Challenges in Diagnosing and Treatment of Pediatric Necrotic Skin Conditions: A Retrospective Study

Introduction

Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are life-threatening, dermal, delayed hypersensitivity reactions [1,2]. They are characterized by necrosis and detachment of the epidermis and mucositis, accompanied by systemic symptoms [1,3]. These conditions are differentiated on the basis of percentage of skin area involvement: for SJS it is less than 10%, in overlap syndrome (SJS-TEN) it is between 10% and 30%, while in TEN it reaches values above 30% [4]. These maladies can occur in both sexes, all races, and all ages, but up to 50% of all cases are patients younger than 18 years [1,5]. Both of these conditions are predicted to affect 2 to 7 per million people annually, and SJS is over 3 times as common as TEN and overlap syndrome combined [6]. Incidence rates per million per year among children are 5.3–6.3, 0.7–0.8, and 0.4–0.5 for SJS, SJS-TEN, and TEN, respectively [7,8]. TEN is associated with mortality among children, with mortality rates reaching up to 16% due to frequent sepsis and multi-organ failure [5,9]. The natural history of SJS/TEN usually encompasses some common features. These include prodromal symptoms such as fever, a sore mouth, general malaise, and upper respiratory tract symptoms several days before dermal lesions appear. Involvement of the mucous membranes, which leads to damaging and hemorrhagic mucositis, can also be attributed as an early sign. Skin symptoms like cutaneous pain present at the onset of the disease evolve into incipient epidermal necrolysis [3].

Skin with lesions is tender to the touch and presents with a positive Nikolsky sign. Blistering is characterized by separation of necrotic epidermis from the underlying dermis, and flaccid bullae formation [3]. The most common causes of SJS/TEN are: drugs (anticonvulsants, antibiotics, and nonsteroidal anti-inflammatory agents), bacterial infections including Mycoplasma pneumoniae, and viral infections including Influenza virus and Epstein Barr virus (EBV) [1]. Dibek and coworkers [10] conducted a study of severe cutaneous adverse drug reactions that included 58 pediatric patients. Among these 58 patients, 35 (60.4%) were diagnosed with SJS/TEN. This group was well characterized – the median age was 9.13 (interquartile range [IQR], 7.05–12.85), and 18 patients (51.4%) were boys. It was shown that the main causative factor of SJS/TEN in this cohort was drugs (31 out of 35 patients, comprising 88.6% of the SJS/TEN cohort). Among antibiotics and antiepileptics, the general picture of the most causative substances were as follows: amoxicillin-clavulanate, ampicillin sulbactam, and clarithromycin caused SJS/TEN in 3, 5, and 2 patients respectively; and carbamazepine and lamotrigine induced the disease in 6 and 3 patients, respectively. Infectious agents played a minor role. Mycoplasma spp. caused 2 cases and herpes simplex virus type 1 caused only 1 case [10].

Care for these patients should be interprofessional. Skin and mucosal surfaces must be examined daily for signs of infection, and trauma to affected areas should be avoided. The use of non-adherent dressings is recommended, and topical anesthetics may be applied on mucosal surfaces [6]. Severe cases require treatment with immunosuppressive therapies and hospitalization in the intensive care unit or the burn unit, while mild cases can be managed with supportive treatment [11]. Due to the rarity and severity of SJS/TEN, no controlled clinical trials have been performed, and currently, there is no standardized treatment regimen that can be applied [4,11].

Discontinuation of the triggering drug is the first and the most important action. Currently, systemic agents like cyclosporine, systemic glucocorticoids, intravenous immunoglobulins (IVIGs), and tumor necrosis factor (TNF) are used in treatment; however, the rationale for their administration is based solely on retrospective case reports and empirical experience [4].

The pediatric population is a group that requires further research concerning clinical manifestations and treatment of these conditions, due to the fact that existing knowledge and procedures are based largely on adult patients and have not been adjusted to pediatric ones [12,13]. The rarity of occurrence of this disease is a reason why there are no unequivocal, scientifically proven treatment recommendations [14]. Therefore, this retrospective study aimed to evaluate the clinical features of necrotic skin reactions in 13 children.

