44 - EPIDERMAL NECROLYSIS (SJS AND TEN) Flashcards
Widespread apoptosis of keratinocytes is provoked by what
Widespread apoptosis of keratinocytes is provoked by the activation of a cell-mediated cytotoxic reaction and amplified by cytokines, mainly granulysin.
Define SJS and TEN
Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are acute life-threatening mucocutaneous reactions characterized by extensive necrosis and detachment of the epidermis and mucosal epithelium (Table 44-1).
In 1922, Stevens and Johnson first reported two cases of disseminated cutaneous eruptions associated with an erosive stomatitis and severe ocular involvement. 1 In 1956, Lyell described patients with epidermal loss secondary to necrosis and introduced the term toxic epidermal necrolysis. 2 Because SJS and TEN share clinical pattern, histopathologic findings, etiology, risk factors, and mechanisms, they are considered as severity variants of one disease entity that differs only in the extent of skin detachment related to the body surface area. 3-5 Therefore, it seems more appropriate to use the designation epidermal (or epithelial) necrolysis for both, as proposed by RuizMaldonado (acute disseminated epidermal necrosis)6 and Lyell (exanthematic necrolysis).
EPIDEMIOLOGY
Epidermal necrolysis (EN) is rare. The overall incidence of SJS and TEN was estimated at 1 to 6 cases per million person-years and 0.4 to 1.2 cases per million person-years, respectively. 8-10 Whereas incidences as high as 5 or 6 cases per million per year derive from medical databases not primarily designed for epidemiologic analysis of rare diseases, 11 incidences of 1 to 2 cases per million per year were calculated using population-based prospective case registry data.12,13
EN can occur at any age, but the risk increases with age leading to the highest incidence in older adults after 65 years of age. 14 Women are more frequently affected, showing a sex ratio of 0.6. Patients infected with HIV and to a lesser degree patients with collagen vascular disease and cancer are at increased risk. 15,16 The overall mortality rate associated with EN is 22% to 27%, varying from approximately 10% for SJS to almost 50% for TEN. 17-19 Increasing age, significant comorbidity, and greater extent of skin detachment correlate with poor prognosis. A prognosis score (SCORTEN) has been constructed for EN, 20 and its usefulness has been confirmed by several teams 21-24 (Table 44-2).
ETIOLOGY
The exact pathophysiology of EN is still unclear; however, drugs are the most important etiologic factors. More than 100 different drugs have been implicated, 25-31 but fewer than a dozen “high-risk” medications account for about half of cases in Europe (Table 44-3), as evidenced by two multinational casecontrol studies. 16,27-30 These high-risk drugs are antibacterial sulfonamides, aromatic antiepileptic drugs, allopurinol, oxicam nonsteroidal antiinflammatory drugs, lamotrigine, and nevirapine. 28-30 The risk seems confined to the first 8 weeks of treatment and most inducing medications revealed the first continuous exposure between 4 and 28 days before reaction onset. 28 Slow dose escalation decreases the rate of mild rashes to lamotrigine and nevirapine, 32,33 but there is no evidence that it decreases the risk of EN. 28,30 Oxcarbazepine, a 10-keto derivative of carbamazepine, which was considered to carry a lower risk, seems to significantly cross-react with carbamazepine. 34 Many nonsteroidal antiinflammatory drugs were suspected to be associated with EN, but there is a substantial difference in risk for EN among them, with oxicam derivatives showing the highest risk, acetic acid derivatives such as diclofenac a moderate, and propionic acid derivatives such as ibuprofen no increased risk. 16,28,35 A significant but much lower risk has also been reported for various groups of antibiotics such as cephalosporins, quinolones, tetracyclines, and aminopenicillins.28 Corticosteroids were significantly associated with an increased relative risk, but confounding could not be excluded. 28 Recent analysis of systematically ascertained registry data on EN using ALDEN (algorithm for causality assessment in EN) 36 in comparison with results of two case-control studies 16,28 revealed that over a period of more than 2 decades, the proportion of validated cases that could be explained by medications with a significant (high and moderate) risk was stable (65%–68%). For roughly one third of the cases, no patent drug cause could be identified, suggesting that many more cases than previously thought have to be considered as “non–drug related” or “idiopathic.”37 The role of infectious agents in the development of EN is much less prominent than for erythema multiforme (see Chap. 43). 