A three-year-analysis of fixed drug eruptions in hospital settings in France

ARTICLE

Auteur(s) : Nesrine Brahimi¹, Emilie Routier², Nadia Raison-Peyron³, Anne-Fleur Tronquoy⁴, Caroline Pouget-Jasson⁵, Stéphanie Amarger⁶, Laurent Machet⁷, Emmanuelle Amsler⁸, Antoine Claeys⁹, Bruno Sassolas¹⁰, Dominique Leroy¹¹, Anne Grange¹², Alain Dupuy¹³, Nadège Cordel¹⁴, Jean-Marie Bonnetblanc¹⁵, Brigitte Milpied¹⁶, Marie-Sylvie Doutre¹⁷, Marie-Thérèse Guinnepain¹⁸, Annick Barbaud⁵, Olivier Chosidow², Jean-Claude Roujeau², Bénédicte Lebrun-Vignes¹⁹, Vincent Descamps¹

¹Department of Dermatology, Bichat Hospital, APHP, 46 rue Henri-Huchard, 75877 Paris
²Henri Mondor Hospital, Créteil
³Saint Eloi Hospital, Montpellier
⁴Charles Nicolle Hospital, Rouen
⁵Nancy Hospital, Nancy
⁶Hôtel-Dieu Hospital, Clermont-Ferrand
⁷Tours Hospital, Tours
⁸Tenon Hospital, Paris
⁹Metz-Thionville Hospital, Metz
¹⁰Morvan Hospital, Brest
¹¹Caen Hospital, Caen
¹²Robert-Debré Hospital, Reims
¹³Saint-Louis Hospital, Paris
¹⁴Pointe-à-Pitre Hospital, Guadeloupe
¹⁵Dupuytren Hospital, Limoges
¹⁶Saint-André Hospital, Bordeaux
¹⁷Haut-Lévêque Hospital, Pessac
¹⁸Pasteur Hospital, Paris
¹⁹Pitié-Salpétriêre Hospital, Paris

accepté le 26 F�vrier 2010

Fixed drug eruption (FDE) is a commonly reported type of cutaneous drug adverse event. This adverse drug reaction has unique properties that allow prompt clinical recognition and proper diagnosis. The most characteristic findings of FDE are recurrence of similar lesions at the same sites and healing with residual hyperpigmentation. The usual short delay between drug intake and the development of FDE helps identify the culprit drug. This adverse drug reaction, which may be considered as a prototype of drug-induced reactions, warrants a special interest to better understand the pathogenesis of adverse drug reactions. Few large in-hospital studies are available. We retrospectively analysed cases of FDE diagnosed in a hospital setting and described their characteristics.

Materials and methods

For each case, the following data were collected: age, sex, drug intake, clinical data (delay between drug intake and development of the cutaneous lesions, number of lesions, localization, residual pigmentation, and severity), available provocation tests results, and pathological findings when available. Practitioners were asked to classify FDE as severe (need for hospitalization, prolonged hospitalization, extensive FDE) or non severe (few or limited lesions).

Because this was a descriptive, non-interventional study, the project did not require any approval by an institutional review board.

Results

We found a correlation between the pattern of lesion distribution and sex. Ninety per cent of men had lesions localised to the genital areas whereas 89% of women had lesions on the limbs (especially on the hands and the feet) (p = 0.0026 Pearson test). The main pathological finding in the FDE was epidermal necrosis (observed in 38 cases). In two cases, lichenoid dermatitis was observed. Fifty five out of 59 patients reported oral or parenteral drug intake within fifteen days of the skin eruption (figure 1). The delay between drug intake and skin eruption was usually around 48 hours. However, it varied from a few hours to fifteen days. In 2 cases, no drug intake was reported. Interestingly, in one case, administration of ginseng was suspected (second episode). In two other cases unidentified eyedrops and influenza vaccine were suspected. Paracetamol, piroxicam, amoxicillin, and carbocystein were the causative drugs in 9, 7, 4, and 4 cases, respectively (table 2). Table 2 shows that non steroidal anti inflammatory drugs (NSAIDs) and analgesics were the most frequent culprit drugs.