Material and Methods

ETHICAL STATEMENT:

The study was approved by the Ethics Committee of the Medical University of Lublin, no. KE/178/02/2024. All procedures were conducted in compliance with applicable guidelines and regulations. Informed consent was obtained from the patients and their caregivers for the publication of visual elements and case reports.

MEDICAL RECORDS AND LABORATORY DATA RETRIEVAL:

The electronic medical records of patients with SJS/TEN diagnosis treated at the University Children’s Hospital in Lublin from January 1^st, 2015, to April 29^th, 2023, were retrospectively analyzed. The diagnosis of SJS/TEN is based on medical history and clinical presentation. The clinical assessment was based on the appearance of the lesions, “shooting target”-type lesions, epidermal exfoliation, or positive Nikolsky’s sign. Attention was paid to the centrifugal progression of the lesions and their location. The time of the appearance of the lesions and the introduction of the causative factor were correlated. No patient required histopathological examination. Patients were included in the study if they had a diagnosis of SJS or TEN using the International Classification of Diseases (ICD-10) codes containing the following phrases: L51.1, L51.2, or L51.3. Exclusion criteria were: lack of definite diagnosis of SJS, TEN, or overlap syndrome in the medical record on discharge, confirmation of an alternate diagnosis, incompatible clinical assessment, or insufficient information. The percentage of skin involvement was calculated using the Lund-Bowder table. Clinically, Stevenson-Johnson syndrome was defined as body surface involvement of less than 10%; TEN (also known as Lyell syndrome) as body surface involvement of more than 30%; and SJS/TEN overlap syndrome as lesions seen on 10% to 30% of the total body surface. Based on medical records and laboratory test results, patients were analyzed in terms of demographic data (sex, age, ethnic origin, comorbidities), length of hospital stay, prodromal symptoms, percentage of affected skin, involvement of mucous membranes, trigger factor, time from causative agent exposure to the onset of symptoms, and treatments used. Demographic factors and clinical data are presented in descriptive manner by number and percentages of the characteristics. Laboratory test results analyzed include full blood count, erythrocyte sedimentation rate index (ESR), C-reactive protein (CRP) levels, procalcitonin levels, detection of IgM, IgA, and IgG antibodies against M. pneumoniae, and detection of Streptococcus pneumoniae antigen by immunochromatographic tests of urine samples.

The medical history of the girl transferred to The Eastern Poland Burn Treatment and Reconstructive Surgery Center in Łęczna was supplemented with a retrospective evaluation of further treatment results.

Due to the small number of reported cases of epidermal necrolysis in children, through our retrospective study we provide an assessment of demographic characteristics, disease factors, treatment, and prognosis in the pediatric group. The limitations of the study were its retrospective nature, single-center design, and small sample size.

STATISTICAL ANALYSIS:

The study presents a descriptive analysis of clinical cases. Continuous variables, such as length of hospitalization and patient age, were summarized using ranges, standard deviations, and means. Categorical data, including mucous membrane involvement, prodromal symptoms, suspected trigger factors, and treatments used, were presented as counts and percentages.

Results

STUDY GROUP CHARACTERISTICS:

We identified 13 patients hospitalized at the University Children’s Hospital in Lublin for necrotic skin reactions. The patients’ characteristics are presented in Table 1.

LATENCY TIME UNTIL DISEASE MANIFESTATION:

According to the children’s prehospital history, the onset of mucocutaneous lesions ranged from 2 to 16 days before the diagnosis. One boy had been under the care of another hospital at the time of symptoms onset. A total of 6 children from the remaining 12 had sought care before presenting at the study center. The average delay between the first contact with medical care and an accurate diagnosis of the lesions was 2.5 days (range: 1–6 days).

DEMOGRAPHIC DATA/STUDY POPULATION:

The average age of the patients was 8.40±5.18 years for TEN diagnosis and 12.93±1.76 years for SJS-TEN. Two patients required re-hospitalization. On admission, the condition of 4 patients (30.8%) was severe, 3 patients (23.0%) had moderate symptoms, and the rest of the patients (46.2%) remained in good condition. The average length of hospitalization was 11.2±4.96 days (range, 3–23 days).

