Catherine A. Harwood1, Jane M. McGregor1 and Charlotte M. Proby2
1Department of Dermatology, The Royal London Hospital, London; Centre for Cutaneous Research, Blizard Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK
2Skin Tumour Laboratory, Division of Cancer Research, Medical Research Institute, Jacqui Wood Cancer Centre, Dundee, UK
The immune system plays a critical role in skin cancer development, progression and destruction. Skin cancers in individuals with compromised immune systems represent a growing challenge in terms of their frequency and diversity as well as their atypical and often aggressive nature. Mortality and morbidity associated with skin tumours in this clinical context are often considerable, their pathogenesis is multifactorial and an evidence base to guide management is lacking in many key areas.
Certain primary immunodeficiencies predispose to skin cancer but the greatest burden of disease is associated with acquired immunodeficiency, including immunosuppressive drug therapy (e.g. following solid-organ and haematopoietic transplantation and for immune-mediated inflammatory disorders), non-Hodgkin lymphoma/chronic lymphocytic leukaemia (NHL/CLL) and HIV infection (Table 146.1).
Table 146.1 Primary and acquired immunodeficiency conditions and associated skin cancers
Immunodeficiency | Condition | cSCC/CIS | BCC | Melanoma | MCC | PCL | KS | Appendageal | DFSP | AFX |
Primary | EV | ++ | + | |||||||
SCID | + | |||||||||
WHIM | + | + | + | |||||||
DOCK8 | + | + | + | |||||||
CVID | + | |||||||||
WAS | + | + | ||||||||
DC | + | |||||||||
Netherton syndrome | + | + | ||||||||
Acquired | HIV | ++ | +++ | + | ++ | ++ | +++ | + | + | + |
NHL/CLL | ++ | + | + | + | + | + | ||||
Solid OTR | ++++ | ++ | + | + | + | ++ | + | + | + | |
HCT | + | + | + | + | ||||||
IMID (IBD, RA) | + | + | + | + | + | + |
Signs + to ++++ gives an approximate frequency of reports in the literature.
AFX, atypical fibroxanthoma; BCC, basal cell carcinoma; cSCC/CIS, cutaneous squamous cell carcinoma/carcinoma in situ; CVID, common variable immunodeficiency; DC, dyskeratosis congenita; DFSP, dermatofibrosarcoma protuberans; DOCK8, dedicator of cytokinesis 8; EV, epidermodysplasia verruciformis; HCT, haematopoietic cell transplant; HIV, human immunodeficiency virus; IMID, immune-mediated inflammatory disorder (IBD, inflammatory bowel disease; RA, rheumatoid arthritis); KS, Kaposi sarcoma; MCC, Merkel cell carcinoma; NHL/CLL, non-Hodgkin lymphoma/chronic lymphocytic leukaemia; OTR, organ transplant recipient; PCL, primary cutaneous lymphoma; SCID, severe combined immunodeficiency; WAS, Wiskott–Aldrich syndrome; WHIM, warts, hypogammaglobulinaemia, infections and myelokathexis.
The range of primary defects of the immune system associated with skin cancer underscores the complexity of the immune repertoire associated with skin cancer development and may provide important insights into its pathogenesis in the general population [1]. The following examples are illustrative of this spectrum of underlying immunological abnormalities.
Epidermodysplasia verruciformis (see Chapter 25). Epidermodysplasia verruciformis (EV) is the most distinctive condition in which skin cancer is associated with a primary immunological deficit. This rare autosomal recessive genodermatosis is characterized by predisposition to persistent infection with a specific genus of human papillomavirus (HPV), namely EV or β-HPV types [2]. Disseminated plane warts, multiple common warts and pityriasis versicolor-like lesions start to appear in early childhood and cutaneous squamous cell carcinomas (cSCCs) develop on UV-exposed sites in up to 60% of patients from the third decade onwards, with no other abnormalities in most patients [2]. Invasive cSCCs are located mainly on the forehead and associated with sun exposure, but onset may be exacerbated by irradiation; they develop slowly, may be locally destructive, but are only rarely metastatic [3]. EV is caused in most cases by mutations in EVER1 (TMC6) and EVER2 (TMC8), adjacent genes located on 17q25 which are involved in the regulation of intracellular zinc distribution, activity of zinc-dependent transcription factors and proliferation [2]. EVER deficiency may alleviate inhibition of expression of the whole genome of HPVs in keratinocytes, and/or impair T-cell immunity against HPVs [3, 4]. EVER2-deficient patients also display mild T-cell abnormalities [5]. Primary T-cell disorders due to deficiency in RHOH and MST1 may cause an EV phenotype [6], as may secondary T-cell defects in conditions such as HIV [7–10].
Severe combined immunodeficiency (SCID) is caused by deficiency and impaired function of T cells and, in some forms, additional reduction or dysfunction of natural killer (NK) cells and/or B cells. It results from diverse molecular defects, e.g. in IL-7R, CD45, IL-2Rγ, JAK3, RAG1, RAG2, ARTEMIS and ADA [1]. Increased cutaneous viral infections are a common feature and include HPV infection [11]. EV-like lesions occur in JAK3- and IL-2-Rγ-deficient patients after bone marrow transplantation [12, 13]. A similar association is reported in other SCID-like immunodeficiences such as WHIM syndrome (warts, hypogammaglobulinaemia, infections and myelokathexis), resulting from mutations of the CXCR4 chemokine receptor and in which vulvar SCC, basal cell carcinoma (BCC) and cutaneous T-cell lymphoma (CTCL) are reported [14]. Autosomal recessive mutations in the DOCK8 (dedicator of cytokinesis 8) gene, cause a combined immunodeficiency syndrome characterized by low T, B and NK cells, elevated serum immunoglobulin E (IgE) levels, depressed IgM levels, eosinophilia, sinopulmonary infections, cutaneous viral infections (HPV, herpes simplex virus and molluscum contagiosum) and malignancy, particularly SCC [15]. Mucocutaneous SCCs occur in 19%, with a young age of onset and are often associated with viral warts. Aggressive CTCL, diffuse large B-cell lymphoma and cutaneous microcystic adnexal carcinoma have also been reported [15, 16].
Common variable immunodeficiency (CVID) is the most prevalent of the primary immunodeficiencies in adults [17]. CVID encompasses a group of genetic disorders characterized by failure of B-cell maturation; the principle defect is in antibody formation, but there are related defects of both humoral and cell-mediated immune responses. It is characterized by hypogammaglobulinaemia, recurrent bacterial infections, autoimmune diseases and malignancy. Particularly lymphoma and gastric cancer, with rare reports of cSCC [18, 19].
Wiskott–Aldrich syndrome is a severe X-linked immunodeficiency caused by mutations in the WASP gene, which encodes key regulators of signalling in haematopoietic cells. Clinical manifestations include immunodeficiency, eczema, autoimmunity and tumour susceptibility, including head and neck SCC and Kaposi sarcoma (KS) [20].
Dyskeratosis congenita is an inherited bone marrow failure syndrome caused by abnormal telomere maintenance resulting from germline mutations in one of nine genes present in approximately 60% of patients. Nail dystrophy, abnormal skin pigmentation and oral leukoplakia are characteristic and patients are at high risk of bone marrow failure, pulmonary fibrosis, liver disease and mucocutaneous SCC [21].
Netherton syndrome due to SPINK5 mutation is an autosomal recessive disorder characterized by congenital ichthyosiform erythroderma, trichorrhexis invaginata, atopy, food allergies and asthma [11]. HPV infection with β-HPV types also occurs and has been associated with keratinocyte skin cancers (KC) in some cases [11, 22].
The total burden of skin cancer associated with acquired immunodeficiency is far greater than that of primary disorders and is predicted to represent an increasingly significant proportion of total skin cancer burden in the future [23]. Although there is no doubting this risk, the exact magnitude has been difficult to quantify because accurate data in the general population on incidence rates of many skin cancers, particularly with non-melanoma skin cancers (NMSC), are incomplete compared with most other tumour types [24] (Table 146.2).
Table 146.2 Reported ranges of skin cancer risk in immunodeficiencies
Skin cancer | OTR | NHL/CLL | HIV | IBD | RA |
SCC | 65–480 | 5–8 | 2.6–4 | 1.23–5.4 | 1.5–1.72 |
BCC | 4.5–10 | 5–8 | 2.1 | 1.2 | – |
Melanoma | 1.35–4 | 1.92–7.74 | 0.81–12.6 | 1.09 | 1.16–2.3 |
Kaposi sarcoma | 40–208 | 3.37–5 | 449–3640 | 2.89 | 2.52 |
Merkel cell carcinoma | 24–182 | 3.4–10.4 | 11.95–13.4 | 4.02 | 1.39–2.42 |
Appendageal | 20–100 | – | 3.26–8.1 | – | – |
Vulva/vagina | 6.9–23.91 | – | 4.41–6.79 | – | – |
Penis | 4.5–25 | – | 3.9–8 | – | – |
Anal | 2.7–14 | 2.44 | 19.63–50 | – | – |
Oral cavity | 2.75–7.5 | 0.9–2.33 | 1.1–2.93 | – | – |
The figures shown are the risk ranges cited in publications which include: Harwood et al. 2013 [50]; Brewer et al. 2013 [33]; Deeken et al. 2012 [27]; Long et al. 2010 [106]; Grulich et al. 2007 [25]; Mercer et al. 2012 [123]; Krynitz et al. 2013 [47]; Jensen et al. 2010 [79]; Lanoy et al. 2009 [26]; Hisada et al. 2001 [43]; Royle et al. 2011 [36]; Shiels et al. 2012 [397]; Clarke et al. 2015 [375]; Hemminki et al. 2012 [373]; Olsen et al. 2014 [518]; Morton et al. 2010 [35]; Mansfield et al. 2014 [399]; Singh et al. 2011 [110]; Crum-Cianflone et al. 2009 [31]; Silverberg et al. 2013 [32].
BCC, basal cell carcinoma; IBD, inflammatory bowel disease; NHL/CLL, non-Hodgkin lymphoma/chronic lymphocytic leukaemia; OTR, organ transplant recipient; RA, rheumatoid arthritis; SCC, squamous cell carcinoma.
