Vol. 125 No. 3 March 2018

Drug-associated hyperpigmentation of the : report of four cases Konstantinos I. Tosios, DDS, PhD, Eleni-Marina Kalogirou, DDS, MSc, and Alexandra Sklavounou, DDS, MSc, PhD

Objective. The aim of this study was to describe 4 patients with oral mucosa hyperpigmentation associated with 4 drug classes and to review the relevant literature. Study Design. Two patients under imatinib and hydroxychloroquine treatment exhibited diffuse palatal hyperpigmentation and 2 patients treated with minocycline and golimumab showed multifocal pigmented macules. In all cases, biopsy was performed. Results. Microscopically, in all cases, there was no increase in the number of melanocytes in the epithelium, and pigment gran- ules were present in the lamina propria. The pigment granules in minocycline- and golimumab-associated hyperpigmentation were seen in the superficial lamina propria and reacted for silver but not iron, whereas in imatinib- and hydroxychloroquine- associated hyperpigmentation, pigment granules were found in the reticular lamina propria and reacted for both silver and iron. A review of the literature found 38 cases of hyperpigmentation of the oral mucosa attributed to minocycline, 23 to imatinib, 1 to hydroxychloroquine without microscopic documentation, and none to golimumab. Conclusions. The temporal relationship between pigmentation and onset of drug effect, resolution following drug withdrawal, and exclusion of other causes support the diagnosis of drug-induced hyperpigmentation. Microscopic examination may be con- tributory to diagnosis, as there are differences among drugs with regard to the distribution of pigment granules and the histochemical reactions of the drugs. (Oral Surg Oral Med Oral Pathol Oral Radiol 2018;125:e54–e66)

Many systemically administered drugs are associated with, hydroxychloroquine, and golimumab) and review the per- among other adverse events, hyperpigmentation of the tinent literature. oral soft tissues, with antibiotics, antimalarials, chemo- therapeutics, hormones, and tranquilizers being the best CASE REPORTS known examples.1 Drug-associated hyperpigmentation of Case 1: Minocycline-associated soft tissues shows nonspecific histologic changes, and hyperpigmentation various mechanisms may be involved in its pathogenesis.2 A 55-year-old female was referred by her dentist for di- They include deposition of the drug substance in tissues, agnosis and management of an area of black pigmentation as seen with heavy metals such as silver, gold, or bismuth, on the gingiva around the lower left canine. It had first or of pigmented metabolites of the drug that chelate with been noticed approximately 5 months earlier, but it had tissue components (i.e., melanin or iron); increase of the been extending recently. The patient’s medical history melanin content of the mucosa directly, by initiation of was significant for cystic acne complicated by Staphy- melanin production by melanocytes, or indirectly, through lococcus infection, which was being treated with 200 mg/d a postinflammatory reaction after a fixed-drug erup- minocycline hydrochloride for the last 8 months, and tion; and accumulation of iron after damage of the vessels 125 mg/d of terbinafine hydrochloride for onychomy- and hemorrhage.2-4 The mechanisms of other drugs remain cosis diagnosed 2 months ago. There was, in addition, unknown. vitamin D deficiency, which was being managed with There is poor evidence for a causal association between vitamin D supplementation. Her recent complete blood a drug and soft tissue hyperpigmentation because most count (CBC) total was within normal limits. She had been reports are case studies.5 Because the dosing regimen and smoking approximately 20 cigarettes per day since she the temporal association between medication intake and was 25 year-old. the occurrence of hyperpigmentation remain unclear, ad- ditional cases may be of clinical use. We describe the clinical, histologic, and histochemi- cal features of 4 white patients with diffuse or multifocal hyperpigmentation of the oral mucosa associated with Statement of Clinical Relevance 4 different classes of drugs (i.e., minocycline, imatinib, Diffuse/multifocal oral mucosal pigmentation may present as a drug-associated adverse effect. Correct Department of Oral Medicine and Pathology, Faculty of Dentistry, National and Kapodistrian University of Athens, Athens, Greece. diagnosis is based on medical history and exclusion Received for publication Jan 12, 2017; returned for revision Oct 14, of other conditions; microscopic examination is con- 2017; accepted for publication Oct 23, 2017. tributory to diagnosis, as there are differences among © 2017 Elsevier Inc. All rights reserved. drugs with regard to pigment granule distribution and 2212-4403/$ - see front matter histochemical reactions. https://doi.org/10.1016/j.oooo.2017.10.006 e54 OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e55

Clinical examination showed hyperpigmentation on the A biopsy on the area of the lower right central incisor facial gingiva distal to the lower right central incisor, in- revealed minocycline-associated hyperpigmentation (see volving both marginal and attached gingivae. The below). Treatment with minocycline and terbinafine hy- pigmentation was of a pale brown color with a dark spot drochloride was terminated a month later, and although (Figure 1A). Areas of pale brown pigmentation were also the patient did not quit smoking, the lesion did not recur, seen on the facial attached gingiva of the maxilla, prox- no additional lesions appeared, and the hyperpig- imal to the upper right first premolar and canine mented areas faded considerably. (Figure 1A, black arrows) and distal to the upper left canine (Figure 1B); on the , proximal to the right Case 2: Imatinib-associated hyperpigmentation first premolar to distal to the lateral incisor (Figure 1A, A 61-year-old male was referred by his hematologist for white arrows); and between the left lateral incisor and assessment of palatal pigmentation. The patient was the canine. The rest of the oral mucosa was normal. No unaware of the presence of the lesion, although he re- ocular, skin, or nail hyperpigmentation was evident or called that approximately 6 months earlier, his regular reported by the patient. dentist had noticed “black macules” on the during

