Drug related clinical pattern in fixed drug eruption

ARTICLE

Fixed drug eruption (FDE) is characterized by recurrent well-defined lesions in the same location each time the drug responsible is taken. Patients often take more than one drug simultaneously, and they often do not remember which drug they have taken. So it is almost always a challenge for the physician to find out the offending agent. The aim of this open prospective study is to investigate the presence of a particular clinical pattern related to the eliciting drug which would help by evaluating FDE cases more accurately, especially those with a multi-drug history.

Patients and methods

Between January 1996 and February 1998, 64 FDE cases were diagnosed in our clinic, and the causative drugs were established by oral provocation as described below.

Oral provocation was performed after written or oral consent of patients, and of parents or legal guardians in children. They were asked not to take any susceptible drugs, antihistamines, immunosuppresives and not to use topical or systemic corticosteroids for at least 4 weeks before and during the provocation tests. Pregnant or lactating women were excluded. Four weeks after the clearance of the lesions an oral challenge with graded doses of the suspected drug was performed starting with 1/8 of a single dose. Doses were increased at 24 hour intervals to one-fourth, one-half and one full dose, if necessary. The test was considered positive if the previously involved FDE sites reactivated with marked violaceous erythema with or without induration, itching or burning. If no reaction occurred with that drug, the next suspected drug-according to the patient's history, was tested in the same manner. If reactivation occurred with the first drug, the next suspected drug was tested only after the complete subsidence of the first reaction.

Patients were asked about the duration of their disease, number of attacks, accompanying diseases, drug intake, atopy, familial occurrence of FDE, skin type and residual pigmentation on previous sites of eruption. Clinical evaluation was performed with regard to localization, morphology and number of lesions and relation to the causative agent. The significance of preferential site involvement due to specific drugs was statistically evaluated by Fisher's exact test. A p value ¾ 0.05 indicated a significant difference between two compared drugs. Statistical analysis was carried out only for drugs causing FDE in 6 or more patients.

Results

The age range of 64 patients (29 female and 35 male) was 6 to 66 years (mean = 34.8 years). Four patients were younger than 12 years old. The occurrence of drug hypersensitivity ranged from 1 week to 10 years (mean = 25.2 months). The number of attacks ranged between one to more than 10 times. At the time of diagnosis, 11 patients (17%) had had their first attack whereas 9 (14%) had had more than 10 attacks. Twenty of 53 patients with recurrent FDE reported an increase in the number and severity of the lesions and/or additional involvement of mucosal surfaces in subsequent attacks. Thirty-six patients had Fitzpatrick skin type III, 83% of them showing residual pigmentation on previous FDE sites. Twenty-seven patients had a Fitzpatrick skin type II, 67% of them had residual pigmentation. One patient had skin type IV and residual pigmentation on previously involved skin.

According to the oral provocation, cotrimoxazole was the cause of FDE in 48 cases (75%), followed by naproxen sodium in 8 cases (12.5%), dipyrone in 6 cases (9.5%), dimenhydrinate in one case (1.5%) and paracetamol in one case (1.5%) (Table I). Sensitivity to more than one drug was not observed. Fifty-eight patients (91%) reacted after ingestion of 1/8 of a single dose, whereas 4 patients reacted after one-fourth dose and 2 patients after one full dose. The time required for the positive reaction to start ranged between 10 minutes and 12 hours (mean: 2.5 hours); it was not related to the type of the drug. In two patients with naproxen-induced FDE, foot lesions in one case and genital lesions in the other case did not react whereas lesions on hand and oral mucosa fully reactivated, respectively. The genital lesion of a patient with cotrimoxazole-induced FDE also did not react whereas a new lesion developed on the wrist, implicating a so-called "wandering" lesion. Dissemination of the lesions or constitutional symptoms did not occur following oral provocation. Patients with generalized FDE showed generalized flare-up of the lesions but systemic reactions did not occur.

Nonpigmenting FDE was diagnosed in 14 cases (22%); six of them with Fitzpatrick skin type III and 8 with skin type II. Cotrimoxazole was the offending drug in 11, dipyrone in 2 cases, and naproxen in one case. Two cases had their second attack whereas the majority of nonpigmenting FDE cases had had 3 to more than 10 attacks (mean: 4.4).

