Guidelines for the diagnosis and treatment of cutaneous squamous cell carcinoma and precursor lesions

Carcinomas are malignant tumours of epithelial origin. Cutaneous carcinomas are primarily of keratinocytic origin (epidermal or follicular keratinocytes) or of adnexal glandular origin. Keratinocytic carcinomas include basal cell carcinomas (BCC) and cutaneous squamous cell carcinomas (SCC). BCC and SCC are by far the most common forms of cancer in humans. Paradoxically, as they are not generally recorded in cancer registers, their importance in terms of public health and their economic impact on healthcare systems are widely underestimated. The relative incidences of BCC and SCC differ according to whether SCC is grouped among lesions with the same oncogenesis or not, namely actinic keratosis (AK) and Bowen's disease (BD) (see below). If AK is included in this group, SCC is the most common form of human cancer. If it is excluded, while being responsible for the majority of deaths attributable to non-melanoma skin cancers (NMSC), SCC is the second most important form of cancer in terms of frequency. Although the majority of SCC cases are not life-threatening, this carcinoma is likely to metastasize, particularly if initial treatment was inadequate. The SFD has drawn up guidelines for the treatment of BCC (2004) and melanoma (2005). The present work is the logical continuation of this initiative. It aims to provide practitioners treating skin cancer patients with a series of recommendations based on scientific evidence or, when this was not available, on expert consensus. Medical textbooks refer to a wide array of clinical and histological forms of SCC. The prognosis for these different forms varies according to whether therapy has been specifically codified or not. In addition, the terminology used to describe these lesions varies, which may lead to confusion and prevent use of a clear decision-making tree. The TNM classification, developed by AJCC/IUAC/UICC, which is used for all skin cancers except melanoma, is not suitable for SCC. It does not take into account the multiple prognostic criteria identified in the literature. Many treatments are currently available, in particular, for SCC precursor lesions, but the criteria of choice and the methods of application are generally far from clear for practitioners. This results in major disparities in terms of therapeutic management. The present guidelines deal with the treatment of SCC and precursor lesions in immunocompetent adults in France, in both ambulatory and hospital settings. The aims are as follows: to clarify the terminology used to describe the different forms of SCC and of their precursors AK and BD; to propose a prognostic classification of SCC that takes into account various clinical and histological factors; to recommend diagnostic and therapeutic measures for SCC based on previously identified prognostic factors; to optimize diagnostic and therapeutic management of AK and BD in accordance with recent data in the literature; to provide an overview of the principles for primary prevention of SCC and precursor lesions (based on the same methods), and for screening of subjects identified as at risk for SCC (other than genodermatosis and immunosuppression). Due consideration has been given to the fact that patients with SCC, AK or BD are generally (very) elderly. This creates problems in terms of screening for lesions, and for amenability to care and treatment (poor compliance to certain treatments, difficulty in carrying out sequential physical treatments, refusal of onerous surgery, or surgery requiring multiple operations). The oncogeriatric dimension of therapy has thus been taken into account. These guidelines do not address the following issues: SCC of the nails or the genital and anal mucosa; SCC in immunosuppressed patients, particularly transplanted patients; SCC in the context of certain genodermatoses. The levels of evidence and grades used are those defined by the HAS [French Health Authority] (Annex 1). The literature on SCC generally carries low levels of evidence and, except where otherwise stated, the recommendations included in these guidelines are of grade C. These guidelines have been created in the form of Recommendations for Clinical Practice in accordance with the ADAPTE method.2 As its name suggests, this method advocates the adaptation to a particular situation – in this case, medical practice in France in 2008 – of one or more guidelines on the same theme, drawn up previously or in other countries. Medical societies concerned with SCC were consulted on the initiative of the French Dermatology Society (SFD), the sponsor, to define the scope of the guidelines, identify work performed on the subject and recommend professional members for the organizing committee (OC), the working group (WG) and the reading