Fig. 1. Multifocal gingival hyperpigmentation in a minocycline-medicated patient. A, Hyperpigmentation of brown color with a dark spot on the facial marginal and attached gingiva distal to the lower right central incisor; areas of pale brown pigmentation proximal to the upper right first premolar and canine (black arrows); and proximal to the lower right first premolar to distal to the lateral incisor (white arrows). B, Pale brown hyperpigmentation on the attached gingiva distal to the upper left canine. C, Micro- scopic examination showed increased deposition of melanin in the cells of the basal and parabasal layer (black arrows) that (D) were positive for methenamine-silver stain. C, Hematoxylin and eosin stain; D, Grocott (Gomori) methenamine-silver stain (orig- inal magnifications C and D, ×400). A high-resolution version of this slide for use with the Virtual Microscope is available as eSlide: VM04566. ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e56 Tosios, Kalogirou and Sklavounou March 2018

Fig. 2. Palate hyperpigmentation in a patient treated with imatinib. A, A symmetric, well-defined, homogeneous gray-black dis- coloration of the hard palate. B, Numerous brown-yellow granules within the reticular lamina propria, among collagen fibers, which were positive for (C) methenamine-silver stain and (D) iron stain. Normal melanin deposition and normal melanocytes number were seen in the covering epithelium (B, hematoxylin and eosin stain; C, Grocott [Gomori] methenamine-silver stain; D, Perls’ iron stain; original magnifications B-D, ×400). A high-resolution version of this slide for use with the Virtual Microscope is avail- able as eSlide: VM04565. routine dental examination, but it did not prompt further symmetric and well defined and did not blanch on pressure investigation. Nineteen years ago, he had undergone bone (Figure 2A). The rest of the oral mucosa appeared normal. marrow transplantation for chronic myeloid leukemia An incisional biopsy revealed imatinib-associated hy- (CML), and for the past 11 years, he was being treated perpigmentation (see below). No modification was made with imatinib mesylate 400 mg/d. According to his at- to his medication, and approximately 31 months later, tending physician, the patient had not received palatal hyperpigmentation was still present, but without hydroxyurea, prednisone, chloroquine, or minocycline in any significant change. the last 11 years. The patient’s most recent CBC was sig- nificant for mild leukopenia and anemia. He was taking Case 3: Hydroxychloroquine-associated lepirudin daily, and he did not smoke. He could not recall hyperpigmentation palatal trauma or any kind of symptomatic irritation to A 53-year-old female was referred by her dentist for eval- this area. No skin or nail pigmentation was evident, uation of an area of black pigmentation on her palate; and he was unaware of any changes in cutaneous the pigmentation was not present on a previous routine pigmentation. examination performed approximately 2 years earlier. The Clinical examination showed that the hard palate had patient was receiving treatment for rheumatoid arthritis a homogeneous gray-black discoloration that was with 400 mg/d hydroxychloroquine for approximately 5 OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e57

Fig. 3. Palate hyperpigmentation in a rheumatoid arthritis patient under hydroxychloroquine treatment. A, A symmetric, diffuse, homogenous blue-brown hyperpigmentation on the hard palate. B, Many fine, brown-yellow granules of approximately the same size in the reticular lamina propria that stained with (C) methenamine-silver stain and (D) iron stain. Note the evident infiltration by erythrocytes in the papillary dermis (B, hematoxylin and eosin stain; C, Grocott [Gomori] methenamine-silver stain; D, Perls’ iron stain; original magnifications B-D, ×400). A high-resolution version of this slide for use with the Virtual Microscope is avail- able as eSlide: VM04564. years and for hypertension with nebivolol hydrochlo- of the increased risk for retinopathy in patients with ride 5 mg daily. She recalled using “prednisone for a long hydroxychloroquine-induced hyperpigmentation,6 this period many years ago” and reported that she did not patient was referred for ophthalmologic evaluation. Ocular smoke. examination results were within normal limits; treat- Clinical examination showed homogeneous, diffuse, ment with hydroxychloroquine was stopped, and the blue-brown pigmentation on the hard palate (Figure 3A) palatal hyperpigmentation faded considerably within the and a small brown macule on the attached gingiva of next 30 months. the lower left canine. The palatal lesion was symmetric and diffuse and did not blanch on pressure. The rest of Case 4: Golimumab-associated hyperpigmentation the oral mucosa was within normal limits. Hyperpig- A 67-year-old woman was referred by her rheumatolo- mentation of the skin of the neck and thorax was also gist for evaluation of numerous macules on the oral present. mucosa, with a request for assessment to rule out mela- An incisional biopsy revealed hydroxychloroquine- . The lesions had been observed by the patient 5 associated hyperpigmentation (see below), and because days earlier, on self-examination. Her last regular dental ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e58 Tosios, Kalogirou and Sklavounou March 2018

Fig. 4. Multifocal oral mucosal hyperpigmentation in a patient treated with golimumab. Dark to light brown, smooth macules with well-defined borders (A) on the right and (B) left buccal mucosa, (C) the hard palate, and (D) the upper close to the vermillion border. E, Microscopic examination from a buccal macule in the golimumab-treated patient revealed increased melanin deposi- tion in the basal and parabasal layers and dark pigmented granules similar to melanin in the superficial lamina propria (black arrows) that stained positive (F) for methenamine-silver stain (E, hematoxylin and eosin stain; F, Grocott [Gomori] methenamine-silver stain; original magnifications E and F × 100). A high-resolution version of this slide for use with the Virtual Microscope is avail- able as eSlide: VM04563. checkup was done approximately 3 months ago, and her Clinical examination showed macules on the buccal dentist did not report any abnormal oral finding. She was mucosa (6 on the right [Figure 4A] and 1 on the left being treated for rheumatoid arthritis with 10 mg/d of [Figure 4B]), hard palate (3 [Figure 4C]), and the upper leflunomide for approximately 11 years and 50 mg per lip close to the vermillion border (1 [Figure 4D]). The month of golimumab for approximately 6 years; macules were homogeneous, were of a dark to light brown levothyroxine sodium for Hasmimoto disease; and color, had smooth and well-defined borders, and mea- nebivolol hydrochloride for hypertension. Her medical sured 0.3 to 0.5 cm. The mucosal texture was normal, history was significant for systemic lupus erythemato- and there was no blanching or hemorrhage on pressure. sus, which had been treated for 3 years with The rest of the oral mucosa was within normal limits. corticosteroids and hydroxychloroquine, but the treat- No pigmentation was reported by the patient or was ment had stopped 8 years ago. She had been smoking evident on her skin or nails. Biopsy was performed on 20 cigarettes per day since she was 31 year-old. 1 palatal and 1 buccal macule. OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e59

