Analysis of 262 Children with Scrub Typhus Infection: A Single-Center Experience

INTRODUCTION

Scrub typhus, a vector-borne rickettsiosis caused by Orientia tsutsugamushi infection, is one of the leading causes of non-malarial febrile illness in Asia. ¹ The disease presents as acute undifferentiated fever with or without eschar. The clinical picture is characterized by fever with chills, myalgia, headache, abdominal pain, vomiting, and capillary leak during the first week of illness. Its nonspecific clinical features and shared seasonal profile with dengue and leptospirosis hamper its early diagnosis, particularly in the tropics. ² The pathophysiology is not entirely clear, though in general, it is assumed to be disseminated vasculitis with primary target being vascular endothelium. Complex interplay of host and pathogen response with combined humoral and cellular immunity contributes to the development of vascular injury to skin, lungs, liver, and brain. ³ The gold standard test for diagnosis, the indirect immunofluorescent assay, is prohibited by exorbitant cost and the need for personnel trained in fluorescent microscopy and is usually limited to research centers. ⁴ Laboratory confirmation, in resource-constrained settings, is commonly accomplished by IgM ELISA (InBios International Kit), with 91% sensitivity and 99% specificity. ^5–7 The disease, in general, portends a mild phenotype with a median mortality of 6% (0–70%) if left untreated. ⁸ Complications of scrub typhus are organ-specific and comprise acute respiratory distress syndrome, acute kidney injury, myocarditis, meningoencephalitis, and disseminated intravascular coagulation. ^9,10 Given the poorly defined symptomatology and life-threatening complications if left untreated, early diagnosis and judicious antibiotic therapy with tetracyclines/chloramphenicol is crucial. The research objective was to study the clinical profile of children with scrub typhus diagnosed by IgM ELISA admitted at a tertiary care teaching institute in South India with special reference to the myriad of typical and atypical manifestations.

METHODOLOGY

This study is a retrospective review of children with clinical and laboratory diagnosis of scrub typhus presenting to the Department of Paediatrics, Pondicherry Institute of Medical Sciences, a tertiary care teaching institute in South India, during the period of January 2012–December 2018. Children younger than 12 years presenting with acute undifferentiated fever, defined as acute febrile illness with onset in the previous 2 weeks with fever of > 38°C for at least 48-hour duration, were screened for eligibility. Children with seropositive IgM ELISA for scrub typhus were enrolled in the study. The InBios International Kit with a sensitivity of 91% and a specificity of 99% was used for diagnostic testing. A sample absorbance value > 0.5 was established as the cutoff criteria based on the local healthy normal sera. Children with classical symptoms and signs suggestive or diagnostic of a particular febrile illness (other than scrub typhus) or with chronic or recurrent febrile illness were excluded. Children with objective evidence of organ dysfunction such as myocarditis, meningoencephalitis, pneumonia, acute kidney injury, acute respiratory distress syndrome, disseminated intravascular coagulation, pancreatitis, and hemophagocytic lymphohistiocytosis were considered as severe scrub typhus. The study was approved by the Institutional Ethics Committee (IEC-RC/16/147). Waiver of parental consent was obtained from the Institutional Ethics committee because this was a retrospective chart review. Clinical features, laboratory investigations, treatment received, and the outcome recorded were obtained. The data were entered in a data recording form and then transcribed to MS Excel data and analyzed using SPSS Statistics for Windows, Version 17.0 (SPSS Inc., Chicago, IL). Descriptive data were represented by frequency, percentage, and mean ± SD. The chi-square test for dichotomous variables and t test and Mann–Whitney U test for continuous variables were applied. The associations of clinical and laboratory variables were studied by multivariate regression and logistic regression. P < 0.05 was considered statistically significant.

