Consensus-Based Recommendations on the Prevention of Squamous Cell Carcinoma in Solid Organ Transplant Recipients

Abstract

IMPORTANCE

There is a paucity of evidence to guide physicians regarding prevention strategies for cutaneous squamous cell carcinoma (CSCC) in solid organ transplant recipients (SOTRs).

OBJECTIVE

To examine the development and results of a Delphi process initiated to identify consensus-based medical management recommendations for prevention of CSCC in SOTRs.

EVIDENCE REVIEW

Dermatologists with more than 5 years’ experience treating SOTRs were invited to participate. A novel actinic damage and skin cancer index (AD-SCI), consisting of 6 ordinal stages corresponding to an increasing burden of actinic damage and CSCC, was used to guide survey design. Three sequential web-based surveys were administered from January 1, 2019, to December 31, 2020. Pursuant to Delphi principles, respondents thoroughly reviewed all peer responses between rounds. Supplemental questions were also asked to better understand panelists’ rationale for their responses.

FINDINGS

The Delphi panel comprised 48 dermatologists. Respondents represented 13 countries, with 27 (56%) from the US. Twenty-nine respondents (60%) were Mohs surgeons. Consensus was reached with 80% or higher concordance among respondents when presented with a statement, question, or management strategy pertaining to prevention of CSCC in SOTRs. A near-consensus category of 70% to less than 80% concordance was also defined. The AD-SCI stage–based recommendations were established if consensus or near-consensus was achieved. The panel was able to make recommendations for 5 of 6 AD-SCI stages. Key recommendations include the following: cryotherapy for scattered actinic keratosis (AK); field therapy for AK when grouped in 1 anatomical area, unless AKs are thick in which case field therapy and cryotherapy were recommended; combination lesion directed and field therapy with fluorouracil for field cancerized skin; and initiation of acitretin therapy and discussion of immunosuppression reduction or modification for patients who develop multiple skin cancers at a high rate (10 CSCCs per year) or develop high-risk CSCC (defined by a tumor with approximately ≥20% risk of nodal metastasis). No consensus recommendation was achieved for SOTRs with a first low risk CSCC.

CONCLUSIONS AND RELEVANCE

Physicians may consider implementation of panel recommendations for prevention of CSCC in SOTRs while awaiting high-level-of-evidence data. Additional clinical trials are needed in areas where consensus was not reached.

Subject to lifelong immunosuppression, solid organ transplant recipients (SOTRs) (especially those with white skin and a history of significant sun exposure) have a high risk of developing cutaneous squamous cell carcinoma (CSCC), with a risk ranging from approximately 20 to 200 times higher than in background populations, resulting in an increased risk of mortality from skin cancer.¹ Physicians seeking to prevent CSCC in SOTRs engage in primary, secondary, or tertiary strategies aimed at reducing skin cancer risk. Primary prevention aims to prevent the onset of disease, whereas secondary and tertiary prevention reduces the morbidity and mortality of a disease that has already occurred.^2,3 Primary prevention in SOTRs may range from patient education, photoprotection, and skin surveillance to treatment of premalignant lesions, such as use of topical^4–11 medications and photodynamic therapy (PDT).^12–17 Oral chemoprevention^18–22 and reduction^23,24 and conversion^25–27 of immunosuppression are options for secondary²⁸ and tertiary prevention.

Despite the advancing literature regarding prevention of CSCC inSOTRs,^15,21,25,29 uncertainty exists regarding best practices for various patient scenarios. Limitations in the current literature for prevention of CSCC in SOTRs include the low number of randomized clinical trials (RCTs) specifically performed in the immunocompromised population and use of surrogate end points. For example, the primary prevention end point of interest for physicians is the development of first CSCC; nevertheless, commonly used end points in the literature are the elimination of actinic keratosis (AK) or the development of subsequent CSCC in patients with a history of CSCC.^{6,8,10,12,16,30} Large, prospective RCTs in this population are challenging because of disease and patient heterogeneity and the latency to onset of CSCC, requiring extended follow-up times. In scenarios where there is high-level evidence regarding benefit,³¹ such as the case for field therapy for AK, treatments may have low adherence rates, thus limiting clinical applicability. Once evidence for a treatment modality is established in SOTRs, as in the case of acitretin for secondary chemoprevention,¹⁸ additional open clinical questions remain: which patients should be treated, when treatment should be initiated, and in what sequence different strategies should be implemented.

