Photodistribution of Blue-Gray Hyperpigmentation After Amiodarone Treatment: Molecular Characterization of Amiodarone in the Skin

Background  For decades, the photodistributed blue-gray skin hyperpigmentation observed after amiodarone therapy was presumably attributed to dermal lipofuscinosis. Using electron microscopy and high-performance liquid chromatography, we identified amiodarone deposits in the hyperpigmented skin sample from a patient treated with this antiarrhythmic agent. Our findings therefore indicate that the hypothesis relating the blue-gray hyperpigmentation to lipofuscin should be challenged.

Observations  A 64-year-old man, skin phototype III, presented with asymptomatic skin hyperpigmentation that had been slowly developing on sun-exposed areas since April 2004. He had been taking amiodarone for 4 years (cumulative dose, 277 g). Electron microscopy did not show lipofuscin pigments in his skin. Conversely, abundant electron-dense membrane-bound granule deposits were observed in most of the dermal cells (fibroblasts, macrophages, pericytes, Schwann cells, and endothelial cells), especially in photoexposed skin. High-performance liquid chromatography confirmed that the skin deposits were composed of amiodarone. These results demonstrate that amiodarone hyperpigmentation is related to drug deposition on photoexposed skin.

Conclusion  Amiodarone-related hyperpigmentation should be considered a skin storage disease that is secondary to drug deposition.

Amiodarone-photodistributed blue-gray skin hyperpigmentation is exceptional. For decades, such hyperpigmentation was attributed to dermal lipofuscinosis, with a granular accumulation of lipofuscin in dermal macrophages.^1,2 It was suggested that the pathogenesis might be related to the basic action of amiodarone on lysosomes and to the extraphototoxic-induced lysosomal damage, which accounted for the specific location of the hyperpigmentation on light-exposed areas.³ This hypothesis was supported by the fact that amiodarone-hyperpigmented skin had a drug and metabolite concentration 10 times higher than that of nonpigmented skin.⁴ To our knowledge, this case represents the first time that amiodarone deposits have been identified in the pigmented skin of a patient treated with this antiarrhythmic agent. Our results clearly demonstrate that amiodarone-induced skin hyperpigmentation is related to drug deposition.

A 64-year-old man, skin phototype III, presented with progressive, asymptomatic skin hyperpigmentation that had been slowly developing on sun-exposed areas since April 2004. His medical history was remarkable for myocardial infarction, ventricular arrhythmia, and heart failure. He had no known history of drug allergy. For several years, he had been treated with acebutolol (200 mg/d), lisinopril (10 mg/d), simvastatin (20 mg/d), and lysine acetylsalicylate (75 mg/d). He had also taken amiodarone (Cordarone) at a dosage of two 200-mg tablets per day, 5 days per week, from 2001 until July 2005. The cumulative dose was estimated to be 277 g over 4½ years. His physical examination revealed photodistributed, blue-gray hyperpigmentation of the face and ears, sparing the area under the nose, all eyelids, nasolabial folds, wrinkles, and submental and postauricular areas (Figure 1). Phototests revealed a polychromatic minimal erythema dose at 600 mJ/cm² (reference value, ≥400 mJ/cm²). The results of the UV-A phototest (13 J/cm²) were negative after 24 hours, with mild hyperpigmentation. Histologic examination of a pigmented skin specimen revealed numerous macrophages accumulated around superficial dermal vessels. The cytoplasm of these cells showed brownish deposits that were positive on periodic acid–Schiff and Fontana stains (Figures 2, 3, and 4).

Electron microscopy of a nonpigmented skin sample showed the presence of a few homogeneous strongly electron-dense granules confined to the upper dermis, whereas a pigmented skin sample revealed numerous similar granules within the thickness of the dermis. These deposits were localized mainly in fibroblasts as well as in other cells, particularly macrophages, endothelial cells, and Schwann cells (Figure 5). At high magnification, they appeared to be surrounded by a membrane. There were no abnormal deposits within the epidermis, hair follicles, or sebaceous glands. No lipofuscin deposits were observed in the pigmented skin sample from our patient.

The first step of the molecular identification of amiodarone deposits in the skin involved the extraction of amiodarone from a skin biopsy specimen. A skin biopsy (punch, 4 mm; 8 mg) was performed on the pigmented skin of the face to demonstrate the presence of drug deposits by extracting the active molecule of amiodarone. Accordingly, the skin specimen was homogenized with 10 mL of methanol and kept at 5°C for 24 hours. The homogenate was then crushed and filtered. The filtrate was evaporated to dryness, dissolved in 2 mL of diethylether, and acidified with 500 μL of 1M hydrochloride. It was shaken for 15 minutes and centrifuged for 10 minutes at 3000 rounds per minute. This last step resulted in an upper organic layer (containing amiodarone) and a lower aqueous layer. The organic layer was dried with anhydrous sodium sulfate and then evaporated to dryness.

