The Use of Tetracyclines for the Treatment of Sarcoidosis

Background  To evaluate the safety and efficacy of minocycline in the treatmentof sarcoidosis, a nonrandomized, open study was performed in patients withcutaneous sarcoidosis.

Observations  Twelve patients with cutaneous sarcoidosis were treated with minocycline,200 mg/d, for a median duration of 12 months. Three patients had extracutaneouslesions at the time of the study. The median follow-up was 26 months. A clinicalresponse was observed in 10 patients, consisting of complete responses in8 patients and partial responses in 2 patients. A progression of skin lesionswas observed in 1 patient, and lesions remained stable in another patient.Adverse effects were minimal, except in 1 patient, who developed hypersensitivitysyndrome. A slight hyperpigmentation occurred in 2 patients at the site ofprevious lesions, which completely disappeared after minocycline use was discontinued.A relapse of skin symptoms occurred after minocycline withdrawal in 3 patients,who further received doxycycline, 200 mg/d, allowing a complete remissionof lesions.

Conclusions  These results support that minocycline and doxycycline may be beneficialfor the treatment of cutaneous sarcoidosis. Randomized controlled studiesare warranted for the evaluation of the true efficacy of tetracyclines inthese patients.

SARCOIDOSIS is a granulomatous, multisystemic disorder of unknown causethat involves predominantly the skin, eye, lungs, and lymph nodes.¹ Therapeutic difficulties originate from the chroniccourse of the disease and the lack of spontaneous regression of lesions. Thischronicity raises the balance between the benefit and the long-term toleranceof therapy. Thus, oral corticosteroids, which have been recognized as themost effective therapy for sarcoidosis, are indicated as the first-line treatmentonly in cases presenting with severe visceral involvement. Oral steroids arenot warranted in cases presenting with less severe involvement, since theirlong-term use is associated with many adverse effects and because relapsesare common during reduction of daily dosage.²In view of the toxicity of steroids, much interest has been recently devotedto the efficacy of corticosteroid-sparing agents in sarcoidosis.³Indeed, several studies^4,5 suggestedthat both chloroquine and hydroxychloroquine sulfate may be beneficial inpatients affected with sarcoidosis involving the skin and extracutaneous organs.However, the efficacy of antimalarial agents is not a consistent finding,thus warranting the need for new alternative drugs. In the present work, wereport the results of an open, prospective study of the efficacy of minocyclinehydrochloride in 12 patients affected with chronic forms of sarcoidosis.

Between 1996 and 1998, 12 patients (9 women and 3 men) who presentedwith histologically proven sarcoidosis involving the skin were enrolled inthe present study. Patients were eligible if they had 1 or more evaluableskin lesions that showed stability or progression for more than 3 months.Patients had no history of hypersensitivity to tetracyclines and were freeof any anti-inflammatory, immunomodulatory, or immunosuppressive therapy fora minimum of 3 months before entering the study. Topical steroid treatmentwas stopped at least 1 month before the study. The essential clinical anddemographic features of the 12 patients are listed in Table 1. The ages of the patients ranged from 16 to 63 years (median,37.5 years; mean, 36.8 years). All patients had a complete physical examination,a chest x-ray examination, and complete laboratory analysis, including peripheralblood cell counts and serum levels of calcium, phosphate, and angiotensin-convertingenzyme. Chest computed tomographic scan was performed in case of abnormalfindings on x-ray examination. Eleven patients presented with multiple skinlesions of sarcoidosis at the onset of the study, and only patient 12 presentedwith a unique subcutaneous nodular lesion involving the frontotemporal area.The cutaneous lesions were typed as papulonodular in 8 cases (cases 1, 3,5, 6, 7, 8, 10, and 11), as plaques in 2 cases (patients 2 and 9), and aslupus pernio in 1 case (case 4), while a unique hypodermal nodule was presentin patient 12. Six patients (cases 1, 2, 8, 10, 11, and 12) had extracutaneousmanifestations at the time of diagnosis of sarcoidosis. Three patients whobelonged to this latter subgroup presented with evaluative extracutaneouslesions at the time of the study: case 1 presented with parenchymal lung involvementassociated with enlarged mediastinal lymph nodes, case 8 was characterizedby an involvement of paranasal sinuses and of cervical lymph nodes, and patient12 presented with enlarged mediastinal lymph nodes (Table 1).