Golimumab treatment was continued, but approxi- identical to those of the oral melanotic macule. In the mately 14 months later, she had to discontinue golimumab former case, there was, in addition, a focally intense in- because of the development of nephrotic syndrome with flammatory infiltration by lymphocytes and plasmacytes. heavy proteinuria, for which she is currently under treat- The granules were positive for methenamine-silver stain ment. The patient reported that no additional oral mucosal (Figures 1D and 4F) and negative for iron stain. lesions developed and that the existing ones tended to In patients with imatinib-associated hyperpigmenta- fade; however, because of her severe medical condi- tion (Figure 2B) and in those with hydroxychloroquine- tion, she did not present for re-examination. associated hyperpigmentation (Figure 3B), there was no increased deposition of melanin in the basal cell layer Microscopic examination and no increase in the number of melanocytes, but in the In all cases, biopsy was performed under local infiltra- latter, erythrocytes were evident in the superficial lamina tion anesthesia, and 5-μm-thick formalin-fixed and propria. Many fine, brown-yellow granules of approxi- paraffin-embedded tissue sections were stained with he- mately the same size were seen within the reticular lamina matoxylin and eosin, Grocott (Gomori) methenamine- propria, among collagen fibers or intracellularly within silver stain, and Perls’ iron stain, by using standard macrophages. They were not consistent with melanin procedures. (Grocott [Gomori] methenamine-silver stains because there was no increased deposition of melanin in melanin black, and Perls’ iron stains ferric iron blue the basal cell layer or with hemosiderin. However, they [Prussian-blue reaction]). The mucosal fragments were stained with methenamine-silver stain (Figures 2C and covered by stratified squamous epithelium. 3C) and iron stain (Figures 2D and 3D), probably rep- In patients with minocycline-associated hyperpigmen- resenting deposition of the drug or drug products that are tation (Figure 1C) and in those with golimumab-associated chelated to iron or express chemical groups of melanin hyperpigmentation (Figure 4E), there was increased de- or ferric iron that react with the respective histochemi- position of melanin in the cells of the basal and parabasal cal stains. Figure 5 presents the main microscopic and layers, as well as in cell processes, but no increase in the histochemical features of drug-associated hyperpigmen- number of melanocytes. Fine, dark-pigmented granules tation of the oral mucosa by class of medication. It should similar to melanin were seen in the superficial lamina be noted that the information about hydroxychloroquine- propria, free among collagen fibers or intracellularly and golimumab-associated hyperpigmentation was based within macrophages. The microscopic features were only on our cases, as microscopic and histochemical

Fig. 5. Main microscopic and histochemical features of drug-associated hyperpigmentation of the oral mucosa by class of med- ication, according to our cases and literature review.3,4,7-12 The information considering hydroxychloroquine- and golimumab- associated hyperpigmentation was based only on our cases because their microscopic and histochemical characteristics have not been previously described. ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e60 Tosios, Kalogirou and Sklavounou March 2018 characteristics of these types of hyperpigmentation have in the 2 patients with rheumatoid arthritis. Therefore, not been previously described. the possibility of another drug playing a causative role in the development of hyperpigmentation cannot be DISCUSSION excluded, although the oral lesion(s) were recognized Diagnosis of drug-associated hyperpigmentation is based by the referring physicians during the use of or shortly on the temporal relationship between hyperpigmenta- after initiating the use of the purportedly offending tion and initiation of drug administration, resolution drug. following drug withdrawal, aggravation upon re-exposure, and exclusion of other etiologies with similar clinical or Minocycline-associated hyperpigmentation histologic features.13 The “diverse and extensive”7 dif- Minocycline is a semi-synthetic tetracycline used as a ferential diagnosis of multifocal or diffuse mucosal broad-spectrum antibiotic in the treatment of various hyperpigmentation has been thoroughly discussed in pre- systemic infections; an anti-inflammatory drug in the vious publications3,7,14,15 and includes racial/normal management of dermatologic diseases, in particular acne pigmentation; smoking-related pigmentation; post- vulgaris; and an immunosuppressive drug in immune- traumatic or postinflammatory hyperpigmentation; mediated diseases, such as rheumatoid arthritis.19 pigmentation associated with systemic diseases such as Hyperpigmentation is an uncommon, but well-documented Addison disease and hemochromatosis; syndromes such adverse effect of minocycline treatment.5 It may affect as Peutz-Jegher syndrome, McCune-Albright syn- various organs, in particular skin and nails, subcutane- drome, and neurofibromatosis; and deposition of ous fat, bones (“black bone disease”) and cartilage, exogenous pigmented material. ocular mucosa, thyroid gland, viscera, and body fluids.20,21 Localization of lesions on the gingiva in the case of Hyperpigmentation of skin, nails or bones was seen in the patient treated with minocycline could be consis- 15% of 333 case reports of adverse events of minocycline tent with normal (racial) pigmentation or smokers’ treatment over a nearly 40-year period.20 Although melanosis.16 However, acute onset of lesions and exclu- pigmentation is considered dose dependent,22,23 the sive marginal gingival distribution conflicts with such a total dose or duration of medication does not affect diagnosis. These conditions could also be included in the intensity of skin hyperpigmentation that may be the differential diagnosis of golimumab-associated pig- associated with other factors, such as exposure to sun- mentation, but in that case, gingival involvement and light, concurrent treatment with other drugs, and genetic early onset would be expected. Although the patients predisposition.23 in two of the cases presented here did not quit smoking, There are at least 3 well-described clinicopathologic no additional lesions appeared, and the hyperpig- types of minocycline-associated skin hyperpigmenta- mented areas either faded considerably or tended to tion that may present alone