CLINICAL MANIFESTATION:

Skin-only lesions were present in 2 children (15.4%), mucosal-only lesions were present in 3 patients (23.1%), and mucocutaneous lesions were present in 8 patients (61.5%). In mixed cases, skin lesions preceded mucosal involvement in 3 children, and in each of these cases, the skin lesions appeared 2 days before mucosal involvement (Figures 1, 2).

In 4 patients of the mixed-lesion patient group, the mucous membranes were the first to undergo necrolysis, and the mucosal lesion appearance preceded skin lesion onset by 2 to 5 days. In 1 patient, due to the already-developed clinical picture and lack of previous history, the order of appearance of lesions remained unassessed. Extreme epidermal necrolysis, with involvement of 100% of the body surface, was diagnosed in 3 patients (23.1%). In the majority of cases (7 out of 13), necrolytic symptoms were preceded by prodromal symptoms. Systemic symptoms preceded the onset of necrotic lesions by an average of 6.6 days (range: 3 to 10 days).

LABORATORY TESTS:

Inflammatory markers were assessed. CRP was examined in all children, and in 6 (46.2%) of them, an increase in CRP levels was noted. In addition, in 6 patients, ESR was measured, revealing an increase in erythrocyte sedimentation in 3 of the children. Procalcitonin assessment was done in 9 of the patients, but only 1 child had raised levels of this marker. Information about inflammatory markers assessment is shown in Table 2.

Blood morphology examination revealed that 8 out of 13 patients exhibited leukocytosis, 4 patients demonstrated lymphopenia (2 had neutrocytosis and 2 had neutrophilia) and 2 patients had leukopenia with neutrocytosis. In 3 cases, no significant abnormalities in blood morphology were registered. The blood morphology results are shown in Table 2.

CAUSATIVE FACTORS:

All potential exposures to substances and infections capable of triggering skin reactions were assessed for the time period of 1 month preceding each patient’s symptoms. A probable or highly probable factor was found in 11 (84.6%) of the analyzed cases. After compilation of the clinical data sets, drugs were identified as the most common cause of the syndromes in question (7 cases, 54.5%). The second most common trigger was infections (45.5%). In 2 cases, the triggering factor could not be definitively characterized. A detailed list of the triggering factors is presented in Table 3.

TREATMENT AND PROGNOSIS:

In the determination of therapy of cutaneous necrotic reaction syndromes, it is crucial to identify and address the triggering factor. To do this during hospitalizations, infections were treated with antibiotics, drug reactions due to non-steroidal anti-inflammatory drugs (NSAIDs) were resolved by changing symptomatic treatment, drug reactions due to antiepileptic treatment were addressed by replacing oxcarbazepine with sodium valproate, and lamotrigine and clarithromycin were discontinued. Depending on the patient’s condition, supportive treatment was also provided. Data regarding the applied treatments are included in Table 3.

Multimodal immunosuppressive treatment was applied, and relied mainly on glucocorticoids as the medication of choice, in 12 (92.3%) subjects. In addition, IVIG was administrated to 4 patients (30.8%), and in 1 case, cyclosporin A was applied. Subjects were treated with immunoglobulins when surface area involvement reached from 80 to 100% (4 patients). Cyclosporine was included as a rescue treatment in a girl with severe progression of symptoms despite standard treatment. A 23-month-old child who suffered from M. pneumoniae infection was treated with cloxacillin, amoxicillin, and clarithromycin, resulting in improvement of symptoms (Table 3). Adverse effects of immunoglobulin infusions are presented in Table 3.

In accordance with an interdisciplinary medical approach, 11 children underwent specialized consultations: 8 of them had a single counseling session, while 3 underwent 2 or more counseling sessions. As to types of advice given, ophthalmology consultation was requested most often (81.8% of cases). In addition, 4 cases benefited from recommendations from an otolaryngologist or dermatologist.