Seven patients (cases 1, 3, 4, 5, 7, 8, and 12) had previously receivedhydroxychloroquine as first-line therapy for a mean of 4.6 months (range,3-6 months), without noticeable regression of lesions. Patients 6 and 9 initiallyreceived chloroquine for 2 and 3 months, respectively, without efficacy. Inpatients 2 and 11, first-line therapy consisted of oral prednisone associatedwith hydroxychloroquine, allowing an initial remission. However, further withdrawalof prednisone was followed by a relapse of skin lesions, without evidenceof extracutaneous involvement. In patient 10, minocycline was given as first-linetherapy, since the patient was affected with myasthenia gravis and diabetesmellitus, 2 conditions that contraindicated antimalarials and oral corticosteroids,respectively. The median duration of sarcoidal cutaneous lesions at the onsetof the study was 15 months (mean, 26.7 months; range, 6-72 months). Patientswere given minocycline, 100 mg twice daily. The clinical response to therapywas evaluated by the same physician, according to the size and number of skinlesions. Patients were examined monthly during the initial 6 months of thestudy, then once every 2 months. The median follow-up was 26 months (mean,29.1 months; range, 12-45 months). A complete response was considered whenall initial lesions completely disappeared, and no occurrence of new lesionswas observed. A partial response was defined by a regression of at least 50%of skin lesions. Either progression or stability or regression below 50% oflesions was considered a lack of response. In patients showing a clinicalrelapse after minocycline withdrawal, doxycycline monohydrate was furthergiven at the daily dosage of 200 mg.

The median follow-up was 26 months (mean, 29 months; range, 12-45 months).A clinical response of sarcoidal cutaneous lesions was documented in 10 of12 patients receiving minocycline, with a duration of response ranging from10 to 41 months (median, 17 months; mean, 21.6 months) (Table 2). A complete clearing of cutaneous lesions was observedin 8 patients (Figure 1 and Figure 2). The mean time to reach maximalresponse of cutaneous lesions from the date of onset of minocycline treatmentwas 3.2 months (median, 3 months; range, 1-6 months); clinical improvementwas noticed as early as 1 month after the onset of minocycline treatment in7 patients. No relapse occurred during treatment. The median duration of minocyclinetreatment was 12 months (median, 13.5 months; range, 1-24 months). Minocyclinecould be withdrawn in 7 patients who achieved complete response of cutaneouslesions (cases 1, 2, 3, 5, 6, 7, and 12). Among this latter subgroup of patients,the complete response was maintained for a mean of 15.3 months (median, 13months; range, 1-33 months) from the date of minocycline withdrawal. Threeof these 7 patients (cases 1, 2, and 3) presented with a relapse of cutaneouslesions that occurred 6, 1, and 13 months after the discontinuation of minocyclinetreatment, respectively. Patients presenting with relapsing lesions were furthertreated with doxycycline 200 mg/d, since the incidence of adverse effects,such as hyperpigmentation and dizziness, is known to be much lower in patientswho receive this latter drug than in those treated with minocycline. In these3 latter patients, a complete remission of cutaneous lesions was observedafter 1 to 3 months of doxycycline therapy, without occurrence of any adverseeffect. Furthermore, a clearing of lung infiltrates was noticed on chest radiographyperformed in patient 1, who was being treated with doxycycline.

A favorable course of intrathoracic involvement correlated with theone of skin lesions in cases 1 and 12. In case 1, a partial regression ofpulmonary infiltrates was noticed after 12 months of minocycline use, anda chest x-ray examination performed 6 months after the onset of second-linetreatment with doxycycline showed no abnormal findings. In patient 12, enlargedmediastinal lymph nodes resolved with minocycline therapy (Figure 3). In 3 patients who presented with lymphopenia (white bloodcell count <1.0 × 10⁹/L) at the time of inclusion, anincrease of blood lymphocyte count was observed with minocycline therapy (Table 3). Furthermore, the angiotensin-convertingenzyme serum level was markedly reduced in 4 patients after 3 months of minocyclinetherapy (Table 3). No evidenceof minocycline toxicity was found in 9 patients. Patient 5 complained of milddizziness, which completely resolved after reduction of the daily dosage ofminocycline to 100 mg. Another patient (case 10) who presented with myastheniagravis and diabetes was considered as a complete responder after 1 month ofminocycline use but developed a drug hypersensitivity syndrome 6 weeks afterthe onset of minocycline treatment. This severe adverse effect was characterizedby a general malaise with fever, a generalized itching rash, a superficiallymphadenopathy, and an interstitial pneumopathy. Biological investigationsrevealed eosinophilia and elevated serum levels of transaminases. All theseabnormalities resolved following treatment with prednisone, 1 mg/kg daily.A slight grayish hyperpigmentation involving the sites of previous sarcoidlesions was observed in another patient and completely resolved 2 months afterminocycline withdrawal (Figure 2B).

