- Email: [email protected]
Improvement of toxic epidermal necrolysis after the early administration of a single high dose of intravenous immunoglobulin
Case report
Improvement of toxic epidermal necrolysis after the early administration of a single high dose of intravenous immunoglobulin Cristobalina Mayorga, PhD*; Maria Jose´ Torres, MD, PhD*; Jose´ Luis Corzo, MD†; Elena Sanchez-Sabate, PhD‡; Javier Alvarez, MD‡; Angel Vera, MD, PhD§; Sinforiano Posadas, PhD‡; Antonio Jurado, MD, PhD†; and Miguel Blanca, MD, PhD‡
Background: Toxic epidermal necrolysis (TEN) is a severe disease often induced by drugs. Treatment is controversial, although intravenous immunoglobulins (IVIGs) have been effective. Objective: To report the case of a child with TEN after lamotrigine treatment, who improved 24 hours after IVIG administration. Methods: Sequential blood and blister fluid samples were obtained for flow cytometry and reverse transcriptase–polymerase chain reaction analyses. Results: The first blood sample, taken before IVIG administration, showed normal levels of lymphocyte subsets and CLA (4.0%) but high levels of activated lymphocytes (CD69) (18.0%). After treatment, the CLA⫹, CD69⫹, and memory cells increased until day 7, decreasing to normal values at days 15 and 30. In the blister fluid samples, taken on day 1, there were high levels of CD8⫹ (70.2%; CD4/CD8 ratio, 1:5), CLA⫹ (18.8%), and CD69⫹ (70%) cells, decreasing 24 hours after IVIG administration. In the blood samples, there was a Th1 cytokine pattern initially, tending to Th0 with time. Perforin, granzyme B, and Fas ligand were only observed before IVIG administration. Conclusions: A single high dose of IVIG interrupted the progression of skin disease and reduced the expression of the apoptotic markers. The immunologic changes, first seen in blister fluid and remaining several days in peripheral blood, indicate that T cells were first recruited to the skin and then recirculated to blood. Ann Allergy Asthma Immunol. 2003;91:86–91.
INTRODUCTION Toxic epidermal necrolysis (TEN) is a severe disease that is usually induced by drugs or viral infections.1 Although the prevalence is low, the mortality rate is 25 to 30%.2 The most frequent drugs involved are sulfonamides, anticonvulsants, and nonsteroidal anti-inflammatory drugs,1,3 with immunologic mechanisms thought to be involved in its pathophysiology.4 Immunohistochemical studies have shown activated CD4 cells in the epidermis and CD8 cells in the dermis,5 with migration to the skin of memory-activated peripheral * Research Unit for Allergic Diseases, Carlos Haya Hospital, Malaga, Spain. † Paediatric Department, Carlos Haya Hospital, Malaga, Spain. ‡ Allergy Service, Hospital La Paz, Madrid, Spain. § Dermatology Service, Carlos Haya Hospital, Malaga, Spain. This study was partly supported by grants from the Spanish Ministry of Health (01/0014, 01/3031, and PIO20640). The specific contribution of each author is as follows: C. Mayorga and S. Posadas performed the laboratory analysis; M. J. Torres, J. L. Corzo, A. Vera, and A. Jurado made the clinical evaluation; J. Alvarez and E. SanchezSabate performed the immunohistochemical analysis; and M. Blanca coordinated all the study and with M. J. Torres and C. Mayorga wrote the article. Received for publication January 30, 2003. Accepted for publication in revised form February 27, 2003.
86
T lymphocytes expressing the skin-homing receptor (CLA⫹CD69⫹CD45RO⫹).6 T cells in TEN secrete perforin and granzyme B and express Fas ligand (Fas-L),7 components of the lymphocyte effector pathway involved in cell death and apoptosis,1 and a relationship between these mediators and the severity of skin diseases induced by drugs has been reported.7,8 Several different treatments have been tried but so far without success.1 Steroids, the most widely used therapy, are controversial and in severe cases may even worsen the disease by increasing the risk of septicemia.2,9 Other drugs, such as cyclosporine, cyclophosphamide, pentoxifylline, and thalidomide, have also been used, but their efficacy has not been confirmed.9 Recovery after the administration of intravenous immunoglobulin (IVIG) has recently been reported,5,10 –14 although a recent study15 performed in 34 patients with StevenJohnson syndrome and TEN does not support the routine use of IVIG treatment in these patients, especially in cases of impaired renal function. The mechanism proposed for the action of IVIG in TEN, inferred from in vitro studies, is the inhibition of Fas-mediated keratinocyte death by naturally occurring Fas-blocking antibodies in the IVIG
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
preparation.8 To our knowledge, the behavior of the immunologic system in TEN has not yet been studied in vivo. We report the case of a patient who experienced rapid recovery from a severe case of drug-induced TEN after a single high dose of IVIG, together with the immunologic parameters, which were monitored before and after treatment. MATERIAL AND METHODS Case Report A 3-year-old boy with cerebral palsy and epilepsy was given lamotrigine (Lamictal; Glaxo-Welcome, Madrid, Spain) (12.5 mg in the morning and 6.25 mg in the evening) to control seizures. Fifteen days after initiating therapy, a maculopapular exanthema appeared on the trunk and legs, progressing rapidly to bullous lesions accompanied by extensive skin scalding and a high temperature (39° C). The mucous membranes were also involved, with erosions on the conjunctival, oropharyngeal, and urogenital mucosae. A diagnosis of TEN, affecting 60% of the body surface area, was made and use of lamotrigine was discontinued. Treatment with intravenous dexamethasone (6 mg/d) was initiated, but because no improvement was observed after 48 hours, the steroids were withdrawn and a single high dose (2 g/kg) of IVIG (Flebogamma; Grifols Institute, Barcelona, Spain) was administered at a rate of 15 mL/h for more than 24 hours. To monitor the participation of the immunologic system during the response to treatment, lymphocyte subpopulations and activation markers were analyzed in parallel in peripheral blood mononuclear cells and blister fluid cells. Blood samples were obtained immediately before IVIG treatment and 3, 7, 15, and 30 days after. Blister fluid samples were obtained before IVIG administration and 3 days after. No further samples were obtained from blister fluid because no further blisters appeared. Flow Cytometry Analysis Flow cytometry was used to study different immunological markers: T-lymphocyte subsets (CD3, CD4, CD8, CD45RO, and CD45RA), the CLA skin-homing receptor, and the cellular activation marker CD69. The production and expression of several cytokines, interleukin (IL)-2, interferon-␥ (IFN-␥), tumor necrosis factor ␣ (TNF-␣), IL-4, and IL-13, were also determined by flow cytometry. This was performed in peripheral blood mononuclear cells, which were isolated on Ficoll-Paque gradients (Pharmacia Biotech Inc, Piscataway, NJ). Six parameters were analyzed on a Facscalibur flow cytometer using Cell Quest software (Becton Dickinson, San Jose, CA). Negative isotype controls were used to verify the staining specificity of the antibodies used. The production and expression of several cytokines were determined in cells stimulated for 4 hours at 37° C with phorbol 12-myristate 13-acetate (Sigma, St. Louis, MO) and ionomycin (Sigma), and the intracellular protein transport of cytokines was disrupted with brefeldin (Sigma).