5 However, cases of EN associated with Mycoplasma pneumoniae infection, viral disease, and immunization have been reported, particularly in children. 38 In many cases a clinical infection is present, but despite extensive laboratory workup, no specific viral or bacterial agent can be detected. 39 Cases of EN have been reported after bone marrow transplantation. Some are an extreme form of acute graft-versus-host disease (GVHD; see Chap. 129); others could be druginduced. The relationship between EN and GVHD is difficult to assess because clinical and histologic skin features are often indistinguishable. 40 Lupus erythematosus (systemic LE or subacute cutaneous LE) is associated with an increased risk of EN. 16,28 In such cases, drug causality is often doubtful and necrolysis might be an extreme phenotype of cutaneous lupus. Therefore, erosive LE has to be considered as a differential diagnosis of EN. 41,42 Finally, radiotherapy in addition to treatment with antiepileptic drugs, such as phenytoin, phenobarbital, or carbamazepine, can trigger EN with lesions localized predominantly at sites of radiation treatment.43,44
Drugs and Recommendations in Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis
PATHOGENESIS
Even if the precise sequence of molecular and cellular events is incompletely understood, several studies provided important clues to the pathogenesis of EN. The immunologic pattern of early lesions suggests a cell-mediated cytotoxic reaction against keratinocytes leading to massive apoptosis. 45,46 Immunopathologic studies have demonstrated the presence of cytotoxic cells, including natural killer T cells (NKT) and drugspecific CD8 + T lymphocytes in early lesions; monocytes and macrophages and granulocytes are also recruited. 47-49 CD94/NKG2C was identified as a killer effector molecule in patients with EN. 50 However, it is generally accepted that specific and nonspecific cytotoxic cells are too few within the lesions to explain fullthickness necrosis of extensive areas of the epidermis and mucous membranes. Amplification by cytokines has been suspected for years, especially for factors activating “death receptors” on cell membranes, especially antitumor necrosis factor (TNF) α and soluble Fas ligand (Fas-L). 47,51 In the past decade, it had been widely accepted that Fas-L was inducing the apoptosis of keratinocytes in EN, 51,52 despite partial evidence and discordant findings. 53-55 An important study has
challenged this dogma by demonstrating the key role of granulysin in EN. 56 The cytolytic protein granulysin was present in the blister fluid of patients with EN at concentrations much higher than those of perforin, granzyme B, or Fas-L. At such concentrations, only granulysin, and to a much lesser degree perforin, were able to kill human keratinocytes in vitro; Fas-L was not. Furthermore, injection of granulysin in the dermis of normal mice resulted in clinical and histologic lesions of EN. 57 Recently, interleukin (IL)-15 could be demonstrated to be associated with severity and mortality in EN. 58 When combined, these results strongly suggest that the effector mechanisms of EN have been deciphered. Cytotoxic T cells develop and are usually specifically directed against the native form of the drug rather than against a reactive metabolite, contrarily to what has been postulated for many years. These cells kill keratinocytes directly and indirectly through the recruitment of other cells that release soluble death mediators, the principal being granulysin and probably also IL-15.56-58
These advances in understanding the final steps of the reaction point to inhibition of release or blockade of granulysin as major aims of therapeutic interventions.
Little is known on the initial and intermediate steps. We still do not understand why very few individuals develop a violent immune response to medications and why effector cells are especially directed to the skin and other epithelia. Actually, most drugs associated with a high risk for EN can also induce a variety of milder and more frequent reactions. Drugspecific CD8 cytotoxic T-lymphocytes were also found in milder skin like maculopapular eruption. 59 Hence, it is tempting to speculate on an abnormal regulation of immune response. Regulatory CD4 + CD25 + T cells have been demonstrated to be potentially important in the prevention of severe epidermal damage induced by reactive cytotoxic T lymphocytes in a mouse model of EN. 60 Similar regulatory cells may play a role in drug eruptions in humans. 61 Altered regulation of the immune response to medications in patients with EN could result from comorbidities that are frequent (eg, cancer, HIV infection, collagen vascular disease), from comedications (eg, corticosteroids), or from genetic background.