The second step involved sample analysis. After skin extraction, the concentrate was reconstituted in methanol and then analyzed by means of high-performance liquid chromatography (HPLC) (isocratic mode, 75% methanol and 25% ammonium formate [25 mM]; column Sunfire C18, 3.5 μm 2.1 × 50.0 mm). To obtain a standard molecule, commercial amiodarone was isolated and purified from the amiodarone tablets and then sequentially added to the extracted sample, with a new HPLC analysis performed after each addition.

High-performance liquid chromatography of the skin sample showed 4 peaks corresponding to 4 retention times: 0.707 minutes, 0.868 minutes, 1.447 minutes, and 8.207 minutes (Figure 6). After each sequential addition of the commercial amiodarone, HPLC analysis revealed a peak increase at 8.207 minutes. The UV spectrum determined at 8.207 minutes showed a perfect match at each peak (Figure 6).

Amiodarone hydrochloride (2-butyl-3-benzofuranyl 4-[2-(diethylamino)-ethoxy]-3,5-diiodophenyl ketone hydrochloride) is an iodinated compound that is widely used in the treatment of cardiac arrhythmias and is known to cause photosensitivity and cutaneous hyperpigmentation. Although amiodarone photosensitivity is quite common and occurs in more than 50% of treated patients, blue-gray cutaneous hyperpigmentation occurs in fewer than 10%.⁵ The clinical features of the photosensitivity response represent a phototoxic reaction to both amiodarone and its major metabolite, mono-N-desethylamiodarone.⁶ Also, it has been shown that amiodarone therapy might induce photoallergy in guinea pigs. However, the photoallergic effect of the drug has generally been masked by its phototoxic potential.¹ Phototoxic reactions can be experimentally elicited with UV-A; the UV-A minimal erythema dose is significantly reduced after 12 months of treatment.⁷ The photoactivating wavelengths are primarily found in the long-wave UV-A spectrum between 350 and 380 nm.⁸ However, phototests may show acute reactions to UV-A and UV-B and significant delayed reactions to UV-A and/or UV-B.^4,9 Zinc oxide–containing preparations appear to be the most effective agents for reducing cutaneous photosensitivity.⁶ Under the regimens commonly used, photosensitivity can be expected to occur after 4 months of continuous treatment and a minimal cumulative dose of 40 g. It appears to be unrelated to the skin type. Photosensitivity gradually decreases and returns to normal between 4 and 12 months after discontinuation of amiodarone therapy.⁷ However, it can sometimes last for several years after drug withdrawal.¹⁰ Amiodarone hyperpigmentation develops mainly in patients with skin type I. It occurs after an average of 20 months of continuous treatment and a minimal cumulative dose of 160 g.⁷ The slow rate of elimination of amiodarone and the high uptake by fat-associated tissues may explain the delayed spontaneous disappearance of cutaneous photosensitivity and the late resolution of the blue-gray discoloration.⁵ In 1 patient, massive amiodarone-induced hyperpigmentation was found to be reversible 33 months after the use of the drug was discontinued.⁷ However, in cosmetically stigmatizing hyperpigmentations, treatment with a Q-switched ruby laser has shown impressive results.¹¹ In our case, photosensitivity toward amiodarone or another drug was ruled out because phototests showed a normal polychromatic minimal erythema dose and a negative UV-A phototest result. Therefore, this case corresponds clinically, histologically, and ultrastructurally to typical amiodarone-photodistributed blue-gray hyperpigmentation, which occurred after 52 months of continuous treatment and a cumulative dose of 277 g.

