrhage in the skin. You can always take a blade and very carefully pare it back. You are much more likely to find this clinical scenario in every day practice than a small acral melanoma.
View this case in the Blog
Desmoplastic Melanoma
The desmoplastic melanoma is both a clinical and a dermatoscopic pitfall. Clinically it is a pitfall because the lesion often is not pigmented. It may have a red slightly scaly superficial appearance but when you feel around the lesion clinically it feels much harder than the surrounding skin and this hardness often extends beyond the overlying confines of the lesion. In some cases though a desmoplastic melanoma may have a superficial spreading melanoma overlying it and this variant can be missed if only a shave biopsy is done of the overlying pigmented lesion.
Suspicion of a desmoplastic melanoma clinically is one reason to do punch biopsies if the lesion is large. Generally we do not like punch biopsies in melanoma but you may need the depth of a punch biopsy to pick up the features of a desmoplastic melanoma. The main dermatoscopic features depend on whether there is an overlying superficial spreading component or not. If there is then the main features you will see will be those of a superficial spreading melanoma with an abnormal thickened network, dots and globules that present in nests particularly at the periphery of the lesion and perhaps some radial streaming or peripheral streaking. There may also be multiple colours. However if there is not a superficial spreading component then these lesions can dermatoscopically look quite bland. They usually have a pink colour and if this is also structureless then this in itself should raise suspicion of a melanoma. Desmoplastic melanoma may also show parallel and anti parallel lines called chrysalids. These represent collagen in the dermis on which the melanoma cells are laid out. These lines are seen in polarised dermoscopy and an example is shown in the image below. Desmoplastic melanoma tends to be a clinical rather than a dermatoscopic diagnosis.
Not all desmoplastic melanomas look like this one. A considerable number have an overlying superficial spreading type melanoma component and a shave biopsy may badly underestimate the thickness of the lesion.
The Half and Half Lesion in Dermoscopy
The half and half lesion is an interesting lesion in dermoscopy. It is always quite striking. Some lesions like the case below are very suggestive on clinical examination of melanoma and they may also raise the issue of being a combined lesion, but it appears that most of these half and half lesions are in fact just slight genetic variations with the same underlying histopathology. The lesion below has two different networks one of which was much darker and thickened and suggested a melanoma had developed. However the network that appears thicker and darker is actually quite regular and there are no other major features of melanoma in this lesion. So if the other examples below show seborrhoeic keratoses you should also look at the section on combined nevi which we will deal with in module 12 (2).
View this case in the Blog
The Recurrent Nevus
This situation usually arises because a pigmented lesion has been shaved and been found to be benign but has not been completely removed. When the nevus cells proliferate again in amongst the scar, the cells can individually look quite suspicious both dermatoscopically and histologically and the pathologist usually has to look at the original histopathology again to be convinced there was not a melanoma in the first place. Hence if you are going to examine pigmented nevi then it is best to do an excisional biopsy with small 1mm to 2mm margins than it is to do a shave or even worse to do a punch biopsy that may inadequately sample the lesion. Recurrent pigmentation that is combined to the scar itself is usually benign. The pigmentation that extends on either side of the scar is always worrying. If there is pigmentation and it is diffuse then several punch biopsies are a reasonable way of approaching the situation. Generally these lesions show multi component patterns with atypical network and irregular linear streaking both in the centre and in the periphery of the lesion. There may be vessels within the scar itself but these will often be described as having a stepladder appearance. Usually the pigmentation is concentrated in the centre of the lesion when it is benign. If the pigmentation is most accentuated at the spreading edge then this is a feature suggestive of a recurrent melanoma.
When these lesions are examined histologically then with the recurrent nevus the proliferation of melanocytes is usually above the scar. Some of these melanocytes may be seen in the overlying epidermis simulating pagetoid spread but the nevus cells in the dermis do not extend beyond the outline of the scar. If the nevus cells are found outside this outline again it is a feature most suggestive of a recurrent melanoma.
View this Blog case
Dr Peter Bourne's excellent presentation on Melanoma diagnostic pitfalls.
I want you to download
Peter's presentation here at this time and just work your way through it. I will cover some of the topics again later in this module. It looks at a Susana Puig paper on melanoma simulants that has some excellent images and then reviews Scott Menzies paper on amelanotic and hypomelanotic melanomas with some good case examples.
When you complete all segments of the Dermoscopy pitfalls module you would be well advised to go over this presentation of Peter's once more. I will link to it again at the end of Module 12.3
