Short communication

Cutaneous squamous cell carcinoma (SCC) is the most common post-transplant tumour in renal transplanted recipients in Sweden. Five percent of the renal transplanted population develop SCC within 10 years of transplantation. Five years after transplantation the incidence of a first SCC is increased 52-fold, and the incidence of all SCCs is increased 121-fold in renal transplant recipients compared with the general population (1). Half of the patients with a first SCC and three-quarters of those with 2 or 3 SCCs are at risk of multiple subsequent SCC and, finally, of metastatic SCC within 10 years after transplantation. These patients are in need of intense skin surveillance. However, the risk factors that indicate the need for intensified skin surveillance have not been completely identified (1–3).

In this prospective clinical observational study at Uppsala University Hospital, Sweden, 73 kidney or simultaneous pancreas and kidney transplanted patients were included. All patients had at least one post-transplant SCC in situ or invasive SCC diagnosed from September 2006 to December 2012 and they were followed for 2 years (4). The aim was to identify the risk factor or risk factors that are the most significant in predicting de novo SCC.

This study was approved by the Regional Ethical Review Board in Uppsala (Dnr 2007/032) and registered with ClinicalTrials.gov (NCT02241564).

The recorded risk factors for de novo SCC, as well as a comparison between patients developing and not developing de novo SCC are presented in Table I. Skin types were assessed according to Fitzpatrick’s classification (5). Sun exposure was assessed based on anamnestic information. Patients with low and medium sun exposure were classified as Group 1 (low risk) and those with a history of active sun bathing (before transplantation or continuously) were classified as Group 2 (high risk). Group 1 had normal skin status according to age or minor skin lesions such as aged skin, telangiectases, elastosis, and some solar lentigines. Group 2 had actinic keratosis, many solar lentigines or more advanced lesions, such as numerous actinic keratoses and field cancerization.

Immunosuppression. All patients except one had calcineurin inhibitor (CNI)-based maintenance immunosuppression, and all patients except 4 (5%) had corticosteroids in combination with CNI at inclusion. Nineteen patients in the study used everolimus as maintenance immunosuppression for > 21 months.

Statistical analyses were performed using software R version 3.1.1 (The R Foundation for Statistical Computing, Vienna, Austria).

The sample size was small compared with the number of potential explanatory variables and it was unlikely a priori that statistical significance alone would be able to identify all prognostic covariates. Therefore, besides Cox regression analysis of all candidate variables, we removed covariates 1 at the time to determine which could be excluded with the smallest deterioration in predictive value.

During follow-up 31 patients (42%) developed de novo SCC, of which 15 (48%) were in situ SCC and 16 (52%) were invasive SCC (Table I).

Patients with more than one previous cutaneous SCC had a 5-fold (1–21) higher risk (R²) of a subsequent skin SCC, and patients with more than two cutaneous SCC had a 14-fold (3–63) higher risk (R²) compared with patients without a previous SCC (p < 0.001) (Table SI). Lower age at baseline SCC contributed more than the remaining risk factors to de novo SCC when the covariates were removed (not shown). Histology of the baseline SCC did not correlate to the histology of the subsequent SCC.

In most skin classifications of transplanted patients, patients with a first skin SCC belong to the higher risk group. It is essential to identify all patients with SCC because SCC can lead to metastasized disease in immuno-compromised patients. SCCs are among the most immunogenic types of cancers, and the risk of subsequent SCCs decreases dramatically when immunosuppression is interrupted (6). mTOR inhibitors probably affect SCCs (7) and SCC is the only tumour type where different features of the cancers have been found in the transplanted population compared with the general population (8). Still, almost half of the patients with a first SCC do not develop a de novo skin SCC within 10 years. This finding means that 95% of all renal transplanted patients in Sweden are not at risk of numerous de novo and later metastasizing SCCs (1).

Although several investigated risk factors are known to contribute to the development of SCC, the detailed description of the skin based on signs of sun damage did not help to predict risk in our model (1, 9, 10).

The finding that the number of previous SCCs correlates with the development of de novo SCCs is in accordance with earlier studies, and seems to be the strongest predictive risk factor (1, 3, 11, 12).

Risk factors of a de novo SCC after having at least one SCC earlier have been investigated in 2 retrospective studies: multiple skin cancers at first dermatological visit, skin type I and renal transplantation before 1984 were the most important contributors to a de novo SCC with a history of at least one previous SCC (3, 11).

One weakness of this study is the limited number of patients, i.e. the parameters shown to contribute to de novo SCCs are probably significant, but the other parameters might be underestimated. For example, we found the histology of the baseline SCC did not correlate with that of a de novo SCC. This finding differs from the results of 2 retrospective studies (11, 12).

mTOR inhibitors have, in earlier studies, been associated with reduced number of de novo SCC (in situ and invasive) and prolonged interval between baseline and de novo SCC (13–15). The protective effect seems, however, to depend on the number of previous SCC lesions; the fewer SCCs the better anti-tumour effect (15). In our study mTOR inhibitors did not contribute to a more advantageous skin SCC development than calcineurin-based immunosuppressive protocols. A bias for this risk variable was that only patients with advanced skin lesions were willing to change the main immunosuppression to mTOR inhibitors.

From a clinical and resource-saving perspective, it is essential to make an early identification of patients most prone to develop de novo SCC. Based on our findings, patients with more than one SCC, particularly in combination with lower age at diagnosis of SCC, should be followed up more intensively than those without a previous SCC. Also, although ultraviolet exposure is a known risk factor for SCC in the skin, our results show that patients with a lower grade of clinical signs of sun damage are also at risk for development of de novo SCC.

This work was supported by the Kidney Foundation and the Bergholm Foundation. The authors gratefully acknowledge assistance from Prof Lars Holmberg, Prof Mats Lambe and Fredrik Sundin at the Regional Tumour Registry in the Uppsala–Örebro region.