Original Article
Lack of Efficacy of Phenytoin in Recessive Dystrophic Epidermolysis Bullosa
N Engl J Med 1992; 327:163-167July 16, 1992
- Abstract
- Abstract
Background.
Recessive dystrophic epidermolysis bullosa is an uncommon, severely disabling, heritable disorder characterized by abnormal fragility of the skin. Open trials have suggested that phenytoin is an effective treatment, and this therapy is now widely used.
Methods.
To determine the efficacy of phenytoin in the treatment of recessive dystrophic epidermolysis bullosa, we performed a randomized, double-blind, placebo-controlled, crossover trial in 36 patients. Each treatment was given for five to seven months, separated by a two-month period. We measured the total number of blisters and erosions on the entire body, the size of three plaques containing blisters and erosions, and the number of blisters and erosions in the three plaques at the beginning and end of each treatment period in each patient.
Results.
Twenty-two patients completed both courses of therapy, seven patients completed one course, and seven patients withdrew before completing a single course. There was no significant difference in disease activity between phenytoin treatment and placebo treatment, as measured by changes in the number of blisters and erosions on the entire body (7 percent decrease vs. 6 percent increase), in the area of three designated plaques (0.4 percent decrease vs. 0.2 percent increase), or in the number of blisters and erosions in the designated plaques (12 percent decrease vs. 31 percent increase).
Conclusions.
Phenytoin is not an effective treatment for patients with recessive dystrophic epidermolysis bullosa. (N Engl J Med 1992;327:163–7.)
Media in This Article
Figure 1
Lesions of Recessive Dystrophic Epidermolysis Bullosa on the Leg of a Three-Month-Old Girl before (Panel A) and after (Panel B) Treatment with Phenytoin, Showing Enlargement of Erosions (Arrow) within a Well-Demarcated Plaque.Figure 2
Plaque Consisting of a Well-Demarcated Area of Blisters (Arrow) and Erosions (Arrowheads) on the Upper Back of a 23-Year-Old Man before (A) and after (B) Treatment with Placebo.Article Activity
- Article
EPIDERMOLYSIS BULLOSA is a heterogeneous group of rare, heritable disorders characterized by abnormal fragility of the skin.1 In affected persons, minor cutaneous trauma causes the formation of slow-healing blisters that are prone to secondary infection and scarring. Studies of the ultrastructural levels at which blistering occurs have led to the recognition of three main types of the disorder: simplex, junctional, and dystrophic. In patients with recessive dystrophic epidermolysis bullosa, blistering occurs below the lamina densa and the number of anchoring fibrils in the upper dermis is decreased.2 , 3 The disease usually begins during infancy, and by adolescence the patients have characteristic mitten-like scarring of the hands and feet and mucosal involvement (e.g., involving the esophagus, eyes, and respiratory tract) that can be severe and life-threatening. As with other types of epidermolysis bullosa, there is no cure.
The collagenase activity of the skin fibroblasts of patients with recessive dystrophic epidermolysis bullosa is increased.4 Phenytoin inhibits collagenase activity in vitro5 and was found to be an effective treatment in two patients with dystrophic epidermolysis bullosa.5 In an open study, treatment with phenytoin for six months reduced the number of blisters and erosions by 40 percent in 12 patients, and their condition generally worsened within four weeks after therapy was discontinued; another 5 patients had no response during therapy.6 These encouraging results led to the widespread use of this treatment in the hopes of managing this disabling and disfiguring disorder. Of the first 125 patients with recessive dystrophic epidermolysis bullosa whose cases were reported to the National Epidermolysis Bullosa Registry, 79 (63 percent) had received phenytoin. Despite its common use in this rare disease, the efficacy of phenytoin has not been studied prospectively. Such studies are difficult to perform in patients with rare diseases, and multicenter cooperation is usually required to enroll enough patients to achieve sufficient power for statistical analysis. To this end, we performed a multicenter randomized, placebo-controlled, double-blind, crossover study of phenytoin in the treatment of recessive dystrophic epidermolysis bullosa.
