Use of intravenous immunoglobulin in toxic epidermal necrolysis and Stevens–Johnson syndrome: Our current understanding

International Immunopharmacology 6 (2006) 543 – 549 www.elsevier.com/locate/intimp

Use of intravenous immunoglobulin in toxic epidermal necrolysis and Stevens–Johnson syndrome: Our current understanding Lars E. French a,*, Jennifer T. Trent b, Francisco A. Kerdel b a

Department of Dermatology, Geneva University Medical School, 24, rue Micheli-du-Crest, CH-1211 Geneva 14, Switzerland b Department of Dermatology, University of Miami, Miami, FL, USA

Abstract Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN, Lyell’s disease, syndrome) are considered to be part of a spectrum of adverse cutaneous drug reactions with increasing severity and extent of skin detachment, ranging from SJS (less than 10% body surface area skin detachment, 1–5% mortality) to TEN (greater than 30% skin detachment, 25–35% mortality). Both SJS and TEN are characterized morphologically by ongoing apoptotic keratinocyte cell death that results in the separation of the epidermis from the dermis. Recent evidence is supportive of a role for the death receptor Fas and its ligand FasL, in the pathogenesis of keratinocyte apoptosis during TEN. This Fas-mediated keratinocyte apoptosis that causes epidermal detachment in TEN can be inhibited in vitro by antagonistic monoclonal antibodies to Fas and by intravenous immunoglobulins (IVIG) which have been shown to contain natural anti-Fas antibodies. Over the last 6 years, numerous case reports and 8 non-controlled clinical studies containing 9 or more patients have analyzed the therapeutic effect of IVIG in TEN. Taken together, although each study has its potential biases, 6 of the 8 studies point towards a benefit of IVIG used at doses greater than 2 g/kg on the mortality associated with TEN. Hopefully, these studies will serve as the basis for designing a prospective controlled trial in the near future; as such, an approach appears the only way to definitively determine the therapeutic potential of IVIG in TEN. D 2005 Elsevier B.V. All rights reserved. Keywords: Intravenous immunoglobulin; Adverse drug reaction; Apoptosis; Fas

1. Summary Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe adverse drug reactions characterized by a low incidence but high mortality. The incidence of SJS is approximately 6 cases per million persons per year, and that of TEN is approximately 2 cases per million persons per year [1]. However, SJS and TEN are idiosyncratic in nature and thereby have the potential to affect any individual * Corresponding author. Tel.: +41 22 3729455; fax: +41 22 3729470. E-mail address: [email protected] (L.E. French). 1567-5769/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.intimp.2005.11.012

taking a medication; the most frequently incriminated drugs being antibiotics, non-steroidal anti-inflammatory drugs, and anticonvulsants. SJS and TEN are currently considered to be part of a spectrum of clinically and pathogenically related druginduced skin diseases with increasing severity. In contrast to the common morbilliform drug rash that is not associated with epidermal detachment, both SJS and TEN are characterized by epidermal detachment ranging from mild, 1–10% of total body surface area (TBSA) in SJS, to moderate, 10–30% TBSA, in what Bastuji-Garin et al. refer to as SJS-TEN overlap [2], and severe, N 30% TBSA in full-blown TEN (Fig. 1). Within this spectrum, the severity or extent of epidermal

544

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

detachment is tightly correlated with the observed rate of mortality, the latter ranging from 1–5% in SJS to 25– 35% in TEN according to the epidemiological studies published in the literature over the past decade. With the objective of being able to precisely predict patient mortality, Bastuji-Garin et al. have recently proposed a scoring system for TEN named the SCORTEN severity of illness score (Table 1) [3]. Seven clinico-biologic parameters, including known adverse prognostic factors such as age greater than 40 years and initial surface of epidermal detachment greater than 10%, are given one point if positive and zero if negative. Computing the sum of the scores for each clinico-biologic parameter results in a bSCORTENQ ranging from 0 to 7, with a score of 0 or 1 predicting a mortality of 3.2%, and a score of 5 or above predicting a mortality of greater than 90% (Table 1). This scoring system, which was developed with a French-based patient cohort, has recently been validated in a U.S.-based patient cohort [4] and is proving to be a valuable tool for predicting patient outcome. Both SJS and TEN are characterized morphologically by ongoing apoptotic keratinocyte cell death that results in the separation of the epidermis from the dermis [5]. Cell death by apoptosis is a tightly regulated physiological process that enables the elimination of unwanted cells without causing an inflammatory response and its consequences. Dysregulation of this type of cell death

Table 1 The SCORTEN scoring system for predicting outcome in TEN Clinical–biologic parameter