or in combination.2,21,23,24 Types fade. In the case of golimumab, the microscopic fea- I and II present mostly with a blue color, mixed with gray tures were consistent with oral melanotic macules that or brown. Type I is localized to sites of skin inflamma- are typically solitary, whereas lack of striate melano- tion, trauma, or scarring and is not influenced by duration nychia conflicted with a diagnosis of Laugier-Hunziker of therapy or cumulative dose, whereas type II is diffuse syndrome.17 A traumatic event or inflammation preced- and affects normal-appearing skin, in particular the skin ing the development of the lesions could not be of the extremities. Type III is also diffuse, has a mostly substantiated—that is, patients with palatal lesions did brown color, and affects normal-appearing sun-exposed not wear a denture that could cause erythematous can- skin. Types II and III are usually associated with long- didiasis, and there were no clinical features of a chronic term high-dose minocycline treatment.23,24 Type IV, which inflammatory disease, such as .8,18 Such is less common, presents as circumscribed hyperpig- conditions would be expected to cause pain and pa- mentation of blue to gray color, particularly within acne tients would remember their occurrence. There was no scars on the back.23 clinical indication of a systemic disease, and the late The microscopic features of minocycline-associated onset of the pigmented lesions was not consistent with skin hyperpigmentation are described mostly in case a genetic syndrome. CBC was within normal limits, reports, which are occasionally contradictory with regard with the exception of the patient treated with imatinib, to the presence of melanin.19,23 Typically, type I and II and in that case, the mild leucopenia and anemia that lesions demonstrate pigment granules in the dermis, were present were not considered contributory to the aligned between collagen fibers or accumulated in mac- development of the pigmented lesion. Finally, the cases rophages distributed around vessels or adnexa that presented here were in patients with chronic diseases react with Fontana-Masson silver staining and Perls’ running a protracted course, where other drugs associ- iron staining, but the staining reaction is not consid- ated with hyperpigmentation could have been used or ered confirmatory for the presence of melanin in all were used—that is, prednisone or hydroxychloroquine cases.24 The study by Argenyi et al. reported that no OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e61 melanosomes are seen on electron microscopic a case of gingival pigmentation,13 the distribution of the examination,2 whereas X-ray energy spectroscopic ex- granules in the basal and parabasal layers of the epithe- amination showed that they contain iron, sulfur, chlorine, lium, as well as within the macrophages in connective and calcium.2 On the basis of those findings, it is thought tissue, was consistent with melanin. The clinical and mi- that the lesions represent oxidized minocycline (type I) croscopic features of the latter case are comparable with or minocycline products (type II) chelated with iron23,25 those of our case, where, in addition, the pigmented gran- because the normally yellow minocycline becomes black ules were positive with methenamine silver staining, but after oxidation.26 In type III lesions, iron is absent, and negative for iron. Although the patients were smokers in there is increased melanin deposition in basal keratinocytes both cases, making smokers’ melanosis a possible di- and adjacent dermal melanophages,24 possibly as a result agnosis, we found that pigmentation was confined to the of minocycline-induced melanin production or forma- marginal gingiva and did not extend to the attached tion of minocycline–melanin complexes.27 Type IV gingiva. Induction of melanin production by minocycline resembles type I histopathologically, but granules react or its metabolites and formation of minocycline-melanin with Fontana-Masson silver staining but not with Perls’ complexes that resist degradation27 cannot be excluded, iron staining.23 but an effect of exposure to sunlight because of the lo- Intraoral minocycline-associated hyperpigmentation is cation of the lesions13 seems unlikely. Previous reports common in alveolar bone and predominantly in the middle have classified oral lesions as consistent with type I or third of the crown or roots of erupted teeth,7,13,25,28 causing II cutaneous hyperpigmentation,7,27 but this classifica- darkening of the overlying mucosa as the black color tion was developed for skin lesions, so more cases should appears through the translucent non-pigmented be studied for clinical utility in oral lesions to become mucosa.7,13,25 Hyperpigmentation of the oral soft tissues evident. in mucosal sites not directly associated with alveolar bone The course of minocycline-induced pigmentation of or teeth, such as the tongue, , and buccal mucosa, as the oral mucosa is unknown, and it is suggested that it well as the gingiva and hard palate,7,13 is very rare and may persist even years after withdrawal of the offend- no case was seen among 331 patients who were treated ing drug.27 Q-switched ruby laser, alexandrite, or for acne with at least 100 mg/d of minocycline for a neodymium:YAG laser have been successfully applied minimum of 6 months.28 In contrast, approximately 10% in the management of this condition.13 of the patients showed blue pigmentation of alveolar bone.28 Treister et al.7 reviewed 33 cases of minocycline- Imatinib-associated hyperpigmentation induced oral hyperpigmentation and added 2 new cases. Imatinib mesylate is a tyrosine kinase inhibitor that targets They noticed, however, that most of the reviewed cases the adenosine triphosphate (ATP) binding site of BCR- could represent staining of the underlying alveolar or ABL tyrosine kinase, as well as the tyrosine kinases of palatal bone because in cases where a mucosal flap was platelet-derived growth factor receptor-β, c-kit, and ABL.30 elevated, the bone was pigmented, whereas when a biopsy It is used in the management of Philadelphia chromosome- was performed, it showed no mucosal deposition of stain- positive CML and gastrointestinal stromal tumors.31 ing material. Our literature review from 2000 to 2016 Cutaneous hypopigmentation was seen in 40.9% of 118 revealed 3 additional cases with gingival involvement,13,27,29 patients with CML treated with imatinib and was attrib- which, in contrast to our case, was associated with hy- uted to inhibition by imatinib of c-kit,30 which, along with perpigmentation of other intraoral13,27 or extraoral sites.13,29 its ligand stem-cell factor (SCF), has a major role in me- Our patient had been treated with a daily dose of 200 mg lanogenesis, melanocyte