group (RG). Mention must be made of the difficulty of recruiting general practitioners to these groups, despite the fact that, in view of the subject, their assistance is vital. In the Spring of 2007, the Dermatology Recommendations Association (aRED), a subgroup of the SFD, created a multidisciplinary OC on behalf of the SFD, comprising doctors in private and public practice, both university and non-university practitioners, from a variety of geographic origins. The OC then set up a WG using the same criteria of professional diversity. Members of the OC and WG were asked to complete a form indicating any conflict of interest regarding management of SCC (Annex 2). The members of the RG were recommended by medical societies, once again with the aim of reflecting the diversity of professional practice. The overall arguments, key points and recommendations were drafted by the WG following the identification and selection of previous guidelines on SCC, contextualization (occasionally critical) of recommendations contained therein and a synthetic update of the literature. Practitioners in the reading group were sent a letter asking for their opinion on the topic, including presentation of the key points and recommendations, in particular, regarding clarity and applicability. The comments made by the RG were analysed by the WG and, whenever possible, taken into consideration in the final draft. Lastly, on the 11 December 2008, the main recommendations were presented and discussed publicly in the presence of practitioners to whom the guidelines were addressed during the Journées Dermatologiques de Paris, the main French national dermatology congress. The low level of evidence in the existing literature underscores the continuing lack of knowledge about optimal management of patients with SCC. These areas represent subjects for future work by the OC and WG (see Perspectives section). I.2.1. ADAPTE method This rigorous and explicit method, recently described and published by an international group,2,3 is designed to enable the adaptation of existing guidelines, and to reduce the time, effort and cost required to create a fresh set of guidelines. Methodological guidelines concerning the use of this method were published online by HAS in March 2007.3 I.2.2. Choice of method for drafting guidelines In the spring of 2007, the SFD Bureau and the aRED decided that the ADAPTE method could be used to draft French guidelines for the management of SCC. Foreign guidelines on this topic, some of them fairly old, were already known, and ADAPTE contextualization was entirely possible, as these guidelines had been published by agencies or medical societies for populations and levels of health infrastructure and organization comparable with those in France. However, as the literature used for the drafting of these guidelines was deemed to be of mediocre quality, the ADAPTE method presented a number of limitations. In addition to this, several questions that have subsequently come to the fore (e.g. the nature of AK or the place of new medical treatments) occupied little or no place in the existing guidelines. Therefore, in addition to adaptation, updating of the literature proved necessary. I.2.3. Definition of the scope of the guidelines The limits of the SCC topic were discussed in conference calls between members of the OC in July 2007. The PIPOH checklist3 used to define the scope and target audience of the guidelines was as follows: P (patient population) = French population of both sexes; I (interventions) = prevention, screening, diagnosis, treatment and monitoring; P (professionals) = specialists responsible for diagnostic and therapeutic management of SCC, general practitioners (GPs), occupational therapists as well as specialists involved in the screening and follow-up; O (outcomes, the evaluation criteria used for the recommendations) = levels of treatment response in terms of remission, local relapse, remote metastasis and mortality, when these parameters were available; H (healthcare setting) = ambulatory or hospital. In addition to diagnosis and curative treatment of SCC, it was decided to include: screening of subjects identified as at risk for SCC (excluding genodermatosis and immunodepression); cutaneous or cutaneous-mucosal sites on the borderline of dermatology: eyelids and vermilion border of the lips; lesions considered precancerous and keratoacanthoma. These lesions are histogenetically related to SCC and have been dealt with in several recent studies of medical and surgical therapies. However, genital and anal sites in both genders were ruled out as these are normally dealt with by gynaecological surgeons, urologists, gastroenterologists or digestive surgeons. Ungual sites were similarly ruled out. In addition, it was decided not to include SCC observed in immunosuppressed organ transplant recipients, as guidelines for the management of these patients were being drawn up under the auspices of the HAS.4 I.2.4. Documentary research In late June to early July 2007, Mrs J. Brugneaux performed a literature search for practical guidelines on SCC and precursor lesions using systematic surveys of medical bibliography databanks (Annex 3) and looking out for guidelines, consensus conferences, articles on decision-making process, systematic reviews, meta-analyses and other national and international evaluation studies. Relevant websites (government agencies, medical societies, etc.) were also explored. Documents not accessible by the standard diffusion circuits ('grey literature') were consulted using every available means. Legal and regulatory texts on this subject were also consulted. Only English- and French-language articles were considered eligible. Initially, the selected references were screened by L. Martin and J.-J. Bonerandi to eliminate all texts unrelated to the subject on the basis of their titles, (e.g. non-cutaneous CE of the head and neck; genital CE) and irrelevant literature (guidelines currently at project level, etc.). The remaining texts (n = 58) were submitted to all CO members in July 2007 to ensure that they were authentic guidelines, didactic articles or authors' opinions. I.2.5. Determination of questions to be covered in the guidelines At a plenary session of the OC held on 19 September 2007, the scope of the guidelines was definitively agreed upon, as were the various topics to be covered therein: clinical, pathological and epidemiological forms of SCC and precursor lesions; prognostic factors for SCC; treatment methods for SCC and precursor lesions; patient management. At 19 September 2007 meeting, a list with the following documents, theoretically amenable to adaptation, was established: Non-melanoma skin cancer: guidelines for treatment and management in Australia. National Health and Medical Research Council. 2002 (NHMRC, Australia). Multiprofessional guidelines for the management of the patient with primary cutaneous squamous cell carcinoma. British Association of Dermatologists/British Association of Plastic Surgeons. 2002 (BAD/BAPS, UK) Guidelines for management of Bowen's disease: 2006 update. British Association of Dermatologists. 2006 (BAD, UK). Guidelines for the management of actinic keratosis. British Association of Dermatologists. 2007 (BAD, UK). Basal cell and squamous cell skin cancers. National Comprehensive Cancer Network. 2007 (NCCN, US). Multiprofessional guidelines for the management of the patient with primary squamous cell carcinoma. National Guideline Clearinghouse. 2007 (NGC, US). Green A, Marks R. Squamous cell carcinoma of the skin (non-metastatic). Clin Evid 2005; 4: 2086–2090. I.2.6. Selection of guidelines for adaptation The suitability of these seven guidelines for adaptation in different areas of practice was assessed by five WG, using the simplified AGREE appraisal instrument5 (Annex 4). The NGC guidelines and the Clinical Evidence article were not used, as the former is a retranscription of the BAD 2002 guidelines, whereas the latter gives no indication of the method used for the bibliography search, and was thus deemed inconsistent with the scope of the present guidelines. The following three SCC guidelines were ultimately selected: Basal cell and squamous cell skin cancers, 2007 (NCCN); Multiprofessional guidelines for the management of the patient with primary cutaneous squamous cell carcinoma, 2002 (BAD 2002); Non-melanoma skin cancer: guidelines for treatment and management in Australia, 2002 (NHMRC), together with Guidelines for the management of actinic keratoses, 2007 (BAD 2007) and Guidelines for management of Bowen's disease: 2006 (BAD 2006). I.2.7. Layout of the guidelines, key points and recommendations Each section was divided into 'items', some of which contained one or more 'key points' or 'recommendations'. An item was comprised of a topic concerning epidemiology, diagnosis or treatment amenable to targeted documentary research (e.g. prevalence of SCC among the French population; use of imiquimod in the treatment of AK). Each item was assigned to a group comprising two or three WG members, based on their experience or interest in the topic. Individual items were either identified as existing in one or more adapted guidelines, or were created anew by the WG. Items are summarized at the end of this section in a summary table showing the adapted guideline(s), where they are also discussed (Table 1). Key points comprise information items of cultural relevance which are possibly not directly related to day-to-day practice. Recommendations refer to the diagnostic or therapeutic management of patients. Key points and, wherever possible, recommendations were graded using the HAS method (Annex 1). In accordance