Methods
Treatment Plan
The study protocol was approved by the appropriate institutional review committees, and informed consent was obtained from either the patients or their parents. The study group consisted of 36 patients, each of whom had unequivocal clinical and histologic evidence of recessive dystrophic epidermolysis bullosa. Thirty-one patients had had the disease since birth. Previous therapy included phenytoin in 20 patients, oral corticosteroids in 5, topical corticosteroids in 10, oral vitamin E in 12, special diets in 6, and oral antimalarial therapy in 1. Each patient was randomly assigned to receive either phenytoin or placebo for five to seven months. After a two-month period of no therapy, the patients received the other treatment. Phenytoin was continued for four months after the patient's blood concentration reached a steady state within the target range of 8 to 20 μg per milliliter. The initial dose of phenytoin was 5 mg per kilogram of body weight per day, given once daily as chewable tablets of 50 mg each. Because the time required to reach the target concentration varied, the entire treatment course ranged from five to seven months. To preserve the double-blind design, the duration of placebo treatment was also varied from five to seven months. The blood phenytoin values were known only to the study coordinator, who instructed the investigators to increase or decrease the dosage of phenytoin. To preserve the double-blind design, blood was also drawn during the placebo treatment period, and the study coordinator instructed the investigators to adjust the dosage of placebo during the first several months of treatment. The patients were examined by an investigator at the beginning and end of each treatment period, and blood samples were obtained every two weeks for blood phenytoin measurements until the target range (8 to 20 μg per milliliter) was achieved; thereafter, blood was drawn at monthly intervals. Measurements were made in plasma or serum samples, depending on the practice at local laboratories. The following methods were used: a homogeneous enzyme immunoassay (Emit phenytoin assay, Syva, Palo Alto, Calif.), an adaptation of the Emit assay performed on the ACA discrete clinical analyzer (Dupont, Wilmington, Del.), and a fluorescence polarization immunoassay performed on the TDx analyzer (Abbott Diagnostics, Abbott Park, Ill.).
The following were determined at the beginning and end of each treatment period: the total number of blisters and erosions on the skin, the size of three chronic plaques (each plaque contained blisters, erosions, or scars), and the total number of blisters and erosions in the three plaques. In addition, the investigators assessed each patient's overall condition using a seven-point scale (in which a score of -3 indicated that the condition was markedly worse, a score of 0 that the condition was unchanged, and a score of +3 that the condition was markedly improved), based on both the overall activity of the skin disease and the patient's well-being, including an assessment of appetite and the level of physical activity. Complete blood counts, a biochemistry profile (blood urea nitrogen and serum creatinine, liver enzymes, bilirubin, albumin, protein, calcium, glucose, iron, total iron-binding capacity, and antinuclear antibody), and urinalysis were performed before and at the end of each treatment period.
Statistical Analysis
The power of the study was calculated on the basis of the variation in blister formation in the original open trial6; the power to detect differences comparable to those found in the open study would exceed 90 percent if 34 patients completed both treatment periods. We enrolled 36 patients, 22 of whom completed both treatment periods. With the final sample, the power to detect differences comparable to those of the open study was 65 percent with a Type I error of 0.05, and more than 80 percent with a Type I error of 0.1, which would be the equivalent of one-tailed testing criteria at a P level of ≤0.05.
A small number of patients had large changes in the size of the plaques or the number of blisters and erosions, with a resultant abnormal distribution of the values. The values were therefore logarithmically transformed. We then calculated the percentage change as the antilog of the sum of the logs of the changes in each value. We used the paired t-test to compare the mean values and the chi-square test to determine the significance in proportions. For nonparametric data such as the physician's global evaluation, we used the Wilcoxon signed-rank test. All P values are two-tailed, and 95 percent confidence intervals were calculated on the basis of the distribution of the results of the Student t-test.