homeostasis, and ultraviolet minocycline for less than 1 year. In other cases, hyper- B–induced pigmentation.30,31 In contrast, cutaneous hy- pigmentation was associated with daily doses of 100 mg26 perpigmentation is a rare adverse event seen in only 3.6% or 200 mg13,27 and appeared even after 2 weeks of of 118 patients with CML receiving imatinib,30 which is treatment.27 dose related and reversible with discontinuation of the There is limited information on the microscopic fea- drug.31 tures of minocycline-associated hyperpigmentation of oral Pathogenetic hypotheses include stimulation of soft tissue.7,13 In a review of 33 cases of intraoral melanogenesis,31 resulting from activation rather than minocycline pigmentation reported up to 2004, all in- inhibition of the c-kit receptor by imatinib8 or because volved alveolar bone, and in 6 cases where biopsy was of a lichenoid reaction that regresses,9 or apoptosis of performed on the gingival hyperpigmentation, no pigment melanocytes and melanin incontinence, which is caused was seen in the mucosa.7 In 2 cases of palatal by increased concentration of imatinib (>12 mol/L).32 pigmentation,7 the granules stained for iron but were not However, accumulation of the drug in macrophages considered morphologically consistent with hemosid- and generation through its processing of chemical groups erin, whereas they did not bleach with hydrogen peroxide, that react with histochemical stains, such as melanin thus excluding the presence of melanin. In contrast, in or lipofuscin,9 or chelation of the drug’s metabolites ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e62 Tosios, Kalogirou and Sklavounou March 2018 with iron and melanin3,9,31 seem more reasonable in another 4 cases, the pigment granules were positive explanations. for both melanin and iron3,12; and in 1 case, only the pres- Pancholi and Taneja33 reviewed 16 cases of palatal pig- ence of iron was recorded.11 In our case, the pigment mentation in patients receiving imatinib,3,4,8-11,14,34-38 in 1 granules reacted for both melanin and ferric iron, al- case in combination with hydroxyurea,37 and added 3 more though there was no increased deposition of melanin in cases; an additional 4 cases have been reported since that the basal cell layer or any hemorrhage. Those findings review.12,39,40 More cases may have been recognized are more consistent with the presentation by the drug but have not been described41; 2 cases of gingival byproducts of chemical groups that react like melanin pigmentation42,43 and 1 case of teeth pigmentation44 have or lipofuscin,9 but the predilection for this phenome- not been documented fully. Sixteen patients with imatinib- non to occur in the palate cannot be explained. associated palatal pigmentation had CML3,4,8-12,34-36,38-40 Hyperpigmentation is not an indication for discon- and 1 each had pelvic fibromatosis,3 leiomyoblastoma,8 tinuation of imatinib in patients with CML, and lesions and acute lymphoblastic leukemia.14 In our case where tend to remain stable, even after 2 years.8 the patient was receiving imatinib 400 mg/d for 11 years, the drug dose was most commonly associated with palatal Hydroxychloroquine-associated pigmentation,33 the lesions were present for at least 6 hyperpigmentation months before diagnosis. The exact time of onset of pig- A common adverse effect of antimalarial drugs that are mentation was uncertain in most cases because the palatal used in the management of rheumatologic or dermato- location and lack of symptoms made their identifica- logic diseases because of their anti-inflammatory and tion incidental, but long-term exposure to drug (5-6 years, immune-modulating properties is hyperpigmentation of range 3 months to 10 years) is suggested.33 In 1 patient, the skin.6 It may occur with all classes of antimalarial hyperpigmentation was seen after just 3 months of treat- drugs, but it is more common with chloroquine47 and ment with imatinib 800 mg/d,34 and the hyperpigmentation quinacrine15,48 and rare with hydroxychloroquine.6,49 In could be probably attributed to increased drug concen- 1 study, it was seen in 67 of 194 patients (35%) treated tration in plasma.32 Our patient did not show any other with chloroquine, but only 2 of 15 patients (13%) treated sign of oral or cutaneous pigmentation. Skin pigmenta- with hydroxychloroquine.50 The median duration of tion coexisted in 2 cases,14,38 was absent in 6 cases3,9,12,37 treatment before the development of cutaneous hyper- and not recorded in 6 cases.4,8,34,36 One patient with acute pigmentation is 6.1 years (3 months to 22 years) and lymphoblastic leukemia had also café-au-lait spots on the the median cumulative dose of the drug 720 g.6 It is skin, but those spots were attributed to the synchronous dose-dependent,6 may decrease over several months after occurrence of neurofibromatosis type I,14 and another the drug is discontinued,51 and can persist for many years patient with CML had melanonychia of the toenails.9 and even improve despite continuation of therapy.6 It In the case presented here, the clinical and micro- manifests as well-defined, brown, blue-gray or dark purple scopic features were consistent with those in previous pigmentation and is commonly associated with toxic reports of imatinib-associated palatal hyperpigmenta- ocular effects, in particular retinopathy that leads to vision tion. Our patient was not treated for CML with other loss.47 drugs that may cause hyperpigmentation, in particular Microscopic examination shows yellow- to brown- hydroxyurea or chloroquine, or with minocycline.3,31 colored pigment granules in macrophages and fibroblasts Besides, oral melanosis associated with hydroxyurea and among collagen fibers47,49; the pigment granules react occurs on sites other than the palate45 and may repre- for iron and some of them for melanin.15,47 In 2 cases of sent postinflammatory pigmentation as a result hydroxychloroquine-associated skin pigmentation, only of hydroxyurea-induced mucositis.3 Hemosiderin melanin was identified,51 and in another 5 cases, melanin deposition—that is, after palatal trauma—could be ex- and ferric iron6 were observed. It is suggested that the pected in a patient with CML who has a predisposition pigment represents either hydroxychloroquine chelated for hemorrhage because of the disease or another to melanin or iron, or melanin whose production is stimu- medication, but in our patient, neither CBC nor the lated by hydroxychloroquine, directly or indirectly through microscopically verified absence of hemorrhage was an increase in the androgen levels caused by the drug.15,47,48 suggestive of such a condition.8 We agree with Mattsson Antimalarial drug–induced oral hyperpigmentation et al.46 that although “it is unwise to categorically state is uncommon and is reported to