VOLUME 91, JULY, 2003
Cells were fixed and permeabilized (Caltag Laboratories, Burlingame, CA) and stained intracellularly with monoclonal antibodies to cytokines (IL-2 and IL-10 were stained with fluorescein isothiocyanate; IFN-␥, TNF-␣, IL-4, and IL-13 were stained with phycoerythrin; all provided by Becton Dickinson). Determination of Cytotoxic Markers The secretion of perforin and granzyme B and the expression of Fas-L were determined by competitive polymerase chain reaction (PCR). RNA Isolation and Complementary DNA Synthesis Total RNA was isolated from cells by the method of Chomczynski and Sacchi.16 The complementary DNA (cDNA) template for reverse transcriptase (RT)–PCR was synthesized from RNA by reverse transcription using a first-strand cDNA synthesis kit (Clontech Laboratories, Palo Alto, CA). In all samples, the quantity of total RNA that was retrotranscribed was 1 g per reaction. For comparison, messenger RNA (mRNA) and cDNA concentrations were normalized to yield equivalent -actin products (housekeeping gene). Results with this control amplification can be used to adjust the amount of the cDNA used in the PCR to use equivalent amounts of cDNA in each reaction. Competitive PCR for Cytotoxic Markers To determine the number of molecules expressing perforin, granzyme B, and Fas-L, we used competitive PCR experiments. In this method, a DNA competitor that contained the same primer template sequences as the target cDNA competes for primer binding and amplification. To quantitate the relative amounts of gene transcripts present in various samples, the individual PCRs were performed by coamplifying the cDNA of interest with an internal PCR control. The competitor DNA fragment (higher molecular weight) was derived from the v-erb B gene (Clontech Laboratories) to which the primer templates had been added. The amplification product of each synthetic competitor differs in size from the original cDNA product. Using the competitor fragment as an internal standard in RT-PCR allows amplification of both the wild-type original DNA and the competitor fragment DNA in the same reaction, using gene-specific primers and separating the products on the basis of size. At similar concentrations of the PCR product, comparable band intensities occur in the ethidium bromide gel electrophoresis. The target cDNA was adjusted to equal concentrations by competitive PCR between -actin cDNA and the competitor fragment.17 Evaluation of competitive RT-PCR experiments was performed on digitalized agarose gels by using image analysis software (TDI’s 1D Manager; Tecnologia para Diagnostico e Investigacion, Madrid, Spain). The detection limit of this method was determined using serial dilutions of a known concentration of positive control cDNA, and after PCR optimal amplification, this was equivalent to 200 molecules of the competitor fragment.
87
Immunohistochemical Studies A punch biopsy specimen was also obtained on day 3 for immunohistochemical analysis. This was obtained with a 4-mm punch (Stiefel Laboratorios, Madrid, Spain). Tissue segments were fixed and embedded in paraffin. Microtome sections (4 m) were processed for hematoxylin-eosin and immunohistochemical staining with the following monoclonal antibodies: CD3, CD4, granzyme B (Novocastra Laboratories, Newcastle upon Tyne, England), CD8, CD45RO, and perforin (Dako, Ely, Cambridgeshire, England), using the automated immunostaining system Techmate 500plus (Dako) with the Envision Plus visualization system (Dako). To perform comparative studies in the transductional expression of different cytokines and cytotoxic markers, samples from 4 children with drug-induced TEN but no IVIG or corticoid treatment, previously followed up in our laboratory, formed a control group. Written informed consent for all the diagnostic procedures was obtained from the parents of all the patients involved. RESULTS The symptoms improved markedly within 24 hours of IVIG administration, with no new bullae or further disease progression, the Nikolsky sign became negative, and the fever abated. Re-epithelialization was complete within 10 days. The levels of the T-lymphocyte subsets, activation marker, and skin-homing receptor in blood and blister fluid samples are shown in Table 1. The first blood sample, taken on day 1 before IVIG treatment, showed normal levels of CD4 and CD8, CLA, and memory (CD45RO) and naive (CD45RA) subsets but a high percentage of activated lymphocytes (CD69). Immediately after IVIG therapy, there was an increase in the percentage of cells expressing CLA, CD69, and CD45RO, reaching maximum levels on day 7. All these markers decreased to normal values by the patient’s age at day 15 or 30. The CD4/CD8 ratio (2:1) remained within normal levels during the entire evaluation period. Ranges for normal values were based on data in 2 groups of children, 1
of 10 children with epilepsy taking lamotrigine and 1 of 9 healthy children not taking any medication. In the blister fluid sample, taken on day 1, there was a high percentage of CD8 cells, with a CD4/CD8 ratio of 1:5. A high percentage of memory cells expressed the CLA marker and activated cells compared with the levels in blood obtained at the same time. Twenty-four hours after IVIG administration (day 3), there was a decrease in levels of CLA, CD8, and CD69 in the blister fluid. Because no further blisters appeared, no more measurements could be taken. Hematoxylin-eosin and immunohistochemical staining of the skin biopsy specimen, taken on day 3, showed necrolysis of keratinocytes in the upper dermis with spongiosis and a scanty cellular infiltrate. CD3 cells were present in perivascular tissue and in the dermoepidermal junction. The number of CD8 cells was moderate and that of the CD4 cells low. There were isolated CD45RO⫹ cells and no CD45RA⫹ cells. The presence of the cytotoxic markers granzyme B and perforin was also moderate (Fig 1). The percentage of T lymphocytes that produce intracellular cytokines in peripheral blood samples is shown in Table 1. The initial measurement, taken in samples before IVIG administration, showed a high percentage of cells that produced IFN-␥ and TNF-␣ and a low percentage of cells that expressed IL-4, corresponding to a Th1 pattern. By day 15 these values had become a Th0 pattern, with a low percentage of cells producing IFN-␥ and TNF-␣ and an increased percentage expressing IL-4. The mRNA expression of the cytokines IFN-␥, TNF-␣, and IL-4 was similar to that seen in the intracellular cytokines, ie, a Th1 pattern during the initial stage of the disease until day 7, then a Th0 pattern at day 15. The mRNA expression of perforin, granzyme B, and Fas-L was only observed in the sample obtained before initiating the IVIG treatment. Comparison of the patterns of these values to the means of 4 children with TEN but no IVIG treatment is shown in Figure 2. The cytokine expression became normal in the index patient treated with IVIG in 15 days (Fig 2A), much