Previous guidelines on the management of CSCC in SOTRs were issued by the International Transplant Skin Cancer Collaborative in 2004.³² The emphasis of these guidelines is on management of existing disease rather than prevention, and their development did not make use of a structured consensus-building approach.

First described in 1948,³³ the Delphi method is a well-established tool to build consensus among experts when such uncertainties exist. It is a structured, iterative process whereby experts are provided several rounds of surveys of increasing specificity designed to encourage the convergence of opinion regarding a problem or question.³⁴ Transparent review of pooled peer responses by participants and opportunities to revise previous responses are core elements of involvement in the Delphi method.³⁵

Because of the paucity of evidence-based data in the prevention of CSCC in the SOTR population, we used the Delphi method among a panel of expert dermatologists to identify consensus-based preventive and treatment recommendations. In this article, we report the development and results of the Strategies for the Prevention of Skin Cancer in Solid Organ Transplant Recipients Delphi study.

Methods

At the September 2018 International Immunosuppression and Transplant Skin Cancer Collaborative meeting, an international group of transplant dermatologists convened to design a Delphi study that would identify consensus-based medical management decisions in the prevention of skin cancer in CSCC. Topics related to management of CSCC, such as excision, Mohs surgery, radiation, imaging, or patient follow-up were specifically excluded because these topics were deemed related to management of active disease and did not play a role in the primary, secondary, or tertiary prevention of CSCC. Prevention of melanoma or other non-CSCC skin cancer was similarly excluded. At this meeting, a key barrier identified by the working group was lack of widely accepted categories to quantify actinic damage and skin cancer burden, which would serve as the basis of the Delphi study. The working group developed a categorical actinic damage and skin cancer index (AD-SCI) that consisted of 6 stages, with each stage representing a common clinical scenario in which medical management of the patient may be altered to decrease the risk of further CSCC development (Table 1). The AD-SCI was also presented at the Skin Care in Organ Transplant Patients Europe 2019 meeting for input from its members. An expert panel was drawn from the membership of 3 international organizations with a focus on skin cancer in SOTRs: the International Immunosuppression and Transplant Skin Cancer Collaborative, the Keratinocyte Carcinoma Consortium Immunosuppression Working Group, and Skin Care in Organ Transplant Patients Europe. Prospective respondents were eligible based on 3 criteria: board certification in dermatology, at least 5 years of experience after residency in treatment of transplant patients, and evidence of active membership in 1 of the 3 aforementioned organizations as defined by review of meeting rosters. Informed consent was not required for this study because all responses were anonymous. This study was approved by the Partners Human Research Committee Institutional Review Board.

Table 1.

Actinic Damage and Skin Cancer Index

Stage	Description
1	No AK; photodamage only (lentigines, poikiloderma, rhytides)
2	Discrete AK
3	Diffuse AK with or without SCCis in a given field
4	First invasive low-risk CSCC
5	Multiple invasive low-risk CSCCsa
6	High-risk CSCCb

Abbreviations: AK, actinic keratosis; CSCC, cutaneous squamous cell carcinoma, SCCis, squamous cell carcinoma in situ.

^aDevelopment of 2 or more CSCCs.

^bAmerican Joint Committee on Cancer stage T3 or Brigham and Women’s Hospital stage T2B tumor.

Steering Committee

Several of us (P.R.M., A.J.P., C.A.H., and C.D.S.) formed the steering committee charged with guiding the Delphi process. The steering committee reviewed and interpreted data from each round, promptly disseminated results to the panel, and formulated subsequent surveys based on results from previous rounds.

Enrollment and Data Collection

Prospective respondents were sent an email that contained a link to a web-based Research Electronic Data Capture (REDCap) survey inviting them to participate. If respondents agreed, they proceeded to the first round of the Delphi study. All 3Delphi rounds were conducted over REDCap, and survey data were housed in REDCap for the duration of the study. Limited demographic information, including country of origin, affiliation, and subspecialty were collected as part of enrollment in round 1.

