Methods and Results Report - Evidence and consensus-based (S3) Guidelines for the Treatment of Actinic Keratosis -International League of Dermatological Societies in cooperation with the European Dermatology Forum.

Abstract

Nast/Werner The following sections represent the methods report of the Evidence and consensus-based (S3) Guidelines for the Treatment of Actinic Keratosis – International League of Dermatological Societies (ILDS) in cooperation with the European Dermatology Forum (EDF). Detailed results of the guidelines development are available in the long version and in the results report of the guidelines, both available online. For clinical guidance on the clinical background, assessment and treatment of actinic keratosis (AK), please consider the long version or the original guidelines publication. Please use the following reference when citing this document: Werner RN, Jacobs A, Rosumeck S, Erdmann R, Sporbeck B, Nast A. Methods and Results Report – Evidence and consensusbased (S3) Guidelines for the Treatment of Actinic Keratosis – International League of Dermatological Societies in cooperation with the European Dermatology Forum. J Eur Acad Dermatol Venereol 2015; 29, e1-e66. These guidelines encompass different clinical aspects related to AK. The primary goal of the guidelines was the development of treatment recommendations appropriate for different subgroups of patients presenting with AK. This was subject to a systematic literature review and a formalized consensus conference of the members of the guidelines’ expert panel. A secondary aim of these guidelines is the implementation of knowledge relating to the clinical background of AK, including recommendations for the histopathological definition of the disease and for the diagnosis and assessment of patients presenting with AK. Clinical background texts were written by the steering group and subgroups of the expert panel, based on a narrative literature review. Some of these aspects (diagnosis, histopathology, assessment of patients with AK) were formally consented as recommendations during the consensus conference. The guidelines were elaborated along adapted recommendations by the WHO guidelines review committee1 and the quality criteria for guidelines as suggested by the Appraisal of Guidelines Research and Evaluation (AGREE II) Instrument2 were incorporated into the methodological development of the guidelines. For the planning and elaboration of the underlying systematic literature review on interventions for AK, the methodology suggested by the Cochrane Handbook for Systematic Reviews of Interventions3, the GRADE working group4 and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement5 was adapted. These evidence- and consensus-based guidelines contain recommendations that were developed to assist clinicians in the care of patients in specific clinical conditions. The recommendations are based on the best available evidence and their development followed a pre-specified, standardized process. Nevertheless, guidelines do not replace the clinicians’ knowledge and skills, since guidelines never encompass therapy specifications for all medical decision-making situations. Guidelines should not be deemed inclusive of all proper methods of care nor exclusive of other methods of care reasonably directed to obtaining the same results. Deviation from the recommendations may be justified or inevitable in specific situations. The ultimate judgment regarding patient care must be individualized and must be made by the physician and patient in the light of all presenting circumstances. Safety aspects that were considered within these guidelines do not represent a comprehensive assessment of all available safety information for the included interventions. They are limited to those aspects chosen for evaluation and the information available in the included clinical trials. Readers must carefully check the information in these guidelines and determine whether the recommendations (e.g. regarding dose, dosing regimens, contraindications, or drug interactions) are complete, correct, up-to-date and appropriate. International guidelines are intended to be adapted to national or regional circumstances (regulatory approval and availability of treatments, health care provider and insurance systems). Particularly, the mode of application of the different treatment options has to be adapted to national approval of the interventions. Thus, the national medical societies associated with the International League of Dermatological Societies (ILDS) will be responsible for the adoption and implementation of the guidelines on a national level. The provision of recommendations that are based on a systematic review of the external evidence and consented by clinical experts during a structured and formalized process aims at improving the medical care of patients presenting with AK. The choice of an adequate evidence-based treatment strategy – adapted to the individual demands – will be facilitated by the provision of recommendations that take into account frequent clinical scenarios. The description of the clinical background, histopathological features and assessment of AK intends to raise awareness of the treatment necessity in a broader range of medical specialties and advance concepts of AK towards a more widely accepted definition. The use of lesion- and field-directed interventions should be optimized by using the most appropriate treatment regarding the extent and type of AK. Along with a clearance of AK lesions and prevention of their recurrence, the provision of evidence-based treatment algorithms intends to decrease the percentage of patients with progression from AK to invasive squamous cell carcinoma (SCC). Adherence to the therapeutic regimen is a basic element for the treatment success. Knowledge on the suggested interventions, including expectable effects, adverse effects, duration and possible alternatives is indispensable in the communication with patients. These evidence-based guidelines can help patients to make informed decisions and, consequently, improve the patient compliance to their therapeutic regimen. The primary goal of these guidelines is to assist health care professionals in the choice of the optimal treatment strategy for their patients with consideration of the severity of the disease and the specific circumstances of the individual patient. Target groups include all health care professionals involved in the assessment and treatment of patients with AK, primarily dermatologists, histopathologists and general practitioners (GP). Due to the international focus of these guidelines and different organizational structures of health care services in different countries, target groups may vary correspondingly. Patients who have AK are mainly adult patients, often of advanced age, and treated in outpatient settings. To take frequent clinical situations into account, different patient subgroups were defined, according to the severity of the disease and the medical history of the patients. The primary focus of these guidelines is the assessment and therapy of patients presenting with single AK lesions, multiple lesions or field cancerization. Patients with concomitant immunosuppression are included as a target group requiring a differential therapeutic approach. There might be significant variability from country to country, not only in regulatory approval and the availability of interventions, but also in terms of health care providers and insurance systems. Thus, these international guidelines are intended to be adapted to the national or regional conditions. Pharmacoeconomic considerations were therefore not considered as part of the reasoning behind the recommendations concerning interventions. These aspects and possible prioritization of certain interventions should be considered when these guidelines are adapted for implementation at a national level. Werner The steering group of the guidelines project was composed by experts in the field of guidelines development. It consisted of members of the Division of Evidence-based Medicine (dEBM) from the Department of Dermatology, Venerology and Allergology, Charité – Universitätsmedizin, Berlin, Germany. The group assisted the guidelines development process with organization of the guidelines process, development of methodology and the conduction of a systematic review of the literature on interventions for AK. Members of the steering group participated in the consensus conference, but were not entitled to vote on recommendations. Members of the expert panel were dermatologists and histopathologists. They were officially nominated by the International League of Dermatological Societies (ILDS). The expert panel members were selected by virtue of their clinical experience and/or research expertise in the field of keratinocytic skin lesions. Participation of general practitioners (GP) was highly desirable and the World Organization of National Colleges, Academies and Academic Associations of General Practitioners/Family Physicians (WONCA) was officially requested to nominate GP members for participation in the expert panel. Unfortunately, no official GP nominations were received. The external review was performed also following the European Dermatology Forum (EDF) Guidelines SOPs. Final approval included ILDS, EDF, (European Academy of Dermatology and Venereology (EADV) and European Union of Medical Specialists (UEMS). Further details of the external review process are described below. An international patient organization to nominate a representative for patients affected by AK could not be identified, and thus patient participation was difficult to realize. Various attempts to include the patient perspective into the guidelines were made: One patient from the Charité – Universitätsmedzin Berlin, Berlin, Germany with large personal experience with different AK treatments was invited to participate in the expert panel. Patient reported outcomes such as Participants’ satisfaction and Participants’ preference were considered as an important outcome and studies reporting on these endpoints were included into the systematic literature review. Patients were invited to take part in the external review and to comment the drafted guidelines document. The expert panel was responsible for the selection of relevant patient subgroups, interventions and outcomes. During the consensus conference, experts were responsible for the appraisal and interpretation of the external evidence supplied by the steering group, considering the overall balance of the benefits and harms of interventions and their clinical expertise. No financial incentives or reimbursement for the participation in the expert panel were administered. A full list of the guidelines steering group and expert panel members is supplied at the beginning of the document. The guidelines project has kindly been supported by the European Skin Cancer Foundation (ESCF). The financial support did not influence the guidelines development. Assessment and synthesis of the evidence were done independently from industrial interest. Key questions to be answered and outcomes were chosen in accordance to consensus of the members from the expert panel. Recommendations on diagnostic means and interventions for the management of AK were exclusively based on the consensus of the members from the expert panel in the consensus conference, according to the clinical expertise and external evidence (systematic literature review of the available data on interventions for AK). A declaration of conflicts of interest (COI) was required for the participation in the guidelines development. The form used to assess the individual interests is presented in the appendix of this document (see chapter 7.1). At the beginning of the formalized consensus conference on the interventions for AK, each member was offered to update his or her declaration. COI were discussed and one member decided to abstain from voting on