Individual score

Scorten (sum of individual scores)

Age N 40 years

Yes = 1, No = 0 Yes = 1, No = 0 Yes = 1, No = 0 Yes = 1, No = 0

0–1

3.2

2

12.1

3

35.3

4 z5

58.3

Malignancy Tachycardia (N120/min) Initial surface of epidermal detachment N 10% Serum urea N 10 mmol/L Serum glucose N 14 mmol/L Bicarbonate b 20 mmol/L

Yes = 1, No = 0 Yes = 1, No = 0 Yes = 1, No = 0

Predicted mortality (%)

90

can, however, cause tissue destructive diseases such as TEN. It is known that certain cells are equipped with cellsurface sensors, named death receptors, that are able to detect the presence of specific extracellular death signals and rapidly trigger cellular destruction by apoptosis. The best studied death receptor signaling pathway to date is that triggered by binding of Fas ligand (FasL) to Fas (Fig. 2). Schematically, multimerization or clustering of Fas

Fig. 1. Clinical features of SJS, SJS-TEN, and TEN. (A) SJS, (B) SJS-TEN overlap, (C) TEN with spots and (D) TEN.

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

545

Fig. 2. The Fas signaling pathway and its molecular control. Fas signaling is triggered upon contact with membrane bound FasL. Subsequent recruitment of FADD and pro-caspase-8/10 leads to upstream caspase (caspase-8/10) autoactivation which initiates apoptosis by subsequent cleavage of downstream effector caspases (caspases-3, -7). An alternative, caspase-independent pathway is initiated upon recruitment of the DDcontaining kinase RIP1.

upon binding of the membrane-bound form of FasL, recruits the bipartite adaptor molecule Fas associated death domain (FADD), composed of an amino terminal death effector domain (DED) and a carboxyl terminal DD. FADD binds to Fas via homophilic DD–DD interactions and recruits the DED-containing caspase-8 (and/ or caspase-10) to the receptor via homophilic DED– DED interactions. Caspase-8 (and/or -10), within this newly formed death-inducing signaling complex (DISC), then proteolytically autoactivates itself and initiates apoptosis by subsequent cleavage of downstream effector caspases (caspases-3, -6, -7). Depending on the cell type and the intensity of the death signal, recruitment of adaptor molecules can be detected within minutes, and apoptosis completed within a few hours. For Fas, there is an additional pathway that can signal apoptosis. It involves recruitment of RIP, a protein that binds the cytosolic domain of Fas and links this receptor to an apoptosis pathway involving activation of Jun N-terminal kinase (JNK) [6].

The mechanisms responsible for enhanced keratinocyte cell death in TEN have long remained unclear; however, we have recently shown that this process is associated with highly increased keratinocyte FasL expression, together with conserved levels of keratinocyte Fas expression in vivo [7]. FasL is also cleaved from keratinocytes and high levels soluble FasL can be detected in the serum and blister fluid. This form of FasL is unlikely to contribute significantly to keratinocyte apoptosis, however, as soluble FasL is 1000-fold less cytotoxic than membrane bound FasL. The molecular mechanism by which culprit drugs upregulate keratinocyte FasL expression in TEN is currently unknown; however, functional experiments performed by overlaying cryostat sections of lesional skin from TEN patients with Fas-sensitive cells as targets have clearly demonstrated that keratinocyte FasL is cytolytically active in TEN. This cytolytic activity could be blocked with monoclonal antibodies that interfere with the interaction of Fas and FasL, thus supporting the

546

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

hypothesis that increased keratinocyte FasL expression is responsible for the keratinocyte apoptosis that characterizes TEN. It has also provided a rationale for molecular strategies to treat TEN (Fig. 3). No specific treatment for SJS or TEN exists to date, and as previously stated, mortality is high. Attempts have been made to decrease mortality in TEN patients through improved supportive care and multiple adjuvant therapies. In 1998, we reported that commercial preparations of intravenous immunoglobulin (IVIG) contain antibodies against Fas that are able to block the binding of FasL

to Fas [7]. Furthermore, intravenous immunoglobulins, by blocking Fas, potently inhibit cell death mediated by recombinant FasL in vitro. In the same report, based on the results of a pilot trial, we suggested that IVIG used in high doses (0.75 g/kg/day for 4 consecutive days), to treat patients with TEN, consistently and rapidly blocked the progression of skin detachment and disease in 10 out of 10 patients [7]. Since then, 7 other studies with 9 or more patients treated with IVIG for TEN have been published [8–14]. While none of these studies are controlled, and the results illustrate that controversy as to the

Fig. 3. Fas-mediated keratinocyte apoptosis in TEN and potential mechanism of inhibition by IVIG. (A) Normal epidermis and (B) toxic epidermal necrolysis: induction of keratinocyte FasL expression and interaction with Fas at the cell surface, leading to keratinocyte apoptosis (C) epidermis during TEN treated by IVIG: predicted inhibition of keratinocyte apoptosis by blockade of Fas by anti-Fas Ab in IVIG.