Two cotrimoxazole-induced FDE cases were brother and sister and they reported that their mother had also cotrimoxazole-induced FDE. Two additional patients reported the occurrence of cotrimoxazole-induced FDE in first and second degree family members, respectively. FDE in one of them was confirmed by rechallenge of the drug.

The most frequently involved localization was the orogenital mucosa (61%). Male genitalia, i.e., glans penis was predominantly affected (52%) and cotrimoxazole was the main causative agent. Oral lesions were mainly located on buccal mucosa, gingiva, inner sites of the lips and less frequently on the tongue. The next frequent sites of involvement were lips and face (48%), followed by trunk and extremities (41%) and acral regions (38%). The sites affected by different drugs are presented in Tables II and III. Fisher's exact test revealed a statistically significant difference only for dipyrone versus cotrimoxazole over trunk and extremities (p = 0.03).

Eighteen patients had a solitary lesion and cotrimoxazole was the offending agent in 16. Five patients had generalized involvement, again with cotrimoxazole as the most offending agent. Two of them showed a generalized bullous reaction. FDE in all 4 children were cotrimoxazole-induced. Two of them showed a generalized involvement of trunk and extremities. One case had a generalized bullous reaction on the skin and orogenital mucosa.

The following unusual clinical patterns were observed: solitary large plaque on the cheek in two cases; both healing without residual pigmentation, linear FDE in one and nonpigmenting symmetrical FDE on the palms in another case due to cotrimoxazole.

Size of the lesions ranged between 1-10 cm. Lesions were mainly of a classical type presenting with erythematous macules and occasional bullae formation. Solitary lesions on the glans penis and oral lesions were exclusively bullous. One cotrimoxazole-induced and another dipyrone-induced case presented with an urticarial type of lesions. Histopathological examination of the urticarial lesions revealed dyskeratosis, hydropic degeneration in the epidermal basal layer, edema in the upper dermis, and a perivascular mononuclear inflammatory infiltration with dense eosinophils.

Discussion

FDE is commonly seen in the 21-40-year age group, although it has been reported at all ages [1]. Although it is believed to occur infrequently in women, the female/male ratio in our FDE series was 1/1.2. Genital mucosa was predominately affected in males whereas extramucosal lesions were predominately observed in females.

Systemic provocation is a safe and still the most reliable method for establishing the etiological agent in FDE [2-4]. In a majority of cases, we achieved positive results with 1/8 of a single dose. No systemic or anaphylactic reaction occurred, even in cases with generalized lesions. In the present study, 75% of FDE cases were cotrimoxazole-induced. Patients were taking this drug for urinary or respiratory tract infections. The prevalence of cotrimoxazole-induced FDE, varies from 11.49% to 71% in different studies, all established by oral provocation [2, 3, 5, 6].

The principal area affected in the present series was the male genitalia with cotrimoxazole as the main cause. In spite of the broad use in many infections in our country, we interestingly did not diagnose any tetracycline-induced FDE which has been reported as a more common cause of genital FDE [2, 5, 7]. Mucosal FDE due to cotrimoxazole has occasionally been documented [3, 8, 9]. Similar to our findings, Kanwar et al. reported cotrimoxazole-induced FDE mainly located on male genitalia [3]. However, in two studies investigating the drug related specific pattern in FDE, tetracycline was the only significant cause of FDE on male genitalia although cotrimoxazole was the most common offender for FDE (36.3% and 32.8%, respectively) [5, 7]. Thankappan et al. reported that cotrimoxazole-induced FDE was mainly located on the lips and on trunk and extremities [5]. Sharma et al. also observed an increased association of FDE on the lips with cotrimoxazole that was not statistically significant [7]. In our series, lips and face were involved in 44% of cotrimoxazole-induced cases whereas this site of the body was involved in 75% of naproxen-induced cases. However, the difference was not statistically significant. Naproxen has only sporadically been reported to cause FDE [10].