HIV infection (see Chapter 31)
HIV/AIDS is associated with a 1.5–2-fold elevated risk of malignancy [25–27]. KS is increased more than 3000-fold compared to the immunocompetent population and, together with NHL and cervical cancer, is an AIDS-defining malignancy [27]. Since the introduction of highly active antiretroviral therapy (HAART), the incidence of AIDS-defining malignancies has fallen threefold and that of non-AIDS-defining malignancies has risen threefold; cSCC/BCC and melanoma are among the most common non-AIDS-defining malignancies, with reported standardized incidence ratios (SIR) of 2.1–4 and 0.81–12.6, respectively [7, 28–30] and BCC is now more common than KS in white skin [31]. In contrast to organ transplant recipients (OTRs) in whom the BCC : cSCC ratio of approximately 4 : 1 in the general population is reversed, this ratio is maintained in HIV with CD4 counts >500 cells/μL, but cSCC predominate when CD4 falls below 200 cells/μL [32]. cSCC arising in the context of an EV-like phenotype has been associated with the immune reconstitution syndrome [9]. Unlike AIDS-defining malignancies, duration of HIV infection rather than degree of immunosuppression, as measured by CD4 count, is the main risk factor for skin cancer [27, 31]. HIV may have direct cellular and molecular effects that contribute to the development of skin cancer, including activation of proto-oncogenes, alterations in cell cycle regulation, inhibition of tumour suppressor genes, induction of microsatellite gene instability and promotion of pro-angiogenesis signaling [7]. HIV-infected patients also have an increased risk of exposure and subsequent infection with other skin cancer associated viruses, including HHV-8 and Epstein–Barr virus (EBV) [25].
NHL is an increasingly common lymphoproliferative malignancy and includes CLL, a clonal B-cell disorder, which accounts for 25% of all leukaemias [33]. These malignancies are associated with innate immunosuppression and defects in both cell and humoral-mediated immune responses, which may be exacerbated by therapy [33, 34]. The overall risk for second malignancies in patients with NHL/CLL is more than doubled compared with the general population [35, 36] and a fivefold to eightfold increase risk of skin cancer has been documented [33, 37], with higher risk in CLL compared with non-CLL NHL [38]. This risk association is reciprocal: the increased risk of subsequent NHL/CLL in patients with melanoma is increased up to 2.7-fold [28] and a similar increased risk for CLL occurs in patients with previous cSCC (SIR 2.3; 95% CI 1.9–2.7) [39]. In CLL, the cumulative skin cancer incidence by 20 years in a US cohort was 43.2% for cSCC and 30.6% for BCC [38]. These skin cancers also behave more aggressively [40, 41, 45]. Mortality from skin cancer was as high as from CLL in one recent study and more advanced CLL stage was predictive of worse skin cancer prognosis [42]. This is particularly evident in Australia where death from skin cancer had the highest standardized mortality ratio (SMR) of all causes in patients with CLL [36]. KS is fivefold more common [43] and melanoma approximately three- to sixfold more common [43, 44]; the melanoma SMR for patients with CLL is 2.8 in the US rising to almost 5 in Australia, where the overall SMR for KC is 17 [36]. Patients with CLL who develop Merkel cell carcinona (MCC) are almost four times more likely to develop metastases [44].
Skin cancer is a well-recognized complication of immunosuppressive drug treatment. Solid-organ transplant recipients, in whom long-term immunosuppressive therapy is required to prevent allograft rejection, represent the largest and most comprehensively studied group. However, the risk after allogenic haematopoetic cell transplantation, in which immunosuppressive drugs are used to prevent graft-versus-host disease (GVHD), and that associated with immunosuppression for immune-mediated inflammatory disorders has been the focus of more recent research.
Organ transplantation is a highly successful treatment for end-stage organ failure, with more than 114 000 organ transplants perfomed worldwide in 2012 45. Survival continues to steadily increase, as does the risk of malignancy [25]. Overall risk for any cancer is twofold to sixfold greater than that of the general population [25, 46–49] with a disproportionate increase in four tumour types: KC, post-transplant lymphoproliferative disorders (PTLD), anogenital malignancy and KS, with smaller but significant increases in hepatocellular and renal cancers and some sarcomas [46].
Spectrum of skin cancers post-transplant KC accounts for more than 95% of all post-transplant skin cancer; cSCCs predominate with an overall SIR ranging from 45 to 480 [47, 50, 51] (see Table 146.2) which is significantly greater for those under 50 years [23, 50] and increases with time post-transplant, reaching more than 200 and 300 in renal and cardiac OTRs, respectively, at 10–20 years post-transplant [47]. BCC are the second most common skin cancer and incidence is approximately fivefold to 10-fold increased, with consequent reversal in the 3–4 : 1 BCC : cSCC ratio usually seen in the general population. Region-specific differences in the frequency of post-transplant BCC alter the extent of this reversal, with lower SCC : BCC ratios reported in Spain and Italy in some studies [52–58], but not all [59]. The BCC : cSCC ratio is also influenced by time from transplantation, since BCC show a more linear increase compared with the exponential rise in SCC [57, 60, 61]. Up to 50% of OTR with cSCC also have BCC [62] and a threefold increased risk of developing internal malignancies [63]. Melanoma incidence is increased 2.1–8-fold [28, 47], KS 40–200-fold [47], appendageal tumours 20–100-fold [47, 395] and MCC up to 60-fold [50, 54, 62, 377, 384], cutaneous lymphoma (PCL) and sarcomas such as dermatofibrosarcoma protuberans (DFSP) and atypical fibroxanthoma (AFX) are also overrepresented, although population-based studies quantifying this risk are few [64, 66, 387, 403].
Tumour burden and accrual The high incidence of skin cancers in OTRs is compounded by their multiplicity, which increases with duration post-transplantation [50, 68–71]. In a UK OTR cohort, almost 30% of all OTRs who had been transplanted longer than 6 months (median 10 years) had developed skin cancer, rising from 10% at 5 years to almost 75% at 30 years. Two-thirds of affected individuals had more than one skin cancer, with an average of six tumours per patient (2–106); a minority of OTRs contributed disproportionately to the total cohort tumour burden, with more than 50% of the total number of cSCCs arising in just 3.4% of individuals [50]. Once the first KC has developed, more than s30% will develop a further KC by 1 year and almost 75% by 5 years [50, 69–71] compared with 14.5% and 40.7%, respectively, in the general population [72]. In an Australian cohort, OTRs developed an average of 3.35 ± 4.29 tumours per year at 20 years post-transplant [73]. In a UK cohort, the time interval between subsequent KC shortened progressively from 24 months to second cancer, 14.7 and 8.4 months to third and fourth, respectively; patients with 10 or more tumours developed a new cancer every 3 months and were at increased risk for metastastic disease. Time to first cancer was reduced from 104.9 to 71 months if there was a history of a pre-transplant KC [50]. The impact of skin cancers on measurements of quality of life in OTRs is unclear: in one US study, the number of skin cancers correlated with higher levels of anxiety although this did not quite reach significance [74], whilst in a cohort from Ireland, skin cancer impacted less on quality of life than certain benign dermatoses associated with transplantation [75].
Geographical or ethnic influences Comparison between studies from different geographical regions is complicated by the diversity in population characteristics, notably skin phototype, as well as environmental factors, including latitude and level of sun exposure [76]. White people living in Australia have the highest incidence of post-transplant KC, affecting more than 80% of those transplanted for 20 years [73, 77, 78], although the SIR may be equivalent in more temperate countries because of their relatively low frequency in the general population [68, 79]. KS is most common in HHV-8 endemic areas such as the Mediterranean and is, for example, more common in the south compared with the north of Italy, reflecting HHV-8 prevalence [49]. It is even more common in sub-Saharan Africa [80] and is the most common post-transplant malignancy in Saudi Arabia [81, 82]. In South Africa, KC was seen only in patients of European origin [80] whereas KS was the most common cancer in non-whites [83]. Japan and Taiwan report a low incidence of KC and KS [84, 85], and although also low in Korea, incidence of KC is still significantly greater than in the general population [86].
Paediatric transplantation The spectum of malignancies developing after paediatric organ transplantation differs to that seen in adult OTR populations; skin cancer is rare [87] but can develop in early adulthood [88]. KC is the most frequent malignancy following renal transplantation and in other organ transplants is the second most common after PTLD [89]. Melanoma is proportionately more common in paediatric OTRs and accounts for 15% of all skin cancers [28].
Type of solid-organ transplant In a population-based study of more than 10 000 OTRs over 20 years, cSCC risk appears to be greatest after cardiac and/or lung transplantation, followed by renal transplantation with risk lowest in liver transplant recipients [47]. Most, although not all, other studies also confirm the risk to be significantly lower in liver compared with renal transplant recipients [60, 90, 91]. Incidence is particularly high after simultaneous pancreas and kidney transplants, with cSCC reported to be 6.2-fold higher than age- and sex-matched renal transplant recipients [92]. However, the number of transplantations does not appear to increase risk [93]. The reasons for the differences according to organ type are not entirely clear, but may relate in part to intensity of immunosuppression [92].
Cause of end-stage organ disease End-stage organ failure itself is associated with a small increased risk of cancer: in one study, SIR for malignancy was 3.27 after renal transplant, 1.35 during dialysis and 1.16 before renal replacement therapy [46]. A Danish registry study also reported an SIR of 4.8 for cSCC among patients with renal failure, but not for cardiac, lung or liver failure [79]. There is some evidence that the cause of end-stage renal disease may have an impact on skin cancer risk with, for example, polycystic kidney disease conferring a higher risk than diabetic renal disease [94], but no association was identified with causes for end-stage liver disease in a French series [60]. Kidney and liver transplantation for HIV-related organ failure may be predicted to significantly increase risk, but early evidence indicates that rates of KS and skin cancer are relatively low, although HPV-related anal neoplasia may be at an increased risk of progression [95]. Pre-transplant immunosuppression did not increase risk of melanoma or KS in an Australian study, although KC was not included in this analysis [96].
Survival after haematopoietic cell transplantation (HCT) for haematological malignancy has increased steadily over the past two decades and secondary solid cancers, including skin cancers are an increasingly important late complication of both conventional myeloablative and non-myeloablative transplants [97–103]. Cumulative incidence estimates in one large study for BCC and SCC at 20 years were 6.5% and 3.4%, respectively [101], and skin cancers occur in both adult and paediatric populations [104].
Inflammatory bowel disease (IBD: Crohn disease and ulcerative colitis), rheumatoid arthritis (RA), psoriasis and systemic lupus erythematosus are associated with an increased risk of skin cancer [105]. Whilst this may be partly due to intrinsic immune dysregulation, most studies have focused on the role of iatrogenic immunosuppression, principally non-biological immunomodulatory drugs (e.g. azathioprine, ciclosporin) and biological response modifiers (BRMs), particularly the anti-tumour necrosis factor (TNF) agents (adalimumab, certolizumab pegol, etanercept, golimumab and infliximab).
