A 59-year-old women was seen in February 1999, presenting interlacing white keratotic lines with an erythematous border located bilaterally on the buccal mucosa and the palate. Additional lesions were found on both sides of the tongue; the latter were of plaque-like form and resembled leukoplakia. Particularly the plaque-like lesions were associated with a moderate burning sensation. Upon further clinical examination, additional genital lichen planus lesions involving the introitus vaginae were detected. Two biopsy specimens were taken, which established the diagnosis of OLP. Tests for hepatitis B and C were negative; laboratory analyses revealed no pathological findings. Mycology swabs did not show significant growth of candida albicans. However, the patient received sympathicolytics for anti-hypertensive therapy. In the following, the sympathicolytics were substituted by Ca-antagonists. After contra-indicating conditions were ruled out, the patient received systemic dapsone therapy in combination with vitamine E. Locally, mometasone furoate monohydrate cream was applied twice daily. These measures largely palliated the symptoms and the clinical aspect was improved, but lesions never cleared completely. After approximately one year without any further improvement, dapsone therapy was discontinued and new biopsy specimens were taken. The histological workup, however, was not specific, but a bullous autoimmune disease could be excluded. In July 2001, systemic acitretin at 0.5 mg/kg body weight was started, but stopped within 6 weeks due to elevated serum lipids. Subsequent biopsy specimens of the genital and oral mucosa were consistent with lichen planus. Due to severe symptoms interfering with normal masticatory function, systemic high dose dexamethasone (100 mg on three consecutive days every 4 weeks) was administered three times, unfortunately without any success. As a consequence, in May 2002 an off-label treatment of the recalcitrant OLP was initiated. Tacrolimus 0.1% ointment (Protopic® 0.1%) was applied twice daily. Substantial pain relief was reported after a few weeks of treatment and most of the lesions ameliorated, with the exception of the plaques on the sides of the tongue. After reduction of the frequency of treatment, a recurrence was observed with an increasing number of ulcerated lesions. Consequently, tacrolimus 0.1% was again administered twice daily. In October 2003 an additional biopsy specimen was taken of the ulcerated plaque on the side of the tongue confirming the diagnosis of OLP without evidence of neoplastic transformation. Albeit the OLP did not resolve by treatment with tacrolimus, the patient remained free of symptoms. Hence, the therapy was maintained; however, it was attempted to decrease the frequency of tacrolimus administration. In April 2005, the plaques on the left side of the tongue appeared increasingly compact and a biopsy specimen confirmed the suspected diagnosis of an oral squamous cell carcinoma (Figure 1A and 1B). After exclusion of systemic metastases, a combination of radiation and chemotherapy was initiated to be followed by surgery.
The pre-malignant potential of OLP has been a controversial issue for the past several decades. Indeed, the reported transformation rates vary from 0 to 9% [4, 10, 11]. Some of the controversy can be attributed to the fact that several studies have focused on the development of oral cancer in cohorts of patients with OLP diagnosed on the basis of different criteria and followed for various periods of time. Despite these differences, the majority of studies have reported a rate of malignant transformation of OLP between 0.5 and 2% over a five year period. Unfortunately, no obvious specific clinical features unequivocally predict the potential for cancer development. In this respect it should be noted, that the notion that ulcerative/erosive OLP lesions are more prone to develop into cancer could not be confirmed. A recent meta-analysis revealed that the different types of OLP developing oral cancer have a rather equal distribution: reticular 33%, plaque 29%, atrophic 13%, and ulcerative/erosive 25% . It has also been proposed, that otherwise benign appearing lesions showing an over-expression of p53 in some of the cells implies the expression of mutated p53 which may be an indicator of potential malignant development. Notably, retrospective analysis of the biopsy specimens revealed such accumulation of p53 over-expressing cells in small clusters (Figure 1C and 1D).
The risk of using tacrolimus topically with respect to carcinogenesis has been of some concern. The reason for this is the fact that systemic long-term treatment with tacrolimus in organ transplant recipients increases the incidence of malignant tumours, particularly squamous cell carcinoma . Recently, there was a first report of the development of a squamous cell carcinoma of the penis after topical use of tacrolimus . In March 2005 the FDA has issued a public health advisory to inform healthcare professionals and patients about a potential cancer risk from the use of tacrolimus which was based on animal studies and case reports in a small number of patients . Causative associations are uncertain, but several patients in whom cancer developed after drug use have been reported. For tacrolimus, 19 cases of cancer were reported, involving 16 adults and 3 children under the age of 16. The cancers were diagnosed 21–790 days after the start of therapy. Nine cases involved lymphomas, and 10 involved skin tumours. The majority of the skin tumours occurred at the site of the drug application. Tumour types included squamous cell carcinoma, cutaneous sarcoma and malignant melanoma. The mechanism is thought to be an inhibition of immune competent cells, which normally survey and prevent malignant and pre-malignant cells from developing into malignant tumours.
It was shown that tacrolimus accelerates carcinogenesis in mouse skin when applied topically after the skin had been pre-treated with a tumour initiator (DMBA) . Because of a reduction in the CD4/CD8 ratio found in the lymph nodes in tacrolimus-treated mice, the authors concluded that the immunosuppressive effect of the drug was responsible for its effect in promoting tumourigenesis. This notion was further substantiated by the finding that the concentration of tacrolimus in the draining lymph node was as high as lymph nodes of animals receiving oral tacrolimus, despite the fact that the serum concentration of tacrolimus in topically treated mice was 50- to 100-fold lower . It should be further noted, that despite of an augmentation of apoptosis in T-cells, tacrolimus was also shown to inhibit apoptosis in non-lymphoid cells [17–20]. Moreover, an influence on proteins of some of the most significant cancer signalling pathways (e.g. Erk and p53) has been demonstrated (Figure 2) [19–22]. Consequently, the carcinogenic potential of tacrolimus may be also due to an direct effect, promoting the transformation of initiated cells.
Treatment with tacrolimus leads to Erk activation in neuronal cells and inhibits the induction of p53 following an apoptotic stimulus in several cell systems [19–22]. In the present case, we actually observed an increased presence of phosphorylated Erk 1/2 within the mucosal epithelium after tacrolimus therapy (Figure 1F and 1G). Moreover, strong phosphoErk signals were obvious in the cells of the tongue carcinoma (Figure 1H). Since the expression levels of p53 are difficult to interpret, as in many tumours p53 accumulates as an inactive mutated protein, we also analysed the expression level of Bax (Figure 1E and 1I to 1K). Bax is a proapoptotic member of the Bcl-2 family and its transcription is directly regulated by p53; hence it may serve as a read out of p53 function. Moreover, it was already reported that tacrolimus prevented an increase in the Bax/Bcl-2 ratio following an apoptotic stimulus in U251 cells, while others found a reduction in mitochondrial Bax with the overall expression level remaining unchanged [19, 20]. Strikingly, we observed a reduction of Bax expression in epithelial cells in some areas of tacrolimus treated mucosa and this reduction was also present in the carcinoma cells (Figure 1I to 1K). It should be further noted, that an even more direct link between tacrolimus and Bcl-2 family proteins results of the finding that the tacrolimus-binding protein FKBP 38 blocks apoptosis, binds to Bcl-2 and targets Bcl-2 to the mitochondria .