Genetic susceptibility plays an important role in the development of EN to a few “high-risk” medications. A strong association was observed in Han Chinese from Taiwan between the human leukocyte antigen HLA-B ∗ 1502 and EN induced by carbamazepine and between HLA-B ∗ 5801 and EN induced by allopurinol. 62,63 B ∗ 1502 association with carbamazepine-related cases was confirmed in several Southeast Asian countries 64,65 but not in Japan and Korea. 66,67 The association between carbamazepineinduced EN and HLA-B ∗ 1502 was not present in European patients without Asian ancestry. 68 On the other hand, HLA-B ∗ 5801 was confirmed to be associated with allopurinol-related EN in Japan 66 and Europe, 69 but the strength of association was lower than in Taiwan. A genome-wide association study on samples of European EN-patients confirmed the involvement of genetic variants located in the HLA region. No other locus reached genome-wide statistical significance in this large sample. If some loci outside HLA play a role in EN, their effect might be very small.
Decisional tree for referral of a patient with epidermal necrolysis.
Clinical history
EN clinically most often begins with unspecific prodromal symptoms such as sore throat, runny nose, cough, headache, fever, and malaise preceding the mucocutaneous lesions by 1 to 3 days. They are followed by the appearance of erythematous macules and atypical targets of the skin that may be confluent and on which blisters occur. Burning or stinging of the eyes, pain when swallowing, or urinating progressively develop, heralding mucous membrane involvement. The majority of cases begin with nonspecific symptoms followed either first by mucous membrane or by cutaneous involvement, but some cases may start with the specific lesions of skin and mucosa. Whatever the initial symptoms are, their rapid progression, the addition of
new signs, severe pain, and constitutional symptoms should alert one to the onset of a severe disease.
A common problem is the fact that medications taken to treat the prodromal symptoms are often accused to have caused the reaction. This mainly concerns antipyretics, analgesics, and secretolytics, sometimes summarized as “cough and cold medicines.” When looking at the use of these medications more closely, they have usually been taken and tolerated before and were started after the onset of prodromal symptoms of EN (“protopathic bias”). Neither of these patterns is typical for exposure to medications causing EN, which have not been used previously and the exposure of which is the first continuous use started 4 weeks to 4 days before reaction onset. Furthermore, these substances do not belong to the drug groups for which epidemiologic studies have estimated an increased risk to induce EN.
CUTANEOUS LESIONS
The eruption is initially symmetrically distributed on the face, the upper trunk, and the proximal part of limbs. 71 The distal portions of the arms as well as the legs are relatively spared, but the eruption can rapidly extend to the rest of the body within a few days and even within a few hours. The initial skin lesions are characterized by erythematous, dusky red, irregularly shaped purpuric macules, which progressively coalesce. Atypical target lesions with dark centers are often observed (Fig. 44-2A). Confluence of necrotic lesions leads to extensive and diffuse erythema. Nikolsky sign, or dislodgement of the epidermis by lateral pressure, is positive on erythematous zones (Figs. 44-3 and 44-4). At this stage, the lesions evolve to flaccid blisters, which spread with pressure and break easily (see Fig. 44-2B). The necrotic epidermis is easily detached at pressure points or by frictional trauma, revealing large areas of exposed, red, sometimes oozing dermis (see Figs. 44-2C and 44-2D). In other areas, epidermis may remain.
Patients are classified into one of three groups according to the total area in which the epidermis is detached or “detachable” (positive Nikolsky): (1) SJS, less than 10% of body surface area (BSA); (2) SJSTEN overlap, between 10% and 30%; (3) TEN, more than 30% of BSA. 3,5 Correct evaluation of the extent of detachment is difficult, especially in zones with spotty lesions. It is helpful to remember that the surface of one hand (palm and fingers) represents a little less than 1% of the BSA.
MUCOUS MEMBRANE INVOLVEMENT
Mucous membrane involvement (nearly always on at least two sites) is observed in approximately 90% of cases and can precede or follow the skin eruption. It begins with erythema followed by painful erosions of the oral, ocular, genital, nasal, anal and sometimes tracheal or bronchial mucosa. This usually leads to impaired alimentation, photophobia, conjunctivitis, and painful micturition. The oral cavity and the vermilion border of the lips are almost invariably affected and feature painful hemorrhagic erosions coated by grayish white pseudomembranes and hemorrhagic crusts of the lips (Fig. 44-5). Approximately 80% of patients have conjunctival lesions, 72,73 mainly manifested by pain, photophobia, lacrimation, redness, and discharge. Severe forms may lead to epithelial defect and corneal ulceration, anterior uveitis, and purulent conjunctivitis. Synechiae between eyelids and conjunctiva often occur. There may be shedding of eyelashes (Fig. 44-5B). Genital erosions are frequent, often overlooked in women, and may lead to synechiae.74,75
What is nikolsky sign
dislodgement of the epidermis by lateral pressure
SYSTEMIC INVOLVEMENT
EN is associated with high fever, pain, and weakness. Visceral involvement is also possible, particularly with pulmonary and digestive complications. Early pulmonary complications occur in approximately 25% of patients and are essentially manifested by elevated respiratory rate and cough, which should prompt strict surveillance. 76,77 Bronchial involvement in EN is not
correlated with the extent of skin lesions or with the offending agent. In most cases, chest radiographs are normal on admission but can rapidly reveal interstitial lesions that can progress to acute respiratory distress syndrome. In all reported cases, when acute respiratory failure developed rapidly after the onset of skin involvement, it was associated with poor prognosis. In the case of respiratory abnormalities, fiberoptic bronchoscopy may be useful to distinguish a specific epithelial detachment in the bronchi from an infectious pneumonitis, which has a much better prognosis.