We found no relationship between the drug and clinical subtypes. However, we found a relationship between localization and drug class. Sixty per cent of patients who had lesions on the face had taken mucolytics and 50% of those with lesions on the genital areas had taken NSAIDs (p = 0.003 Pearson test). There was no relationship between FDE severity and drug class (p = 0.192 Pearson test). We found that the causative drug class was associated with the age of the patients: antibiotics and paracetamol induced FDE occurred in patients older than 60 years and younger than 50 years, respectively (p = 0.0293 analyse of variance Anova).

Patch tests were inconsistently informative: they were positive in twelve cases and negative in seven. Whether they were performed on involved or normal skin was inconsistently mentioned. Among negative patch tests, 3 were done on involved skin. Among positive patch tests, 2 were done on involved skin. Interestingly, in one case where patch tests were performed on both normal and involved skin, patch testing was only positive on involved skin.
Table 1 Patients’ demography and FDE characteristics (n = 59)

Comments

This study confirms the traditional FDE findings, including the short delay between drug intake and FDE development (48 hours), recurrent episodes involving the same sites, clinical characteristics (plaques with bullous lesions, pigmented residual lesions). But it also highlights some unusual and important characteristics. In light skinned patients, FDE is often unpigmented. In our series, 12 cases out of 59 were unpigmented. We must emphasize that FDE is named in French “érythème pigmenté fixe”. This inappropriate term may lead to misdiagnosis and needs to be revised. The proportionally high number of recurrent episodes (n = 34) illustrates the fact that the diagnosis of FDE is often delayed. This finding may be explained by the low recognition rate of FDE by general practitioners.

Our series confirms that this “benign” adverse drug reaction may be severe. One must remember that FDE may be a differential diagnosis of toxic epidermal necrolysis. In our hospital based study, 13/59 cases were considered by the physician as severe. The recruitment of patients in a hospital setting is an evident bias. Our results should be interpreted based on the manner patients were recruited: patients with the severe form were admitted for FDE, while those with the benign forms developed FDE during hospitalization for another disease.

The most common causative drug was paracetamol. Paracetamol-induced FDE has already been reported [4-7]. Interestingly such cases often affected children. Accordingly, FDE in our 4 youngest patients was induced by paracetamol. The NSAIDs are common causes of FDE. Few cases of FDE induced by piroxicam have been reported [8, 9]. In our study, we registered seven cases for piroxicam and two further cases for naproxen and tenoxicam. Amoxicillin-induced FDE (four cases) had also been previously reported in the literature [10-13]. Carbocystein was also a frequent culprit drug: four cases were found. Since these four drugs (paracetamol, NSAID, amoxicillin, carbocystein) are frequently prescribed in combination, it may be at times difficult to pinpoint the real culprit drug. In two cases no culprit drug was identified. Food contaminants or additives may have been implicated. In one case, ginseng intake was initially suspected and further confirmed following a second episode.

Patch testing on involved skin is considered to be of great value for identification of the culprit drug in FDE. Intra epidermal memory T cells are considered to be a major actor in FDE, and are responsible for recurrent events. One case perfectly illustrates this classical view: patch testing with paracetamol was positive only on affected lesional skin whereas this same test was negative in normal skin. But on the whole, our results do not support the relevance of patch testing site. Indeed, positive patch tests were observed on normal skin and negative tests on involved skin.