Results
The characteristics of the 36 patients with recessive dystrophic epidermolysis bullosa who entered the study are shown in Table 1Table 1
Clinical Characteristics of Patients with Recessive Dystrophic Epidermolysis Bullosa, According to the Extent of Their Participation in the Clinical Trial.*. Among them, 22 patients completed both periods of treatment, 7 completed only one treatment period (5 were lost to follow-up, 1 was thought to be pregnant, and 1 could not swallow pills because of dysphagia), and 7 others dropped out before completing any treatment period (2 patients were unable to tolerate phenytoin because of central nervous system effects including seizures, 3 did not want to undergo phlebotomy, and 2 were lost to follow-up).Table 2Table 2
Clinical Characteristics of the 22 Patients with Recessive Dystrophic Epidermolysis Bullosa Completing Treatment with Both Phenytoin and Placebo, at the Beginning of Each Treatment Period.* lists the clinical characteristics at the beginning of each treatment period of the 22 patients who completed both courses of treatment. Among the 22 patients, there were no significant differences between treatment periods in the changes in the number of blisters or erosions on the entire body, in the area of the designated plaques, or in the number of blisters or erosions in the designated plaques (Table 3Table 3
Changes in Numbers of Blisters and Erosions and Areas of Plaques in Patients with Recessive Dystrophic Epidermolysis Bullosa during Treatment with Phenytoin and Placebo.*). The total number of blisters and erosions decreased by 7 percent during treatment with phenytoin and increased by 6 percent during treatment with placebo. The area of the three designated plaques decreased by 0.4 percent during phenytoin treatment and increased by 0.2 percent during treatment with placebo, and the number of blisters and erosions in the designated plaques decreased by 12 percent and increased by 31 percent, respectively. Lesions characteristic of the disease are shown before and after treatment with phenytoin in Figure 1Figure 1Lesions of Recessive Dystrophic Epidermolysis Bullosa on the Leg of a Three-Month-Old Girl before (Panel A) and after (Panel B) Treatment with Phenytoin, Showing Enlargement of Erosions (Arrow) within a Well-Demarcated Plaque. and before and after treatment with placebo in Figure 2Figure 2Plaque Consisting of a Well-Demarcated Area of Blisters (Arrow) and Erosions (Arrowheads) on the Upper Back of a 23-Year-Old Man before (A) and after (B) Treatment with Placebo..At the end of the phenytoin-treatment period, the overall condition of the 22 patients was compared with their condition at the beginning of the period and was rated as definitely worse in 2, slightly worse in 2, unchanged in 7, slightly improved in 6, and definitely improved in 2 (data were missing for 3 patients). The respective numbers at the end of the placebo-treatment period were 2, 4, 11, 1, and 3 (data were missing for 1 patient). None of the patients were considered to have had either marked increases or marked decreases in the severity of their symptoms at the end of any period. Among the 18 patients (36 courses) whose condition was assessed at the end of both periods, 4 were considered to have had similar results at the end of both treatment periods, 6 to have improved at the end of the placebo-treatment period, and 8 to have improved at the end of the phenytoin-treatment period (P = 0.38 by the Wilcoxon signed-rank test).
No abnormalities in the complete blood count, biochemistry profile, or urinalysis were attributable to phenytoin therapy. Three patients had ataxia, vomiting, nystagmus, or dizziness when their blood concentrations of phenytoin were 25.88, 28.0, and 30.8 μg per milliliter. To determine whether the response depended on the blood phenytoin concentration, we performed a separate analysis of the response of 12 patients who completed both treatment periods and had blood phenytoin concentrations of at least 6 μg per milliliter for at four months. The responses of these 12 patients were similar to those of the total group of 22 patients (Table 3).
Discussion
In contrast to the results of open studies, in which treatment with phenytoin reduced the number of blisters and erosions by more than 40 percent in 12 of 17 patients6 and 14 of 22 patients7 with recessive dystrophic epidermolysis bullosa, phenytoin had no significant therapeutic effect in this controlled trial. Since some patients in the open trials who responded to phenytoin had blood drug concentrations as low as 0.1 to 4 μg per milliliter,6 differences in the dosage are unlikely to explain the differences in results. Several patients in our study reported subjective improvement in the severity of blistering during treatment, but the improvement actually occurred during placebo therapy. Epidermolysis bullosa has a variable course, and the severity of blistering can be worsened by factors such as increased physical activity, a high ambient temperature, and infection, all of which are difficult to control in open studies.