Table 1. Flow Cytometry Measurements of Antigen Surface Markers and Intracellular Cytokines in Blood and Blister Fluid during a Case of Toxic Epidermal Necrolysis Treated with a Single High Dose of Intravenous Immunoglobulin Antigen Surface Markers
Intracellular Cytokines
Day 1* 1† 3* 3† 7* 15* 30*
CD4
CD8
CLA
CD45RO
CD45RA
CD69
IFN-␥
TNF-␣
IL-2
IL-4
IL-10
IL-13
57.1‡ 15.5‡ 68.56 14.87 58.18 66.47 62.5
30.7‡ 70.75‡ 28 85.5 39.66 29.84 37.5
4.04‡ 18.8‡ 3.01 15.7 10.5 3.85 1.13
25.3‡ 85.5‡ 26.49 95.6 40.12 39.69 25.36
78.89 11.1 87.15 9.9 76.63 81.35 60.25
18.03 70 19.42 67.31 47.58 8.87 1.78
21.8
41.3
1.8‡
3.2†
7.4‡
0.8‡
18.6
36.87
3.98
3.7
8.4
1.1
22.1 3.8 2.5
1.99 1.62 0.6
0.8 7.94 5.4
2.26 2.8 0.2
0.55 1.3 1.5
9.94 3 2
Abbreviations: IFN-␥, interferon-␥; IL, interleukin; TNF-␣, tumor necrosis factor ␣. * Blood samples. † Blister fluid samples. ‡ Data before intravenous immunoglobulin treatment.
88
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
Figure 1. Skin biopsy specimen taken on day 3 and processed for hematoxylin-eosin (H-E) (A; original magnification ⫻20) and immunohistochemical stains (B-H): lymphocyte subpopulations CD3 (B), CD4 (Novocastra Laboratories, Newcastle upon Tyne, England) (2° C), and CD8 (Dako, Ely, Cambridgeshire, England) (D; original magnification ⫻20), CD45RO and CD45RA (Dako) (E-F; original magnification ⫻40), and the cytotoxic markers perforin (Dako) (original magnification ⫻40) and granzyme B (Novocastra) (original magnification ⫻20) (G-H).
Figure 2. Competitive polymerase chain reaction (PCR) for cytotoxic marker data for our case of toxic epidermal necrolysis (TEN) and for the mean of 4 other similar control patients with TEN. A, Number of molecules of the cytokines in our patient at different times. B, Mean number of cytokines in the 4 controls at different points in time. C, Number of molecules of the cytotoxic markers in our patient at different times. D, Mean number of cytotoxic markers in the 4 controls at different points in time. IFN-␥ indicates interferon-␥; TNF-␣, tumor necrosis factor ␣; IL, interleukin; and IVIG, intravenous immunoglobulin.
sooner than the 30 days required in the group of 4 untreated children (Fig 2B). Comparison of the secretion of perforin and granzyme B and the expression of Fas-L showed a sharp decrease in the patient treated with IVIG (Fig 2C) compared with the other patients (Fig 2D). DISCUSSION Although several agents can induce TEN, most studies indicate that the condition is drug related.1 Increasing evidence suggests that immunologic mechanisms take part during the
VOLUME 91, JULY, 2003
onset and progression of the disease.6 Cell apoptosis causes eruption of mucous membranes, detachment of the epidermis, and severe systemic symptoms, and there may be involvement of the liver or other organs.9 Although the exact mechanism of TEN is unknown, it has been suggested that a drug or drug-related metabolites trigger an immunologic response in the skin, with homing of peripheral T cells to the skin and development of a severe skin response. Cell death and subsequent skin necrosis and detachment of the epidermis are caused by 2 lymphocyte activation pathways: perforin–
89
granzyme B and Fas–Fas-L interaction. In addition, TNF-␣ has been reported to contribute to the pathologic process.9 In the case reported herein, there was a high percentage of activated memory T cells, mainly cytotoxic cells (CD8), expressing the skin-homing receptor CLA, mostly present in the skin blister fluid, although the low initial levels in peripheral blood increased at subsequent determinations. These data are similar to those reported by others in blister fluid.6,18 The clinical course of TEN usually lasts several weeks and often takes months before complete resolution. Skin detachment, with the appearance of blisters, usually occurs early and continues for a week or longer, their extension contributing to the prognosis. The immunologic abnormalities parallel the course of the disease, and increased levels of activated CLA⫹ T cells have been detected 15 to 20 days after initiation of symptoms.6 In our case, these abnormalities tended to reach normal values sooner (by day 15). A number of treatment options have been proposed. Corticoids have been shown to be effective in milder forms of TEN19 but have not always proved beneficial.1 A recent report20 showed that in severe reactions to drugs, such as Stevens-Johnson syndrome or TEN, no differences in the expression of IFN-␥, TNF-␣, perforin, granzyme B, or Fas-L were observed between patients treated with or without corticoids. Thalidomide, a powerful inhibitor of TNF-␣, given 400 mg/d for 5 days, was no more beneficial than placebo and has even been associated with increased mortality.9 In our case, there was a rapid improvement in the skin lesions with disappearance of skin detachment and the Nikolsky sign became negative. Perforin, granzyme B, and Fas-L were only detected in the peripheral blood mononuclear cells before administration of IVIG. This rapid improvement in the skin lesions was not observed in the controls, in whom the expression of cytokines and cytotoxic markers persisted longer, until day 30. Recently, intravenous gamma globulin, 0.4 g/kg per day for 5 consecutive days, resulted in complete healing by day 18, but no detailed analysis was made during the initial days of treatment.10 Another report5 showed that a daily dose of 0.75 g/kg for 5 consecutive days produced resolution of the blisters after 3 days, with improvement of the lesions on day 7 and complete re-epithelialization after 10 days. In our patient, the dose was 2 g/kg for 1 day. This resulted in marked clinical improvement accompanied by a decrease in the immunological markers CLA, CD8, and CD69 in the blister fluid. Even though our patient had received a 2-day course of steroids beforehand, we speculate about the beneficial role of IVIG, since corticoids have not generally proved effective in TEN. As far as we are aware, this is the first study to monitor various immunologic markers in a patient with TEN treated with IVIG. The improvement in both clinical and immunologic markers suggests the beneficial effect of a single high dose of IVIG for the control of TEN. Although this study concerns only one case, the results are important for understanding the immunologic mechanism, but further clinical
90
studies are needed to confirm the beneficial role of an early high dose of IVIG in TEN. ACKNOWLEDGMENT We thank Ian Johnstone for help with the English-language version of the text.