recommendations concerning methyl-aminolevulinic acid photodynamic therapy (MAL-PDT) due to conflicting interests. The expert panel did not see any substantial conflicts of interest and there were no further comments or remarks. COI of each person involved in the guidelines development are presented in the appendix of the results report (see chapter 8.1). The selection of key questions to be answered by guidelines depends on the definition of subgroups of patients, the selected interventions and their comparators, and finally on the outcomes to be considered. The respective decisional steps for the preparation of the systematic literature review were performed via electronic mail contacts and consented in an online kick-off conference with the members of the expert panel. A consensus of ≥75% of the members of the expert panel served as relevant cut-off for the confirmation of each decided aspect and its inclusion in the systematic literature review and in the formalized consensus conference. Different subgroups of patients presenting with AK, requiring differential therapeutic approaches were defined in order to address the demands of clinical practice. The definitions were based on suggestions of the steering group and clinical expertise of the expert panel members. The defined categories served as basis for separate assessment of the interventions during the systematic literature review and the formalized consensus conference. For details on the chosen subgroups of patients see chapter 3. Multiple lesion- and field-directed interventions are available for the treatment of AK. The options are further extended by the availability of different formulations and treatment schemes. For the selection of the relevant interventions to be included in the guideline, all members of the expert panel were consulted. Interventions could be chosen from a list supplied by the steering group or be proposed by each member of the expert group. The fact that certain interventions were not included does not necessarily imply that it may not be an appropriate treatment for AK. For details on the interventions selected for evaluation see chapter 4. The evaluation of the interventions was based on efficacy, cosmetic, patient reported and safety outcomes. Expert panel members were asked to rate outcomes with respect to their relevance for clinical decisions concerning the choice of treatment of AK. Rating was performed on a scale from 1 to 9 with 1 representing irrelevant and 9 representing critical outcomes. Mean values of the ratings from the experts served to rank the importance of the selected outcomes when grading the available evidence. A mean score of 7–9 rated an outcome as critical for a decision, 4–6 rated an outcome as important but not critical for decision-making, and a mean score of 1–3 indicated that the respective outcome was of limited importance6. The selection of outcomes to be considered was additionally based on the availability of reported outcomes in the available evidence. For details on the chosen outcomes and their rating see chapter 5. A systematic search for existing guidelines and systematic reviews on Interventions for AK in Medline, Embase, the Cochrane Library, the Guidelines International Network (G-I-N) database, and National guidelines clearinghouse was conducted at the beginning of the project. Relevant hits were evaluated independently by two assessors (SR, RNW) using the Appraisal of Guidelines for Research and Evaluation II (AGREE II) tool2 or the SIGN Methodology Checklist 1: Systematic Reviews and Meta-analyses7. A relevant and recent high quality systematic review was identified8. More details are presented in the results part (see results report of the guidelines). The identified Cochrane Review was used as basis of the body of evidence. A systematic literature search was performed to update the included Cochrane review using the databases Cochrane Library, Medline, Medline in Process and Embase, and covered the periods from March 2011 through the date of the search (January 25th, 2013). The search strategies corresponded to the strategies used in the Cochrane review8. Detailed electronic search strategies for the different databases are presented in the appendix (see chapter 7.2). Titles and abstracts of the update search were individually checked for eligibility by two independent assessors (RNW, BS). Full texts of potentially relevant studies were similarly checked for eligibility by two independent assessors (RNW, AJ). In the case of disagreement during the screening of abstracts and full texts, a third assessor (AN) was involved and the conflict solved by discussion. Criteria for the eligibility of studies for inclusion in the systematic review were similar to those of the Cochrane review on interventions for actinic keratosis.8 Eligible studies for inclusion were RCTs (including parallel group and intra-individual designs as well as crossover trials) reporting on participants with a clinical or histological diagnosis of at least one AK lesion at baseline. Randomization had to refer to participants or to body parts of participants (e.g. left vs. right side), not to individual AK lesions. Publication language was not restricted. Studies reporting on participants with a particular predisposition for developing sun exposure-related skin lesions (e. g. Xeroderma pigmentosum, Albinism) were excluded. Additional criteria were defined by the expert panel concerning the selection of interventions and the selection of outcomes to be considered. For a list of the selected interventions and outcomes please see chapters 4 and 5. Data collection of the update search results was done independently by two assessors (RNW, AJ), using a standardized data extraction form (Microsoft® Excel worksheet). The original Review Manager9 file from the Cochrane review8 was kindly made available by the authors. This file was updated along the selected eligibility criteria and the update search by two independent assessors (AJ, SR/RNW). Discrepancies of the extracted data were reviewed and discussed. Included studies were categorized according to the AK severity in the participants at baseline. As there is no pre-existing, widely accepted method for classification of AK severity, the subgroups of patients as defined by the expert panel were used. The studies were categorized on the basis of the inclusion criteria of each individual trial. If disease severity as inclusion criterion was not reported or if the inclusion criteria of the trial overlapped the defined categories of patients, studies were classified in accordance to the mean AK lesion counts and standard deviation at baseline. If studies could not be classified into a singular category, the data were taken into account for both respective patient subgroups and GRADE quality ratings with respect to directness were adapted. The available evidence and its quality were summarized according to the system recommended by the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) working group4 for each available outcome in each comparison. Using the GRADE profiler10, GRADE evidence profiles were developed for each available comparison of interventions, based on the rated outcomes (see chapters ‎2.3.1 and ‎5). The quality of the evidence for each key question was categorized into one of four categories, from ‘very low’ to ‘high’.11 Table 1 summarizes the different quality levels of evidence and the approach used to grade the quality of evidence as suggested by the GRADE working group.11 1. Limitations to study quality 2. Inconsistency 3. Indirectness 4. Imprecision 5. Reporting bias 1. Large effect 2. Dose–response 3. All plausible confounding would have reduced the demonstrated effect The following criteria, as defined by the GRADE working group were applied to decrease or increase the quality ratings for each key question, intervention and outcome: Limitations to the study quality: The Cochrane risk of bias tool3 was used to assess limitations to the study quality on a study level. The following domains were assessed: random sequence generation, allocation concealment, incomplete outcome data, selective reporting, blinding of participants and personnel, blinding of outcome assessment and other sources of bias. Overall study quality depended on the limitations of the contributing studies. A downgrading of 1 (‘serious limitations’) or 2 points (‘very serious limitations’) was possible.12 Inconsistency: Overall quality of evidence was downgraded by 1 point (‘important inconsistency’), when the study results were heterogeneous with respect to the direction or the size of the effect. The main criteria for downgrading were: widely varying point estimates across the studies, minimal or no overlap of the confidence intervals (CI), large I² (I² is a statistical test quantifying the variation in the point estimate between the studies).13 Inconsistency could not be assessed in case of only one contributing study. Indirectness: When differences between the effect size in the populations recruited for the study participation and the patient subgroup to make a recommendation for were expected (due to significant and important differences in the studied populations to the target population), overall study quality was downgraded by 1 (‘some’) or 2 points (‘major uncertainty about the directness’).14 Here, study quality was downgraded, when the study inclusion criteria or the patient characteristics at baseline did not match exclusively one of the predefined patient subgroups. Imprecision: The main criterion for determining the precision of the pooled effect size is the width and position of the 95% confidence interval (CI)15: the overall study quality was downgraded for imprecision if the CI was very wide (range of >100), crossed the threshold of minimal important difference (defined as the line of no effect ±0.25) or if the CI crossed the line of no effect and the threshold of minimal important difference. For continuous outcomes such as the mean reduction in AK lesion counts, the minimal important difference was calculated as the line of no effect ±0.5*SD of the control group. Publication bias: When publication bias was expected to influence the size or direction of the effect, study quality was downgraded by 1 point16. Due to the low number of contributing trials for each comparison, no formal testing (e.g. visual characterization of funnel plots) could be performed. Large effect/evidence of a dose–response gradient/confounders that would have decreased the effect: Rating up the quality of evidence due to the mentioned reasons is generally recommended only to be applied to results from observational studies or nonrandomized trials17. As the systematic literature search was restricted to randomized controlled trials, no upgrading of the overall study quality was performed. The quality of the evidence was evaluated by two assessors (AJ, SR) after discussion of each aspect. In case of dissent of the assessors, a third assessor (RNW) was involved and the conflict solved as a majority decision. Comments to justify the ratings are supplied in case of downgrading. For each intervention or comparison of interventions, a short text summarizing the available evidence and a GRADE summary of findings table is presented (see results report and long version of these guidelines). The summary of findings (SoF) tables encompass a detailed summary of the findings and their interpretation18. Data are presented as risk ratios (dichotomous outcomes)19 or mean differences (continuous outcomes)20. The risk ratio (RR) refers to the relative risk of an event occurring in the interventional group compared with the control group. For continuous data (e.g. the mean reduction in AK lesions counts), the mean difference relative to the control group is presented. All recommendations were consented during the consensus conference, moderated by Alexander Nast, MD, head of the steering group and certified moderator for the German