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

efficacy of IVIG still exists, interesting information can be extracted from their analysis (Table 2). In 1999, Stella et al. reported their experience with the use of IVIG in 9 patients suffering from SJS and TEN with an average epidermal detachment of 16% TBSA (range: 4–37%). IVIG treatment was started on average 6.1 days after disease onset. Eight patients healed and 1 patient died of septic shock and multiple organ failure, thus resulting in a reported mortality rate of 11%. Interruption of further epidermal detachment was observed on average 4.8 days after the onset of IVIG treatment. Three of the 8 surviving patients required transient mechanical ventilation due to pulmonary failure, and 1 other patient required dialysis due to acute renal failure. The largest of the studies summarized in Table 2 was a multicenter study of 48 patients with TEN treated with IVIG at a mean total dose of 2.7 g/kg (doses ranging from 0.65–5.8 g/kg divided over 1–5 days), started on average 7.3 days after the onset of disease [9]. Success was concluded based on an 88% survival rate, but also the rapid (mean of 2.3 days) cessation of skin and mucosal detachment in 89.6% of patients. Mortality was associated with a lower dose of IVIG, longer time of onset to IVIG use, coexisting underlying chronic conditions, older age, and greater body surface area involved. Two other studies published simultaneously by Trent et al. and Bachot et al. used SCORTEN to compare the effect of IVIG on expected mortality and reported contradictory results [10,11]. Trent et al. analyzed in a retrospective non-controlled manner 16 patients with a mean TBSA of epidermal detachment of 43% and Bachot et al. studied in a prospective non-controlled manner 34 patients with a mean TBSA of epidermal detachment of 19%. Whereas, Trent et al. found an 84.4% reduction in predicted mortality in patients treated with IVIG, the Bachot study reported an increase in

547

mortality (24% predicted vs. 32% observations). However, it is of note that the latter study included several SJS patients at admission, 2 of which were considered as TEN by the third day, thus possibly leading to a lower SCORTEN predicted mortality. Also, 32 of 34 patients in the Bachot study were treated with IVIG originating from the same batch, and we have reported that large batch-to-batch variability in anti-Fas activity can be observed in IVIG, with rare batches even completely lacking activity [9]. Finally, differences in doses of IVIG used in the 2 studies (2 g/kg in the Bachot study vs. 4 g/kg in the Trent study) and/or time from onset to the use of IVIG (4.1 in the Bachot study vs. 3.4 days in the Trent study) may have contributed to the different outcomes. The study by Campione et al. describes 10 patients suffering from TEN with a mean TBSA of epidermal detachment of 44% that were treated with 400 mg/kg/ day on 5 consecutive days (2g/kg total dose), starting on average 3 days after disease onset [12]. According to the calculated SCORTEN for this patient cohort at admission, the predicted mortality rate was 35%, and the observed mortality, after IVIG therapy, was 10%. The authors also report that in 9 of their patients, clinical improvement could already be observed after the first infusion of IVIG. Shortt et al. report the results of a retrospective noncontrolled analysis of 32 patients with TEN evaluated to have a mean TBSA of epidermal detachment of 65%,16 of whom received IVIG at a total dose of 2.8 g/kg and 16 others received standard care only [13]. A mortality rate of 25% was observed in the IVIG-treated group vs. 38% mortality rate in the standard care group. Although the authors correctly state that the observed difference in mortality is not significant ( p = 0.364), significance may have been achieved had the study groups been larger.