The treatment resulted in improvement in 12 patients. In 1 case, hospitalization was terminated at the mother’s request before the therapeutic goal was fully achieved, and 1 patient did not benefit from the therapeutic management. In the other patients, the time required for remission of symptoms ranged from 6 to 17 days (10 on average). A fulminant course of cutaneous necrosis was present in a girl without health improvement, despite applied therapy. On the day of admission, her overall state had already been worsening for the previous 2 days, with 100% of her body surface affected, including the scalp (Figures 3–6). Interestingly, her condition had not been accompanied by systemic symptoms, and the inflammatory markers remained within normal limits (Table 2). The only deviation was her ESR, which was 37 mm/h. The presumed trigger may have been lamotrigine. Immunosuppressive treatment was attempted. Unfortunately, the patient’s condition progressed (Table 3). After 2 days of hospitalization, the decision was made to transfer the child to the Eastern Center for Burn Treatment and Reconstructive Surgery in Łęczna, to optimize the treatment. It is worth noting that the child’s condition continued to deteriorate and in a short time period she required transfer to the intensive care unit. She had to undergo a partial resection of her small intestine due to necrotic lesions, but was discharged home in good overall condition after about 30 days.

Discussion

PEDIATRIC AND ADULT NECROTIC SKIN REACTIONS:

No mortality was observed in the patient cohort included in this study. This outcome is in line with the accepted knowledge that children with necrotic skin reactions present with less severe symptoms than adults, which is reflected in higher mortality rates, reaching 34% in adults with necrotic skin reactions [8,19–23]. Despite TEN progression requiring intensive care with subsequent transfer to a burn unit in 1 patient, the resolution of her clinical condition was eventually positive. To sum up, all of the patients had positive treatment outcomes.

In comparison with adults, recurrence is more frequent in children, reaching 18%, vs 7% in adults [5,24]. In our study, recurrences occurred in 2 patients (15.4%). In both of these cases, the recurrences were induced by the same triggering factors as the initial episode of the disease. In the first case, recurrence was due to subsequent exposition to ibuprofen, while in the second case, the recurrence was caused by M. pneumoniae reinfection within 2 years of the initial infection. This clearly shows how crucial it is that well-developed counseling for pediatric patients and their parents is provided. In particular, such counseling should provide information about increased risk of subsequent illness if exposed to the etiological factor(s). This flow of information ought to be applied in all SJS/TEN cases, whether caused by pharmacological substances or by infections, allowing informed avoidance of factor(s) that could cause recurrences in the future [17].

TRIGGERS OF NECROLYTIC SKIN REACTIONS IN THE CONTEXT OF CURRENT RESEARCH:

In the current study, the most common causative factor was drugs (54.4%), mostly NSAIDs (23.1%), followed by antiepileptic drugs (15.4%) and antibiotics (7.7%). This is in line with general information found in the literature: that administration of antibiotics and anticonvulsants is widespread among patients suffering from necrolytic skin reactions [5,10,25]. Moreover, paracetamol and ibuprofen are also considered well-known causative agents of this disease group [26]. The second main agent causing SJS/TEN in the cohort of patients enrolled in the present study was infections (45.5%).

Careful analysis of the temporal coincidence of exposure to the triggering agent and the onset of symptoms was conducted. Despite this, the cause of the condition remained ambiguous in 2 patients. The first patient (a girl) had been diagnosed with pediatric inflammatory multisystem syndrome (PIMS) 24 days before the appearance of mucocutaneous lesions, which was the only criterion matching our temporal analysis. The second patient had skin lesions for about 2 weeks before reporting to the hospital. Flu-like symptoms, fever, and gastrointestinal distress preceded skin and mucosal presentations. Interestingly, this patient had previously been to Egypt. Due to the lack of sufficient data regarding both the history of medications administered and negative laboratory tests, it was not possible to confirm the triggering factor unequivocally and we can only speculate that an unspecified infection may have been the culprit. In 1 case, there was no significant exposure to any agent capable of causing an epidermal necrotic reaction. Progression of necrotic lesions after admission to the hospital occurred in 5 out of 13 patients, while in the rest of the group, no further escalation was noted.