REFERENCES 1. Roujeau JC, Stern RS. Severe cutaneous adverse reactions to drugs. N Engl J Med. 1994;331:1272–1285. 2. Becker DS. Toxic epidermal necrolysis. Lancet. 1998;351: 1417–1420. 3. Guillaume JC, Roujeau JC, Revuz J, Penso D, Touraine R. The culprit drugs in 87 cases of toxic epidermal necrolysis (Lyell’s syndrome). Arch Dermatol. 1987;123:1166 –1170. 4. Miyauchi H, Hosokawa H, Akaeda T, Iba H, Asada Y. T-cell subsets in drug-induced toxic epidermal necrolysis: possible pathogenic mechanism induced by CD8-positive T cells. Arch Dermatol. 1991;127:851– 855. 5. Paquet P, Jacob E, Damas P, Pierad GE. Treatment of druginduced toxic epidermal necrolysis (Lyell’s syndrome) with intravenous human immunoglobulins. Burns. 2001;27: 652– 655. 6. Leyva L, Torres MJ, Posadas S, et al. Anticonvulsant-induced toxic epidermal necrolysis: monitoring the immunologic response. J Allergy Clin Immunol. 2000;105:157–165. 7. Posadas SJ, Padial A, Torres MJ, et al. Delayed reactions to drugs show levels of perforin, granzyme B and Fas-L to be related to disease severity. J Allergy Clin Immunol. 2002;109: 155–161. 8. Viard I, Wehrli P, Bullani R, et al. Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science. 1998;282:490 – 493. 9. Wolkenstein P, Latarjet J, Roujeau JC, et al. Randomised comparison of thalidomide versus placebo in toxic epidermal necrolysis. Lancet. 1998;352:1586 –1589. 10. Magina S, Lisboa C, Gonc¸alves E, Conceicao F, Leal V, Mesquita-Guimaraes J. A case of toxic epidermal necrolysis treated with intravenous immunoglobulin. Br J Dermatol. 2000;142: 191–192. 11. Stella M, Cassano P, Bollero D, Clemente A, Giorio G. Toxic epidermal necrolysis treated with intravenous high-dose immunoglobulins: our experience. Dermatology. 2001;203: 45– 49. 12. Trent JT, Kirsner RS, Romanelli P, Kerdel FA. Analysis of intravenous immunoglobulin for the treatment of toxic epidermal necrolysis using SCORTEN: the University of Miami experience. Arch Dermatol. 2003;139:39 – 43. 13. Prins C, Kerdel FA, Padilla RS, et al. Treatment of toxic epidermal necrolysis with high-dose intravenous immunoglobulins: multicenter retrospective analysis of 48 consecutive cases. Arch Dermatol. 2003;139:26 –32. 14. Tristani-Firouzi P, Petersen MJ, Saffle JR, Morris SE, Zone JJ. Treatment of toxic epidermal necrolysis with intravenous immunoglobulin in children. J Am Acad Dermatol. 2002;47: 548 –552. 15. Bachot N, Revuz J, Roujeau JC. Intravenous immunoglobulin treatment for Stevens-Johnson syndrome and toxic epidermal
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
16. 17. 18.
19.