Previous electron microscopy studies of amiodarone-pigmented skin demonstrated 6 distinctive morphological types of intracytoplasmic inclusions in many dermal cell types. The pathogenesis may be related to the action of the drug on cell membranes, local metabolic damage, and accumulation of the drug on the lysosomes, with acceleration of the physiological aging cell process.¹² In a previous report, the presence of high concentrations of iodine, which was observed on electron probe analysis, suggested that the cutaneous deposits are made up of amiodarone itself or a metabolite.⁵ Our results confirm this hypothesis. After the extraction procedure that was performed on the hyperpigmented skin of our patient's face, HPLC of the skin sample showed 4 peaks corresponding to 4 retention times: 0.707 minutes, 0.868 minutes, 1.447 minutes, and 8.207 minutes. Later on, after each sequential addition of commercial amiodarone (extracted from Cordarone tablets), HPLC revealed a clear increase of the peak at 8.207 minutes. This finding suggests that the molecule corresponding to the retention time of 8.207 minutes and the commercial amiodarone that was added are the same compound. To be more precise, the UV absorption spectrum of each peak was determined at 8.207-minutes. These UV absorption spectra were perfectly identical (Figure 6). Therefore, in this case, the molecule extracted from the skin, which showed a peak at 8.207 minutes, is amiodarone. We were not able to identify the nature of the other 3 molecules corresponding to the retention times of 0.707, 0.868, and 1.447 minutes. These molecules may represent amiodarone photoproducts/metabolites, skin components such as melanin, or cutaneous deposition of other drugs taken by the patient. The recognition of these molecules is not easy because amiodarone metabolism is more complex than has generally been accepted. It was observed that mono-N-desethylamiodarone may further be cleared by hydroxylation, dealkylation to di-N-desethylamiodarone, and deamination to deaminated amiodarone.¹³ In our case, HPLC analysis could neither exclude nor confirm the presence of lipofuscin. However, electron microscopy showed that lipofuscin pigment was absent in our patient's skin. This finding indicates that the hypothesis relating the blue-gray hyperpigmentation to lipofuscin should be challenged. Also, direct evidence of massive amiodarone deposits in the hyperpigmented skin on electron microscopy provides a strong argument in favor of a direct pathogenic role for amiodarone. There are numerous reasons to question the role of lipofuscin as a causative factor in amiodarone hyperpigmentation. Lipofuscin is a naturally occurring autofluorescent lipopigment that accumulates in aging cells as a normal part of senescence; it is called the wear-and-tear or aging pigment. Because this material exhibits fluorescence, lipofuscin has been described by its spectral properties, with an excitation between 320 and 480 nm and an emission wavelength between 460 and 630 nm, with a peak at 580 nm corresponding to yellow and not blue fluorescence.^14,15 Conversely, electron microscopic examination of the sun-exposed skin of patients without amiodarone discoloration shows pigment deposits similar to those already described in patients with amiodarone hyperpigmentation in exposed and nonexposed skin.⁷ Finally, the presence of amiodarone deposits in the skin, with or without lipofuscin, is able to induce the blue-gray hyperpigmentation. This pigmentation could be explained by the Tyndall effect, in which dermal pigment, whether melanin, iron, or other pigment, is perceived as blue, gray, or blue-gray.

In conclusion, our results confirm that amiodarone blue-gray hyperpigmentation should be considered a skin storage disease that is secondary to drug deposition. A comparative study with nonpigmented, photoprotected skin would need to be carried out to find out whether the other 3 unidentified molecules (0.707 minutes, 0.868 minutes, and 1.447 minutes) on HPLC are photoproducts of amiodarone.

Correspondence: Alfred Ammoury, MD, Service de Dermatologie, Centre Hospitalier Universitaire de Toulouse, Hôpital Purpan, 31059 Toulouse CEDEX, France (docalf@yahoo.com).

Accepted for Publication: June 10, 2007.

Author Contributions: Dr Ammoury had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Ammoury, Michaud, Paul, Bazex, Chouini-Lalanne, and Marguery. Acquisition of data: Ammoury, Michaud, Paul, Prost-Squarcioni, Alvarez, Launay, Bazex, Chouini-Lalanne, and Marguery. Analysis and interpretation of data: Ammoury, Michaud, Paul, Prost-Squarcioni, Alvarez, Lamant, Launay, Bazex, Chouini-Lalanne, and Marguery. Drafting of the manuscript: Ammoury. Critical revision of the manuscript for important intellectual content: Ammoury, Michaud, Paul, Prost-Squarcioni, Alvarez, Lamant, Launay, Bazex, Chouini-Lalanne, and Marguery. Administrative, technical, or material support: Michaud and Chouini-Lalanne. Study supervision: Ammoury, Paul, Bazex, Chouini-Lalanne, and Marguery.

Financial Disclosure: Dr Ammoury has had a consultancy agreement with Novartis Pharma Basel Switzerland since November 2006. Dr Paul was an employee of Novartis Pharma Basel Switzerland (until March 2006); has received honoraria and grants from Solgel, Pierre Fabre, and Novartis; and is a coinventor on several patents (WO2006053699, WO/2004/087170, WO/2004/087141, WO/2004/087143, WO/2004/0871118, WO2001/095890).

Previous Presentation: This study was presented as a poster at the Journées Dermatologiques de Paris Congress; December 5-9, 2006; Paris, France.