In IBD, several observational studies document an increased risk of skin cancer. A large retrospective, nested case–control study of KC in patients with IBD showed a 60% excess risk compared with controls; thiopurine use was associated with an adjusted odds ratio (OR) of 3.56, rising to 4.27 for use for more than 1 year and an OR of 2.18 for persistent BRM use in Crohn disease [106]. A meta-analysis also confirmed an increased risk of KC with thiopurine use in IBD, with a hazards ratio (HR) of 2.28 [107] and a retrospective cohort study of more than 14 000 patients with IBD showed a similar HR of 2.1 for KC with thiopurine use, but no increase for melanoma; KC risk increased with duration of thiopurine exposure and returned to pre-exposure levels when thiopurines were stopped [108]. A prospective observational cohort study in France confirmed this increased risk with HR 5.9 for ongoing thiopurine exposure, which was almost doubled for patients >65 years of age compared with those <65 years of age, but in this study the increased risk continued even after stopping thiopurines [109]. A smaller cohort study showed an equally high risk of cSCC with thiopurines (HR 5.4), with an increased risk of BCC (HR 1.2) in those not on thiopurines [110, 111]. A study from South Africa found the KC risk was highest in white patients [112]. Not all studies confirm an association and the smaller risk of KC in patients with IBD compared with OTRs probably reflects the use of intermittent monotherapy in IBD, in contrast to the two- or three-drug regimens used to prevent graft rejection [113]. In an observational study, IBD was also associated with an excess risk of melanoma with BRM therapy, particularly in Crohn disease, with an OR of 1.88 [114]. In contrast, a meta-analysis of 22 randomized controlled trials (RCTs) failed to provide conclusive evidence of increased skin cancer risk with BRMs in IBD, although as these trials did not extend beyond 12 months, a longer term risk cannot be excluded [115].
In RA, observational cohort studies have also shown an increased risk of cSCC in patients treated with azathioprine for more than 1 year [116] and a significant increase in both KC and melanoma (ORs, 1.5 and 2.3, respectively) with BRM use [117], but no increased risk of BRMs over other therapies [118]. However, systematic reviews and meta-analyses of skin cancer risk associated with BRMs have provided conflicting results. A twofold increased risk of KC was identified with adalimumab, infliximab and etanercept [119]; a similar excess risk for KC (OR, 1.45) and melanoma (OR1.79) was also found in a separate study [120]; a trend for increased KC was identified in another meta-analysis [121]; but no increased risk was seen with the newer agents certolizumab and golimimab [122]. Data from the British Society of Rheumatologists Biologics Register identified an increased risk of KC of approximately 1.72 and 1.83 in patients, respectively, on disease-modifying antirheumatic drugs and BRMs compared with the general population, but no evidence of differing excess risk between these drug groups [123]; an observation confirmed in another meta-analysis of 63 RCTs with 9 BRMs [124]. Similarly, no statistically significant association with anti-TNF therapy was identified in a meta-analysis of psoriasis trials [125]. It is possible that discrepancies between studies may be due to methodological differences and duration of follow-up is likely to be an important confounding factor; the lack of statistical association in some of these large trials contrast with the case reports and cohort studies in which rapid development of KC and recurrence of melanoma are described [126].
The pathogenesis of skin cancer arising in the context of immunosuppression is likely to be multifactorial and current evidence suggests a complex interplay primarily between UV radiation (UVR), altered immune surveillance, drugs and oncogenic viruses, with likely additional roles for host genetic susceptibility factors, chronic inflammation and donor-derived cells.
As in the general population, UVR is an important risk factor (see Chapter 9); most immunosuppression-associated KC is more prevalent in regions of high ambient solar radiation, 75% occur on photoexposed body sites, are more common in those with fair skin phototype and a history of chronic UV exposure and, in particular, childhood sunburn [127, 128, 129]. UVR increases skin cancer risk by local reduction of tumour immune surveillance and is mutagenic; targeted gene sequencing of p53 in OTR SCCs, PTCH in OTR BCCs and whole exome sequencing studies of cSCC from OTRs have shown a high prevalence of characteristic UV-induced mutations [130–132]. Although there are no clear differences in the genetic changes present in tumours from immunocompromised individuals compared with the general population [133, 134], some of the drugs used in these patients, e.g. ciclosporin, azathioprine and voriconazole, may also interact with UVR and directly or indirectly enhance its carcinogenic effects, as may HPV (see below).
The parallels between malignancy in OTRs and patients with HIV/AIDS support the likely importance of immunosuppression and reduced tumour immune surveillance or ‘immunoediting’ per se as an important contributory factor to the increased skin cancer risk [25, 126, 135]. In general, the incidence of OTR KC is proportional to the level of immunosuppression [136, 137] and associated with lower peripheral CD4 counts [138]. Intensity of immunosuppression also influences risk of post-transplant melanoma and KS [46, 139]. Reduction of immunosuppression in OTRs reduces the rate of subsequent accrual of skin cancers [140, 141], particularly those such as KS, which are virus related [142]. Synergy between viral oncogenesis and immune dysregulation, e.g. by enhanced viral replication or integration, may provide additional or alternative mechanisms in immunosuppression-related skin carcinogenesis [143]. The immunophenotype also differs in OTR cSCC, with higher numbers of circulating regulatory T cells predictive for new cSCC development [144]. In HIV, the risk of KS but not KC was proportionate to the absolute CD4 count in one study [31], although the risk of cSCC but not BCC may increase with lower CD4 counts [32]. NHL/CLL is associated with innate immune dysregulation involving complex defects of both humoral and cell-mediated immunity, which, independent of treatment-related risk factors, may be sufficient to account for the increased skin cancer risk [33, 38].
In addition to systemic immune dysregulation, the local tumour microenvironment also plays a critical role in carcinogenesis [145] and there is evidence for a unique immune microenvironment in OTR cSCC. The density of inflammatory infiltrate appears to be reduced [146, 147] and the combination of reduced CD4+ T-cell infiltration [148], decreased cytotoxic CD8+ T cells [144, 147, 149] and increased regulatory T cells described in some [144, 149] but not all [147, 148] studies, is predicted to lead to a ‘permissive’ tumour microenvironment with decreased immune surveillance. In addition, impaired antigen presentation though reduced CD123+ plasmacytoid dendritic cells [147] and increased exposure to IL-22, may accelerate tumour growth [149] and potentially contribute to the aggressive nature of some OTR cSCC.
Current immunosuppressive drug regimens usually use a combination of agents with differing modes of action at specific sites of the T-cell activation cascade [150]. In transplantation, immunosuppressive protocols consist of two phases: a perioperative induction phase (using for example OKT3, antithymocyte globulin, basiliximab, daclizumab) is followed by a long-term maintenance phase (e.g. ciclosporin, tacrolimus, azathioprine, mycophenolate mofetil (MMF), sirolimus, everolimus) [51]. Characteristics of the immunosuppressive regimen including duration, use of induction therapy and type of maintenance therapy may all be important risk factors for skin cancer, but establishing the degree of risk conferred by individual drugs is challenging, given the large number of potential confounding factors, including variations in individual dosage and regimens depending on the type of transplant and tolerability [50, 150, 151]. There is some evidence that azathioprine may pose an increased risk over ciclosporin and corticosteroids [152, 153], but this has not been confirmed in all studies [58, 154]. Similarly, some studies [59, 155–157] but not all [158] have shown a reduced risk with MMF compared with azathioprine. MMF, however, has been linked to BCC risk in cardiac transplant recipients [159]. Tacrolimus has a relative protective effect compared with ciclosporin in some studies [58] but not all [59, 155, 158]. There is clear evidence, however, that mammalian target of rapamycin (mTOR) inhibitors (rapamycin/sirolimus, everolimus) confer reduced skin cancer risk [160] and this is discussed in more detail later.
The overall level of immune suppression may be more important than the effects of specific drugs. Both duration and dose intensity of immunosuppressive drug therapy appear to be relevant: triple versus dual versus monotherapy and higher versus lower dose ciclosporin regimens are associated with increased risk [136, 137, 161, 162], whereas less intensive immunosuppression in liver transplant recipients may account for their significantly lower rates of skin cancer compared with other OTRs [91]. Similarly, the lower skin cancer risk in HCT recipients reflects the generally shorter duration of immunosuppressive drug use post-transplant compared with OTRs [163]. However, even prolonged use of single agent oral corticosteroids and azathioprine is associated with a twofold to fourfold increased risk of KC [30, 106, 164, 165].
In addition to induction of decreased immunosurveillance, certain immunosuppressants have direct effects on carcinogenesis and tumour progression [150]. Thiopurines (e.g. azathioprine) and calcineurin inhibitors (CNIs, e.g. ciclosporin, tacrolimus) demonstrate synergistic interactions with UVB and UVA, which may, for example, promote UV-induced DNA damage and/or inhibit DNA repair. In contrast, mTOR inhibitors have direct anticarcinogenic properties including suppression of angiogenesis, autophagy-mediated DNA repair and promotion of memory T-cell function [150, 166–171]. Examples of direct pro- and anticarcinogenic mechanisms are summarized in Table 146.3 [150, 166, 169, 172–197].
Table 146.3 Direct pro- and antitumour effects of immunosuppressive drugs
Drug | Mechanism | Reference |
Ciclosporin | Reduced repair of UV-induced DNA damage | Heman et al. [172] Sugie et al. [173] Yarosh et al. [174] Thoms et al. [175] |
Increased TGF-β production | Hojo et al. [176] Maluccio et al. [177] |
|
Enhanced UVB-induced inflammation and angiogenesis | Duncan et al. [178] | |
Induction of oncogene ATF3 and suppression of p53-dependent senescence | Wu et al. [179] | |
Reduced apoptotic response to UV by MPTP inhibition | Norman et al. [180] | |
Activation of AKT by PTEN suppression | Han et al. [181] | |
Activation of TAK1/TAB1 signalling | Xu et al. [182] | |
Augmented EMT by TGF-β1 signalling | Walsh et al. [183] | |
Reduced NER by downregulation of XPA and XPG | Kuschal et al. [184] | |
Potentiation of oncogenic ATF3 by UVA | Dziunycz et al. [185] | |
Azathioprine | Reduced repair of UV DNA damage | Kelly et al. [186] de Graaf et al. [187] |
Incorporation of metabolite 6-thioguanine into DNA and generation of mutagenic oxidative DNA damage with UVA | O'Donovan et al. [188] | |
Photosensitizes skin to UVA in vivo | Perrett et al. [189] Hofbauer [190] |
|
Protein oxidation and damage to the DNA repair proteome by 6-thioguanine and UVA with impaired NER | Gueranger et al. [191] | |
mTOR inhibitors | Inhibits rather than promotes cancer | Campistol et al. [192] Kauffman et al. [193] Mathew et al. [194] |
Suppression of angiogenesis by reduction of VEGF | Guba et al. [166] | |
Antiproliferative | Aissat et al. [195] | |
Increased autophagy-mediated DNA repair | Saha et al. [196] | |
Increased AKT1 and reduced AKT2 | Sully [169] | |
Reduced EGFR expression | Liu et al. [197] | |
Promotion of memory T-cell function | Jung et al. [150] |
AKT, Ak strain transforming; ATF3, activating transcription factor 3; EGFR, epidermal growth factor receptor; EMT, epithelial-mesenchymal transition; MPTP, mitochondrial permeability transition pore; NER, nucleotide excision repair; PTEN, phosphatase and tensin homologue; TAK1/TAB1, transforming growth factor β-activated kinase 1/TAK1 binding protein 1; TGF, transforming growth factor; VEGF, vascular endothelial growth factor; NER XPA/XPG, nuclelotide excision repair genes xeroderma pigmentosa-A/-G.