necrolysis: a prospective noncomparative study showing no benefit on mortality or progression. Arch Dermatol. 2003;139: 33–36. Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156 –159. Siebert PD, Larrick J. Competitive PCR. Nature. 1992;359:557–558. Le Cleach L, Delaire S, Boumsell L, et al. Blister fluid T lymphocytes during toxic epidermal necrolysis are functional cytotoxic cells which express human natural killer (NK) inhibitory receptors. Clin Exp Immunol. 2000;119:225–230. Criton S, Devi K, Sridevi PK, Asokan PU. Toxic epidermal
necrolysis: a retrospective study. Int J Dermatol. 1997;36: 923–925. 20. Posadas SJ, Leyva L, Torres MJ, et al. Subjects with allergic reactions to drugs show in vivo polarized patterns of cytokine expression depending on the chronology of the clinical reaction. J Allergy Clin Immunol. 2000;106:769 –776. Requests for reprints should be addressed to: Miguel Blanca, MD, PhD Laboratorio de Investigacio´n, Plaza Hospital Civil Pabello´n 5 So´tano, 29009 Malaga, Spain E-mail: [email protected]
Answers to CME Questions—Annals of Allergy, Asthma, and Immunology, July, 2003 Fiocchi A, Martelli A, De Chiara A, Moro G, Warm A, Terracciano L: Primary dietary prevention of food allergy. Ann Allergy Asthma Immunol. 2003;91:3-13. 1. b 2. e 3. c 4. d 5. d
VOLUME 91, JULY, 2003
91
Improvement of toxic epidermal necrolysis after the early administration of a single high dose of intravenous immunoglobulin Cristobalina Mayorga, PhD*; Maria Jose´ Torres, MD, PhD*; Jose´ Luis Corzo, MD†; Elena Sanchez-Sabate, PhD‡; Javier Alvarez, MD‡; Angel Vera, MD, PhD§; Sinforiano Posadas, PhD‡; Antonio Jurado, MD, PhD†; and Miguel Blanca, MD, PhD‡
Background: Toxic epidermal necrolysis (TEN) is a severe disease often induced by drugs. Treatment is controversial, although intravenous immunoglobulins (IVIGs) have been effective. Objective: To report the case of a child with TEN after lamotrigine treatment, who improved 24 hours after IVIG administration. Methods: Sequential blood and blister fluid samples were obtained for flow cytometry and reverse transcriptase–polymerase chain reaction analyses. Results: The first blood sample, taken before IVIG administration, showed normal levels of lymphocyte subsets and CLA (4.0%) but high levels of activated lymphocytes (CD69) (18.0%). After treatment, the CLA⫹, CD69⫹, and memory cells increased until day 7, decreasing to normal values at days 15 and 30. In the blister fluid samples, taken on day 1, there were high levels of CD8⫹ (70.2%; CD4/CD8 ratio, 1:5), CLA⫹ (18.8%), and CD69⫹ (70%) cells, decreasing 24 hours after IVIG administration. In the blood samples, there was a Th1 cytokine pattern initially, tending to Th0 with time. Perforin, granzyme B, and Fas ligand were only observed before IVIG administration. Conclusions: A single high dose of IVIG interrupted the progression of skin disease and reduced the expression of the apoptotic markers. The immunologic changes, first seen in blister fluid and remaining several days in peripheral blood, indicate that T cells were first recruited to the skin and then recirculated to blood. Ann Allergy Asthma Immunol. 2003;91:86–91.
INTRODUCTION Toxic epidermal necrolysis (TEN) is a severe disease that is usually induced by drugs or viral infections.1 Although the prevalence is low, the mortality rate is 25 to 30%.2 The most frequent drugs involved are sulfonamides, anticonvulsants, and nonsteroidal anti-inflammatory drugs,1,3 with immunologic mechanisms thought to be involved in its pathophysiology.4 Immunohistochemical studies have shown activated CD4 cells in the epidermis and CD8 cells in the dermis,5 with migration to the skin of memory-activated peripheral * Research Unit for Allergic Diseases, Carlos Haya Hospital, Malaga, Spain. † Paediatric Department, Carlos Haya Hospital, Malaga, Spain. ‡ Allergy Service, Hospital La Paz, Madrid, Spain. § Dermatology Service, Carlos Haya Hospital, Malaga, Spain. This study was partly supported by grants from the Spanish Ministry of Health (01/0014, 01/3031, and PIO20640). The specific contribution of each author is as follows: C. Mayorga and S. Posadas performed the laboratory analysis; M. J. Torres, J. L. Corzo, A. Vera, and A. Jurado made the clinical evaluation; J. Alvarez and E. SanchezSabate performed the immunohistochemical analysis; and M. Blanca coordinated all the study and with M. J. Torres and C. Mayorga wrote the article. Received for publication January 30, 2003. Accepted for publication in revised form February 27, 2003.
86
T lymphocytes expressing the skin-homing receptor (CLA⫹CD69⫹CD45RO⫹).6 T cells in TEN secrete perforin and granzyme B and express Fas ligand (Fas-L),7 components of the lymphocyte effector pathway involved in cell death and apoptosis,1 and a relationship between these mediators and the severity of skin diseases induced by drugs has been reported.7,8 Several different treatments have been tried but so far without success.1 Steroids, the most widely used therapy, are controversial and in severe cases may even worsen the disease by increasing the risk of septicemia.2,9 Other drugs, such as cyclosporine, cyclophosphamide, pentoxifylline, and thalidomide, have also been used, but their efficacy has not been confirmed.9 Recovery after the administration of intravenous immunoglobulin (IVIG) has recently been reported,5,10 –14 although a recent study15 performed in 34 patients with StevenJohnson syndrome and TEN does not support the routine use of IVIG treatment in these patients, especially in cases of impaired renal function. The mechanism proposed for the action of IVIG in TEN, inferred from in vitro studies, is the inhibition of Fas-mediated keratinocyte death by naturally occurring Fas-blocking antibodies in the IVIG
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
preparation.8 To our knowledge, the behavior of the immunologic system in TEN has not yet been studied in vivo. We report the case of a patient who experienced rapid recovery from a severe case of drug-induced TEN after a single high dose of IVIG, together with the immunologic parameters, which were monitored before and after treatment. MATERIAL AND METHODS Case Report A 3-year-old boy with cerebral palsy and epilepsy was given lamotrigine (Lamictal; Glaxo-Welcome, Madrid, Spain) (12.5 mg in the morning and 6.25 mg in the evening) to control seizures. Fifteen days after initiating therapy, a maculopapular exanthema appeared on the trunk and legs, progressing rapidly to bullous lesions accompanied by extensive skin scalding and a high temperature (39° C). The mucous membranes were also involved, with erosions on the conjunctival, oropharyngeal, and urogenital mucosae. A diagnosis of TEN, affecting 60% of the body surface area, was made and use of lamotrigine was discontinued. Treatment with intravenous dexamethasone (6 mg/d) was initiated, but because no improvement was observed after 48 hours, the steroids were withdrawn and a single high dose (2 g/kg) of IVIG (Flebogamma; Grifols Institute, Barcelona, Spain) was administered at a rate of 15 mL/h for more than 24 hours. To monitor the participation of the immunologic system during the response to treatment, lymphocyte subpopulations and activation markers were analyzed in parallel in peripheral blood mononuclear cells and blister fluid cells. Blood samples were obtained immediately before IVIG treatment and 3, 7, 15, and 30 days after. Blister fluid samples were obtained before IVIG administration and 3 days after. No further samples were obtained from blister fluid because no further blisters appeared. Flow Cytometry Analysis Flow cytometry was used to study different immunological markers: T-lymphocyte subsets (CD3, CD4, CD8, CD45RO, and CD45RA), the CLA skin-homing receptor, and the cellular activation marker CD69. The production and expression of several cytokines, interleukin (IL)-2, interferon-␥ (IFN-␥), tumor necrosis factor ␣ (TNF-␣), IL-4, and IL-13, were also determined by flow cytometry. This was performed in peripheral blood mononuclear cells, which were isolated on Ficoll-Paque gradients (Pharmacia Biotech Inc, Piscataway, NJ). Six parameters were analyzed on a Facscalibur flow cytometer using Cell Quest software (Becton Dickinson, San Jose, CA). Negative isotype controls were used to verify the staining specificity of the antibodies used. The production and expression of several cytokines were determined in cells stimulated for 4 hours at 37° C with phorbol 12-myristate 13-acetate (Sigma, St. Louis, MO) and ionomycin (Sigma), and the intracellular protein transport of cytokines was disrupted with brefeldin (Sigma).