Gastrointestinal tract involvement is less commonly observed, with epithelial necrosis of the esophagus, small bowel, or colon manifesting as profuse diarrhea with malabsorption, melena, and even colonic perforation. 78,79 Renal involvement has been reported. Proteinuria, microalbuminuria, hematuria, and azotemia are not rare. Proximal tubule damage can result from necrosis of tubule cells by the same process that destroys epidermal cells. 80 Glomerulonephritis is rare.81
Laboratory findings
There is no laboratory test to support the diagnosis of EN. Laboratory examinations are essential for evaluation of severity, prognosis, and daily management as for all life-threatening conditions in intensive care units (ICUs).
Evaluation of respiratory rate and blood oxygenation are among the first steps to take in the emergency department. Any alteration should be checked through measurement of arterial blood gas levels. Serum bicarbonate levels below 20 mM indicate a poor prognosis.20 They usually result from respiratory alkalosis related to the specific involvement of bronchi and more rarely from metabolic acidosis.
738
Massive transdermal fluid loss is responsible for electrolyte imbalances, hypoalbuminemia, and hypoproteinemia, and mild and transient renal insufficiency and prerenal azotemia are common. Increased blood urea nitrogen level is one marker of severity. Anemia is common, and mild leukocytosis as well as thrombocytopenia may occur. Neutropenia is often considered to be an unfavorable prognostic factor but is too rare to have a significant impact on SCORTEN. Transient peripheral CD4 + lymphopenia is nearly always seen and is associated with decreased T-cell function. Mild elevation in levels of hepatic enzymes and amylase (most probably of salivary origin) are frequent but without impact on prognosis. A hypercatabolic state is responsible for inhibition of insulin secretion or insulin resistance, which results in hyperglycemia and occasionally overt diabetes. A blood glucose level above 14 mM may be a marker of severity.20 Other abnormalities in laboratory values may occur, indicating involvement of other organs and complications such as sepsis.
Histologic appearance of epidermal necrolysis.
Skin biopsy for routine histologic and possibly immunofluorescence studies are strongly recommended, especially if there are alternative diagnoses to consider (see Fig. 44-3). The biopsy should be taken from a fresh lesion, preferably form the erythematous margin and not directly out of a blister because the latter carries the risk that epidermis and dermis are completely separated or partly lost. If similar lesions are present on the legs and elsewhere, the biopsy should preferentially not be taken from the legs below the knee. In the latter localization, changes could be superimposed by stasis dermatitis, complicating the correct histopathologic diagnosis.82
In the early stages, epidermal involvement is characterized by sparse apoptotic keratinocytes in the suprabasal layers (Fig. 44-6). In fully developed stages of the disease, when epidermal detachment and epidermolysis appear, in addition to the already mentioned changes subepidermal vesiculation secondary to extensive vacuolar alteration and confluent necrosis of keratinocytes develops. 82,83 Apoptosis of epithelial cells may involve sweat glands and hair follicles. A moderately dense mononuclear cell infiltrate of the papillary dermis is observed, mainly represented by lymphocytes, often CD8 + and macrophages.83,84 Occasional eosinophilic granulocytes may be found. Variable numbers of extravasated erythrocytes are present. As the process develops rapidly, usually neither melanophages nor siderophages are found in the upper part of the dermis. The cornified layer remains unchanged. Results of direct immunofluorescence test are negative. Histopathology of involved mucous membranes show a similar pattern but are rarely performed.85