Акценти от 2-ри ден

Updated epidemiology of melanoma, genetics of melanoma, Hard-to-treat melanomas, Influence of female hormones, Immunotherapy for immunocompromised patients, Comparison of international clinical practice guidelines in oncology

SYMPOSIUM: Genetics of melanoma

Chairs: E. NAGORE - S. PUIG

In the next session, on advances in the genetics of melanoma, Joan Anton PUIG reported on the relevance and feasibility of circulating tumour DNA assays in advanced melanoma. This detection of mutated DNA, released into the blood by vascular tumours, is already routine in lung cancer monitoring. The amount of mutated DNA (for example BRAFV600E) decreases with effective treatment and increases prematurely in the event of acquired resistance. However, as assay techniques are increasingly sensitive (especially digital PCR), and the majority of benign nevi are also mutated BRAFV600E, patients with dysplastic nevi syndrome may sometimes have a "false" positive assay, fortunately at much lower rates than metastatic stages. Thresholds of "normality" will therefore need to be established based on the number of nevi and the sensitivity of each technique.

SYMPOSIUM: Difficult to treat subtypes of melanoma

Chairs: C. CARRERA - L. KANDOLF-SEKULOVIC

Update on treatment of lentigo maligna: C. CARRERA

In the afternoon, a session about melanomas that are "particularly difficult to treat" piqued my curiosity. The first speaker, Cristina Carrera, showed how confocal microscopy, which is unfortunately only accessible in some equipped and trained centres, can be of great help for Dubreuilh melanoma, at different stages of treatment:

- During the diagnosis stage, in the event of a partially achromic lesion, confocal microscopy makes it possible to confirm and map the melanoma with greater certainty than the clinical exam and the dermoscopy.

- During the preoperative stage, it helps better define the margins. A preoperative confocal microscopy examination would modify the initially planned surgical margins in more than 70% of cases.

- When surgery is ruled out, following a topical treatment (excluding off-label, such as imiquimod or ingenol mebutate) or radiation therapy, confocal microscopy can manage the response in the entire treated area.

Update on treatment of mucosal melanoma: L. KANDOLF-SEKULOVIC

The next speaker, Lidija Kandolf-Sekulovic, addressed the issue of mucosal melanoma treatment. Adjuvant radiotherapy is recommended for mucosal melanoma of the head (oral and nasal cavities), but not in other melanomas (anogenital). In anogenital melanoma, complete excision is typical. A minimum margin of 1 mm, except in penile melanoma, or a minimum margin of 5 mm seems preferable. During the stage of metastatic mucosal melanoma, immunotherapy is often the first-line therapy because BRAF mutations are rare (but possible: 3-16%, with 89% BRAFV600E). As anti-PD-1 monotherapy is not very effective in mucosal melanoma, it is recommended to begin, from the start if possible, dual anti-PD-1 and anti-CTLA4 therapy (37% of responses, vs 23% with anti-PD-1 alone, and 26% with KIT inhibitors).

Management of melanoma in pregnancy: M. PASEK

The next speaker, Marek Pasek, addressed the issue of female hormones' impact on melanoma. He began by sweeping aside myths about hormones: there is no increased risk of melanoma from either the pill or hormone replacement therapy. Data on pre-IVF hormones are insufficient to reach a conclusion. But it is true that the ERβ protein (oestrogen receptor β) is often expressed in melanoma.

When age and Breslow depth are equal, the prognosis of melanomas in pregnant women is the same for non-pregnant women. On the other hand, it is necessary to avoid the sentinel lymph node technique (particularly in the 1st trimester, increased teratogenic risk). In the event that melanoma is suspected, an excision with a margin of 3-5 mm is recommended first, under local anaesthesia of course, and using the sentinel lymph node technique for melanomas over 0.8 mm in Breslow depth should be postponed to after pregnancy. Imaging tests such as scans or PET scans are contraindicated. Ultrasounds are preferable, even for thick melanomas or clinical metastases, or whole-body MRI without gadolinium injection. Surgery under general anaesthesia should be avoided especially during the first trimester (increased risk of miscarriage), but should still be discussed depending on the vital risk incurred by the mother.