VOLUME 91, JULY, 2003
Cells were fixed and permeabilized (Caltag Laboratories, Burlingame, CA) and stained intracellularly with monoclonal antibodies to cytokines (IL-2 and IL-10 were stained with fluorescein isothiocyanate; IFN-␥, TNF-␣, IL-4, and IL-13 were stained with phycoerythrin; all provided by Becton Dickinson). Determination of Cytotoxic Markers The secretion of perforin and granzyme B and the expression of Fas-L were determined by competitive polymerase chain reaction (PCR). RNA Isolation and Complementary DNA Synthesis Total RNA was isolated from cells by the method of Chomczynski and Sacchi.16 The complementary DNA (cDNA) template for reverse transcriptase (RT)–PCR was synthesized from RNA by reverse transcription using a first-strand cDNA synthesis kit (Clontech Laboratories, Palo Alto, CA). In all samples, the quantity of total RNA that was retrotranscribed was 1 g per reaction. For comparison, messenger RNA (mRNA) and cDNA concentrations were normalized to yield equivalent -actin products (housekeeping gene). Results with this control amplification can be used to adjust the amount of the cDNA used in the PCR to use equivalent amounts of cDNA in each reaction. Competitive PCR for Cytotoxic Markers To determine the number of molecules expressing perforin, granzyme B, and Fas-L, we used competitive PCR experiments. In this method, a DNA competitor that contained the same primer template sequences as the target cDNA competes for primer binding and amplification. To quantitate the relative amounts of gene transcripts present in various samples, the individual PCRs were performed by coamplifying the cDNA of interest with an internal PCR control. The competitor DNA fragment (higher molecular weight) was derived from the v-erb B gene (Clontech Laboratories) to which the primer templates had been added. The amplification product of each synthetic competitor differs in size from the original cDNA product. Using the competitor fragment as an internal standard in RT-PCR allows amplification of both the wild-type original DNA and the competitor fragment DNA in the same reaction, using gene-specific primers and separating the products on the basis of size. At similar concentrations of the PCR product, comparable band intensities occur in the ethidium bromide gel electrophoresis. The target cDNA was adjusted to equal concentrations by competitive PCR between -actin cDNA and the competitor fragment.17 Evaluation of competitive RT-PCR experiments was performed on digitalized agarose gels by using image analysis software (TDI’s 1D Manager; Tecnologia para Diagnostico e Investigacion, Madrid, Spain). The detection limit of this method was determined using serial dilutions of a known concentration of positive control cDNA, and after PCR optimal amplification, this was equivalent to 200 molecules of the competitor fragment.
87
Immunohistochemical Studies A punch biopsy specimen was also obtained on day 3 for immunohistochemical analysis. This was obtained with a 4-mm punch (Stiefel Laboratorios, Madrid, Spain). Tissue segments were fixed and embedded in paraffin. Microtome sections (4 m) were processed for hematoxylin-eosin and immunohistochemical staining with the following monoclonal antibodies: CD3, CD4, granzyme B (Novocastra Laboratories, Newcastle upon Tyne, England), CD8, CD45RO, and perforin (Dako, Ely, Cambridgeshire, England), using the automated immunostaining system Techmate 500plus (Dako) with the Envision Plus visualization system (Dako). To perform comparative studies in the transductional expression of different cytokines and cytotoxic markers, samples from 4 children with drug-induced TEN but no IVIG or corticoid treatment, previously followed up in our laboratory, formed a control group. Written informed consent for all the diagnostic procedures was obtained from the parents of all the patients involved. RESULTS The symptoms improved markedly within 24 hours of IVIG administration, with no new bullae or further disease progression, the Nikolsky sign became negative, and the fever abated. Re-epithelialization was complete within 10 days. The levels of the T-lymphocyte subsets, activation marker, and skin-homing receptor in blood and blister fluid samples are shown in Table 1. The first blood sample, taken on day 1 before IVIG treatment, showed normal levels of CD4 and CD8, CLA, and memory (CD45RO) and naive (CD45RA) subsets but a high percentage of activated lymphocytes (CD69). Immediately after IVIG therapy, there was an increase in the percentage of cells expressing CLA, CD69, and CD45RO, reaching maximum levels on day 7. All these markers decreased to normal values by the patient’s age at day 15 or 30. The CD4/CD8 ratio (2:1) remained within normal levels during the entire evaluation period. Ranges for normal values were based on data in 2 groups of children, 1
of 10 children with epilepsy taking lamotrigine and 1 of 9 healthy children not taking any medication. In the blister fluid sample, taken on day 1, there was a high percentage of CD8 cells, with a CD4/CD8 ratio of 1:5. A high percentage of memory cells expressed the CLA marker and activated cells compared with the levels in blood obtained at the same time. Twenty-four hours after IVIG administration (day 3), there was a decrease in levels of CLA, CD8, and CD69 in the blister fluid. Because no further blisters appeared, no more measurements could be taken. Hematoxylin-eosin and immunohistochemical staining of the skin biopsy specimen, taken on day 3, showed necrolysis of keratinocytes in the upper dermis with spongiosis and a scanty cellular infiltrate. CD3 cells were present in perivascular tissue and in the dermoepidermal junction. The number of CD8 cells was moderate and that of the CD4 cells low. There were isolated CD45RO⫹ cells and no CD45RA⫹ cells. The presence of the cytotoxic markers granzyme B and perforin was also moderate (Fig 1). The percentage of T lymphocytes that produce intracellular cytokines in peripheral blood samples is shown in Table 1. The initial measurement, taken in samples before IVIG administration, showed a high percentage of cells that produced IFN-␥ and TNF-␣ and a low percentage of cells that expressed IL-4, corresponding to a Th1 pattern. By day 15 these values had become a Th0 pattern, with a low percentage of cells producing IFN-␥ and TNF-␣ and an increased percentage expressing IL-4. The mRNA expression of the cytokines IFN-␥, TNF-␣, and IL-4 was similar to that seen in the intracellular cytokines, ie, a Th1 pattern during the initial stage of the disease until day 7, then a Th0 pattern at day 15. The mRNA expression of perforin, granzyme B, and Fas-L was only observed in the sample obtained before initiating the IVIG treatment. Comparison of the patterns of these values to the means of 4 children with TEN but no IVIG treatment is shown in Figure 2. The cytokine expression became normal in the index patient treated with IVIG in 15 days (Fig 2A), much