In both studies, dipyrone-induced FDE was mainly located on the trunk and extremities [5, 7]. Sharma et al. found out that this was a statistically significant association [7]. In our series, the most common offender for lesions on the trunk and extremities was also dipyrone. However, a definite significant difference could only be shown between dipyrone and cotrimoxazole but not between dipyrone and naproxene for this location. On the other hand, Thankappan et al. reported that dipyrone was found to cause solitary lesions in the majority of cases; mainly located over the back below the scapula [5]. This is not consistent with our findings. We observed that dipyrone caused multiple/generalized FDE lesions (Table I).

Dimenhydrinate and paracetamol are less common causes of FDE [11-14]. Thankappan reported 9 cases with paracetamol-induced FDE, all with an extensive involvement of trunk and extremities [5]. Paracetamol-induced FDE in our case was also located on trunk.

Generalized FDE has frequently been reported in association with phenazone derivatives and barbiturates [15]. In the present series, it was unusually associated with cotrimoxazole or dipyrone. Generalized involvement was also observed in 2 of 4 children with cotrimoxazole-induced FDE. Although FDE occurs only rarely in children, generalized bullous FDE seems to be relatively frequent [16]. Acetylsalicylic acid and hyoscine butyl bromide were the commonest drugs incriminated [16]. None of the cases was associated with cotrimoxazole in previous reports.

Several less common clinical types of FDE with urticarial, nodular, eczematous and erythema-multiforme like lesions [1, 17], and unusual appearance such as symmetrical pigmenting [18], nonpigmenting [19, 20], wandering [21], pulsating [22] and linear lesions [23] have been reported. We also observed urticarial lesions with histopathologically marked eosinophilic infiltration in 2 cases due to cotrimoxazole and dipyrone.

Nonpigmenting lesions were present in nearly 1/4 of the cases independent of the skin type. One could speculate that it is related more to the number of attacks, rather than the drug involved, whether lesions of FDE are pigmenting or non-pigmenting. Repeated attacks would then almost always result in pigmentation. In contrast to that, most of the patients with nonpigmenting lesions in the present series had repeated attacks. Furthermore, cotrimoxazole was the offending drug in most of them including symmetrical nonpigmenting palmar lesions in one case, and solitary nonpigmenting large plaque on the cheek in two cases. As far as we know, nonpigmenting lesions have been reported as a rare finding in FDE, occurring almost always after pseudoephedrine [19, 20]. Nonpigmenting FDE due to cotrimoxazole or its constitutes has not been reported before. Shelley and Shelley [19] regard nonpigmenting FDE as a separate entity showing a symmetric distribution in contrast to the common asymmetrical pigmenting type. In our opinion, it is questionable if this is a rare and separate entity because we observed both symmetrical and solitary nonpigmenting FDE mainly after cotrimoxazole, but also after dipyrone or naproxen.

Cotrimoxazole-induced wandering lesion was diagnosed in one case that was also an unusual finding. Wandering FDE has previously been reported as a mucocutaneous reaction to acetaminophen [21].

Linear FDE in the present series has recently been reported as the second linear FDE in the literature [24].

In the present series, four cases with cotrimoxazole-induced familial FDE were observed; three of them were confirmed by rechallenge with the drug. There is only one report in the literature of a familial FDE due to cotrimoxazole [25]. Familial cases have recently been reported with pyrazolones, and a genetic susceptibility to FDE has been linked to HLA-B22 [26].

CONCLUSION

In conclusion, the present study shows that cotrimoxazole is the major etiological agent in FDE in Turkey, followed by naproxen and dipyrone. In spite of the similar broad use of tetracyclines in our country none of our cases was tetracycline induced. It is apparent that each community has to establish its own list of the most frequent FDE inducers as drugs used in various countries differ from each other. Direct questioning of these "top" FDE inducers could help in evaluating the FDE cases, especially those with more than one suspected drug in the history. Cotrimoxazole preferably involves male genitalia as solitary penile lesions, but unfortunately, a definite relation with clinical pattern and the causative agent could not be established. Furthermore, a genetic background of cotrimoxazole-induced FDE has to be investigated, which is also supported by familial cases.