with the ADAPTE method,6 items identified in previous guidelines were evaluated by WG members to determine the degree of concordance between the data analysed and the conclusions set out in the arguments, and between these conclusions and the recommendations proposed. Any divergence with regard to prior guidelines is indicated at the start of the paragraph. In most cases, updating of bibliographical references proved necessary, and this was carried out by Mrs J. Brugneaux (January 2008) (Annex 3). Finally, actual adaptation (drafting of the arguments, key points and recommendations) involved summarizing the proposals set out in the various guidelines and drafting the arguments suitable for medical practice in France in 2008. An initial version of the arguments and the proposed recommendations was reread and discussed in plenary sessions by the WG on 11 March 2008 and by the OC on 25 March 2008. Updating of the bibliography was suspended at this point. Opinions differed within the WG regarding a number of relevant items (prognostic classification of primary SCC, value of routine histological analysis of excision margins and hierarchical classification of therapeutic choices for precancerous lesions). In the absence of consensus within the adapted guidelines and of literature providing an adequate level of evidence, these divergent opinions gave rise to numerous e-mail exchanges and meetings until a pragmatic consensus was reached within the WG and the CO. The successive versions of the arguments attest to changes in viewpoints regarding these items. The WG stressed the need for clinical trials designed to obtain factual information which would settle these divergences. A complete argumentation framework and an initial version of the short text comprising the key points and recommendations were sent to readers during summer 2008. Final versions of the documents and of a highly synthetic pocket-sized flyer were prepared in autumn 2008. The CO and the WG are well aware that their editorial choices simply reflect medical and scientific knowledge concerning SCC and precursor lesions up to spring 2008. Follow-up on bibliography data and annual meetings of the WG are scheduled to ensure prompt modification of the guidelines following the publication of relevant diagnostic, prognostic or therapeutic information regarding SCC and precursor lesions (see Perspectives). The term squamous cell carcinoma encompasses all malignant epithelial tumours with predominantly malpighian differentiation. SCCs include primary malignant skin tumours with malpighian differentiation, and are distinct from other primary epithelial skin tumours such as BCC. The term thus covers a number of different clinico-anatomical entities, some of which only differ in terms of clinical presentation or degree of aggressiveness. The inclusion of AK and keratoacanthoma under SCC by some authors and in certain reference works is currently disputed.6 The adapted guidelines, i.e. the three guidelines dedicated to SCC (NHMRC,7 NCCN8 and BAD9) and those specifically focused on AK and BD,10,11 restrict themselves to a summary description of these entities without discussing clinico-anatomical forms and nomenclature. The WG felt it was necessary to adopt a position on this subject and therefore, the ADAPTE method was not used in drafting this chapter, for which a specific bibliographical analysis was performed. This chapter contains epidemiological overviews and proposals with regard to terminology and classification. It also describes the populations targeted by these guidelines. II.1.1. Environmental factors Exposure to sun Sunlight is the principal environmental factor, and evidence for its role in SCC relies on the appearance of lesions on the areas of skin most exposed to sun, a greater prevalence of lesions among fair-skinned subjects, a latitude gradient for populations with the same skin phototype and a higher incidence of the disease among patients working outdoors.12 The occurrence of SCC is associated with total cumulative lifetime UV dose. The most commonly affected sites include the face, back of the hands and forearms. In a Spanish study, more than 92% of SCC cases occurred in these areas.13 UVB (290–320 nm) and UVA (320–400 nm) play a role in carcinogenesis. For most SCC, UV-induced mutations are observed in the P53 gene.14 Artificial sources of UV have also been incriminated.15 PUVA therapy in excess of 200 sessions is associated with the onset of SCC. Sources of this type of radiation in tanning salons are not harmless and must be added to other risk factors.16 According to the 2005 report by the French Environmental Health Safety Agency (AFSSE), 'Evaluation of risks associated with exposure to UV radiation', the risk of skin cancer (carcinoma or melanoma) is increased by a factor 1.10 with 30 sessions per year over