Chronic forms of sarcoidosis raise the issue of the long-term toleranceof the treatment, since individual patients may require lifetime maintenancetherapy to avoid relapses of skin lesions.⁶Although it is widely admitted that severe forms of the disease, ie, neurosarcoidosis,severe uveitis, and forms with cardiac involvement, require high-dose corticosteroids,the optimal therapeutic regimen for cases without severe visceral involvementhas not been clearly delineated.⁷ Indeed, thebalance between benefit and toxicity is of primary importance in cases thatmay be associated with a severe functional and/or esthetic prejudice.⁸ Although corticosteroids used at intermediate dosage,ie, 0.5 mg/kg daily, may bring a short-term benefit in these patients, theirprolonged use is associated with significant toxic effects.²Consequently, several agents that show potential corticosteroid-sparing propertieshave been recently investigated in patients affected with sarcoidosis. Amongthese drugs, both chloroquine and hydroxychloroquine have also been beneficialin patients affected with cutaneous or extracutaneous sarcoidosis.^4,5,9 However, responses to antimalarialtherapy may be incomplete, and resistance is not a rare finding, thus warrantingthe need for alternative nonsteroidal therapies. More recently, the efficacyof methotrexate has been reported in disseminated forms of the disease.^10,11 However, taking into account the risk-benefitratio of this drug, its use in cutaneous forms of the disease has to be questioned.

In recent years, interest has been focused on some classes of antibiotics,such as tetracyclines, that, beside their anti-infectious effects, exhibitpotent immunomodulating properties.^12,13Thus, minocycline has been shown to inhibit T-cell proliferation in vitro.¹⁴ Furthermore, both minocycline and doxycycline havebeen shown to inhibit granuloma formation in vitro.¹⁵These results obtained in vitro have been the rationale basis for the successfuluse of minocycline and doxycycline in granulomatous dermatoses, such as silicone-inducedsubcutaneous granulomas¹⁶ and granulomatouscheilitis.¹⁷ The results of the present studysupport the efficacy of both minocycline and doxycycline in patients affectedwith cutaneous sarcoidosis following a long-term course. Furthermore, sincelesions involving the lung parenchyma and mediastinal lymph nodes remittedin 2 patients, it is likely that minocycline and doxycycline are also efficientin treating extracutaneous lesions. In patients affected with sarcoidosis,it remains to be determined whether therapeutic efficacy of minocycline anddoxycycline result either from an anti-infectious mechanism acting on a putativemicrobial agent or from anti-inflammatory or immunomodulatory properties.The changes of angiotensin-converting enzyme serum level and peripheral bloodlymphocyte counts observed with minocycline therapy support the latter hypothesis.Interestingly, immunomodulatory effects and/or anti-infectious effects ofminocycline have been advocated to explain the therapeutic efficacy of thisdrug in inflammatory disorders other than sarcoidosis, mostly rheumatoid arthritis.¹⁸ However, supporting the hypothesis that anti-infectiouseffects of tetracyclines might be involved in their efficacy on cutaneouslesions of sarcoidosis, recent publications reported the presence in sarcoidallymph nodes of DNA products specific for mycobacterial¹⁹and propionibacterial acnes,²⁰ this latter bacterialspecies being known for its sensitivity to tetracyclines. On the other hand,we have recently reported striking remissions of noninfectious granulomatousdisorders with minocycline therapy, such as silicone-induced cutaneous granulomas,showing that the anti-inflammatory properties of tetracyclines documentedin vitro may be relevant in vivo.¹⁶

Minocycline toxicity was minimal in the present study, except in 1 blackpatient with myasthenia gravis and insulin-requiring diabetes mellitus, whodeveloped a drug hypersensitivity syndrome. Since the incidence of minocycline-relatedhypersensitivity has been claimed to be higher in patients with autoimmunedisorders than in control patients and in black patients than in white patients,²¹ it would be warranted to exclude these conditionsfrom future trials using minocycline.²² Indeed,the incidence of the main adverse effects related to minocycline therapy,such as drug hypersensitivity syndrome, hyperpigmentation, and dizziness,appears to be much lower in patients who receive doxycycline.^23,24Thus, in future trials, it will be of great interest to evaluate more extensivelythe therapeutic efficacy of doxycycline in patients who present with granulomatouscutaneous disorders.

It is noteworthy that among 9 patients who received single-agent therapywith chloroquine or hydroxychloroquine for sarcoidosis before entering intothe minocycline study, 7 did not show any response, suggesting that tetracyclinesshould be considered an alternative therapy in cases that show a resistanceof sarcoidal lesions to antimalarial drugs. However, because sarcoidosis lesionsmay be self-resolving, and the results of the present study were obtainedin a limited, open trial, caution should be used when making conclusions concerningthe true efficacy of tetracyclines in sarcoidosis. Indeed, prospective placebo-controlledstudies are warranted to comparatively evaluate the efficacy and toxicityof minocycline and doxycycline in patients affected with chronic forms ofsarcoidosis.

Accepted for publication July 27, 2000.

We are indebted to Laurence Ollivaud, MD, Manuelle Viguier, MD, andThibault Tandeau de Marsac, MD, for clinical assistance, Olivier Vérola,MD, for histopathological examination of skin biopsy specimens, and to Anne-MarieZagdanski, MD, for interpretation of chest x-ray films.

Corresponding author and reprints: Hervé Bachelez, MD, PhD,Institut de Recherche sur la Peau, Hôpital Saint-Louis, 1 Avenue ClaudeVellefaux, 75475 Paris Cedex 10, France (e-mail: herve.bachelez@sls.ap-hop-paris.fr).