occur on the hard that imatinib mesylate is responsible, … we do suggest palate, gingiva, lips, and buccal mucosa.15,47 In a retro- that the drug is very strongly implicated.” spective study, pigmentation of the buccal mucosa or The microscopic features in previous reports of palatal the hard palate was seen in 10 patients (5%) using melanosis are contradictory: In 4 patients, the pigment chloroquine, but none in hydroxychloroquine,50 and there granules were positive for melanin and negative for are rare case reports of hyperpigmentation in patients iron,4,8,9 as is also described in skin hyperpigmentation31; with systemic treated with OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e63 hydroxychloroquine, involving skin and the palate6 or before starting treatment with TNF-α inhibitors.58,63 In only the gingiva.52 A case of pigmentation of the lower contrast, nonbiologic disease-modifying antirheumatic gingiva in a patient with Sjögren’s syndrome treated drugs, such as methotrexate, that are typically used in with hydroxychloroquine53 is more consistent with racial rheumatoid arthritis are not a strong risk factor for pigmentation or smoker’s melanosis. The clinical, mi- melanoma.63 croscopic, and histochemical features in patients receiving The mechanism underlying the development of mela- hydroxychloroquine6 are similar to those of patients noma in patients treated with TNF-α inhibitors is taking chloroquine phosphate,47,54 suggesting that the not known.61 Direct association with the ensuing drug may be implicated in this phenomenon. There is immunosuppression56,61 because of the disease or the in- no other report on the microscopic findings of hyper- hibitory role of TNF in tumor surveillance63 is supported pigmentation associated with the drug in the oral mucosa. by the association between organ transplantation or HIV In our case, the pigment granules reacted for both melanin and the risk for melanoma.61,63 There is, in addition, the and ferric iron, as is described in some skin lesions.6 possibility of increased detection of melanoma in this These findings are also quite similar to those in our group of patients as a result of better medical monitoring.61 case of imatinib-associated hyperpigmentation, with the Another explanation is that TNF-α inhibitors may in- exception of the presence of hemorrhage that could act crease skin photosensitivity or directly influence as a source of iron. However, the reaction for melanin melanocytes, an effect that is under investigation because and the absence of granules consistent with hemosid- it may prove useful in the management of vitiligo.55 In erin can be explained only by the assumption that products particular, melanocytes both produce TNF-α and express of the drug that is locally processed by macrophages TNFR1 and TNFR2 receptors in response to stress, and express chemical groups that react like melanin or TNFR1 may mediate apoptosis through a caspase lipofuscin.9 3–dependent pathway in a variety of cells, including Resolution may be seen months to years after drug melanocytes.55 In neonatal human melanocytes, TNF-α withdrawal.48 Thirty months after the palate biopsy, our inhibits proliferation and melanogenesis by decreasing patient was lost to follow-up, so further information is tyrosinase activity in a dose-dependent manner,65 and it not available. may promote melanocyte apoptosis.55 At the same time, however, TNF-α may stimulate the secretion of mela- nogenic factors, such as stem cell factor, hepatocyte Golimumab-associated hyperpigmentation growth factor, basic fibroblastic growth factor, and Golimumab is a fully human monoclonal antibody that endothelins, as is seen in postinflammatory or post- blocks tumor necrosis factor (TNF)-α signaling. It acts traumatic pigmentation.66 Therefore, it may be as a decoy receptor that binds both soluble and membrane- hypothesized that TNF-α inhibitors directly promote me- bound TNF-α and prohibits it from binding and acting lanocyte survival and melanogenesis, but this may be an through its true receptors TNFR1 and TNFR2.55 Because oversimplification because the interplay of TNF-α with TNF-α is a polypeptide hormone that plays a crucial role other cytokines that affect melanocytes is a complex in many rheumatologic, gastrointestinal, and dermato- process.65 logic diseases, its inhibitors are used in the treatment of In our case, a temporal relationship with golimumab respective diseases, in particular psoriasis, psoriatic ar- administration could not be established with certainty thritis, rheumatoid arthritis, ankylosing spondylitis, and because the lesions were asymptomatic. In addition, as Crohn’s disease.56,57 a result of the development of nephrotic syndrome, the TNF-α inhibitors are associated with an increased patient could not present for re-examination, although risk for the development of skin melanoma in patients she thought that withdrawal from the drug had a posi- with rheumatoid arthritis56,58-60 or inflammatory bowel tive effect on the pigmented lesions. Increased melanin disease,61 although there is no increased risk for cancer deposition without increased number of melanocytes, ob- overall.58,60,62,63 This observation was originally made in served on microscopic examination, is consistent with patients taking infliximab64 and later confirmed for 3 the effect of TNF-α on melanocytes. The patient was also other TNF-α inhibitors that are currently in use, namely, treated with hydroxychloroquine plus corticosteroids, but golimumab, etanercept, and adalimumab, but not for that treatment had been terminated 8 years before the cetrolizumab.56 Studies have shown that patients with patient observed the dark macules, and in the mean- rheumatoid arthritis treated with TNF-α inhibitors have time, no pigmentation was seen during a number of an increased risk of developing invasive melanoma or a routine dental examinations. Although there are no reports second primary melanoma, relative to patients treated of soft tissue pigmentation caused by golimumab, ex- with nonbiologic drugs, whereas there is as an in- clusion of other diseases and the possibility that the drug creased risk of recurrent melanomas in patients with may stimulate melanogenesis65 implicate it in the devel- rheumatoid arthritis who have a history of melanoma opment of hyperpigmentation. ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e64 Tosios, Kalogirou and Sklavounou March 2018