The effects of BRMs, including anti-TNF agents, are discussed earlier.
There are conflicting data on whether antiretroviral drugs affect the risk of non-AIDS-defining cancers, including skin cancers [198], with no clear patterns emerging [27].
Data are conflicting on whether chemotherapy for CLL/NHL affects the incidence of skin cancers [38]. Although it has been suggested in CLL that cytotoxic chemotherapy contributes to skin cancer risk in CLL [199], more recent studies have shown that chemotherapeutic regimens probably do not influence the development of secondary malignancy, indicating a relationship with skin cancer that is unlikely to be primarily iatrogenic [35, 200, 201].
Certain drugs commonly used in immunocompromised individuals may also affect skin cancer risk. Voriconazole is a triazole antifungal often used in the treatment and prophylaxis of invasive fungal infections such as aspergillosis in solid-organ (particularly lung) and HCT recipients. There are numerous case reports and series of its association with cSCC, which may be multiple and aggressive [202, 203]. Retrospective studies have identifed it as an independent risk factor for cSCC in lung transplant recipients [204–207]. In a French series of 19 cases, the majority of patients affected were immunosuppressed and cSCC developed after a mean of 35 months [208]. A multistep photo-induced process was observed in most cases, with acute phototoxicity in the first year, actinic keratosis (AK) in the second/third year and cSCC by the third year onwards, suggesting that voriconazole phototoxicity, possibly enhanced by immunodeficiency, was responsible for a carcinogenic effect [203, 208]. Statins have immunomodulatory effects, which have led to concerns that they may increase the risk of KC [209]. However, no effect on risk was identified in a recent meta-analysis [210]. Thiazide diuretics were associated with a modestly increased risk of KC in a population-based case–control study [211]. Non-steroidal anti-inflammatory drugs have been proposed to have a possible protective effect against skin cancer, confirmed in one recent systematic review [212], but not in a large meta-analysis [213].
The most common immunosuppression-associated malignancies are those due to known or suspected oncogenic viruses [25]. In the skin, these include KS (HHV-8) (see Chapter 139), post-transplant lymphoproliferative disorders (EBV) and, most recently, Merkel cell carcinoma (Merkel cell polyomavirus, MCPyV) (see Chapter 145) [252]. These pathogens and their multiple mechanistic pathways of viral oncogenesis are discussed in more detail elsewhere, but these mechanisms may be further complicated in the setting of immuncompromise, for example by interactions with immunosuppressive drugs [143]. This section focuses on the role played by HPV in immunosuppression-related skin malignancy (see Chapter 25).
HPV has long been proposed to contribute to the pathogenesis of cSCC, but its role remains controversial [215, 216]. HPV is a double-stranded DNA virus and more than 170 types are recognized [216, 217]. High-risk mucosal HPV types (principally α-HPV types 16, 18) are the main carcinogens responsible for anogenital SCC [218] and recognition of this has culminated in preventative vaccination against HPV-16 and -18 [219]. α-HPVs also cause a proportion of head and neck SCC (HNSCC) [220] and periungual SCC [221]. Anogenital dysplasia, head and neck SCC and periungual SCC are all more frequent in long-term immunosuppressed patients. In EV, β-HPVs are detected in over 90% of cSCC and appear to act as co-carcinogens with UVR [3]. In contrast to high-risk α-HPVs, β-HPVs rarely integrate into the host genome and most β E6 and E7 proteins do not target p53 or Rb [222]. However, research by many groups has provided evidence of alternative mechanisms by which β-HPV types in the skin may have pro-carcinogenic effects in cooperation with UVR (Table 146.4) [223–247]. In particular, functional studies have shown that specific β-HPV oncoproteins abrogate UV-induced apoptosis, delay DNA repair and overcome cell cycle arrest [248], interfere with NOTCH tumour suppression [242], enhance dermal invasion [241] and β-HPV types 8 and 38 E2, E6 and E7 proteins are oncogenic in transgenic mice [245–247].
Table 146.4 Possible roles for β papillomaviruses in skin carcinogenesis
Mechanism | Cellular target | β papillomavirus type |
Abrogation of cell cycle check points | pRB p16 p53 p53, pRb |
HPV-38 [223] HPV-5, -8 [224] HPV-38 [225] HPV-49 [226] |
Prolong cell lifespan | Telomerase | MHPV-38 [227] |
Inhibition of DNA repair | XRCC1 Unknown P300, ATR |
HPV-8 [228] HPV-5 [229] HPV-5, -8 [230] |
Genome destabilization (aberrant mitosis and dysregulated centrosome duplication) | P300, p53 | HPV-5, -8, -38 [231] |
Inhibition of apoptosis | Bak | HPV-5, -8, -20, -38, -76, -92, -96 [232, 233] HPV-8, -20 [234] |
Bax | HPV-5, -8, -38 [235, 236] | |
P300, p53 | HPV-38 [237] HPV-8 [238] |
|
NF-kB TIP60 HIPK2, p53 |
HPV-23 [239] | |
Evasion of host immune surveillance | TAP-1 IL-8 |
HPV-8 [240] HPV-5, -8 [241] |
Interfere with NOTCH tumour suppression | NOTCH signalling in epidermis and mesenchyme | β papillomaviruses [242] |
Repression of TGF-β signalling pathway | SMAD3 | HPV-5 [243] |
Enhanced dermal invasion | MMP-1, MMP-8, MT-1-MMP | HPV-8 E7 protein [244] |
Oncogenic in transgenic mice | – | HPV-8 E2, E6 and E7 proteins expressed from a keratin promoter [245–247] |
HPV, human papillomavirus; IL, interleukin; MMP: matrix metalloproteinase; NF-κB, nuclear factor κ-light-chain-enhancer of activated B cells; TGF, transforming growth factor.
A role for HPV in non-EV immunosuppression-related cSCC is plausible, given the association with oncogenic viruses of the most overrepresented malignancies in immunosuppressed individuals, the widespread cutaneous HPV infection seen in many immunodeficiencies and the HPV-related histological features in some cSCCs [146]. However, an oncogenic role for HPV remains unproven [215, 216, 249–253]. More than 100 studies have investigated the epidemiological relationship between HPV and cSCC, particularly in OTRs, and most show a significant correlation between the presence of the virus and cSCC development. However, interpreting this literature is complicated by the heterogeneity of studies in terms of patient populations, sampling and HPV detection methods (which vary markedly in sensitivity and ability to detect the large number and diversity of potentially relevant HPV types) [215, 216, 249–253]. In addition, β-HPVs are ubiquitous in normal skin, with a probable reservoir in hair follicles; individuals are colonized with a specific and persistent profile of multiple β-HPV types early in life [254] and β-HPV seroprevalence increases with age [255] but not UV exposure [256]. Against this background, epidemiological studies have confirmed the presence of β-HPV DNA in >90% of OTR SCC, a higher prevalence than in immunocompetent tumours and significantly higher than in normal skin [257]. In situ hybridization has also identified viral gene expression in tumours [258] although viral load is usually <1 viral copy per cell and higher in AK compared with cSCC [259]. There is an association between cSCC and seroprevalence for β-HPV types in both OTR and immunocompetent populations [260–262], although serology does not always correlate with HPV DNA presence in skin [263]. Concordant detection of HPV DNA in hair follicles and antibodies for the same virus type may be more relevant, with an overall OR of 1.6 for SCC in one large study [262]. In addition, a positive seroresponse to β-HPVs at the time of transplantation was predictive for subsequent KC risk, with a hazard ration of 2.9 [264]. In contrast to anogenital cancer, no clear hierarchy of specific HPV types has been defined, but there is a trend towards β-HPV types oncogenic in EV, notably HPV-5, -8, -36 and -38, which fits well with functional studies [262].
More recently, whole transcriptome sequencing has revealed low transcriptional activity of HPV in tumours, evidence that perhaps HPV is not involved in cutaneous oncogenesis and is merely a marker of immunosuppression [214, 265]. Alternatively, HPV may be acting through a ‘hit and run’ mechanism, involved in initiation rather than promotion or maintenance of oncogenesis, with HPV-induced perturbation of cellular DNA repair or apoptosis predisposing keratinocyte stem cells to UV-induced damage [215, 216, 248, 253]. Clarification of the part played by HPV may provide future directions for more targeted therapy. HPV vaccination is one such strategy and has been suggested as a possible approach to prevention of post-transplant HPV-associated epithelial malignancies, although as current vaccines protect only against mucosal HPV types, there is a less convincing rationale for their use in cSCC compared with anogenital and HNSCC [253, 266]. However, preliminary experimental data from mouse models have supported vaccination against cutaneous HPV types as a viable approach to cSCC prevention, even in the setting of immunosuppression [267, 268] although immunogenicity of HPV vaccines may be suboptimal in immunocompromised patients [269].
Germline single nucleotide polymorphisms (SNPs) may be associated with increased risk of skin cancer. Of those investigated in relation to immunosuppression-associated skin cancer, pigmentation genes are the most extensively studied [270]. In Norwegian OTRs, variation in the key signalling regulator MC1R but not other pigmentation-associated genes (the MC1R antagonist ASIP, and downstream melanization regulatory genes TYR and TYRP1) had a significant impact on cSCC risk, independent of conventional risk phenotypes including hair colour and skin phototype [271]. A common polymorphism of p53 results in either a proline or arginine at residue 72 of exon 4 and these polymorphic alleles have distinct biochemical and functional properties, including their ability to signal apoptosis following DNA damage. A significant association was identified between p53 codon 72 arginine homozygosity in cSCC from OTR but not immunocompetent individuals [272]. A correlation has been identified between polymorphisms in detoxifying enzymes glutathione-S-transferase [273, 274] and IL-10 polymorphisms and cSCC risk in OTR [275]. COX2 gene regulatory region variants appear to be associated with risk of OTR KC, although differ in cSCC compared with BCC [276]. Association of folate pathway-related methylenetetrahydrofolate reductase MTHFR:C677T gene polymorphism and risk of OTR cSCC has been observed [277] and MTHFR polymorphisms also associate with aberrant OTR cSCC methylation [278]. Haplotypes containing T(1686)-T(3944) alleles but not polymorphisms in the proximal 5′ regulatory region of the PTCH1 gene were shown to be associated with an increased BCC risk in Italian OTRs [279]. Vitamin D receptor (Intron8G/T, [280]), EGFR +61 A-G [281] and Toll-like receptor 4, 7 and 8 polymorphisms [280] are not associated with transplant KC. Polymorphisms in DNA repair genes have been investigated, particularly in relation to azathioprine exposure; MSH2 and MLH1 protein expression was not altered in OTR cSCCs and there was no difference in expression between cSCCs from OTRs and immunocompetent patients and no association between MSH2 polymorphism genotype frequency and OTR skin cancer status [282]. Despite these data, none of the known biomarkers are yet sufficiently robust to use as part of a skin cancer prediction algorithm in OTRs.