Finally, Céleste Lebbé enlightened us on the obscure subject of immunotherapy in immunosuppressed patients, who were excluded from clinical trials from the outset. Data are still scarce, but it seems that ipilimumab causes less transplant rejection than anti-PD-1 (fewer responses, unfortunately). 7/15 kidney transplant patients who received anti-PD-1 rejected their transplant, 6 of whom resumed dialysis. Transplant rejection is also expected in liver transplant patients (4/7 patients treated, 3 of whom died). However, in this small series of liver transplant patients, 2/7 experienced complete remission. This terrible benefit/risk ratio should therefore be seriously discussed with each patient. To increase our knowledge of the risks and benefits of transplant patients receiving systemic melanoma treatment, Céleste Lebbé and Julie Delyon have set up an ideally European register in which each of us is invited to participate.

Finally, in a small cohort of patients afflicted with autoimmune disease prior to introducing anti-PD-1 treatment, 38% experienced flare-ups of a pre-existing autoimmune disease. Exacerbations are therefore not systematic. I hope this summary of the day was clear and I'll tell you more tomorrow!

Dr Oriol YÉLAMOS PENA

Dermatologist, Hospital Clínic & Centro Médico Teknon, Barcelona, Spain

BREAKOUT SESSION: Skin cancer guidelines: International structure & cooperation

Chairs: C. GARBE - J. THOMPSON

In the US: D. COIT

NCCN guidelines are meant to help physicians not necessarily in comprehensive cancer centers, and are actually a guide, to help clinicians but they are recommendations, not obligations. Guidelines sometimes are difficult to make, especially where no clear evidence exists. In these cases, consensus is achieved with voting. It is important to consider that the guidelines are written among different specialists with different backgrounds and beliefs. Importantly, no conflicts with the industry exist when elaborating the guidelines. If there are conflicts, a given member may be removed. When evidence exists regarding a given issue, this evidence is used, and then the gaps are filled with consensus among experts. This has been important, since if something does not have a strong evidence, the likeliness of non-reimbursement is big. Guidelines are also translated into different languages such as Spanish. To summarize, the NCCN guidelines are meant to set the foundation of oncologic care in the USA trying to cover the current evidence, reality of the population as well as the reimbursement system in the USA.

In Australia: J. THOMPSON

Guidelines are meant basically to give appropriate advice to patients, that clinicians have the best available evidence and when evidence doesn’t exist experts may give their expert opinions to help the rest. The first guidelines by the Australian Cancer Network were written in 1997, so Australia’s guidelines are relatively new. In 2008 the second cancer guidelines were written together with New Zealand, but it was very complicated since they had a different reality (no PET-CT machines in NZ back then for example). In 2014 new guidelines were written down without New Zealand and were available in 2018 online (not printed in paper) at wiki.cancer.org.au where useful information for patients, physicians and everyone was available online. At the end of each chapter of the guidelines, there is stated which is the level of evidence. However, for the melanoma there are not a lot of grade A evidence, and most of them are B and C. However, since the guidelines are online, they are to amend (for example the no need for completely lymph node dissection).

In Germany: C. GARBE

The S3 Melanoma guidelines adapted recommendations from the Australian, French, US and other guidelines. Also, systematic reviews were performed. After the guidelines were made, two consensus meetings were held so around 60 people did the consensus. So, the recommendations not only have the level of recommendation but also they add the percentage of agreement consensus. The problem with this system is that takes more time to develop the guidelines. Hence, it would be nice to share the effort and unify classifications regarding levels of evidence.

Differences and recommendations: A. PFLUGFELDER

The different guidelines have many differences in the way they are presented, how they are labeled, and how they deal with different realities.