CONCLUSIONS 5. Krause W. Drug-induced hyperpigmentation: a systematic review. Correct diagnosis of drug-associated pigmentation is J Dtsch Dermatol Ges. 2013;11:644-651. crucial for patient management and to avoid confusion 6. Jallouli M, Frances C, Piette JC, et al. Hydroxychloroquine- induced pigmentation in patients with systemic lupus erythematosus: with other systemic diseases and conditions, but it also a case-control study. JAMA Dermatol. 2013;149:935-940. protects the patient from unnecessary anxiety, testing, and 7. Treister NS, Magalnick D, Woo SB. Oral mucosal pigmentation other adverse effects.7 The last is particularly true for chlo- secondary to minocycline therapy: report of two cases and a review roquine, where cutaneous and mucosal hyperpigmentation of the literature. Oral Surg Oral Med Oral Pathol Oral Radiol Endod may be a marker for retinopathy that is irreversible and . 2004;97:718-725. 8. Mattsson U, Halbritter S, Morner Serikoff E, Christerson L, 47,54 may lead to blindness. Hard palate pigmentation Warfvinge G. Oral pigmentation in the hard palate associated with without involvement of the soft palate is a strong indi- imatinib mesylate therapy: a report of three cases. Oral Surg Oral cation of a drug-induced adverse effect.48 Biopsy is Med Oral Pathol Oral Radiol Endod. 2011;111:e12-e16. indicated for lesions developing on mucosal sites, such 9. Steele JC, Triantafyllou A, Rajlawat BP, Field EA. Oral mucosal as the hard palate, where melanoma usually develops17,67 hyperpigmentation and horizontal melanonychia caused by imatinib. Clin Exp Dermatol. 2012;37:432-433. and in patients taking drugs, such as golimumab, in which 10. Hindocha A, Clark M. Palatal hyperpigmentation in a patient with case there is an increased risk for the development of chronic myeloid leukaemia on imatinib mesylate therapy. BrJOral melanoma.56,58-61 Maxillofac Surg. 2014;52:e115. In conclusion, drug-associated hyperpigmentation of 11. Loudon JA, Coleman HG, Allen CM, Schifter M, Ng T, Al- the oral mucosa is microscopically characterized by a Horani G. Rare oral manifestation of chronic myelogenous leukemia and its targeted therapy: a case report and literature review. Univ normal number of melanocytes in the epithelium and the J Med Dent. 2012;1:79-85. presence of pigment granules within the lamina propria, 12. Romeo U, Palaia G, Fantozzi PJ, Tenore G, Bosco D. A rare case among collagen fibers, and within macrophages. An in- of melanosis of the hard palate mucosa in a patient with chronic creased deposition of melanin in the basal cell layer and myeloid leukemia. Case Rep Dent. 2015;2015:817094. the presence of granules in the superficial lamina propria 13. LaPorta VN, Nikitakis NG, Sindler AJ, Reynolds MA. Minocycline- associated intra-oral soft-tissue pigmentation: clinicopathologic are strongly indicative of minocycline- and golimumab- correlations and review. J Clin Periodontol. 2005;32:119-122. associated pigmentation. In the former, the histochemical 14. Resende RG, Teixeira RG, Vasconcelos Fde O, Silva ME, Abreu findings considering the reaction of the granules for silver MH, Gomez RS. Imatinib-associated hyperpigmentation of the or iron are conflicting, but in the latter, the granules react palate in post-HSCT patient. J Craniomaxillofac Surg. 2012;40: for silver but not for iron. In contrast, lack of deposi- e140-e143. 15. Kleinegger CL, Hammond HL, Finkelstein MW. Oral mucosal hy- tion of melanin in the basal cell layer and the presence perpigmentation secondary to antimalarial drug therapy. Oral Surg of granules in the reticular lamina propria point to Oral Med Oral Pathol Oral Radiol Endod. 2000;90:189- imatinib- and hydroxychloroquine-associated hyperpig- 194. mentation. In the former, the granules react for silver and 16. Meleti M, Vescovi P, Mooi WJ, van der Waal I. Pigmented lesions variably for iron, whereas in the latter, they react for both of the oral mucosa and perioral tissues: a flow-chart for the di- agnosis and some recommendations for the management. Oral Surg silver and iron, and there is hemorrhage. It should be noted Oral Med Oral Pathol Oral Radiol Endod. 2008;105:606- that our report of the microscopic and histochemical char- 616. acteristics of hydroxychloroquine- and golimumab- 17. Fernandes D, Ferrisse TM, Navarro CM, Massucato EM, Onofre associated hyperpigmentation is the first in the literature, MA, Bufalino A. Pigmented lesions on the mucosa: a wide range Oral Surg Oral Med Oral Pathol Oral Radiol so our findings should be further verified. of diagnoses. . 2015; 119:374-378. 18. Basso FG, Boer CC, Correa ME, et al. Skin and oral lesions as- The excellent technical assistance of Mrs. Maria Manou, MTL, sociated to imatinib mesylate therapy. Support Care Cancer. 2009; MSc, and Mr. Eugene Danas, MTL, MSc, is acknowledged. 17:465-468. 19. Alkhatib AA, Sessoms S. The tarnished tongue. Am J Med. 2006; 119:832-834. REFERENCES 20. Smith K, Leyden JJ. Safety of doxycycline and minocycline: a sys- 1. Scully C, Bagan JV. Adverse drug reactions in the orofacial region. tematic review. Clin Ther. 2005;27:1329-1342. Crit Rev Oral Biol Med. 2004;15:221-239. 21. Friedman IS, Shelton RM, Phelps RG. Minocycline-induced hy- 2. Argenyi ZB, Finelli L, Bergfeld WF, et al. Minocycline-related perpigmentation of the tongue: successful treatment with the cutaneous hyperpigmentation as demonstrated by light microsco- Q-switched ruby laser. Dermatol Surg. 2002;28:205-209. py, electron microscopy and X-ray energy spectroscopy. J Cutan 22. Goulden V, Glass D, Cunliffe WJ. Safety of long-term high- Pathol. 1987;14:176-180. dose minocycline in the treatment of acne. Br J Dermatol. 1996; 3. Li CC, Malik SM, Blaeser BF, et al. Mucosal pigmentation caused 134:693-695. by imatinib: report of three cases. Head Neck Pathol. 2012;6:290- 23. Mouton RW, Jordaan HF, Schneider JW. A new type of 295. minocycline-induced cutaneous hyperpigmentation. Clin Exp 4. Wong M, Sade S, Gilbert M, Klieb HB. Oral melanosis after ty- Dermatol. 2004;29:8-14. rosine kinase inhibition with Imatinib for chronic myelogenous 24. Bowen AR, McCalmont TH. The histopathology of subcutane- leukemia: report of a case and review of the literature. Dermatol ous minocycline pigmentation. J Am Acad Dermatol. 2007;57: Online J. 2011;17:4. 836-839. OOOO REVIEW ARTICLE Volume 125, Number 3 Tosios, Kalogirou and Sklavounou e65