Association of Scientific Medical Societies (AWMF). Formal consensus methodology (nominal group technique) was used to agree upon the recommendations21. All expert panel members without critical conflicts of interest were entitled to vote on the recommendations. The consensus conference was performed as an online consensus conference, using a regular telephone conference for the sound and the online platform Adobe® Connect™ for the presentation of the evidence data from the systematic literature review and voting on recommendations. The results from the systematic literature review (summary of findings tables and textual summaries) were supplied to the members of the expert panel prior to the consensus conference. During the consensus conference, the results of the systematic literature review were presented for each intervention prior to the discussion and voting on the recommendation for the respective intervention. When evaluating the evidence, the balance of benefits and harms, considering the predefined ranking of the importance of the outcomes, and the quality of the evidence were taken into consideration. Besides the evidence from the systematic review of the literature, expert opinion and experience was included, particularly if the body of evidence was insufficient and if further aspects such as time and costs, additional side-effects, quality of life, resource use, etc. had to be considered. Additional reasoning was required to be discussed and explicitly stated in the case of aberration from the external evidence. To simplify the identification of consented recommendations, all consented recommendations are highlighted throughout the guidelines documents (tables). In order to avoid ambiguity, a standardized language was used to classify the direction and strength of each recommendation. Based on the GRADE approach, five strengths of recommendations were differentiated: strong recommendations for or against the use of an intervention, weak recommendations for or against the use of an intervention, and no recommendation.22 The strength is expressed by the wording and symbols as shown in Table 2. The strength of a recommendation had to be based on the quality of the evidence as shown above (high/moderate/low/very low) and the balance of expected undesirable and desirable outcomes.23 If expert opinion without external evidence was incorporated into the reasoning for making a certain recommendation, the rationale was provided. For each recommendation, the quality of consensus in terms of percentage of agreement was measured and documented. Three levels of consensus were defined and distinguished. A ‘strong consensus’ (agreement of at least 90% of the expert panel members participating in the conference) was generally aimed at. In cases where only lower values of agreement were achieved, these were defined as ‘consensus’ (75–89% agreement) or ‘weak consensus’ (50–74% agreement). Before publication, the guidelines draft underwent an extensive internal and external review. Internal review was accomplished at the beginning of the guidelines development to confirm the selection of key questions (kick-off conference), prior to the consensus conference for a preliminary review of the results from the systematic literature review, after the consensus conference to confirm the completed recommendations, and after the external review to confirm changes before publication. The external review took place from 24th of March through 5th of May 2014. All ILDS member societies, the European Dermatology Forum (EDF), European Union Of Medical Specialists (UEMS), European Academy of Dermatology and Venereology (EADV), and European Association of Dermato-Oncology (EADO) were officially invited. Furthermore, the Skin Cancer Foundation, American Cancer Society and European Skin Cancer Foundation were invited to participate in the external review. The review took place using an open-access Internet platform (www.crocodoc.com), and comments could directly be integrated in the guidelines documents. The comment function was open to every interested individual. In total, 103 comments were posted on the online platform (38 on the short version of the guidelines and 65 on the long version of the guidelines). We received nine additional letters from different institutions. Each comment was assessed individually and categorized according to the required consequences. A document summarizing all comments, individual responses and their handling is available at the Division of Evidence based Medicine (Charité – Universitätsmedizin Berlin, Berlin, Germany). The guidelines were approved by the ILDS, the EDF, the EADV and the UEMS. During the phase of external review, the members of the expert panel piloted the drafted guidelines within their own practices and were encouraged to comment on the practicability and results during the second internal review. International guidelines are intended to be adapted to the national circumstances of each health system. Therefore, a formalized piloting of the recommendations will have to take place in each country and the national societies are responsible for the planning, realization, and evaluation of piloting projects. International guidelines are intended to be adapted to national or regional circumstances (regulatory approval and availability of treatments, health care provider and insurance systems). Thus, the national medical societies associated to the ILDS will be responsible for the adaption and implementation of the guidelines on a national level. In order to assist implementation, additional material such as a short version of the guidelines will be supplied. The original guidelines publication and a long version of the guidelines, this methods report and the results report including detailed data on the methodology and results will be published online. Evaluation strategies with respect to the awareness of the treatment necessity amongst patients and physicians, the treatment adhesion and treatment succ