Radiation therapy may contribute to skin carcinogenesis in immunocompromised individuals. Increased melanoma risk persists for more than 20 years after radiotherapy for Hodgkin lymphoma and, as not all tumours arise within the irradiated field, radiation may be having an additional systemic effect [283]. In HCT, total-body irradiation conditioning regimens increase BCC risk; those exposed to radiation at age less than 10 years show significantly greater risk than older individuals [101].
Acute GVHD after HCT is an independent risk factor for cSCC and chronic GVHD increases the risk of both SCC and BCC [101–103, 163], with a RR of 5.8 in one study [103]. A large case–control analysis demonstrated that the severity and duration of chronic GVHD was a significant risk factor as was use of azathioprine, particularly in combination with ciclosporin and prednisolone [163].
The presence of donor-derived cells has been reported as potentially pathogenic in HCT-related oral malignancies [284] and in OTR KC [285, 286] and KS [287], although the mechanisms involved remain unclear.
Skin tumours in the setting of immunosuppression may have atypical presentations and altered clinical courses [33, 50, 128, 288].
Epidemiological research, including prospective cohort studies, have identified clinical features predicting OTRs at greatest risk for developing skin cancer which are clinically robust, allowing stratification of patients at highest risk of future skin malignancies (Table 146.5) [50, 51, 127, 262, 289–291].
Table 146.5 Risk factors for skin cancer development in solid-organ transplant recipients
Risk factors for time to first skin cancer | Risk factors for total number of skin cancers | Other patient-related risk factors | Transplant-related risk factors |
Duration of immunosuppression Age at transplant Sunburn pre-transplant Ethnicity (white versus non-white) Skin cancer pre-transplant |
Duration of immunosuppression Age at transplant Sunburn pre-transplant Chronic UV exposure Skin phototype Male Number of keratotic lesions (actinic keratoses and verrucokeratotic lesions) |
Smoking – inconsistent Alcohol – inconsistent Genetic polymorphisms CD4 count β human papillomavirus DNA/serology concordance |
Allograft type Cause of end-stage organ disease |
Based on references cited in this chapter, including the following studies: Bouwes Bavinck et al. [127], Casabonne et al. [289], Joly et al. [290], Bouwes Bavinck et al. [291], Harwood et al. [50] and Proby et al. [262].
Cutaneous SCC are more than 150-fold more common in OTRs, with lower levels of increased risk, ranging from 1.5 to 8, observed in other immunocompromised groups (see Table 146.2). They are predominantly located on UV-exposed sites [50], but are more common on non-head and neck sites in immunocompromised compared with immunocompetent individuals [146, 293]. Although diagnosis is usually made clinically, appearances may be atypical [40, 146, 294] diagnostic accuracy may be relatively low and a high index of suspicion is required [294]. Pain is a useful symptom of invasive malignancy in this context [67, 295]. Differential diagnoses include Bowen disease, AK and other rare skin tumours such as appendageal malignancies. Infections may also simulate cSCC, in particular viral warts (which may be clinically and histologically atypical), chronic herpes simplex and atypical mycobacterial infections. Keratoacanthomas (see Chapter 142) are regarded as spontaneously resolving, well-differentiated cSCC-like lesions [296]. They appear karyotypically simpler than cSCC [297, 298] and have other genetic differences [299–302]. However, clinicopathological differentiation from cSCC is not straightforward and they tend to be managed as well-differentiated cSCC in the setting of immunosuppression.
In terms of histology, differentiation status of OTR cSCC is not significantly different to immunocompetent populations in most studies [146, 303], but there are reports of increased frequency of spindle cell morphology, reduced inflammatory infiltrate, evidence of HPV infection [146], increased perineural and lymphatic invasion [303], increased acantholysis and increased depth of invasion [293]. However, these features are not consistent across all studies, and while they may account for the observed increase in local recurrence and metastasis [303], it is plausible that differences in overall prognosis may instead reflect the significantly greater overall tumour burden in individual OTRs, rather than increased aggression of most individual tumours [146]. In a retrospective study, the cumulative incidence of local recurrence was 19% for stage T1 SCC at 5 years, and 54% for stage T2 SCC at 3 years [304]. There is an estimated metastatic risk of up to 7% for cSCC in OTRs, more than twice that in the general population [50, 62, 303–305]; prognosis for metastatic disease is worse in immunocompromised individuals [306] with an overall 5-year survival of 14–39% and median 3-year survival of 56% in OTRs [307]. In transit metastases are more common in OTRs [308], with scalp representing a particularly high-risk site [309]. OTRs from countries with a high incidence of KC, such as Australia, experience greater overall morbidity and mortality compared with recipients from more temperate climates [310]. However, although the SMR for KC was almost 50 in one study of Australian liver and heart/lung transplant recipients [311], an even higher SMR for cSCC has been reported in the UK [50].
Cutaneous SCC in patients with HIV and CLL are also more aggressive; rates of local recurrence after Mohs micrographic surgery are higher than expected and may be partly the result of factors such as dense lymphocytic infiltrate obscuring tumour margins and perineural invasion [312, 313]. Locoregional recurrences are sevenfold more common than in the general population and metastases occur in 18% [33, 38, 314–316].
The true SIR of AKs and Bowen disease in immunocompromised individual OTRs is not known, although prevalence was approximately 50% in one French OTR series [317]. Whether individual dysplastic lesions progress more frequently or rapidly to cSCC compared with the general population is also unclear, although ‘field cancerization’ is a common problem in OTRs and other immunocompromised individuals [296, 318] (Figure 146.1) and refers to the presence of multiple genetic abnormalities in a tissue as a result of exposure to a carcinogen – in the case of the skin, UVR. It manifests clinically as areas of confluent AK and Bowen disease, sites at which cSCC preferentially develop [318]. In one study, cSCC risk was significantly increased in OTR with AK and field cancerization and higher than the corresponding risk of cSCC in OTR with AK but no field cancerization [319].
AK may be contiguous with multiple plane warts in areas such as the dorsum of the hand and forehead, and it may be difficult to distinguish viral and dysplastic lesions from each other without diagnostic biopsy (Figure 146.2). Certain histological features are more common in OTR compared with immunocompetent AK, including mitotic activity, parakeratosis and verrucous change [320].
Porokeratosis is considered to have pre-malignant potential and is well documented in association with immunosuppression [321]. Up to 10% of OTRs are affected [322, 323]. It may cause diagnostic confusion [296] and may rarely progress to cSCC, including metastatic disease [324, 325].
BCC are up to 10-fold more common in the setting of immunocompromise. Anatomical location differs between BCC and cSCC in OTRs: in one UK cohort study, 22.2% BCC versus 8.5% cSCC were truncal whereas 5.2% BCC versus 36.3% cSCC occurred on the hands/forearms [50]. Superficial BCCs on the upper trunk are also observed more frequently in OTRs and HIV [146, 326], and whilst this may indicate pathogenetic differences, it may also be partly the result of closer surveillance of these patients compared with the general population [327]. High-risk BCC (infiltrative/morphoeic, micronodular and basosquamous BCC) do not appear to be more common in OTRs [146], but are more frequent in HIV [328]. Histologically, in addition to an overrepresentation of superficial morphology, inflammatory infiltrate is reduced, squamous differentiation is increased and HPV changes are more frequent in OTRs but tumour depth, perineural and vascular invasion are similar to control tumours [146, 326]. Common differential diagnoses include molluscum contagiosum, particularly in HIV, and sebaceous gland hyperplasia, which is more common in OTRs, often in association with CNI use [329]. Recurrence rates for BCC are significantly increased after conventional and Mohs surgery in both HIV and CLL, although this is not reported in OTRs [38, 146, 303, 315, 316, 330, 331].
Given the strong influence of the immune system on melanoma pathogenesis and progression, a high incidence of melanoma in immunocompromised individuals might be expected; although risks of up to 12 times that of the general population have been reported, most studies report SIRs of 2–5 which is significantly lower than for cSCC [28, 51, 128, 332, 333]. Melanoma arises in three clinical scenarios in OTRs: as recurrence of pre-transplant melanoma, by transmission from the organ donor and, most commonly, de novo post-transplant [28]. In a multicentre European study of 100 patients, only post-transplant melanomas >2 mm (T3 and T4) Breslow thickness had a significantly worse prognosis compared with matched American Joint Committee on Cancer melanoma database controls [334]. However, a study from North America comparing outcomes of 638 OTRs with those expected in control subjects showed worse overall survival, regardless of Breslow thickness [335], as did an Australian study [336]. Despite historic data suggesting recurrence rates approaching 20% in patients with pre-transplant melanoma [337], most recent studies have not shown a worse outcome, although numbers are small [334, 335, 338]. Melanoma is the second most common donor-derived malignancy after renal cancer [339], but the prognosis is much worse, with rapid development of metastatic disease in almost 80% and death in 66% of cases reported to date, including a melanoma resected 32 years previously [340–346]. A past history of melanoma is usually regarded, therefore, as an absolute contraindication to organ donation in most cases [342]. In HIV, melanoma is approximately 2.6 times more common [25, 29] and has a more aggressive clinical course [348]. Melanoma risk is increased approximately 2–7 times in CLL and overall survival is worse, with a SMR of more than 7 in Australia [28, 35, 36, 201, 335, 349].
Risk factors for post-transplant melanoma in OTRs appear to be similar to those in the general population [62, 334]. A particularly important risk factor may be increased numbers of melanocytic naevi. This has been observed in paediatric organ transplant recipients [350, 351] and in individuals with HIV [352]. Also reported is the entity of eruptive melanocytic naevi (EMN), which describes the rapid simultaneous appearance of multiple melanocytic naevi, often hundreds in number, on previously uninvolved sun-exposed skin. Although also reported in otherwise healthy individuals, it arises most often in association with bullous dermatoses, BRAF inhibitors, α-melanocyte-stimulating hormone (MSH) agonists and, in particular, with immunosuppression. It has been described in OTRs [353, 354], HIV [355, 356], in patients immunosuppressed with thiopurines or biological agents (infliximab, etanercept, and alefacept) for conditions such as IBD, psoriasis and myasthenia gravis [357–359] and post-chemotherapy [360]. EMN have been observed to regress upon withdrawal of immunosuppression [361]; generalized eruptive lentiginosis has also been reported in similar groups of patients, although less frequently [358, 519]. BRAF V600E gene mutations have been detected in EMN in the setting of immunosuppression [362]. However, no clear progression of EMN to melanoma has been described [363].