Earlier research has revealed that both chemical medications and infectious agents have the potential to induce SJS/TEN [3]. This stresses the need for careful and detailed diagnostics for presumable microorganism presence, especially Mycoplasma, with mandatory nasopharyngeal serology testing and assessment by dedicated polymerase chain reaction (PCR) [27]. Our data indicate that infection with M. pneumoniae was responsible for both the initial SJS/TEN overlap syndrome and subsequent recurrence in 1 patient. This is in line with the well-established observation that infectious agents are culprits of persistent recurrences in a great number of cases and, as a trigger, are difficult to avoid, in contrast to medical substances administration, which can be promptly halted [28]. It must be noted, though, that it is sometimes hard to confirm that a single drug is either fully responsible for enhancement of prodromal symptoms development or is an individual potential cause of SJS/TEN [3].

All 4 female patients in our cohort suffered from TEN, while male patients were diagnosed with TEN, SJS-TEN overlap syndrome, or SJS (4, 4, and 1 patient, respectively). Antoon et al [11] reported that TEN occurred more frequently in female patients (54.8%), whereas SJS was more common in male patients (61.2%), in a study of 898 pediatric hospitalizations. In comparison with adult patients, this sex ratio is reversed, which could be explained by different biological predisposition. In addition, a higher frequency of SJS/TEN in White children was noticed. As for medications, IVIG was prescribed for a total of 229 (25.5%) patients, 167 (18.6%) received steroids only, and in 153 patients (17.04%), both steroids and IVIG were administrated. Recurrences were relatively common, accounting for 9.85% of patients readmitted within 1 month, and for 18.9% of patients readmitted within 6 months [11].

Due to the rarity of SJS and TEN in children, there is a paucity of information concerning treatment in this group of patients. The treatment regimen was generalized based on observations made in adults. The choice of care was selected according to symptoms and each patient’s general condition [11]. In severe TEN or SJS cases, patients need intensive therapy and/or treatment in a burn treatment center [29].

In the present study, 12 out of 13 patients received both antibiotics and glucocorticoid treatment. Cyclosporine was included in 1 case, due to progression of the disease. This was supported by information provided by 2 case reports [30,31]. Adverse reactions were observed in 2 children (15.4%).

Multidisciplinary cooperation during the diagnostic and treatment process was very important. Other studies have also emphasized this aspect of patient care, to achieve a successful outcome [17,29]. For the patients in the present report, the cooperatively working disciplines included dermatology, intensive care, ophthalmology, and mental health. In our retrospective analysis, 11 out of 13 (84.6%) patients had at least 1 specialist consultation. It is worth mentioning that mental health counseling with long-term follow-up may be beneficial for some patients, providing psychological support [32]. In some cases, it may prevent posttraumatic stress disorder and pharmacophobia. Moreover, it may alleviate discomfort regarding cosmetic consequences of the disease [33,34].

To summarize our observations, improvement and a condition of good health were achieved in all patients. Therefore, we would like to share our experiences, especially in light of the absence of established recommendations for this clinical problem in pediatric populations.

LIMITATIONS OF THE STUDY AND FUTURE PERSPECTIVES:

There were several limitations of this study. First of all, retrospective analysis of a limited number of patients may blur the general picture of incidence and character of SJS/TEN in Polish pediatric populations. Further, the presented study was single-centered, which additionally limited the sample size included in the research. Due to this fact, the study did not take into account cases of patients treated in different hospitals. For these reasons, the real number of recurrences, as well as the characteristics of possible patients who were not enrolled in the study, may have been omitted. All those factors may have influenced the final outcome of the current research.

Conclusions

Early recognition of SJS and TEN, and the use of effective treatment, is crucial for the prevention of mucocutaneous complications, and for a generally successful outcome and prevention of morbidity and mortality. All healthcare practitioners should be aware of these conditions and the need to properly refer patients with blistering cutaneous reactions.

The current study indicates a diagnosis delay of 2.5 days (range 1–6 days). This suggested a relatively prompt response to first presentations of the disease by physicians, which could have contributed to the good treatment results. The most common causative factors in our study were drugs, mostly NSAIDs, and infections. In the majority of cases, disorders were preceded by prodromal symptoms. Necrolytic symptoms were mainly present on mucous membranes and skin simultaneously.

We used antibiotics and anti-inflammatory therapy in all patients, and this approach achieved resolution of symptoms. In the absence of information from randomized studies, we share our experiences here.