Table 1. Flow Cytometry Measurements of Antigen Surface Markers and Intracellular Cytokines in Blood and Blister Fluid during a Case of Toxic Epidermal Necrolysis Treated with a Single High Dose of Intravenous Immunoglobulin Antigen Surface Markers
Intracellular Cytokines
Day 1* 1† 3* 3† 7* 15* 30*
CD4
CD8
CLA
CD45RO
CD45RA
CD69
IFN-␥
TNF-␣
IL-2
IL-4
IL-10
IL-13
57.1‡ 15.5‡ 68.56 14.87 58.18 66.47 62.5
30.7‡ 70.75‡ 28 85.5 39.66 29.84 37.5
4.04‡ 18.8‡ 3.01 15.7 10.5 3.85 1.13
25.3‡ 85.5‡ 26.49 95.6 40.12 39.69 25.36
78.89 11.1 87.15 9.9 76.63 81.35 60.25
18.03 70 19.42 67.31 47.58 8.87 1.78
21.8
41.3
1.8‡
3.2†
7.4‡
0.8‡
18.6
36.87
3.98
3.7
8.4
1.1
22.1 3.8 2.5
1.99 1.62 0.6
0.8 7.94 5.4
2.26 2.8 0.2
0.55 1.3 1.5
9.94 3 2
Abbreviations: IFN-␥, interferon-␥; IL, interleukin; TNF-␣, tumor necrosis factor ␣. * Blood samples. † Blister fluid samples. ‡ Data before intravenous immunoglobulin treatment.
88
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
Figure 1. Skin biopsy specimen taken on day 3 and processed for hematoxylin-eosin (H-E) (A; original magnification ⫻20) and immunohistochemical stains (B-H): lymphocyte subpopulations CD3 (B), CD4 (Novocastra Laboratories, Newcastle upon Tyne, England) (2° C), and CD8 (Dako, Ely, Cambridgeshire, England) (D; original magnification ⫻20), CD45RO and CD45RA (Dako) (E-F; original magnification ⫻40), and the cytotoxic markers perforin (Dako) (original magnification ⫻40) and granzyme B (Novocastra) (original magnification ⫻20) (G-H).
Figure 2. Competitive polymerase chain reaction (PCR) for cytotoxic marker data for our case of toxic epidermal necrolysis (TEN) and for the mean of 4 other similar control patients with TEN. A, Number of molecules of the cytokines in our patient at different times. B, Mean number of cytokines in the 4 controls at different points in time. C, Number of molecules of the cytotoxic markers in our patient at different times. D, Mean number of cytotoxic markers in the 4 controls at different points in time. IFN-␥ indicates interferon-␥; TNF-␣, tumor necrosis factor ␣; IL, interleukin; and IVIG, intravenous immunoglobulin.
sooner than the 30 days required in the group of 4 untreated children (Fig 2B). Comparison of the secretion of perforin and granzyme B and the expression of Fas-L showed a sharp decrease in the patient treated with IVIG (Fig 2C) compared with the other patients (Fig 2D). DISCUSSION Although several agents can induce TEN, most studies indicate that the condition is drug related.1 Increasing evidence suggests that immunologic mechanisms take part during the
VOLUME 91, JULY, 2003
onset and progression of the disease.6 Cell apoptosis causes eruption of mucous membranes, detachment of the epidermis, and severe systemic symptoms, and there may be involvement of the liver or other organs.9 Although the exact mechanism of TEN is unknown, it has been suggested that a drug or drug-related metabolites trigger an immunologic response in the skin, with homing of peripheral T cells to the skin and development of a severe skin response. Cell death and subsequent skin necrosis and detachment of the epidermis are caused by 2 lymphocyte activation pathways: perforin–
89
granzyme B and Fas–Fas-L interaction. In addition, TNF-␣ has been reported to contribute to the pathologic process.9 In the case reported herein, there was a high percentage of activated memory T cells, mainly cytotoxic cells (CD8), expressing the skin-homing receptor CLA, mostly present in the skin blister fluid, although the low initial levels in peripheral blood increased at subsequent determinations. These data are similar to those reported by others in blister fluid.6,18 The clinical course of TEN usually lasts several weeks and often takes months before complete resolution. Skin detachment, with the appearance of blisters, usually occurs early and continues for a week or longer, their extension contributing to the prognosis. The immunologic abnormalities parallel the course of the disease, and increased levels of activated CLA⫹ T cells have been detected 15 to 20 days after initiation of symptoms.6 In our case, these abnormalities tended to reach normal values sooner (by day 15). A number of treatment options have been proposed. Corticoids have been shown to be effective in milder forms of TEN19 but have not always proved beneficial.1 A recent report20 showed that in severe reactions to drugs, such as Stevens-Johnson syndrome or TEN, no differences in the expression of IFN-␥, TNF-␣, perforin, granzyme B, or Fas-L were observed between patients treated with or without corticoids. Thalidomide, a powerful inhibitor of TNF-␣, given 400 mg/d for 5 days, was no more beneficial than placebo and has even been associated with increased mortality.9 In our case, there was a rapid improvement in the skin lesions with disappearance of skin detachment and the Nikolsky sign became negative. Perforin, granzyme B, and Fas-L were only detected in the peripheral blood mononuclear cells before administration of IVIG. This rapid improvement in the skin lesions was not observed in the controls, in whom the expression of cytokines and cytotoxic markers persisted longer, until day 30. Recently, intravenous gamma globulin, 0.4 g/kg per day for 5 consecutive days, resulted in complete healing by day 18, but no detailed analysis was made during the initial days of treatment.10 Another report5 showed that a daily dose of 0.75 g/kg for 5 consecutive days produced resolution of the blisters after 3 days, with improvement of the lesions on day 7 and complete re-epithelialization after 10 days. In our patient, the dose was 2 g/kg for 1 day. This resulted in marked clinical improvement accompanied by a decrease in the immunological markers CLA, CD8, and CD69 in the blister fluid. Even though our patient had received a 2-day course of steroids beforehand, we speculate about the beneficial role of IVIG, since corticoids have not generally proved effective in TEN. As far as we are aware, this is the first study to monitor various immunologic markers in a patient with TEN treated with IVIG. The improvement in both clinical and immunologic markers suggests the beneficial effect of a single high dose of IVIG for the control of TEN. Although this study concerns only one case, the results are important for understanding the immunologic mechanism, but further clinical
90
studies are needed to confirm the beneficial role of an early high dose of IVIG in TEN. ACKNOWLEDGMENT We thank Ian Johnstone for help with the English-language version of the text.