KS is an AIDS-defining disease in HIV and was the most common skin malignancy in the pre-HAART era, and although its incidence has now dramatically reduced, it remains a significant problem in areas where access to HAART is limited [364, 365]. KS is also seen in the context of iatrogenic immunosuppression, particularly in patients from geographical areas of high HHV-8 seroprevalence [366]. In OTRs, it is usually due to reactivation of latent virus [367, 368], although post-transplant acquisition, for example through blood transfusion or via the donor organ (particularly after liver transplantation), is also recognized [369]. In African patients, presentation may be similar to African endemic KS with oedema of the legs (usually unilateral) preceding the appearance of typical purple/red papules, plaques and nodules on the skin. Upper gastrointestinal endoscopy and lung imaging/bronchoscopy may be required to exclude visceral disease. In one series from the UK, almost 15% of African OTRs were affected [50]. The role of HHV-8 serological screening of both recipients and donors is not yet established [370]. However, in a large prospective French series, 13% of recipients who were HHV-8 positive pre-transplant went on to develop KS, as did 3.1% of recipients who received an organ from an HHV-8 positive donor, with no differences in survival and allograft loss compared to HHV-8 negative donors and recipients [367, 369].
Immunocompromised individuals represent approximately 10% of all Merkel cell carcinoma (MCC) patients [371] and have an early age of onset and a more aggressive course [44, 143, 372–377], with a MCC-specific survival of 40% at 3 years compared with 74% in the general population in one large series [371]. In OTRs, the SIR is up to 60, incidence increases with age and time post-transplant, the majority of tumours are on UV-exposed sites and affected patients frequently have multiple other skin cancers [128, 375]. Epidemiological evidence from a US registry-based cohort study suggested a possible synergistic effect of ciclosporin or azathioprine with UV exposure [375] and the poorer prognosis is stage independent [378, 379]. In HIV, SIR was 11.95 in a registry linkage study from the US; the age of onset was significantly younger than in the general population and increased with age and UV exposure [380], with an average survival of 18 months [381]. Compared with healthy controls, MCPyV prevalence and viral load is increased in both OTR and HIV-infected individuals, which could partly explain the increased MCC risk [382]. It is difficult to predict how HAART will affect MCC incidence rates; prolonged life expectancy could lead to a rise in MCC incidence rates, whereas immune restoration due to HIV suppression could help to control cutaneous MCPyV replication and thus lower the risk of MCC development [382]. In a Scandanavian registry study, the SIR for MCC in NHL/CLL was 17.9 and, conversely, that for CLL after a diagnosis of MCC was 15.7 [377] and this reciprocal relationship was also observed in a cohort from Israel [383]. Prognosis for MCC appears to be worse in patients with CLL with an SMR of 3.8 in one population-based study [44]. Although MCPyV has some lymphotropism, it is unlikely that it is directly involved in CLL genesis, as has previously been speculated [384].
Systemic lymphomas are more common in immunocompromised individuals and may present with cutaneous involvement but PCLs are rare. In OTRs, primary CTCLs and cutaneous B-cell lymphomas (CBCL) without nodal or visceral disease at presentation are an uncommon manifestation of PTLD, but their incidence is increased compared with the general population [62, 385–388]. In a multicentre European study of 35 OTRs with PCL, the spectrum of disease was similar to that seen in the general population with 69% T-cell and 31% B-cell lymphomas, the majority of the latter being EBV positive [387]. This contrasts to earlier studies in which B-cell lymphomas were reported to be more common [388]. Prognosis of CD30-positive CTCL was worse than post-transplant mycosis fungoides and its counterpart in the immunocompetent population [387]. Lymphoma is the most common malignancy in HIV [365] and PCLs are also more common in HIV, often of T-cell origin and include CD30-positive anaplastic T-cell lymphoma, which may have a worse prognosis than in the general population [25, 45, 389, 390]. CTCL is rare in HIV but appears to have a conventional clinicopathological presentation. However, a non-clonal lymphoproliferative disorder closely simulating CTCL which is frequently CD8-positive but rarely progresses to true lymphoma (pseudo-CTCL or atypical cutaneous lymphoproliferative disorder) is also recognized [390–392]. PCLs in NHL/CLL are also overrepresented [35, 393].
Tumours of eccrine, apocrine, follicular and sebaceous appendages, particularly sebaceous carcinoma and eccrine porocarcinoma, are up to 100-fold more common in OTRs [378, 394, 395] and are also reported in HIV [396, 397] and NHL/CLL [398, 399]. Sebaceous carcinoma has been linked to dysregulated DNA mismatch repair similar to that observed in Muir–Torre syndrome (MTS): this may be as a result of chronic azathioprine exposure in OTRs [395] and HIV infection has been described to exacerbate skin tumour development in known MTS [400]. Establishing the diagnosis of appendageal tumours prospectively may be challenging: these tumours often arise in individuals who have many other skin cancers and may closely simulate more common skin cancers [395].
Sarcomas other than KS also appear to be overrepresented [401]; the true extent of the increased incidence is difficult to assess given the rarity of these tumours in the general population [378]. AFX and its deeper variant, undifferentiated pleomorphic sarcoma, have been reported in OTRs, HIV and NHL/CLL, with increased rates of recurrence and metastasis in some but not all studies [66, 113, 402–407]. DFSP has been reported in OTRs [64, 378, 408, 409]. Cutaneous leiomyosarcomas have been reported in HIV and OTRs and are frequently EBV positive [401]. Angiosarcomas also occur in both HIV and OTRs and in the latter appear to have a particular predilection for arteriovenous fistula sites [378, 401].
There are few RCTs and limited non-randomized prospective and retrospective studies to guide decision-making in the management of skin cancers arising in immunocompromised individuals. In the setting of organ transplantation, efforts have been made to obtain consensus expert opinion in particularly important skin cancer management areas [141, 410, 411] and a multidisciplinary approach with close dialogue between dermatologists, transplant clinicians, oncologists, surgeons and other relevant health care professionals plays a key part in delivering comprehensive care [412–415]. Many of these same principles apply to the management of skin cancer in other immunocompromised patient groups including CLL [399] and HIV [288].
Table 146.6 Suggested management protocol for primary skin cancers in organ transplant recipients based upon published literature summarized in the text
Stage | Therapeutic considerations | Surveillance |
Pre-transplant | Risk assessment Education Photoprotection Treatment of pre-cancerous lesions |
|
Post-transplant | Baseline risk assessment Education Photoprotection |
Within 6–12 months of transplantation (risk stratified by age at transplantation, skin phototype, number of sunburns – see Figure 146.3) |
No lesions | Education, photoprotection | Surveillance according to risk levels 1–5 (see Figure 146.3) |
AK/Bowen disease/FC | Photoprotection Lesion-directed therapy (e.g. cryotherapy, surgery) Field-directed therapy (e.g. topical 5-FU, imiquimod; diclofenac gel; ingenol mebutate gel; PDT) |
12 months |
Early skin cancer | ||
cSCC | Low risk: surgery (excision, curettage/cautery) Treatment of AK/Bowen disease/FC High risk: surgery (excision, Mohs micrographic surgery); consider sentinel node biopsy, adjuvant radiotherapy, revision of immunosuppression in appropriate circumstances (evidence that mTOR inhibitors more effective if introduced after first cSCC) |
4 months (see Figure 146.3) |
BCC | Infiltrative, nodular: surgery (excision, Mohs’ micrographic surgery) Superficial: surgery (excision, curettage/cautery); non-surgical (cryotherapy; 5-FU; imiquimod; PDT) |
6 months (see Figure 146.3) |
Moderate risk skin cancer (>3–5 KC; early KS) | Treatment of individual lesions as indicated above Rigorous treatment of AK/Bowen disease/FC Consider revision of immunosuppression (reduction or switch to mTOR inhibitor) with transplant clinicians Consider systemic retinoids |
3 months (see Figure 146.3) |
High risk skin cancer (>10 SCC; melanoma; extensive KS; Merkel cell carcinoma; certain appendageal tumours) | Aggressive treatment of individual lesions as indicated above Strong indication to revise immunosuppression (reduction or switch to mTOR inhibitor) with transplant clinicians Strong indication for systemic retinoids |
2 months (see Figure 146.3) |
Adapted from Harwood et al. [50].
AK, actinic keratosis; BCC, basal cell carcinoma; PDT, photodynamic therapy; FC, field cancerization; KC, keratinocyte cancer; KS, Kaposi sarcoma; cSCC, cutaneous squamous cell carcinoma; 5-FU, 5-fluorouracil.
There is no evidence to suggest that low-risk invasive skin tumours and premalignancies (e.g. superficial and nodular BCC, Bowen disease, low-risk SCC) require significantly different management approaches to the general population although the index of suspicion for possible malignancy should be high and the threshold for biopsy correspondingly low, particularly in areas of field cancerization [128, 414, 415].
Excision is the most appropriate option for the majority of tumours although rates of recurrence may be higher even after Mohs micrographic surgery [33, 410] and optimal excision margins have not been defined. Curettage and electrocautery may offer satisfactory clearance rates for selected low-risk tumours, including well-differentiated SCC, and may be considered in certain clinical situations for reasons of cosmesis and convenience [416].
Non-surgical approaches including cryotherapy, photodynamic therapy (PDT), imiquimod cream and 5-fluorouracil cream (see later) may have a therapeutic role in selected cases, particularly in patients with multiple low-risk malignancies in whom repeated surgery is otherwise required. Although response rates may be lower, e.g. for PDT [417, 418], there is no evidence that these agents carry additional significant risk.
Melanoma, MCC, certain appendageal tumours and high-risk SCC and BCC are a particular management challenge in immunocompromised individuals, but there is little evidence to guide optimal treatment, which currently largely parallels that in the general population.
There is also limited evidence on the role of staging investigations such as sentinel lymph node biopsy [419, 420]. Sentinel node biopsy as a staging procedure has been reported in selected OTR with melanomas [411] and may provide prognostic information in other OTR skin cancers, including high-risk cSCC and MCC [410], although there are few data to guide selection of appropriate patients [421]. There are some data to suggest that positron emission tomography–computed tomography (PET-CT) may be promising in nodal staging of cSCC in patients with CLL [41].