Consensus Building and Information Gathering

In accordance with the Delphi method,³⁶ the first round administered consisted of open-ended questions tailored to each AD-SCI stage of CSCC development. Multiple-choice format answers that comprised plausible management options identified by the steering committee were provided. Responses from round 1 were used to formulate rounds 2 and 3, whereby more directed questions were asked. The steering committee plan for consensus-building subsequent to the open-ended round 1 is shown in the Figure. Supplemental questions, aimed at better understanding respondents’ rationale for their choices (eg, why for some modalities, such as nicotinamide, respondents did not reach consensus despite randomized data supporting their use) were asked at the completion of each round; participants were unable to change their answer choices for that round after review of supplemental questions. Participants were also invited to offer open-ended feedback at the end of each Delphi round. All administered surveys can be found in the eAppendix in the Supplement.

Figure.

Strategy for Building Consensus in Strategies for the Prevention of Skin Cancer in Solid Organ Transplant Recipients

Definition of Consensus and Threshold for Panel Recommendation

The percentage agreement of a panel of experts in response to a query or statement is a common measure of consensus in Delphi processes.³⁷ A wide range of consensus thresholds are found in the literature, ranging from 51% to 97%; 75% is cited as a median value.³⁷ In this study, a threshold of 80% or higher was used to delineate consensus, evaluate the responses from each round, and develop subsequent rounds (Figure). A near-consensus category of 70% to less than 80% agreement was also defined. The final panel recommendations included any management strategy that achieved consensus or near-consensus (70%–100%). If at any time in the survey a management strategy was selected by fewer than 10% of panelists, it was not included as an option in subsequent rounds.

Definition of Low- and High-Risk CSCC and Rate of CSCC Formation

For the purposes of the Delphi study, a low-risk CSCC was defined as an American Joint Committee on Cancer (AJCC)³⁸ or Brigham and Women’s Hospital (BWH)³⁹ T1 tumor. A high-risk CSCC was defined as an AJCC T3 or BWH T2b tumor (because higher stages, AJCC T4 and BWH T3, are rare).With respect to patients diagnosed with multiple CSCCs, a low rate of formation was defined as 1 dermally invasive CSCC diagnosed per year, whereas a high rate of formation was defined as 10 dermally invasive CSCCs diagnosed per year.

Statistical Analysis

Descriptive statistics were performed for each question on completion of each round. Results from preceding rounds were emailed to the panel for review at least 1 week before the subsequent round. All statistical analysis was performed using Excel, version 16.42 (Microsoft Corp).

Results

Three Delphi rounds were completed as planned. Seventy-four prospective expert dermatologists met the predetermined inclusion criteria and were invited participate. Of these, 50 (68%) responded to round 1. Subsequently, 48 respondents completed round 2 and 46 completed round 3. Forty-eight of the initial 50 respondents (96%) participated fully in 2 of 3 rounds and were considered final members of the panel. Baseline demographic characteristics of the final panelists are given in Table 2. Thirteen countries were represented on the panel, with 27 panelists (56%) located in the US. Twenty-nine (60%) were Mohs surgeons.

Table 2.

Panel DemographicCharacteristics^a

Characteristic	No. (%) of panelists (N = 48)
Length of experience, y
5–<10	18(38)
10–<15	11 (23)
15–≤20	9(19)
>20	10(21)
Mohs surgeon	29 (60)
Medical dermatologist/cutaneous oncologist	19(40)
Country of practice
US	27 (56)
Europe	18 (38)
UK	4(8)
Spain	4(8)
The Netherlands	3 (6)
Belgium	1 (2)
Turkey	1 (2)
Switzerland	1 (2)
Norway	1 (2)
Italy	1 (2)
France	1 (2)
Austria	1 (2)
Brazil	1 (2)
Australia	2 (4)

^aReflected are panelists who completed more than 2 of 3 Delphi survey rounds.

The Delphi panel was able to make specific management recommendations across 5 of the 6 AD-SCI stages (Table 3). Consensus was not achieved, and no management recommendation was made in the scenario in which an SOTR develops their first low-risk CSCC.