Most guidelines have similar recommendations on the margins of excision. For melanoma in situ the majority of guidelines agree on 5mm margin excision, except the German guidelines that say it is OK to just remove the tumor. More differences occur when the Breslow thickness is > 1mm, since some guidelines consider 1cm and others 2cm of margin excision. Regarding sentinel lymph node biopsy, there is important variability but overall most guidelines say that SLNB should be offered. Conversely, all guidelines recommend complete node dissection when clinically positive nodes. Regarding node dissection in positive SLNB this has been changing since MSLT-II showed that no improvement is obtained if complete node dissection is performed if positive SLNB. However, most guidelines suggest that one has to discuss this option with the patient. Regarding adjuvant therapy, there is no consensus yet and most guidelines suggest that it may be discussed with the patient.

Harmionisation of terminology and wikification: T. EIGENTLER

The problem with different guidelines is that some are written down faster (USA) with less consensus (like the German). So, it would be good to have a fast way to obtain very robust evidence. The solution but being to have an active community with moderators, quality manages and authors all together in a platform. This doesn’t exist in Oncology but it exists in the IT world, where tools that expedite process are available, such as GitHub: you can propose changes, request reviews, visualize differences, comment in context and protect branches.

PLENARY LECTURES:

Chairs: F. M. CAMACHO - G. PELLACANI

Total body mapping, non-invasive diagnosis and AI systems in skin cancer detection: A. HALPERN

Regarding total body photography one of the most advanced systems is the 3D system from Vectra. However, although this system is very advanced, this is not available in most practices. Therefore, one approach that may be widespread is artificial intelligence. The challenges of AI are the fact that AI algorithms need information which includes images but also metadata. Then, it is important to have big databases with data to feed AI algorithms. One of the efforts to provide data is the International Skin Imaging Collaboration (ISIC). ISIC encompasses several institutions worldwide which use dermoscopy daily. These institutions provide images to a global image archive. However, there are challenges in images in dermatology since no standards exist on how to take a picture, how to store them, there are no standards in terminology…

The ISIC archive has different aims such as sharing images, education, perform studies… Currently, the ISIC archive has 23 906 public images. Also, these images have been used to develop AI algorithms in multiple conference challenges such as the ISBI or the MICCAI meetings. These algorithms are actually performing very well and equally to dermatologists. Actually, only expert dermoscopists do better than some of the algorithms.

Then, the bottom-line is that rather than the question whether AI is going to be in our practices or isn't a real question, the question is when. AI will actually not replace dermatologists but will help triage lesions that then need to be evaluated, because it is still important to check lesions since exceptions and difficult cases exist (for example, spitzoid lesions in children should not be removed whereas in adults they should be removed). In addition, other non-invasive technologies (confocal microscopy, tape stripping, impedanciometry…) may help in also identifying high risk lesions.

How does UV induce skin cancer?: A. VIRÓS

UV radiation plays a complex role in melanoma. Virós et al performed a series of experiments and saw that when stimulating the skin of mice with UV they stimulated the genesis of nevi, but also the genesis of melanomas. The type of mutations was different whether the UV exposure was chronic or acute: epithelial melanocytes that receive intermittent (acute) sun exposure and when melanoma arises in these ase they are mostly driven by BRAF mutations, in which p53 acts as a cooperator. However, the other melanomas arising in chronically sun exposed melanomas are mainly driven by NRAS. NRAS melanomas have a high tumor burden with many UV-related mutations, whereas BRAF melanomas have a low tumor burden with few UV-related gene mutations. Zaidi et al (Nature 2011) showed that early exposure to UV increases the likelihood of melanoma. As a summary, it seems that melanomas arising in sun-exposed areas are linked to high DNA damage, whereas the melanomas occurring in less or not sun-exposed are linked to inflammation.

SYMPOSIUM: New imaging diagnostic

Chairs: J. MALVEHY - I. ZALAUDEK

What's new in dermoscopy of skin cancer: I. ZALAUDEK

If we create population-based screening strategies, we may save one live from melanoma in 100 000 people. Therefore, screening everyone is difficult, expensive and most probably not useful. However, it is more important to screen the people with high risk factors such as atypical mole nevus, red hair… However, these people are a minority! The majority of people who develop melanoma do not have these risk factors. So we still need to know more about risk factors.