RESULTS

During the period of January 2012–December 2018, records of 41,365 children presenting to the pediatric services with fever were reviewed, and a total of 262 children with undifferentiated fever and IgM ELISA positivity for scrub typhus were diagnosed. Of these, 170 were boys (65%) and more than half of the children were between the age-group of 1–5 years (Table 1). The youngest infant diagnosed was aged 5 months, and 13 children (5%) presented during infancy. Fever was universal, and 128 children presented within the first 5 days of fever (44%). Generalized lymphadenopathy (93.5%) and hepatomegaly (70%) were the most common clinical findings. Painless eschar was identified in 31% of children, with greater predilection for groin and axilla (Figure 1). Nineteen children had developed a diffuse maculopapular rash. The laboratory profile of children with scrub typhus is illustrated in Table 2. Thrombocytopenia defined by a platelet count of less than 100 × 10⁹/L of blood was striking in one-third of children. Of the 262 children, 25 children were found to have severe scrub typhus and 18 children required intensive care stay; nine required ventilator support for pneumonia, and six required neurocritical care monitoring for meningoencephalitis. Chest radiograph suggestive of bronchopneumonia was noted in 11 children, and hemophagocytic lymphohistiocytosis was identified in three children (Table 3). There was no mortality during the study period. Multivariate analysis revealed elevated aspartate aminotransferase as a significant predictor of severity of scrub typhus ([OR 3.9], P value 0.005). Other variables such as fever > 5 days, eschar, leukocytosis, thrombocytopenia, hypoalbuminemia, and hyponatremia were insignificant. Overall, 239 children received doxycycline therapy, and 23 received azithromycin therapy. Severe forms of scrub typhus with complications received a 14-day therapy. Rapid defervescence within 48 hours was noted in 87% of the children. Table 4 compares the profile of children based on eschar, a clinical hallmark of scrub typhus. Clinical evidence of lymphadenopathy was found to have a significant association with eschar in children with scrub typhus (P < 0.005).

View Full Size

Well-established eschar in the axilla (A) and groin (B).

Citation: The American Journal of Tropical Medicine and Hygiene 104, 2; 10.4269/ajtmh.20-1019

• Download Figure
• Download figure as PowerPoint slide

DISCUSSION

In this 6-year retrospective study, the authors describe the profile of pediatric scrub typhus and share the single-center experience of one of the largest data in children from Indian subcontinent. IgM ELISA was used for the diagnosis of scrub typhus in our study. Weil Felix is the cheapest and the most easily available method for scrub typhus serological diagnosis and has been used hitherto in several studies. Its low sensitivity, dubious false positivity in Proteus infections and its utility only in the second week of illness are the major drawbacks and have been supplanted by other serological methods such as IgM ELISA, PCR, and indirect immunofluorescence assay. Indirect immunofluorescence assay is the gold standard test for diagnosis but is limited by expensive fluorescent microscopy and expert training. ⁴ IgM ELISA testing with optimal sensitivity (91%) and specificity (99%), like in our study, is the practical alternative in resource-limited settings. ⁷

In our study, boys and children younger than 5 years were likely to be infected with scrub typhus, similar to that reported by Ganesh et al. ¹¹ This could be justified by the increased outdoor play in the vector-ridden vegetation by children. In addition, 13 infants (5%) were diagnosed with scrub typhus. This is in consonance with studies by Dinesh et al. and Palanivel et al. who had described infantile onset scrub typhus in approximately 10% of their cohorts. ^12,13 This comparatively higher percentage reported by the two authors might be due to the less number of children recruited in their study. Given the restricted outdoor activity in infants, Narayanasamy et al. ¹³ suggested the role of mite-ridden rodents or domesticated dogs in the transmission. However, a secondary route of exposure with mites hidden in garments brought in from outdoors is a reasonable explanation. ¹⁴

Scrub typhus infection is the leading treatable cause of non-malarial febrile illness in Asia. Over the last decade, it has become a seemingly ubiquitous part of Indian subcontinent with several published reports from distinct geographical locations. A review of the published literature on pediatric scrub typhus from India over the last decade was carried out by the authors ^11–21 (Table 5). For the purpose of this review, citations were identified through PubMed search limited to the past 10 years (2010–2019) using the search terms (including variations), “scrub typhus,” “children,” and “India,” combined with study filters for cohort studies and original research. Acute undifferentiated fever lasting for more than 5 days with hepatomegaly and lymphadenopathy is one of the compatible clinical features of scrub typhus infection and has been reiterated from our analysis. However, in areas of high endemicity such as India, the threshold for diagnosis has to be lowered, as exemplified from our study, where 44% of children were identified within the first 5 days.