25. Odell EW, Hodgson RP, Haskell R. Oral presentation of 47. de Andrade BA, Fonseca FP, Pires FR, et al. Hard palate hyper- minocycline-induced black bone disease. Oral Surg Oral Med Oral pigmentation secondary to chronic chloroquine therapy: report of Pathol Oral Radiol Endod. 1995;79:459-461. five cases. J Cutan Pathol. 2013;40:833-838. 26. Meyerson MA, Cohen PR, Hymes SR. Lingual hyperpigmenta- 48. Lerman MA, Karimbux N, Guze KA, Woo SB. Pigmentation of tion associated with minocycline therapy. Oral Surg Oral Med Oral the hard palate. Oral Surg Oral Med Oral Pathol Oral Radiol Pathol Oral Radiol Endod. 1995;79:180-184. Endod. 2009;107:8-12. 27. Filitis DC, Graber EM. Minocycline-induced hyperpigmentation 49. Cho EB, Kim BC, Park EJ, et al. Hydroxychloroquine-induced involving the oral mucosa after short-term minocycline use. Cutis. hyperpigmentation. J Dermatol. 2012;39:859-860. 2013;92:46-48. 50. Skare T, Ribeiro CF, Souza FH, Haendchen L, Jordao JM. An- 28. Eisen D. Minocycline-induced oral hyperpigmentation. Lancet. timalarial cutaneous side effects: a study in 209 users. Cutan Ocul 1997;349:400. Toxicol. 2011;30:45-49. 29. Herbers AH, Gerlag PG. Bluish-grey pigmentation of finger- 51. Puri PK, Lountzis NI, Tyler W, Ferringer T. Hydroxychloroquine- nails, gingiva, teeth and peri-oral region. Minocycline. Neth J Med. induced hyperpigmentation: the staining pattern. J Cutan Pathol. 2004;62:58, 65. 2008;35:1134-1137. 30. Arora B, Kumar L, Sharma A, Wadhwa J, Kochupillai V. Pig- 52. Veraldi S, Schianchi-Veraldi R, Scarabelli G. Pigmentation of the mentary changes in chronic myeloid leukemia patients treated with following hydroxychloroquine therapy. Cutis. 1992;49:281- imatinib mesylate. Ann Oncol. 2004;15:358-359. 282. 31. Alexandrescu DT, Dasanu CA, Farzanmehr H, Kauffman L. Per- 53. Kalampalikis A, Goetze S, Elsner P. Isolated hyperpigmentation sistent cutaneous hyperpigmentation after tyrosine kinase inhibition of the oral mucosa due to hydroxychloroquine. J Dtsch Dermatol with imatinib for GIST. Dermatol Online J. 2008;14:7. Ges. 2012;10:921-922. 32. Wang Y, Zhao Y, Liu L, et al. Inhibitory effects of imatinib mesylate 54. Gallo CB, Luiz AC, Ferrazzo KL, Migliari DA, Sugaya NN. Drug- on human epidermal melanocytes. Clin Exp Dermatol. 2014;39: induced pigmentation of hard palate and skin due to chronic 202-208. chloroquine therapy: report of two cases. Clin Exp Dermatol. 2009; 33. Pancholi N, Taneja P. Intraoral hyperpigmentation due to imatinib 34:e266-e267. mesylate. A review of the literature. Oral Surg. 2016;9:206- 55. Webb KC, Tung R, Winterfield LS, et al. Tumour necrosis factor- 214. alpha inhibition can stabilize disease in progressive vitiligo. Br J 34. Lewis DM. Diffuse pigmentation of the palate. J Okla Dent Assoc. Dermatol. 2015;173:641-650. 2009;100:24-25. 56. Nardone B, Hammel JA, Raisch DW, Weaver LL, Schneider D, 35. Yu YH, Shere Y, Vigneswaran N. Oral and maxillofacial pathol- West DP. Melanoma associated with tumour necrosis factor- ogy case of the month. Palatal melanosis associated with imatinib alpha inhibitors: a Research on Adverse Drug events And Reports mesylate therapy. Tex Dent J. 2012;129:764-765, 786-768. (RADAR) project. Br J Dermatol. 2014;170:1170-1172. 36. Roeker LE, Wolanskyj AP. Imatinib-associated melanosis of the 57. Silva LC, Ortigosa LC, Benard G. Anti-TNF-alpha agents in the palate. Am J Hematol. 2014;89:564. treatment of immune-mediated inflammatory diseases: mecha- 37. Khoo TL, Catalano A, Supple S, et al. Hyperpigmentation of the nisms of action and pitfalls. Immunotherapy. 2010;2:817-833. hard palate associated with imatinib therapy for chronic myeloid 58. Ramiro S, Gaujoux-Viala C, Nam JL, et al. Safety of synthetic leukemia with a genetic variation in the proto-oncogene c-KIT. and biological DMARDs: a systematic literature review inform- Leuk Lymphoma. 2013;54:186-188. ing the 2013 update of the EULAR recommendations for 38. Song HS, Kang HY. Imatinib mesylate-induced hyperpigmenta- management of rheumatoid arthritis. Ann Rheum Dis. 2014;73: tion of the nose and palate. Ann Dermatol. 2014;26:532- 529-535. 533. 59. Wolfe F, Michaud K. Biologic treatment of rheumatoid arthritis 39. Lyne A, Creedon A, Bailey BM. Mucosal pigmentation of the hard and the risk of malignancy: analyses from a large US observa- palate in a patient taking imatinib. BMJ Case Rep. 2015;pii: tional study. Arthritis Rheum. 2007;56:2886-2895. bcr2015209335. doi:10.1136/bcr-2015-209335. 60. Mariette X, Matucci-Cerinic M, Pavelka K, et al. Malignancies 40. Matziari E, Poulopoulos A, Zormpa M, Mesemanoli Z, Andreadis associated with tumour necrosis factor inhibitors in registries and D, Kolokotronis A. Imatinib mesylate related oral mucosal prospective observational studies: a systematic review and meta- pigmentation: two unusual cases. Oral Dis. 2016;22:49. analysis. Ann Rheum Dis. 2011;70:1895-1904. 41. Torres-Pereira CC. Are we ready to blame imatinib for palatal pig- 61. Long MD, Martin CF, Pipkin CA, Herfarth HH, Sandler RS, mentation? Oral Surg Oral Med Oral Pathol Oral Radiol Endod. Kappelman MD. Risk of melanoma and nonmelanoma skin cancer 2011;112:287. author reply 288. among patients with inflammatory bowel disease. Gastroenterol- 42. McPherson T, Sherman V, Turner R. Imatinib-associated hyper- ogy. 2012;143:390-399, e391. pigmentation, a side effect that should be recognized. J Eur Acad 62. Moulis G, Sommet A, Bene J, et al. Cancer risk of anti-TNF- Dermatol Venereol. 2009;23:82-83. alpha at recommended doses in adult rheumatoid arthritis: a meta- 43. Singh N, Bakhshi S. Imatinib-induced dental hyperpigmentation analysis with intention to treat and per protocol analyses. PLoS in childhood chronic myeloid leukemia. J Pediatr Hematol Oncol. ONE. 2012;7:e48991. 2007;29:208-209. 63. Raaschou P, Simard JF, Holmqvist M, Askling J, Artis Study Group. 44. Agrawal P, Singh O, Nigam AK, Upadhyay S. Imatinib-induced Rheumatoid arthritis, anti-tumour necrosis factor therapy, and risk dental hyperpigmentation in chronic myeloid leukemia in an adult of malignant melanoma: nationwide population based prospec- female. Indian J Pharmacol. 2015;47:685-686. tive cohort study from Sweden. BMJ. 2013;346:f1939. 45. Kumar B, Saraswat A, Kaur I. Mucocutaneous adverse effects of 64. Khan I, Rahman L, McKenna DB. Primary cutaneous melanoma: hydroxyurea: a prospective study of 30 psoriasis patients. Clin Exp a complication of infliximab treatment? Clin Exp Dermatol. 2009; Dermatol. 2002;27:8-13. 34:524-526. 46. Mattsson U, Halbritter S, Morner Serikoff E, Christerson L, 65. Swope VB, Abdel-Malek Z, Kassem LM, Nordlund JJ. Interleukins Warfvinge G. Are we ready to blame imatinib for palatal pigmen- 1 alpha and 6 and tumor necrosis factor-alpha are paracrine in- tation? Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2011; hibitors of human melanocyte proliferation and melanogenesis. J 112:288. Invest Dermatol. 1991;96:180-185. ORAL AND MAXILLOFACIAL PATHOLOGY OOOO e66 Tosios, Kalogirou and Sklavounou March 2018

66. Feller L, Masilana A, Khammissa RA, Altini M, Jadwat Y, Lemmer Reprint requests: J. Melanin: the biophysiology of oral melanocytes and physio- Konstantinos I. Tosios, DDS, PhD logical oral pigmentation. Head Face Med. 2014;10:8. Department of Oral Medicine and Pathology 67. Smith MH, Bhattacharyya I, Cohen DM, et al. Melanoma of the Faculty of Dentistry oral cavity: an analysis of 46 new cases with emphasis on clini- National and Kapodistrian University of Athens cal and histopathologic characteristics. Head Neck Pathol. 2016; 2 Thivon Street, 11527 Athens 10:298-305. Greece [email protected]