Table 2 Summary of clinical study results on the effect of IVIG Author

Viard Science 1998

Stella Dermatol 2001

Prins Arch Derm 2003

Trent Arch Derm 2003

Bachot Arch Derm 2003

Campione Acta Derm 2003

Shortt J Bum Care Reh 2003

Brown J Burn Care Reh 2004

Study

Prosp Non-Cont 10 39 28.5 3 g/kg – – 0/10 (0%)

? Non-Cont 9 ? 16 2.8 g/kg – – 1/9 (11%)

Retrosp Non-Cont 48 43 45 3 g/kg – – 6/48 (12%)

Retrosp Non-Cont 16 43 43 4 g/kg 5.8 (36%) – 1/16 (6%)

Prosp Non-Cont 34 47 19 2 g/kg 8.2 (24%) – 11 (32%)

Prosp Non-Cont 10 49 49 2 g/kg 3.5 (35%) – 1 (10%)

Retrosp Non-Cont 32 53 65? 2.8 g/kg 6 (38%) – 4 (25%)

Retrosp Non-Cont 45 45 49 1.6 g/kg – 6/21 (29%) 10/24 (42%)

Patients Age Detachment Dose IVIG Predicted deaths Deaths without IVIG Deaths with IVIG

Only studies reporting series of more than 9 patients were analyzed.

548

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

Finally, the retrospective non-controlled study by Brown et al. concerned 45 TEN patients with a mean TBSA of epidermal detachment of 45%, 21 of whom received standard care, including surgical debridement in a burn unit, and the remaining 24 received similar standard care along with IVIG at a significantly lower than usual total dose of 1.6 g/kg [14]. In this study, mortality rate was 42% in the IVIG group vs. 29% in the standard care group. Mortality in the IVIG-treated group was also higher than that predicted by SCORTEN (38%). These results are somewhat surprising, and several features of this study are quite unusual. First, the standard care protocol is different from that used in most centers with experience in the management of TEN, in that wound debridement, a practice that is no longer recommended, is performed on all patients at admission. Second, the mortality rate of 42% observed in the IVIG-treated group of patients is higher than that reported in epidemiological studies of TEN and higher than that historically observed with standard care in most experienced centers. This suggests that some individual or combined aspect of the management protocol for patients in the IVIG group of this study may differ from that of the other studies assessing the effect of IVIG in TEN. The association of IVIG treatment and debridement is certainly one difference between this and other studies. Third, the total dose of IVIG used is the lowest of all studies described herein as only 1.6 g/kg total dose of IVIG was administered. Fourth, the onset of IVIG treatment was quite late (average of 7 days after onset of disease). Taken together, although all studies published to date are non-controlled, 6 of the 8 studies summarized in Table 2 point towards a benefit of high-dose IVIG upon the mortality rate associated with TEN. It should be noted that in the 2 studies that showed no benefit on mortality rate, the total dose of IVIG used was 2 g/kg or less, whereas in 5 of the 6 studies showing a benefit of IVIG on mortality rate, the total dose of IVIG used was greater than 2 g/kg. This apparent dose dependence is further suggested when mortality rate for the 3 lowdose vs. 4 high-dose studies is compiled. In the studies using high-dose IVIG (N2 g/kg), a 59% reduction between expected (SCORTEN) of control group (34 F 4%) and observed (14 F 8%) mortality rate is detected, whereas in the low-dose IVIG studies, only a 3% reduction between expected (31 F 7%) and observed (30 F 19%) mortality rate was detected. It appears, therefore, that total doses of 3–4 g/kg are most likely to be effective. Higher doses are not warranted. In 2003, we reported that an analysis of anti-Fas activity within different batches of IVIG revealed sig-

nificant differences. This analysis was performed on a large array of batches of 3 commercialized forms of IVIG (SandoglobulinR, ZLB Bioplasma Inc.; Gammagard S/D, Baxter Healthcare Corporation; EndobulinR and Endobulin S/DR, Baxter Healthcare Corporation). The Fas-inhibitory activity (1/IC50) of IVIG batches was found to be extremely variable, ranging from zero to a maximum of 2 (average 0.75), and was shown to be due to differences in anti-Fas Ab content within IVIG batches. The clinical significance of these differences is not clear at present. Also, it is difficult in practice to determine the anti-Fas activity of a given batch when treating a given patient. Our opinion is that these differences may contribute to the lower efficacy of IVIG observed in TEN when low doses of IVIG are used, but further work would be required to affirm this with certainty. Use of IVIG for the treatment of SJS has also been reported, but conclusions are more difficult to draw as the mortality rate associated with this condition is usually low (1–5%). A total of 3 studies exist to date, 2 of which concern pediatric patients. In adults, Prins et al. have reported a series of 12 patients treated with a mean dose of 0.6 g/kg/day for an average of 4 days [15]. An objective response to IVIG infusion was observed within a mean of 2 days in 100% of treated patients with an overall survival rate of 100% and total skin healing in an average of 8.3 days. In pediatric patients, a first study by Morici et al. in which 7 children received IVIG (2 g/kg total dose) and 5 children received supportive care only the authors reported a significant reduction in the duration of fever for the IVIG-treated (8 days) vs. supportive care (14 days) group [16]. A second study by Metry et al. retrospectively analyzed 7 children treated with IVIG (2 g/kg total dose) and reported the absence of new blisters within 24–48 h of the initiation of IVIG, good outcome in all patients, and the absence of side effects [17]. Therefore, high-dose IVIG seems effective in blocking the progression of SJS and reducing the time to complete skin healing. In conclusion, studies of the pathogenesis of TEN suggest that destruction of the epidermis, in the context of the ongoing drug-related hypersensitivity reaction, is due to massive Fas-mediated keratinocyte cell death by apoptosis. IVIG was shown to inhibit Fas–FasL interaction and cell death in vitro and thus provides a rationale for use in humans. Since that discovery, 7 large non-controlled clinical studies have been published, and lower than expected mortality rates have been reported in 6 of these. Although each study has its potential biases, and the 7 studies are not strictly com-