Our study has several limitations. Only 22 of the 36 patients who were enrolled completed both treatments. The results, however, indicate that at most the improvement associated with phenytoin therapy was small. As in all multicenter studies, the patients were examined by different investigators, which may have introduced some variation in the assessment of the responses to therapy. Seventeen patients were examined by the same investigator throughout both treatment periods, but four patients were examined by two investigators, and one patient by three investigators. The same limitation may have applied to the use of local laboratories for the measurement of blood phenytoin concentrations. The occurrence of symptoms of phenytoin toxicity in some patients may also have compromised the blinded nature of the study. Finally, we cannot exclude the possibility that the recessive form of dystrophic epidermolysis bullosa may represent a heterogeneous group of disorders, one or more of which may respond to phenytoin therapy. This possibility is supported by the finding that the clinical response of some patients correlated with the ability of phenytoin to inhibit collagenase in cultures of fibroblasts from the patients.6 In another study,8 only 4 of 18 patients had elevated collagenase activity in cultured skin fibroblasts, suggesting that the disorder is heterogeneous. We did not measure collagenase activity and therefore cannot comment on any correlation between it and the response to treatment in our patients. Increased collagenase activity has also been reported in skin fibroblasts obtained from patients with junctional epidermolysis bullosa,9 , 10 enhancing the possibility that such increases may not be unique to recessive dystrophic epidermolysis bullosa or may not be its principal defect, but rather may reflect the participation of collagenase in the normal process of wound healing.11 Although our results do not exclude the possibility that some patients with recessive dystrophic epidermolysis bullosa may respond to phenytoin treatment, they do indicate that phenytoin offers no overall benefit when compared with placebo for the treatment of this disorder.
Supported in part by grants (FD01261, AG00282, and AG04993) from the Public Health Service, by a General Clinical Research Center grant (M01RR00102) from the National Institutes of Health to Rockefeller University Hospital, and by the Pew Trusts.
Presented in part at the International Epidermolysis Bullosa Symposium, London, January 19–20, 1989, and the 1989 Tricontinental Meeting of the European Society for Dermatologic Research, the Japanese Society for Investigative Dermatology, and the Society for Investigative Dermatology, Washington, D.C., April 26–30, 1989.
*The members of the Epidermolysis Bullosa Study Group are listed in the Appendix.
Source Information
From the Laboratory for Investigative Dermatology, Rockefeller University, New York (D.C.-B., A.N.L., D.M.C.), and the Department of Dermatology, Beth Israel Hospital, Harvard Medical School, Boston (R.S.S.). Address reprint requests to Dr. Carter at the Laboratory for Investigative Dermatology, Rockefeller University, 1230 York Ave., New York, NY 10021.
Appendix
*The following were members of the Epidermolysis Bullosa Study Group: E.A. Bauer, Stanford University, Stanford, Calif.; M. Bond, University of Tennessee, Memphis; N.B. Esterly, Children's Memorial Hospital, Chicago; J.-D. Fine and L.S. Osment, University of Alabama, Birmingham; S. Gellis, New England Medical Center, Boston; R.-R. Isseroff, University of California, Davis; A.W. Lucky, University of Cincinnati, Cincinnati; S.B. Mallory, Arkansas Children's Hospital, Little Rock; J.S. Storer, Tulane University, New Orleans; V.P. Sybert, Children's Orthopedic Hospital and Medical Center, Seattle; and M. Varghese, Rockefeller University, New York.