Table 3.

Consensus-Based Medical Management Recommendations for the Prevention of CSCC in Solid Organ Transplant Recipients by AD-SCI Stage^a

Group	Recommendation
Stage 1
Photodamaged skin only	Preventive measures: education, sun protection strategies, sunscreen, and/or skin surveillance (C)b
Stage 2
Discrete AKs
Scattered AKs (2a)	Thin: cryotherapy should be used as first-line treatment (C); if this fails, cryotherapy should be repeated (NC); oral chemoprevention should not be initiated (C)c Thick: cryotherapy should be used as first-line treatment (C); if this fails, lesion directed therapy should be repeated (C); oral chemoprevention should not be initiated (NC)
Grouped AKs (2b)	Thin: field therapy should be used as first-line treatment (NC); oral chemoprevention should not be initiated (C)b Thick: lesion-directed therapy (C) using cryotherapy (NC) followed by field therapy (C); a fluorouracil-based modality (NC) should be used as first-line treatment for field therapy; oral chemoprevention should not be initiated (NC)b
Stage 3
Field cancerization	Lesion-directed therapy (C) followed by field therapy (C) using a fluorouracil-based modality (NC) should be used as first-line treatment; speaking with transplant team regarding immunosuppression should not be initiated (C)b
Stage 4
First invasive low-risk CSCCd	No consensus achieved
Stage 5
Multiple invasive low-risk CSCCsd
Low rate (1 CSCC per year) (5a)	Oralchemoprevention should be initiated (NC)
High rate (10 CSCCs per year) (5b)	Oralchemoprevention with acitretin should be initiated (C); speak with transplant team regarding modification of immunosuppression (C)
Stage 6
High-risk CSCCe
As first CSCC (6a)	Oral chemoprevention with acitretin should be initiated (NC); speak with transplant team regarding modification of immunosuppression (NC)
After multiple low-risk CSCCs (6b)	Oral chemoprevention with acitretin should be initiated (C); speak with transplant team regarding modification of immunosuppression (C)

Abbreviations: AD-SCI, Actinic Damage and Skin Cancer Index; AK, actinic keratosis; C, consensus; CSCC, cutaneous squamous cell carcinoma; NC, near consensus.

^aRecommendations made if near consensus (70 to <80%) or consensus (≥80%) was achieved.

^bPreventive measures were recommended for all AD-SCI stages but are only listed for stage 1 because of table space constraints.

^cNegative consensus or negative near consensus indicates panelists responded negatively when asked about this intervention.

^dAmerican Joint Committee on Cancer or Brigham andWomen’s Hospital stage T1 tumor.

^eAmerican Joint Committee on Cancer T3 or Brigham andWomen’s Hospital stage T2B tumor.

Supplemental Information

When asked about the association of the various immunosuppressive medications with skin cancer formation, 42 of 48 respondents (88%) reported that azathioprine was the most associated with skin cancer formation, whereas 33 of 48 respondents (69%) thought that sirolimus was the least associated. A total of 40 of 46 respondents (87%) identified fluorouracil-based therapy as the most effective field agent, but only 10 of 46 (22%) reported that it resulted in the best adherence. Only 2 of 46 respondents (4%) reported that PDT was the most effective field agent, but 34 (74%) reported that its use results in the best adherence. Compliance concerns were identified as the major barrier (78%) to broader use of fluorouracil-based therapy despite RCT data demonstrating its benefit in immunocompetent patients.³¹ A total of 33 respondents (72%) reported that lack of data specific to SOTRs precluded widespread recommendation of nicotinamide for CSCC prophylaxis, despite RCT data⁴⁰ demonstrating its benefit in immunocompetent patients and its safety in renal transplant patients.²¹ A total of 30 respondents (65%) reported that adverse effects limited widespread implementation of early mammalian target of rapamycin (mTOR) conversion in SOTRs despite RCT data^25,41 demonstrating its benefit in reducing CSCC risk in this population.