a 10-year period and by 1.39 with 100 sessions. Other exogenous risk factors include arsenic, pesticides, hydrocarbons, local factors The main factor is the skin which is The risk is higher in patients with a for In a a in that to is associated with a risk of SCC, AK and keratoacanthoma. Other risk factors The following factors are involved in than of and have been particularly in the genital and anal but also in in the of in areas that are both exposed and to and in and immunocompetent patients However, the presence of is to risk factors are in organ sun and Squamous cell carcinoma a number of malignant primary epithelial skin tumours with malpighian differentiation, and is distinct from BCC. The key factor onset of SCC, Bowen's disease (BD) or actinic keratosis (AK) is the total lifetime of UV whether or The risk of BD or SCC is affected by skin which is A number of which may be specific to particular have been such as for actinic and SCC of the as well as for genital or anal SCC. medical also to and keratosis of common lesions, particularly among subjects, in areas exposed to the sun such as the face, the back of the hands and the in subjects, and is associated with other of (e.g. etc.). is based on the clinical and is by lesions of more to the than to the with a of or with of and or In practice, AK or of AK are from multiple which come together to form the actinic is the of This particular is by the presence of as a second factor and by the of SCC in this is no histological of may be or of the The is to and is or An on with the and is a is based on the presence of various keratinocytic of by changes in the or In these in and do not the or the skin The term carcinoma in is only used when these together in a and the reference works of and The WG that these are simply of the same and, as do not classification as distinct These are not for France. The published prevalence for adults over between and in populations of the but between and in populations of the In of between and and of patients over have at one If AK may or to SCC. The of of such lesions over a period has been to be between and of A in this particular that a patient with AK has a risk of one of these AK into SCC within According to several from to of AK to SCC within of the genital form a group of AK to their and the role of in their A of gynaecological literature has the of subsequently and as well as cases observed are under the term or or according to whether they the or more than of the epithelial the is the term SCC or Bowen's disease is the between AK and lesions and that the of lesions and be in the same as their in the genital or using the term keratinocytic The WG felt that was little value in to criteria for entities with divergent and clinical It considered that were no for the clinical name of by the medical with a name that is clear to practitioners. may to grade the level of epithelial in their This to be were it not for its impact in terms of treatment and health etc.). of a that all forms of AK are in fact SCC, together with carcinoma, keratoacanthoma and other these entities are by the of mutations in However, a number of epidemiological studies have that the of AK may take three different or to SCC. The for AK is In the by Marks in the the over a period was than per The guidelines AK as precancerous However, the and BAD guidelines do not adopt a position concerning the nature of simply that most SCC to but that the risk of of AK to SCC is thus the choice of therapy for most AK described in to carcinomas of the a a and and mutations the of new primary tumours and of local the it has been that UV radiation is associated with the and of carcinogenesis. have been to be more common in areas than in have been at AK excision These clinical namely the of SCC and and only on and the of of AK in certain areas (e.g. In practice, these could treatment of the of the affected than treatment of However, no studies as the value of such an Bowen's disease Bowen's disease is an SCC. The prevalence and incidence of this disease in France and are most affected by the disease in published cases are in the of and are predominantly of the cutaneous as a which is generally or with a It is generally in areas of covered lesions In the and BD may be or and of the or forms are associated with The is with a and of and present at all by they do not the basal of BD with or have been and may be required to determine whether they are In the diagnosis with 3) is based primarily on of clinical medical presence of histological of and, in certain cases, identification of any associated This risk of has been in a and based on several studies. The risk to be between and for cutaneous and for and clinical presentation the appearance on the of an The risk to be greater than that of the common SCC AK is an of AK is associated with exposure to