Although eschar is indicative of scrub typhus infection, its absence does not rule out the disease. The prevalence of eschar is extremely variable, ranging from < 1 to 92% in the available literature. ^22–24 It was noted in 31% of the children in our study. Eschar is highly evolving and develops at the site of mite bite; it initially begins as a central vesicle with surrounding erythema and overlaying scales and finally heals as necrotic lesion resembling a burn of a cigarette toward the end of the first week. ²⁵ Dark skin tone, uncharacteristic manifestations of eschar, and concealed areas of the body like groin, axillary skin folds, and behind ear lobule are few of the challenges in clinical examination. We compared the clinical features of children with scrub typhus based on the eschar and found a significant association between clinical evidence of lymphadenopathy and eschar positivity. Accordingly, lymphadenopathy may be considered as a clinical pointer toward eschar, and a detailed serial examination is warranted. Also, clinical identification of eschar has diagnostic implications: PCR testing of the eschar swabbing is a rapid and reliable bedside tool in the early phase of the disease. ^26,27

Scrub typhus is, in general, a simple febrile illness with admirable outcome if treated early, as evidenced by nil mortality in our 262 children. This is in concordance with previous studies from India where mortality rates predominantly ranged from 0% to 7.5%. Furthermore, only 25 children (9.5%) from our study developed severe forms of scrub typhus marked by organ dysfunction. This is remarkably low when compared with earlier studies from the same region. Manish et al. ¹⁸ demonstrated an exceedingly high complication rate following scrub typhus infection with high incidences of myocarditis (34%) and acute kidney injury (20%) in their cohort of 35 children.Rarity of the disease in the region, which is not a hilly area, low index of suspicion for early diagnosis in 2011 when the study was conducted, and their small sample size can explain the high complication rate noted by Manish et al. Our results, however, are comparable with the study of 358 children from Chennai, 150 km from our center, by Ganesh et al. ¹¹ who reported a 14.5% complication rate. Furthermore, Ganesh et al. identified elevated aspartate aminotransferase liver enzyme levels and thrombocytopenia as predictors of complication due to scrub typhus. An elevated aspartate aminotransferase liver enzyme level as a determinant of severe scrub typhus has been re-established by our study. Intrahepatic sinusoidal endothelial vasculitis and direct cytopathic liver damage contribute to this elevation. Unexplained transaminitis without jaundice in a febrile child should raise the suspicion for scrub typhus infection in endemic areas. ²⁸ Other predictors of severe scrub typhus such as hypoalbuminemia, hyponatremia, encephalopathy, and leukocytosis proposed by Naryananasamy et al. ²⁹ were not recognized in our study.

Likewise, three children developed hemophagocytic lymphohistiocytosis following scrub typhus in our cohort. Pancytopenia and hyperferritinemia hinted at the diagnosis and intravenous azithromycin without chemotherapy was effective. Data on hemophagocytic lymphohistiocytosis, as a life-threatening complication of scrub typhus in children, are grossly deficient and limited to occasional case series. Jin et al. ³⁰ reported the beneficial outcome in 15 children with scrub typhus associated HLH with intravenous antibiotics (azithromycin or chloramphenicol) without the need for 2004 HLH protocol chemotherapy.

Large sample size, robust back up of data from a single center, and analysis of predictors of severe scrub typhus are the strengths of the study. Our study has few limitations such as its retrospective design and evaluation of only hospitalized children with modest representation of the community.

In conclusion, our study accentuates the varied spectrum of tsutsugamushi infection in children including a 9.5% complication rate. Elevated aspartate aminotransferase enzyme levels is a predictor of severity. Meticulous examination for eschar is merited in children with lymphadenopathy. Zero mortality in our cohort signifies the excellent outcome with judicious first-line antibiotics.