L.E. French et al. / International Immunopharmacology 6 (2006) 543–549

parable, it appears from comparative analysis that IVIG at total doses of more than 2 g/kg is a safe and potentially useful treatment for TEN. Hopefully, the cumulative results obtained in a non-controlled manner to date will serve as the basis for designing a prospective controlled trial in the near future. Such an approach appears to be the only way to definitively determine the therapeutic potential of IVIG in TEN. References [1] Roujeau JC, Stern RS. Severe adverse cutaneous reactions to drugs. N Engl J Med 1994;331:1272 – 85. [2] Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of cases of toxic epidermal necrolysis, Stevens–Johnson syndrome, and erythema multiforme. Arch Dermatol 1993;129:92 – 6. [3] Bastuji-Garin S, Fouchard N, Bertocchi M, Roujeau JC, Revuz P, Wolkenstein P. SCORTEN: a severity-of-illness score for toxic epidermal necrolysis. J Invest Dermatol 2000;115:149 – 53. [4] Trent JT, Kirsner RS, Romanelli P, Kerdel FA. Use of SCORTEN to accurately predict mortality in patients with toxic epidermal necrolysis in the United States. Arch Dermatol 2004; 140:890 – 2. [5] Paul C, Wolkenstein P, Adle H, Wechsler J, Garchon HJ, Revuz J, et al. Apoptosis as a mechanism of keratinocyte death in toxic epidermal necrolysis. Br J Dermatol 1996;134:710 – 4. [6] Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti S, et al. Fas triggers an alternative, caspase-8-independent cell death pathway using the kinase RIP as effector molecule. Nat Immunol 2000;1:489 – 95. [7] Viard I, Wehrli P, Bullani R, Schneider P, Holler N, Salomon D, et al. Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science 1998;282:490 – 3.

549

[8] 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 – 9. [9] Prins C, Kerdel FA, Padilla RS, Hunziker T, Chimenti S, Viard I, 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. [10] 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. [11] Bachot N, Revuz J, Roujeau JC. Intravenous immunoglobulin treatment for Stevens–Johnson syndrome and toxic epidermal necrolysis: a prospective noncomparative study showing no benefit on mortality or progression. Arch Dermatol 2003; 139:33 – 6. [12] Campione E, Marulli GC, Carrozzo AM, Chimenti MS, Costanzo A, Bianchi L. High-dose intravenous immunoglobulin for severe drug reactions: efficacy in toxic epidermal necrolysis. Acta Derm Venereol 2003;83:430 – 2. [13] Shortt R, Gomez M, Mittman N, Cartotto R. Intravenous immunoglobulin does not improve outcome in toxic epidermal necrolysis. J Burn Care Rehabil 2004;25:246 – 55. [14] Brown KM, Silver GM, Halerz M, Walaszek P, Sandroni A, Gamelli RL. Toxic epidermal necrolysis: does immunoglobulin make a difference? J Burn Care Rehabil 2004;25:81 – 8. [15] Prins C, Vittorio C, Padilla RS, Hunziker T, Itin P, Forster J, et al. Effect of high-dose intravenous immunoglobulin therapy in Stevens–Johnson syndrome: a retrospective, multicenter study. Dermatology 2003;207:96 – 9. [16] Morici MV, Galen WK, Shetty AK, Lebouef RP, Gouri TP, Cowan GS, et al. Intravenous immunoglobulin therapy for children with Stevens–Johnson syndrome. J Rheumatol 2000;27: 2494 – 97. [17] Metry DW, Jung P, Levy ML. Use of intravenous immunoglobulin in children with Stevens–Johnson syndrome and toxic epidermal necrolysis: seven cases and review of the literature. Pediatrics 2003;112:1430 – 6.