- References
References
1
Lin
CrossRef | Web of Science | Medline2
Briggaman
CrossRef | Web of Science | Medline3
Tidman
CrossRef | Web of Science | Medline4
Bauer
CrossRef | Web of Science | Medline5
Eisenberg
CrossRef | Web of Science | Medline6
Bauer
Full Text | Web of Science | Medline7
Cooper
CrossRef | Web of Science | Medline8
Winberg
CrossRef | Web of Science | Medline9
Kero
CrossRef | Web of Science | Medline10
Kero
Medline11
Bauer
CrossRef | Web of Science | Medline
- Citing Articles (17)
Citing Articles
1
Irene Lara-Corrales, Patricia C. Parkin, Derek Stephens, Jill Hamilton, Gideon Koren, Miriam Weinstein, Ronald G. Sibbald, Elena Pope. (2011) The efficacy of trimethoprim in wound healing of patients with epidermolysis bullosa: A feasibility trial. Journal of the American Academy of Dermatology
CrossRef2
Jemima E. Mellerio, Jacqueline E. Denyer. 2011. Epidermolysis Bullosa. , 118.1-118.34.
CrossRef3
J.-D. Fine, S. M. Burge. 2010. Genetic Blistering Diseases. , 1-37.
CrossRef4
S. M. Langan, H. C. Williams. (2009) A systematic review of randomized controlled trials of treatments for inherited forms of epidermolysis bullosa. Clinical and Experimental Dermatology 34:1, 20-25
CrossRef5
Matthias Titeux, Valérie Pendaries, Laure Tonasso, Audrey Décha, Christine Bodemer, Alain Hovnanian. (2008) A frequent functional SNP in the MMP1 promoter is associated with higher disease severity in recessive dystrophic epidermolysis bullosa. Human Mutation 29:2, 267-276
CrossRef6
Noah Scheinfeld. (2006) Nephrogenic Fibrosing Dermopathy. American Journal of Clinical Dermatology 7:4, 237-247
CrossRef7
Jennifer T. Trent, Robert S. Kirsner. (2003) Epidermolysis Bullosa: Identification and Treatment. Advances in Skin & Wound Care 16:6, 284-290
CrossRef8
Ysabel M Bello, Anna F Falabella, Lawrence A Schachner. (2003) Management of epidermolysis bullosa in infants and children. Clinics in Dermatology 21:4, 278-282
CrossRef9
P. J. Rutherford, D. C. Davidson, S. M. Matthai. (2002) Dietary calcium in galactosaemia. Journal of Human Nutrition and Dietetics 15:1, 39-42
CrossRef10
Shan Pai, M. Peter Marinkovich. (2002) Epidermolysis Bullosa. American Journal of Clinical Dermatology 3:6, 371-380
CrossRef11
Daniel Berg, Maryam Asgari. (2001) Evidence-Based Medicine for Dermatologic Surgeons: Concepts in Critical Appraisal of Information. Dermatologic Surgery 27:6, 511-514
CrossRef12
Sree R. K. Karukonda, , , Timothy Corcoran Flynn, , Erin E. Boh, , , Elizabeth I. McBurney, , Glenn G. Russo, , Larry E. Millikan, , . (2000) The effects of drugs on wound healing--part II. Specific classes of drugs and their effect on healing wounds. International Journal of Dermatology 39:5, 321-333
CrossRef13
Andrew N. Lin. (1997) Pyloric Atresia and Epidermolysis Bullosa. Pediatric Dermatology 14:5, 406-408
CrossRef14
Christopher B. Olsen, Laurence F. Bourke. (1997) Recessive dystrophic Epidermolysis bullosa. Two case reports with 20-year follow-up. Australian Dental Journal 42:1, 1-7
CrossRef15
Györgyi Talas, Titus S.T. Adams, Mark Eastwood, Guillermo Rubio, Robert A. Brown. (1997) Phenytoin reduces the contraction of recessive dystrophic epidermolysis bullosa fibroblast populated collagen gels. The International Journal of Biochemistry & Cell Biology 29:1, 261-270
CrossRef16
M G. S. DUNNILL, R. A. J. EADY. (1995) The management of dystrophic epidermolysis bullosa. Clinical and Experimental Dermatology 20:3, 179-188
CrossRef17
Katz, Stephen I., . (1992) Inherited and Acquired Blistering Diseases. New England Journal of Medicine 327:3, 196-197
Full Text
