This article does not give you positive features to look for but merely rates the significance of absent positive features you would hope to find! It emphasises the importance ofhe ugly duckling sign and the history of recent change.(IMCC)
Dermoscopic Features of Difficult Melanoma
MARIA A. PIZZICHETTA, MD**Centro di Riferimento Oncologico, CRO, Aviano, Italy; , IGNAZIO STANGANELLI, MDSkin Cancer Unit, Ravenna-Niguarda Hospital, Milan, Italy; , RICCARDO BONO, MDIstituto Dermopatico dell'Immacolata, IRCCS, Roma, Italy; , H. PETER SOYER, MD§§Department of Dermatology, Medical University of Graz, Graz, Austria, SERENA MAGI, SCDSkin Cancer Unit, Ravenna-Niguarda Hospital, Milan, Italy; , VINCENZO CANZONIERI, MD**Centro di Riferimento Oncologico, CRO, Aviano, Italy; , GIUSEPPE LANZANOVA, MDSkin Cancer Unit, Ravenna-Niguarda Hospital, Milan, Italy; , GIORGIO ANNESSI, MDIstituto Dermopatico dell'Immacolata, IRCCS, Roma, Italy; , CESARE MASSONE, MD§§Department of Dermatology, Medical University of Graz, Graz, Austria, LORENZO CERRONI, MD§§Department of Dermatology, Medical University of Graz, Graz, Austria, AND RENATO TALAMINI, SCD**Centro di Riferimento Oncologico, CRO, Aviano, Italy; , ON BEHALF OF THE ITALIAN MELANOMA INTERGROUP (IMI)*Centro di Riferimento Oncologico, CRO, Aviano, Italy; Skin Cancer Unit, Ravenna-Niguarda Hospital, Milan, Italy; Istituto Dermopatico dell'Immacolata, IRCCS, Roma, Italy; §Department of Dermatology, Medical University of Graz, Graz, Austria
Address correspondence and reprint requests to: Maria Antonietta Pizzichetta, MD, Division of Medical Oncology C–Preventive Oncology, Centro di Riferimento Oncologico, Via Pedemontana Occidentale 12, 33081 Aviano, Italy, or e-mail: pizzichetta@cro.it
The authors have indicated no significant interest with commercial supporters.
Abstract
BACKGROUND
The dermoscopic diagnosis of cutaneous melanoma (CM) may be difficult because some CM lack specific dermoscopic features for melanoma diagnosis.
OBJECTIVE
To evaluate whether a diagnosis of CM could be achieved using the classic dermoscopic melanoma-specific criteria, we conducted a retrospective multicenter study of 508 CM samples.
METHODS
All the dermoscopic images were analyzed to identify the dermoscopic criteria found in dermoscopically difficult melanomas (DDM) and to examine the possible relation of dermoscopic diagnosis with respect to the difficulty of the dermoscopic diagnosis and the melanoma thickness.
RESULTS
A significant percentage of melanomas, 89 of 508 (17.5%), were DDM. The criteria leading to a significant increased risk of DDM were presence of streaks [odds ratio (OR), 2.26; 95% confidence interval (CI), 1.15–4.47), absence or presence of regular pigmentation (OR, 3.41; 95% CI, 1.70–6.85), absence of a blue-whitish veil (OR, 4.04; 95% CI, 2.33–6.99), absence of regression structures (OR, 4.31; 95% CI, 2.42–7.66), and the presence of hypopigmentation (OR, 2.61; 95% CI, 1.49–4.58).
CONCLUSION
A significant number of melanomas defy even dermoscopic diagnosis. Only a meticulous comparative and interactive process based on an assessment of all the individual's other nevi ("ugly ducking" sign) and a knowledge about recent changes can lead to the recognition of DDM.
5 comments:
This article and the recent cases on the blog spot are disturbing. The question is "what do we do about it". If ivory tower academics are stating that they miss a substantial number of melanomas, what is your average bog standard skin cancer doc to do. (let alone the average Oz GP who is said to correctly diagnose his 1-2 a year!).
Obviously willy nilly removal of all remotely abnormal pigmented lesions is one possible answer but not a course I am prepared to follow. Equally photography of all lesions and continuous monitoring is not practical in our setting. I believe that we should continue to do as most of us are, i.e. excise the doubtful and monitor the not sure. I dont think I have missed many melanomas as I have yet to be sued! Equally I am sure I have not excised lesions similar to some of the level 1 melanomas recently posted on the blog. Accordingly, I look forward eagerly to 2 presntations by Dr Ashfaaq Marghoob in the coming skin cancer fest, namely:-
1. "Does biologically benign melanoma exist?"
2. "Does in-situ melanoma always become invasive?'
For those of us using the Skin Cancer Clinic Blog the experience is sometimes disheartening. Everytime Cliff puts a presentation up I can feel a three putt coming on! I still refuse to double guess him!
I firmly believe we should desist from excessive surgical excissions in these circumstances.
Perhaps this merely reflects that whilst dermoscopic criteria have been refined for most types of invasive MM, there is yet to be defined a set of dermoscopic criteria for the full range of MM in-situ, especially.
I don't think that, with experience, going outside the standard dermoscopic criteria to excise is wrong if you are able to demonstrate a reasonable NNT (<10:1)and that you are diagnosing a fair number of MMis.
If I strictly applied Menzie's criteria in the last month in my practice I would have "missed" most of the 8 new MM seen in my practice (6 MMis, one 0.25mm, one 0.45 mm). Still had a NNT of about 2.5:1.
In this paper, I also found the lack of concordance beween pathologists worrying....
I have expounded this theory before on the blog, but I do believe that the various deermoscopy algorithms are not in fact rules, but rather teaching aids. Once you know what you're doing, you no longer really do it by the rules. I think we make our minds up quite quickly if a lesion is benign or malignant, then look for "features" to support this view. And for lesions that elicit an initial reaction of "I don't like it", we find a justification to excise even in the absence of classical features.
I would certainly be very interested to know for this study, what led to the decisions to excise these "featureless" melanomas. And for those detected by serial imaging, what led to the decision to image in the first place.
I also wonder if MMIS inevitably progresses, or whether it is in any way analagous to the situation with breast screening and DCIS in particular; I beleive around 20-30% more malignancies are found in the screened group than the unscreened, with the clear impication that many malignancies do not progress, and that there is an inevitable element of overtreatment in the screening group.
We really don't know that all these MMIS will inevitably become invasive, indeed I have a conviction (totally unsupported by any evidence) that a significant number do not progress.
I find it amazing that discordant opinions on histology review cast doubt on the diagnosis of 25% of the featureless "melanomas", yet this was brushed aside with a comment that a kappa of .75 is "good"...
Having reread the whole article and digested the comments, particularly with reference to Alans comments in his first paragraph which is the way I practice. The only "featureless melanoma" picture shown in the full article (albeit not very clear pictures) is one that most bloggers would have had in the pot straight away (eccentric area of pigmentation, probable radial streaming and multiple colours). From the recent posts on the blog, it seems inescapable to me that some people have a much more highly developed UDR (ugly duckling recognition) factor(otherwise known as the Rosendahl sign) than others especially me. I feel we need to concentrate on how the UDR factor is developed and used. In other words is the UDR factor a product of time spent with individual patients, use of monitoring, type of equipment etc or is it a God-given gift. Secondly, and this is harder, how much difference does it make?
Post a Comment