The most original finding of our study is the link between the anatomical distribution of lesions and gender. FDE involved the genitalia in men and the limbs in women. We have no clear explanation for this finding. One possibility is that involvement of genitalia in women is harder to recognize. It was also proposed that FDE could be a sexually inducible reaction: post-coital FDE were reported in patients whose spouses were taking offending drugs [14]. As previously mentioned, half of the patients who had lesions on the genital areas had taken NSAIDs. These drugs are eliminated by the urinary tract. In the same way that FDE may be caused by drugs present in vaginal fluids, the presence of drugs in the urethra may cause lesions on the genitalia. Such an association between FDE location and drugs has already been described for tetracycline and analgesics, which preferentially involve the genitalia, and limb and trunk, respectively [15, 16]. The authors provided no explanation for this finding.

In our series, carbocystein-induced FDE (four cases) preferentially affected the face, a region near the site of local inflammation in these patients treated for cough. Based on FDE pathogenesis, which is a lymphocyte CD8-mediated reaction, it has been proposed that the offending drug may induce local reactivation of memory T cell lymphocytes localized in epidermal and dermal tissues and initially targeted by the viral infection [17].

Acknowledgements

References

2 Moore N, Biour M, Paux G, et al. Adverse drug reaction monitoring: doing it the French way. Lancet 1985; 2: 1056-8.

3 Fiszenson-Albala F, Auzerie V, Mahe E, et al. A 6-month prospective survey of cutaneous drug reactions in a hospital setting. Br J Dermatol 2003; 149: 1018-22.

4 Patel RM, Marfatia YS. Clinical study of cutaneous drug eruptions in 200 patients. Indian J Dermatol Venereol Leprol 2008; 74: 430.

5 Gómez Martínez M, Fernández de Miguel C, Domínguez Lázaro A, Laguna Martínez JJ, Orta Martiartu M, Lozano Ayllón M. Fixed exanthema due to paracetamol. J Investig Allergol Clin Immunol 1996; 6: 131-2.

6 Silva A, Proença E, Carvalho C, Senra V, Rosário C. Fixed drug eruption induced by paracetamol. Pediatr Dermatol 2001; 18: 163-4.

7 Galindo PA, Borja J, Feo F, Gómez E, Encinas C, García R. Nonpigmented fixed drug eruption caused by paracetamol. J Investig Allergol Clin Immunol 1999; 9: 399-400.

8 Gül U, Gönül M, Heper AO. Piroxicam-induced fixed drug eruption. Eur J Dermatol 2008; 18: 96-7.

9 Cuerda Galindo E, Goday Buján JJ, García Silva JM, Martínez W, Verea Hernando M, Fonseca E. Fixed drug eruption from piroxicam. J Eur Acad Dermatol Venereol 2004; 18: 586-7.

10 Jiménez I, Antón E, Picáns I, Sánchez I, Quiñones MD, Jerez J. Fixed drug eruption from amoxycillin. Allergol Immunopathol 1997; 2: 247-8; (Madr).

11 Arias J, Férnandez-Rivas M, Panadero P. Selective fixed drug eruption to amoxycillin. Clin Exp Dermatol 1995; 20: 339-40.

12 Gil García JF, Lérida Arias MT, del Pozo Hernando LJ, García-Pando AC. Fixed drug eruption due to amoxicillin. Med Clin 1994; 102: 438; (Barc).

13 Shuttleworth D. A localized recurrent pustular eruption following amoxycillin administration. Clin Exp Dermatol 1989; 14: 367-8.

14 Zawar V, Kirloskar M, Chuh A. Fixed drug eruption – a sexually inducible reaction? Int J STD AIDS 2004; 15: 560-3.

15 Schmutz JL, Barbaud A, Trechot P. Fixed drug eruption and carbocysteine. Ann Dermatol Venereol 2005; 132: 403.

16 Sharma VK, Dhar S, Gill AN. Drug related involvement of specific sites in fixed eruptions: a statistical evaluation. J Dermatol 1996; 23: 530-4.

17 Shiohara T, Mizukawa Y. Fixed drug eruption: a disease mediated by self-inflicted responses of intraepidermal T cells. Eur J Dermatol 2007; 17: 201-8.