It is usually recommended that high-risk OTR tumours require ‘more aggressive’ surgery [410], but this has not been clearly defined. For example, in high-risk cSCC, optimal excision margins are not established and the circumstances in which micrographic surgery is preferable to conventional surgery have not been rigorously evaluated [410, 421]. For melanoma, there is better evidence for recommended surgical excision margins in the general population, but no specific validation in OTRs [411].
In the general population, radiotherapy may be indicated for selected inoperable primary cSCC and BCCs, positive SCC excision margins not amenable to further surgery or as adjuvant therapy for extensive perineural invasion in cSCC and for MCC, although evidence is limited [422]. Radiotherapy may have a similar role in OTRs, although is not the first option for most primary tumours given the high incidence of multiple primary tumours and the long-term risk of second skin malignancies [410, 423].
Consensus expert opinion recommends reduction of immunosuppression should be considered for high-risk tumours in OTRs, but when and how this should be performed remain uncertain. The International Transplant Skin Cancer Collaborative-Skin Care in Organ Transplant Patients, Europe (ITSCC-SCOPE) collaborative has published recommendations for reductions in immunosuppression based on the type of high-risk tumour as well as the type of donor organ [141, 411]. In KS, long-term tumour remissions can be achieved by the reduction (or discontinuation) of immunosuppression [150]. In addition, conversion from CNIs to mTOR inhibitors has been reported to induce dramatic KS regression together with recovery of HHV-8-specific T cells and has been proposed as a first line intervention for patients with KS [368, 424]. However, this may not always prove successful and some patients may be refractory or progress after partial response suggesting that overall reduction in immunosuppression is important in addition to the antitumour effects of mTOR inhibitors in mediating reduction of KS [425, 426].
Current management for advanced and metastatic skin tumours in immunosuppression is similar to the general population in the absence of specific clinical studies, although with the additional strategy of immunosuppression reduction, conversion to mTOR inhibitors or, in some cases, complete withdrawal of immunosuppression [141, 411, 427, 428]. Thresholds for considering each approach and their sequencing have not been validated and tend to be on a case-by-case basis.
There are currently no standards of care for the management of regionally advanced and metastatic cSCC in OTRs [428, 429]. Surgery and radiotherapy approaches are currently broadly similar to those used in the general population [423, 427]. Chemotherapy in this setting has included use of systemic 5-fluorouracil (capecitabine), cisplatin, paclitaxel and retinoids, with transplant-directed dosage adjustment and close monitoring of allograft function [427, 428, 430]. Epidermal growth factor receptor (EGFR) is overexpressed and amplified in SCC; EGFR inhibitors are showing promise in metastatic head and neck SCC [431] and there are case reports of benefit from cetuximab, a chimeric human–mouse monoclonal IgG1 anti-EGFR antibody, in OTR cSCC [432], although fatal pulmonary toxicity has occurred in lung transplant recipients [433]. EGFR tyrosine kinase inhibitors such as gefitinib and erlotinib have also been used [434, 435]. Although use of these targeted therapies in immunocompromised individuals has not been specifically evaluated in clinical trials, some ongoing studies allow their recruitment, e.g. OTRs with aggressive and/or metastatic cSCC are eligible for inclusion in phase II trials of erlotinib prior to surgery or radiation and a phase II study of dasatinib, a multi-kinase inhibitor that targets BCR/Abl and Src family tyrosine kinases 436. Systemic retinoids may also play a role in the management of advanced cSCC, as may revision of immunosuppression and/or switch to mTOR inhibitors (see later).
Targeted BRAF and MEK inhibitors and immunomodulatory therapy with checkpoint inhibitory anti-CTLA4 and anti-PD-1/PD-L1 agents have had a significant impact on the management of melanoma in the general population in recent years, but have not been specifically evaluated in immunocompromised individuals. The checkpoint inhibitors pose a theoretical risk of precipitating organ rejection or GVHD, but ipilimumab has been used safely to treat melanoma after solid-organ and haematopoietic cell transplantation without doing so [437–439] and has also been used in HIV [440]. Use of BRAF inhibitors requires the presence of a targetable BRAF mutation and one study has suggested that their rate in OTR melanomas is less frequent than in control lesions [441]. In terms of conventional chemotherapies for advanced skin cancers, concerns arise when combining with HAART and in OTRs, as there is the possibility of overlapping toxicities, chemotherapy effects on immune status and potential drug–drug interactions [427, 428].
A patient with a past history of skin cancer being considered for transplantation or other iatrogenic immunosuppression is an increasingly common clinical scenario. For most, the benefits will outweigh the risks associated with further skin cancers or possible recurrence and metastasis, but decisions should be made in consultation with a transplant clinician on a case-by-case basis [94, 442]. Patients with pre-transplant KC develop their first post-transplant KC in the same time interval as between first and second cancers in those without pre-transplant KC [50]. The role of mTOR immunosuppression and systemic retinoids in delaying this onset are not established, but close surveillance is warranted [50]. A past history of primary melanoma is perhaps the most challenging situation and the evidence base to guide decision making is limited [28, 334, 335]; expert consensus has proposed that in situ melanoma should not be considered a contraindication to transplantation; for melanoma Breslow <2 mm transplantation is generally deemed an acceptable risk after 2 years and after 5 years for Breslow >2 mm [411]. In cases of metastatic KC or more advanced primary melanoma, transplantation is almost always contraindicated [411].
Possible strategies to prevent immunosuppression-related skin cancer include primary prevention of de novo malignancies, preventing progression of pre-malignant lesions and preventing second and subsequent primary tumours.
European, North American and other expert consensus guidelines recommend advice for strict photoprotection post-transplant including avoidance of sunburn, intentional tanning and unnecessary UVR exposure, especially between 1100 h and 1500 h, April to October in the UK. Sunscreens with broad spectrum, high factor UVB and UVA protection are recommended, together with use of broad-brimmed hats, sunglasses and protective clothing [410, 443, 444]. There is convincing evidence that sunscreens reduce AK, SCC and melanoma but not BCC in the general population [436, 445–447], but it is less clear whether their use post-transplant has the same effect. A prospective open non-randomized trial of liposomal sunscreen in OTRs showed significant reduction at 24 months in AK and cSCC but not BCC [448]. Results from a European RCT are awaited (www.clinicaltrials.gov; identifier NCT01532453), but these data do provide some support for the adoption of photoprotective measures, at least in OTRs at high risk for developing skin cancer. One possible adverse effect of adhering to photoprotective measures is reduction in vitamin D levels, which may already be compromised in OTRs [450]. In the liposomal sunscreen study, 25-hydroxyvitamin D levels were lower in the sunscreen group [448] and monitoring of vitamin D levels in OTRs may be advisable when rigorous photoprotection is being advocated [450].
It is assumed that AK and Bowen disease are precursors of cSCC, providing a rationale for treating AK in order to prevent cSCC, although this remains unproven. In the general population estimated rates of progression of up to 0.075% per year per lesion rising to 0.53% with a past history of skin cancer [451] 65% of SCC arise in AK [452] and 7.7 AK predict a 10% risk of transformation to cSCC within 10 years [453]. No equivalent data exist for immunocompromised individuals, although it is likely that rates of progression, particularly in areas of field cancerization, may be accelerated compared with the general population.
Cryotherapy and surgery are used as lesion-directed treatments for individual AKs/Bowen lesions. In one study, excision and grafting of areas of severe field carcinogenesis within which cSCC were developing was reported as a successful strategy for OTR cSCC prevention [454]. More often, areas of field cancerization are treated with ‘field-directed’ therapies including topical agents (5-fluorouracil cream, imiquimod cream, diclofenac gel and ingenol mebutate gel) and PDT. It is not clear how frequently such field treatments should be used to maintain clearance and whether treatment reduces cSCC risk. Field-directed treatments are often combined with lesion-directed therapies, with a low threshold for biopsy of any persistent lesions to exclude invasive malignancy [128, 415].
5-Fluorouracil (5% and 0.5%) an inhibitor of thymidylate synthetase, is cytostatic to proliferating cells. As a 5% cream it has been used widely for many years, although there are few controlled studies of its efficacy in treating AK in either the general population or OTRs [455]. It is usually applied once or twice daily for 3–4 weeks. In two studies in the general population, it achieved 70% clearance of AK at 6 months [456] and 78% clearance sustained for 12 months [457]. Lower clearance rates have been reported in OTRs [458].
Imiquimod (5% and 3.75%) is an immunomodulator belonging to the family of imidazoquinolines and is an agonist for toll-like receptor 7, activates the innate immune system and generates a Th1 cytotoxic T-cell response. RCTs in the general population demonstrated 64–84% clearance of AK [459, 460]. In a multicentre RCT recruiting 43 OTRs in six countries, complete response was observed in 62.1% and partial response in 80% [461]. In a smaller RCT of 20 OTRs, 50% clinical and 37% histological improvement were obtained and there were fewer cSCC in treated skin over 12 months, although this did not reach statistical significance [462]. Although there is a theoretical risk of immunostimulation and allograft rejection [415], limited use appears to be safe in OTRs and most clinicians limit imiquimod in OTRs to 25–50 mg/day [461, 462].
Diclofenac 3% in 2.5% hyaluronic acid gel inhibits cyclo-oxygenase-2 pathways. In an RCT of 32 OTR with multiple AK treated twice daily for 16 weeks, complete clearance was obtained in 41% but recurrent disease was noted at 24 months in 55% [448].
Ingenol mebutate is a biologically active macrocyclic diterpene ester extracted from the sap of the Euphorbia peplus plant. It has recently been licensed for treatment of AK and has the advantage of requiring only two to three applications. There are currently no clinical trials reported of its use in immunocompromised individuals.
Photodynamic therapy has been assessed in a number of studies in OTRs. In an RCT of aminolaevulinic acid (ALA) PDT with red light, 88% of OTRs responded at 4 weeks, reducing to 48% at 48 weeks, compared with 94% reducing to 72% response, respectively, in immunocompetent patients (P <0.05). This difference possibly reflects either more persistent lesions in OTRs or increased recurrence rates. Response was lowest on the hands/arms, a feature also noted with use of topical agents [463]. In a separate RCT in 17 OTRs using methyl aminolevulinate (MAL) PDT, 76% complete clearance was observed at 16 weeks [464], with similar rates observed for facial AKs in another study, but with significantly lower rates of clearance (72% versus 40%) on acral sites [465]. In a small within-patient RCT comparing MAL-PDT with topical 5-fluorouracil, PDT was found to be significantly more effective, with superior cosmesis and patient satisfaction, but also more pain [458]. Several RCTs have directly investigated PDT in cSCC prevention [417, 418, 466] including one study in OTRs [467]. Although all have shown reductions in AK, this is not always sustained and only one has shown a possible activity in cSCC prevention [466].