REFERENCES 1. Roujeau JC, Stern RS. Severe cutaneous adverse reactions to drugs. N Engl J Med. 1994;331:1272–1285. 2. Becker DS. Toxic epidermal necrolysis. Lancet. 1998;351: 1417–1420. 3. Guillaume JC, Roujeau JC, Revuz J, Penso D, Touraine R. The culprit drugs in 87 cases of toxic epidermal necrolysis (Lyell’s syndrome). Arch Dermatol. 1987;123:1166 –1170. 4. Miyauchi H, Hosokawa H, Akaeda T, Iba H, Asada Y. T-cell subsets in drug-induced toxic epidermal necrolysis: possible pathogenic mechanism induced by CD8-positive T cells. Arch Dermatol. 1991;127:851– 855. 5. Paquet P, Jacob E, Damas P, Pierad GE. Treatment of druginduced toxic epidermal necrolysis (Lyell’s syndrome) with intravenous human immunoglobulins. Burns. 2001;27: 652– 655. 6. Leyva L, Torres MJ, Posadas S, et al. Anticonvulsant-induced toxic epidermal necrolysis: monitoring the immunologic response. J Allergy Clin Immunol. 2000;105:157–165. 7. Posadas SJ, Padial A, Torres MJ, et al. Delayed reactions to drugs show levels of perforin, granzyme B and Fas-L to be related to disease severity. J Allergy Clin Immunol. 2002;109: 155–161. 8. Viard I, Wehrli P, Bullani R, et al. Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science. 1998;282:490 – 493. 9. Wolkenstein P, Latarjet J, Roujeau JC, et al. Randomised comparison of thalidomide versus placebo in toxic epidermal necrolysis. Lancet. 1998;352:1586 –1589. 10. Magina S, Lisboa C, Gonc¸alves E, Conceicao F, Leal V, Mesquita-Guimaraes J. A case of toxic epidermal necrolysis treated with intravenous immunoglobulin. Br J Dermatol. 2000;142: 191–192. 11. Stella M, Cassano P, Bollero D, Clemente A, Giorio G. Toxic epidermal necrolysis treated with intravenous high-dose immunoglobulins: our experience. Dermatology. 2001;203: 45– 49. 12. Trent JT, Kirsner RS, Romanelli P, Kerdel FA. Analysis of intravenous immunoglobulin for the treatment of toxic epidermal necrolysis using SCORTEN: the University of Miami experience. Arch Dermatol. 2003;139:39 – 43. 13. Prins C, Kerdel FA, Padilla RS, et al. Treatment of toxic epidermal necrolysis with high-dose intravenous immunoglobulins: multicenter retrospective analysis of 48 consecutive cases. Arch Dermatol. 2003;139:26 –32. 14. Tristani-Firouzi P, Petersen MJ, Saffle JR, Morris SE, Zone JJ. Treatment of toxic epidermal necrolysis with intravenous immunoglobulin in children. J Am Acad Dermatol. 2002;47: 548 –552. 15. Bachot N, Revuz J, Roujeau JC. Intravenous immunoglobulin treatment for Stevens-Johnson syndrome and toxic epidermal
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
16. 17. 18.
19.
necrolysis: a prospective noncomparative study showing no benefit on mortality or progression. Arch Dermatol. 2003;139: 33–36. Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156 –159. Siebert PD, Larrick J. Competitive PCR. Nature. 1992;359:557–558. Le Cleach L, Delaire S, Boumsell L, et al. Blister fluid T lymphocytes during toxic epidermal necrolysis are functional cytotoxic cells which express human natural killer (NK) inhibitory receptors. Clin Exp Immunol. 2000;119:225–230. Criton S, Devi K, Sridevi PK, Asokan PU. Toxic epidermal
necrolysis: a retrospective study. Int J Dermatol. 1997;36: 923–925. 20. Posadas SJ, Leyva L, Torres MJ, et al. Subjects with allergic reactions to drugs show in vivo polarized patterns of cytokine expression depending on the chronology of the clinical reaction. J Allergy Clin Immunol. 2000;106:769 –776. Requests for reprints should be addressed to: Miguel Blanca, MD, PhD Laboratorio de Investigacio´n, Plaza Hospital Civil Pabello´n 5 So´tano, 29009 Malaga, Spain E-mail: [email protected]
Answers to CME Questions—Annals of Allergy, Asthma, and Immunology, July, 2003 Fiocchi A, Martelli A, De Chiara A, Moro G, Warm A, Terracciano L: Primary dietary prevention of food allergy. Ann Allergy Asthma Immunol. 2003;91:3-13. 1. b 2. e 3. c 4. d 5. d
VOLUME 91, JULY, 2003
91