Discussion

The Strategies for the Prevention of Skin Cancer in Solid Organ Transplant Recipients Delphi process was able to provide consensus-based management recommendations regarding optimal prevention strategies of CSCC in SOTRs across 5 of the 6 AD-SCI stages. Areas of consensus may aid physicians in establishing best practices regarding prevention of CSCC in SOTRs in the setting of limited high level of evidence data in this population. The panel recommended routine skin surveillance and sunscreen use for all patients. No studies on the impact of routine skin surveillance in SOTRs have been performed to date, but targeted screening in high-risk populations may be effective in melanoma⁴² and is in keeping with prior guidelines³² in transplant recipients. The panel’s recommendation for regular sunscreen use in SOTRs is in accordance with literature providing some evidence for a reduction in the incidence of AK and CSCC in this population with regular use.²⁹

Although lesion-directed therapy with cryotherapy was favored for scattered AK, scenarios that involve anatomically grouped AK or field cancerized skin led to a recommendation for the initiation of field therapy with fluorouracil. However, the recommendation for fluorouracil was based on near but not full consensus of the panel. Indeed, no particular field agent reached the 80% or greater consensus threshold at this AD-SCI stage, reflecting a degree of uncertainty in the panel on this point. A prior review by Blomberg et al⁴³ similarly identified uncertainty regarding the optimal field agent for management of AK and prevention of CSCC in SOTRs as a knowledge gap in the literature. This finding is in contrast to recent studies in immunocompetent patients, demonstrating that field therapy with fluorouracil is superior to PDT, ingenol mebutate, and imiquimod in a large, double-blind RCT for treatment of AK,³¹ reduces the need for lesion-directed therapy,⁴⁴ and prevents CSCC.⁴⁵ Although similar large RCTs are lacking for SOTR specifically, small split-patient studies in SOTRs found that PDT outperformed fluorouracil³⁰ and imiquimod¹⁶ in the treatment of AK. When the panel was presented with data supporting the use of fluorouracil-based therapy in immunocompetent patients (see Supplemental Information section), 40 of 46 respondents (87%) reported that they believed fluorouracil was the most effective field agent, but adherence concerns limited broader use (reported by 36 respondents [78%]). Acitretin was the sole oral chemoprevention agent recommended by our panel, put forth in scenarios in which SOTRs were developing CSCC at a high rate or had developed a high-risk CSCC. This recommendation is supported by RCTs in renal transplant recipients.^18,19 The preferential use of acitretin in SOTRs with advanced CSCC disease as seen in our panel is somewhat discrepant with the existing literature, in which a benefit to acitretin was reported in SOTRs who had a history of, on average, less than 1 nonmelanoma skin cancer per year diagnosed during 5 years and as little as 1 nonmelanoma skin cancer in the prior 5 years.¹⁹ Furthermore, even low-rate CSCC formation portends poor outcomes. In a previous analysis,⁴⁶ development of 10 or more CSCCs during a 10-year period (approximately 1 per year) in a group of mostly immunosuppressed patients was associated with a 26% risk of nodal metastasis, suggesting that early and aggressive preventive intervention is indicated in this population. To date, no RCTs have compared the benefit of acitretin in SOTRs with differential (high vs low) rates of CSCC formation.

Finally, the panel recommended initiating discussions with transplant physicians regarding immunosuppression modification in patients with advanced CSCC (AD-SCI stages 5 and 6) disease but otherwise did not make a recommendation as to the best immunosuppression modification strategy to pursue. Specifically, the panel did not recommend discussing reduction of immunosuppression or conversion to mTOR inhibition, 2 evidence-based^24,25,41 mechanisms of secondary and tertiary prevention of CSCC in SOTRs, with transplant physicians. In these scenarios, 58% to 67%, depending on the specific query, of the panel preferred to defer this decision to transplant physicians. Although the absence of such a recommendation runs contrary to previous expert statements^32,47,48 and some RCTs have demonstrated that reduction of immunosuppression²⁴ or conversion to mTOR-based immunosuppression^25,41 may decrease CSCC formation, this discrepancy is instructive regarding the balance expert physicians must use in management of SOTRs: patients with reduced immunosuppression are at risk for inferior graft survival,⁴⁹ and conversion to mTOR may affect overall survival.⁵⁰