Several approaches to secondary prevention have been reported, including revision of immunosuppression (both reduction and conversion to mTOR inhibitors) and use of systemic retinoids. However, evidence regarding the threshold at which to initiate these approaches and how to sequence them is lacking and currently undertaken on an individualized basis.
Revision of immunosuppression Most skin cancers are significantly less frequent when immunosuppression is reduced or stopped following graft failure [139], but there is limited evidence regarding when, and to what extent, immunosuppression should be reduced whilst ensuring allograft survival. An expert consensus survey reported by the ITSCC and SCOPE proposed guidelines for thresholds at which alteration of immunosuppression should be considered which take into account the number of tumours and the type of transplanted organ [141]. Possible approaches include reduction of immunosuppressive drug dosage or discontinuing of one drug in a multidrug regimen, often azathioprine [128]. In the absence of controlled trials, this is ultimately a multidisciplinary decision made on a case-by-case basis.
Conversion to mTOR inhibitors An alternative strategy for revision of immunosuppression is conversion from CNIs to mTOR inhibitors, which have both immunosuppressive and antitumour properties. Retrospective registry data analyses, observational series and post hoc analyses of immunosuppression RCTs have demonstrated a reduced incidence of malignancy in OTRs receiving mTOR inhibitors as part of their immunosuppressive drug regimen [193, 194, 468, 469]. Conversion to mTOR inhibitors specifically to reduce or prevent skin cancer was first reported in OTRs with KS [192, 424, 470, 471] and subsequently for other skin cancers [472], with reduction in the thickness and peritumoural vascularization of cSCC [473]. Four RCTs and a retrospective case series have now confirmed such a switch significantly reduces the incidence of post-transplant cSCC [474–478], and an RCT of everolimus in cardiac transplant recipients is ongoing (www.clinicaltrials.gov; identifier NCT0079918). The published data suggest that conversion to mTOR inhibitors should be undertaken early: reduction in cSCC accrual was most significant for OTRs with only one previous cSCC, with a non-significant reduction of subsequent cSCCs in patients who had multiple cSCCs before conversion to sirolimus [475, 477, 480]. A phase II randomized study to evaluate the effectiveness of converting from CNIs to sirolimus in treating existing cSCC in OTRs is in progress (www.clinicaltrials.gov; identifier NCT01764607). The effect on BCC reduction is less significant, possibly because of differential expression of phosph-mTOR in BCC compared with cSCC [482] and MCC was actually increased in one large registry-based study [375]. The adverse event profile of these agents means that this strategy is not appropriate for all OTRs [475]. A recent systematic review of 21 RCTs confirmed a 56% reduction in KC and 40% reduction in malignancies overall in patients converted to mTOR inhibitors, but the authors also identified an increased risk of death (HR 1.4) with no reduction in malignancy-associated death, bringing into question the role for their use in reducing or preventing malignancy [483].
Retinoids Experimental data suggest that systemic retinoids exert chemopreventive effects during the promotion and progression stages of carcinogenesis [484, 485]. The mechanism is not fully characterized but retinoids alter gene transcription through their actions on cellular retinoid receptors and are antiproliferative, antiapoptotic, immunomodulatory, promote keratinocyte differentiation and arrest growth and replication of HPV [486]. Topical retinoids may have some activity in OTRs [487]. Systemic retinoids have resulted in reductions in cSCC in OTRs in case series [65, 488–491] and three RCTs have confirmed significant reduction in AK and/or cSCC [492–494]. In most studies, the main adverse effects have been cheilitis, xerosis and hyperlipidaemia and may be dose limiting [65, 492, 493, 495]. Since rebound flares of cSCC may occur upon their discontinuation, systemic retinoid chemoprevention should be viewed as long term [65, 485, 492, 494, 495]. Currently, there are no US Food and Drug Administration (FDA) or European Medicines Agency (EMA) approvals for their use in skin cancer chemoprevention, so no formal prescribing guidelines exist. There is a theoretical risk of allograft rejection given their immunomodulatory properties [485], and consensus opinion recommends starting at low dose (e.g.10 mg/day acitretin) and escalating as tolerated to an effective maintenance dose (e.g. up to 30 mg/day acitretin) [65]. Further research is needed to clarify indications for their initiation, as well as the tolerability and efficacy of optimal dosing regimens [485].
T4 endonuclease V (T4N5) is an enzyme involved in the repair of DNA damage after exposure to UVR. Clinical efficacy of a topical liposomal formulation in reducing AKs has been demonstrated in patients with xeroderma pigmentosum [496]. Results of a phase II study of T4N5 lotion looking at safety and efficacy in preventing KC in OTRs are awaited (www.clinicaltrials.gov; identifier NCT00089180).
Capecitabine is an oral prodrug of 5-fluorouracil. In two observational case series of OTRs with multiple and advanced KC in whom it was used at low dose, it significantly reduced incidence rates of cSCC, BCC and AK and associated toxicity, although resulting in discontinuation in 60%, was deemed to be manageable [497, 498].
Afamelanotide is a chemical analogue of α-MSH, a peptide hormone produced by keratinocytes and melanocytes in response to UVB and induces synthesis of melanin in melanocytes as part of the tanning response. Afamelanotide is more potent and longer acting than natural α-MSH and is administered by subcutaneous pellet. Results of a multicentre randomized double-blind placebo-controlled phase II trial to investigate its efficacy in reducing AKs and SCC in OTRs are awaited (www.clinicaltrials.gov; identifier NCT00829192).
Preliminary data exploring the cost–benefits of screening for skin cancer in the general population are available [499], but there is no equivalent information relating to immunocompromised populations [500]. Despite this, regular skin cancer surveillance in OTRs is increasingly undertaken: in 2006, at least annual surveillance was provided by 66% of UK centres and 97% of centres in Australia, with most examinations being undertaken by non-dermatologists in the UK [501]. Given the limited evidence base, the following is a proposed approach to OTR skin cancer surveillance and draws upon expert consensus opinion and prospective cohort data [50, 70, 308, 410, 414, 444, 502–504].
Following transplantation, baseline assessment of individual skin cancer risk and education focusing on photoprotection, self-skin examination and early detection of suspicious lesions is justified for all patients based on current evidence [51, 502], although the timing of initial screening may be stratified according to risk [504]. Such health promotion advice is better recalled and implemented and skin cancer awareness is improved if provided in the setting of a specialist skin clinic [505] and may be further optimised by strategies such as use of written and audiovisual material [449, 506, 507]. Subsequent repeated reinforcement of educational messages both improves recall and adherence to advice [479, 508, 509]. Optimal timing of delivery of this information is unclear; in one study patients expressed a preference for receiving information on malignancy pre-transplant which was then regularly repeated post-transplant [510] whilst in another, post-transplant intervention was preferred [344, 520].
Many consensus guidelines recommend that all OTRs should have a full skin examination by a specialist at least every 12 months [410, 444, 502, 503, 511]. There are, for example, more than 140 000 OTRs in the US [128], 21 000 in the UK, 7000 in Australia and, if extended to other immunocompromised patient groups, the challenge for health care resources would be considerable. Recent data suggest that such surveillance may be better optimised and deployed if targeted by risk stratification, with age at transplant, skin type and previous history of sunburn being the most significant predictive risk factors (Figure 146.3) [73, 504, 512]. Based on accrual of skin cancers established in a UK cohort, five risk categories were defined and follow-up intervals assigned aimed at ensuring cumulative KC remained below 5% (Figure 146.3) [50]. Once pre-malignancy or skin cancer has developed, closer surveillance is justified [50, 69, 70, 481] and in the UK, the National Institute for Health and Care Excellence (NICE) has recommended that immunosuppressed individuals with pre-cancerous lesions or invasive skin cancer should be seen in dedicated clinics for this purpose [513], an approach also being adopted in other countries [128, 502]. Those with low-risk lesions (e.g. AK, Bowen disease) are likely to benefit from at least annual review and following a first skin cancer, surveillance should be more frequent: based on ensuring a cumulative KC risk below 15%, initial intervals of 4–6 months depending on skin cancer type are justified [50], reducing to 3–4 monthly or more frequently in those with multiple or very high-risk tumours (Table 146.7) [50, 69, 70, 308, 410, 414, 415, 502, 503].
Table 146.7 Suggested surveillance protocol in organ transplant recipients with skin cancer
Risk level | Surveillance intervals |
First squamous cell carcinoma | 4, 8, 12 months; then annually if no further cancers |
First basal cell carcinoma | 6, 12 months; then annually if no further cancers |
2nd/3rd cancer | 3, 6, 9, 12 months; then annually if no further cancers |
10th cancer | 2 monthly |
Based on Harwood et al. 2013 [50].
Potential risk reduction strategies such as photoprotection and treatment of AK may be most effective if initiated in the pre-transplant period but have yet to be validated [410, 442, 503]. Risk stratification including, for example, determination of HHV-8 status, may inform appropriate tailoring of transplant immunosuppressive drug regimens.
It has been recommended that patients with CLL be referred for full-body skin evaluation within 6 months of diagnosis but, as for OTRs, there are few data demonstrating the clinical utility of such a screening approach. Nonetheless, 22% of those screened in one cohort had developed skin cancer within 6 months of CLL diagnosis, which the authors cited as evidence to support routine screening [399]. Whether or not routine surveillance is justifiable in CLL and other patient groups such as those with IBD [109, 111, 514], it is likely that heightened awareness of skin cancer risk amongst both patients and health care providers through targeted education initiatives may ultimately improve early detection and treatment of immunosuppression-associated skin cancer.
Several special interest groups have formed in recent years and focus on education (for patients, carers and health care providers), prevention and treatment. In Europe, SCOPE [515] has coordinated a number of research initiatives and, with ITSCC [516], has produced expert consensus guideline documents for the management of post-transplant skin cancer. In the UK, BSSCII (British Society for Skin Care in Immunosuppressed Individuals) [517], a group affiliated to the British Association of Dermatologists, aims to fulfil a similar role with a broad remit across all immunocompromised patient groups. The AT-RISC Alliance (After Transplantation – Reduce Incidence of Skin Cancer) [516] is linked to ITSCC and has developed educational resources for both patients and health care providers.
Skin cancers in immunocompromised patients represent a significant burden of disease for affected individuals and an escalating challenge for health care providers and resources. Risk stratification together with appropriate counselling, surveillance, access to rapid diagnosis and treatment and preventative strategies may reduce the incidence and impact of these skin cancers in the future. However, the evidence base in many of these key areas is limited and further research is urgently required. Such high-risk patients present a model of accelerated skin carcinogenesis and are a relevant population in whom to further investigate pathogenesis, therapeutic and preventative approaches. Indeed, it is likely that many future advances in improving patient outcomes for skin cancer, particularly cSCC, will come from research in immunocompromised patient populations.