Article accepted on 31/1/00

REFERENCES

1. Sehgal VN, Gangwani OP. Fixed drug eruption. Current concepts. Int J Dermatol 1987; 26: 67-74.

2. Sehgal VN, Gangwani OP. Genital fixed drug eruptions. Genitourin Med 1986; 62: 56-8.

3. Kanwar AJ, Bharija SC, Singh M, Belhaj MS. Ninety-eight fixed drug eruptions with provocation tests. Dermatologica 1988; 177: 274-9.

4. Breathnach SM. Management of drug eruptions: Part II. Diagnosis and treatment. Australas J Dermatol 1995; 36: 187-91.

5. Thankappan TP, Zachariah J. Drug-specific clinical pattern in fixed drug eruptions. Int J Dermatol 1991; 30: 867-70.

6. Gómez B, Sastre J, Azofra J, Sastre A. Fixed drug eruption. Allergol et Immunopathol 1985; 13: 87-91.

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

8. Gaffoor PM, George WM. Fixed drug eruptions occurring on the male genitals. Cutis 1990; 45: 242-4.

9. Pandhi RK, Kumar AS, Satish DA, Bhutani LK. Fixed drug eruptions on male genitalia: clinical and etiologic study. Sex Transm Dis 1984; 11: 164-6.

10. Gonzalo Garijo MA, Bobadilla González P. Cutaneous reaction to naproxen. Allergol Immunopathol (Madr) 1996; 24: 89-92.

11. Schnyder UW, Wutrich B, Mitarbeiter B. Multiples fixes Arzneimittelexanthem auf Dimenhydrinat. Dermatologica 1982; 165: 292-5.

12. Hogan DJ, Rooney ME. Fixed drug eruption due to dimenhydrinate. J Am Acad Dermatol 1989; 20: 503-4.

13. Wilson HT. A fixed drug eruption due to paracetamol. Br J Dermatol 1975; 92: 213-4.

14. Thomas RH, Munro DD. Fixed drug eruption due to paracetamol. Br J Dermatol 1986; 115: 357-9.

15. Kauppinen K, Stubb S. Fixed eruptions: causative drugs and challenge tests. Br J Dermatol 1985; 112: 575-8.

16. Kanwar AJ, Bharija SC, Belhaj MS. Fixed drug eruptions in children: a series of 23 cases with provocative tests. Dermatologica 1986; 172: 315-8.

17. Sowden JM, Smith AG. Multifocal fixed drug eruption mimicking erythema multiforme. Clin Exp Dermatol 1990; 15: 387-8.

18. Sehgal VN, Gangwani OP. An unusual symmetrical fixed drug eruption. Indian J Dermatol Venerol Lepr 1985; 51: 172-3.

19. Shelley WB, Shelley ED. Nonpigmenting fixed drug eruption as a distinctive reaction pattern: examples caused by sensitivity to pseudoephedrine hydrochloride and tetrahydrozoline. J Am Acad Dermatol 1987; 17: 403-7.

20. Alanko K, Kanerva L, Mohell-Talolahti B, et al. Nonpigmented fixed drug eruption from pseudoephedrine. J Am Acad Dermatol 1996; 35: 647-8.

21. Guin JD, Haynie LS, Jackson D, et al. Wandering fixed drug eruption. A mucocutaneous reaction to acetaminophen. J Am Acad Dermatol 1987; 17: 339-402.

22. Parish LC, Witkowski JA. Pulsating fixed drug eruption due to tetracycline. Acta Derm Venereol (Stockh) 1978; 58: 545-7.

23. Sigal-Nahum M, Konqui A, Gaulier A, Sigal S. Linear fixed drug eruption. Br J Dermatol 1988; 118: 849-51.

24. Özkaya-Bayazit E, Baykal C. Trimethoprim induced linear fixed drug eruption. Br J Dermatol 1997; 137: 1028-9.

25. Bhargava P, Kuldeep CM, Mathur NK. Polysensitivity and familiar occurrence in fixed drug eruption. Int J Dermatol 1997; 36: 234-40.

26. Pellicano R, Ciavarella G, Lomuto M, Di Giorgio G. Genetic susceptibility to fixed drug eruption: evidence for a link with HLA-B22. J Am Acad Dermatol 1994; 30: 52-4.