What seems to impact melanoma detection is dermoscopy: with dermoscopy we can identify more melanomas, diagnose them earlier and thinner. However, we still have not seen a decrease in the mortality of melanoma. So, probably what we detect now is melanomas that will not kill our patients, whereas the aggressive ones we cannot identify with dermoscopy (ie nodular melanoma). Instead of teaching the ABCDE to patients we need to tell them that anything that is new needs to be checked, since nodular melanomas are rapidly growing aggressive tumors. In fact, most nodular melanomas, including thin nodular melanomas (Breslow < 2mm) are identified by the patient! So in these cases dermoscopy or doctors are relatively useless, and therefore we need to teach people which are the signs of rapidly growing melanomas, rather than teaching the ABCDE rule. In addition, smartphones and artificial intelligence seem the way to identify aggressive melanomas.

Digital imaging and beyond: J. MALVEHY

Nowadays we live in the digital era. Currently still many physicians ask if imaging is important or relevant, but the truth is that one image is more useful than written descriptions. Therefore, clinicians should obtain images of the lesions they explore. It is important for these reasons to obtain pictures, and if we don’t patients will do. In fact, patients take pictures of their lesions, so we need to be beyond them.

Also, up to 60% of melanomas derived to a skin cancer unit in Barcelona come without pictures taken by their physicians. Also, when they take pictures, there is a lot of variability on how the pictures are taken, which cameras are used… So, an additional problem is that we need to establish standards in how pictures are taken, managed and stored. Dermatology is not like radiology where DICOM standards exist. Hence, it is important to have standards not only for clinical photography but for dermoscopy, total body photography and even confocal and other imaging techniques.

In terms of what’s the future is total body dermoscopy (using cross polarized light), artificial intelligence algorithms which will triage the lesions relevant… But physicians will still need to check special sites and challenging cases.

Confocal microscopy and OCT in skin cancer diagnosis: G. PELLACANI

Reflectance confocal microscopy (RCM) and OCT are very good tools but which one to get is different according to what do you want to use it for.

If you need something to diagnose a disease that you need high magnification, like melanoma, you then need RCM since it has cellular resolution. RCM has very high diagnostic accuracy and reduces the number needed to excise to diagnose one skin cancer.

If you need a device to diagnose BCC, you can diagnose it with OCT (mostly the margins, and you can also measure the thickness) but also with RCM (confirms the diagnosis but won’t allow you to see the margins very well).

For actinic keratosis is may be good to combine the two technologies, RCM and OCT. In difficult cases, RCM will confirm the diagnosis, and with OCT you can evaluate the thickness and whether the dermal-epidermal junction is disrupted then you can suspect that maybe you’re facing an early SCC.

Electrical impedanciometry in melanoma detection: R. HOFMANN-WELLENHOF

Impedanciometry uses different electrical signals that react differently depending whether a lesion is benign or malignant. Depending on how disarray the tissue is, the machine renders a score that will be classified in different categories. The score ranges from 0 to 10, being the higher scores more likely to being melanoma. Generally, the cut-off to distinguish benign from malignant is around 3. The diagnostic accuracy seems very good but as with any other diagnostic tool one has to integrate all the information (clinical, dermoscopic…) before deciding whether to excise a lesion or not, since false positives and false negatives can occur.

AWORKSHOP: Dermatologic ultrasound in skin cancer patients

Chairs: F. ALFAGENE - P. GIAVEDONI

Dermatologic ultrasound: Practicalities and state of the art in skin oncology: P. GIAVEDONI

Ultrasonography (US) can be very useful in dermatology since it’s relatively easy to use and cheaper than other devices. Dr Ximena Worstman was the first one in 2004 to really use ultrasonography for the skin, using also color Doppler. Later, in Spain Dr Alfajeme started spreading its use since 2013. Due to the exponential growth of publication, the DERMUS group wrote a consensus paper in which the terminology and the equipment needs are discussed.