The panel did not reach any consensus management recommendation for prevention for an SOTR who develops a first low-risk CSCC. Notably, panel recommendations in this scenario were not affected when asked about the specific organ transplanted (ie, abdominal vs thoracic). As discussed above, the panel did not incorporate the results of the RCT by Euvrard et al⁴¹ demonstrating benefit of early conversion to sirolimus in renal transplant recipients (most patients in this study were enrolled after a first CSCC) into its recommendations. Sixty-five percent of respondents reported that widespread use of mTOR conversion in SOTRs is limited by adverse effects. The lack of consensus regarding management for SOTRs who develop a first low-risk CSCC reflects clinical equipoise and should be the subject of further investigation.

The panel did not make a recommendation for use of nicotinamide or capecitabine in any of the stages presented. The absence of a recommendation for nicotinamide is notable in the context of a double-blind RCT in immunocompetent patients⁴⁰ demonstrating benefit in prevention of AKs and CSCCs. A smaller RCT²¹ in 22 SOTRs did not find a significant benefit from nicotinamide but was underpowered to do so because the study was terminated early because of low enrollment. However, the study demonstrated safety, with no major adverse effects noted. In supplemental questioning, 72% of the panel reported that the lack of efficacy data specifically for SOTR limited their use of nicotinamide. Given the low cost, high safety, and demonstration of CSCC reduction in non-SOTRs, nicotinamide administration may be an area for further consideration and expanded study. Efficacy of capecitabine has been reported in case series in SOTRs,²² but it has not been studied in a randomized manner for chemoprevention. More than half of the panel (52%) reported that they did not have routine access to capecitabine in their practice.

Limitations

Our study has some limitations, including those intrinsic to the Delphi process. Our chosen criteria for consensus were strict (≥80% agreement), and thus more areas of consensus may have been reported with a lower standard. Distinctions were drawn to distinguish patient scenarios, such as the difference between forming low-risk CSCC at a low rate (1 CSCC per year) vs high rate (10 CSCCs per year), which may leave ambiguity, for example, in how an SOTR with multiple low-risk CSCCs per year above the low but below the high threshold should be best managed. Similarly, we explored the association of a first low-risk and high-risk CSCC with prevention strategies but did not inquire as to how an intermediate-risk CSCC (AJCC T2 or BWH T2a) would affect management. The AD-SCI is based on expert opinion and is not a validated instrument. Finally, because the panel comprised an international group of transplant dermatologists, the availability of certain prevention strategies varied, which may have affected survey responses.

Conclusions

This Delphi process resulted in recommendations for management strategies in the prevention of CSCC in SOTRs in 5 of the 6 AD-SCI stages and included cryotherapy for scattered AKs, field therapy for AKs in 1 anatomical area (augmented by cryotherapy if thick), combination lesion directed to hyperkeratotic lesions followed by field therapy with fluorouracil for field cancerized skin, and initiation of acitretin treatment and discussion of immunosuppression reduction or modification for patients developing 10 CSCCs per year or a tumor with 20% or greater risk of nodal metastasis. These recommendations reflect consensus among expert transplant dermatologists and the incorporation of limited and sometimes contradictory evidence into real-world clinical experience across a range of CSCC disease severity. No consensus was reached regarding management of SOTRs who have a first low-risk CSCC; future studies and clinical trials aimed at assessing optimal management in this clinical scenario are much needed, and involvement of transplant medicine colleagues should be considered in such investigations. These recommendations will assist physicians in implementing prevention strategies for management of CSCC in SOTRs while awaiting high level-of-evidence data to guide best practices.

Key Points.

Question

What are the recommended medical interventions for skin cancer prevention in solid organ transplant recipients?

Findings

On the basis of the results of the Delphi study, cryotherapy is recommended as first-line therapy for actinic keratosis, with the exception of thin actinic keratoses grouped in 1 area. In that scenario and for field cancerization, field therapy should be performed.

Meaning

Oral chemoprevention and discussion with the transplant team regarding modification of immunosuppression should be initiated when the patient develops multiple low-risk cutaneous squamous cell carcinoma (>10 tumors per year) or a high-risk cutaneous squamous cell carcinoma.