Probes used range from 7.5 to 22 Mhz. After selecting the appropriate probe one has to select the adequate preset depending which areas are explored. It is important to use a large amount of gel and not to press a lot since some structures can disappear. Sometimes it can be useful to use a separator disk, but before it is necessary evaluating the lesion only with gel. Later, we need to know which parameters are more important in mode B: gain (contrast of grays), depth, and focus of the image. Later, it is important to perform the US in a systematic way: measure the 3 axis (transversal, longitudinal and depth), then… (in ppt).

Artifacts need to be known: posterior shadowing (typical of calcified or hyperkeratotic lesions), posterior reinforcement (typical from cysts or melanoma M1), mirror images (image is reflected because of the presence of the bone, typical when we evaluate the scalp.).

Doppler: blue means that blood is moving away, whereas red means that the blood is approaching. However, US can be limited for example to diagnose melanoma since melanin is not visible with US leading to low sensitivity.

Multiparametric ultrasonographic assessment of non melanoma skin cancer: R. BADEA

US is a fast, non-invasive method that is very useful in dermatology. US can use different modes (more info in the slide if necessary):

- Gray mode - Color Doppler

- Power Doppler

- Elastography

- CEUS (contrast US)

Malignant non-melanoma tumors are hypoechoic, with irregular margins, hypervascular (especially with arterial vessels) with low resistivity index and hyperemic surrounding tissue. In elastography they are rigid and when contrast is used in CEUS mode they have enhancement in the arterial phase (30 seconds) and longer (20-30 seconds) wash out.

At the end, the diagnosis is rendered after using all the different modes.

Other applications of US are:

1) to diagnose early recurrences or try to identify residual tumor just after surgery,

2) to evaluate what is under scars in lymph node areas.

Ultrasonography for the follow up and prognosis of melanoma: L. FERRÁNDIZ

The management has changed dramatically with the advent of new therapies but also because of the results of the MSLT-II, which showed that there was not an increase in overall survival with complete lymph node dissection. Therefore, instead, it was suggested that ultrasonography would be the method to follow these patients.

In their unit they perform US before SLNB staging, in patients who do not want to undergo or cannot undergo SLNB, then during follow-up in patients who do not undergo SLNB, and finally preoperatively or when guiding treatment such as T-VEC.

The equipment used generally is a linear transducer with frequencies between 7.5 to 14 MHz. Pathologic lymph nodes are hypoechoic. Why? Normally lymph nodes have a hyperechoic center with a hypoechoic medulla. Hence, anything that disturbs this should be suspected to be malignant. Also, since normal lymph nodes are ovoid, if the shape is not ovoid also suspect something pathologic.

Any lymph nodes should be suspected for malignancy if:

- that presents with medulla displacement.

- spherical silhouette (Solbiati index < 1.5)

o Solbiati index is the division between the long and short axis of a lymph node.

- Hump structure

- Hypoechoic islets

- Central perfusion absent (vessels get to a normal node from the center, like a kidney.)

- Peripheral perfusion

- Parenchymal perfusion

Some tricky lesions are vessels, but then is when we use Doppler. Also, seromas can be hypoechoic but inside show septae which are hyperechoic.

New diagnostic approaches in dermatologic ultrasound of skin cancer: elastography, US contrasts and ex-vivo: F. ALFAGEME

Elastography allows us to “touch” deep structures and know if they are soft or hard. Collagenic structures are generally harder than fatty ones. Applications of elastography are not only to diagnose cancers (they tend to be harder.) but also to delineate tumor margins. Also, elastography can help distinguish different BCC subtypes: morphea form is harder/more rigid compared to others.

Contrasts can also help identify malignancy, since contrast has a slow wash out.

Ex-vivo ultrasonography can be used for example to delineate the BCC margins after surgery, in a very fast and cheap way.