This cohort described the main clinical and laboratory manifestations and treatment used in 57 Brazilian children and adolescents with PIMS-TS who were hospitalized and followed up in five tertiary pediatric rheumatology centers. The most frequent manifestations were fever, gastrointestinal and mucocutaneous. Almost 80% of the patients showed a KD phenotype, either the classic or the incomplete form. All of them presented elevated inflammatory activity markers and most presented altered blood counts, especially anemia and lymphopenia. A laboratory detection of SARS-CoV-2 was found in most patients and after treatment with IVIG and/or glucocorticoids, the patients had an excellent prognosis, with no deaths.
In this study, we used the case definition criteria of the RCPCH as, by their criteria, a laboratory evidence of SARS-CoV-2 and previous contacts with COVID-19 are not mandatory and since not all of our patients had access to laboratory tests for the diagnosis of SARS-CoV-2 [9].
In our cohort, PIMS-TS was observed in all age groups, predominantly between 6–12 years. This finding was also observed in American cohorts and in a systematic review article [19,20,21]. These data confirm one of the main epidemiological differences between PIMS-TS and KD, which is the age range affected [13, 14]. In KD, the vast majority of children are under 5 years of age, with a median of 3 years [16]. Several publications have described a slight predominance of males over females, which was not observed in our sample. In KD, the greater involvement of males is already well established [16].
The two main comorbidities described in the main multisystem inflammatory syndrome registries are obesity and asthma, which were also described in this study, although less frequently [19,20,21,22]. Mortality from acute COVID-19 infection is rare in pediatrics and is associated with the presence of comorbidities, which in PIMS-TS rarely occurs, as it usually affects previously healthy children.
The presence of fever was described in all patients in our cohort, as well as a high frequency of mucocutaneous manifestations, mainly rash, conjunctivitis and hyperemia of the oral mucosa. These are classic findings in KD, and are the main clinical similarities between PIMS-TS and KD [13, 14]. However, as described in other case series from different countries, we also found in our cohort the presence of abdominal manifestations, mainly abdominal pain [4,5,6,7,8]. In some series, some of these patients even underwent an exploratory laparotomy to clarify the abdominal pain [6, 8]. While fever and mucocutaneous manifestations were the main similarities observed between PIMS-TS and KD, the presence of abdominal pain was the main difference between the two diseases, since abdominal pain is rare in KD and is shown to be a frequent manifestation in PIMS-TS [13, 14].
Although both diseases have cardiovascular manifestations, the type and severity of involvement are different between PIMS-TS and KD [13, 14]. As observed in other series, we also observed in our PIMS-TS patients a higher frequency of myocarditis and a lower frequency of coronary artery changes, which are the main feature of KD [8, 19, 20]. Myocardial involvement associated with shock in KD, known as KD shock syndrome, is rare, affecting less than 10% of patients with KD [17]. In other words, in PIMS-TS, a myocardial involvement is common and severe, whereas a coronary artery involvement is mild and transient [23,24,25].
At the beginning of the pandemic, several case reports and series compared PIMS-TS and KD because of the similarities between the two entities [4, 6, 7]. In an initial series of 10 patients in the province of Bergamo, half of the patients met the criteria for the classic form of KD and the other half for the incomplete form [5]. As the pandemic progressed, the frequency of complete KD was described as more frequent, as occurred in our sample [7, 26]. Although most patients met the classification criteria for KD, PIMS-TS has a wide phenotypic range beyond KD: ranging from multiple organ dysfunction with shock and myocarditis to a febrile phenotype, with more than one mild systemic involvement (especially mucocutaneous or gastrointestinal) and elevation of inflammatory markers [6].
All the patients presented elevated inflammatory markers, especially CRP. In addition, most presented increase in ferritin, LDH, and aminotransferases, which supports PIMS-TS as a hyperinflammatory syndrome. Interestingly, although D-dimer was increased in 96% of the patients in our cohort, no thrombotic events were observed. The frequency of thrombotic events ranged from 0 to 7% in some series [20, 27]. In some inflammatory conditions the level of D-dimer was elevated without the occurrence of thrombotic events [28].
The main changes found in the blood count were anemia and lymphopenia. While anemia reflects an inflammatory activity of PIMS-TS, lymphopenia is the main hematologic feature of the SARS-CoV-2 infection [29]. Unlike the classic finding of thrombocytosis found in KD, only 12% of the patients in our cohort showed an elevation of platelets [16]. However, we observed thrombocytopenia in 30% of cases. This dissociation between the frequency of thrombocytosis classically found in KD and the higher frequency of thrombocytopenia in PIMS-TS has also been observed in other case series and also corroborates that KD and PIMS-TS are two distinct clinical entities [5, 6].
We observed in our cohort that three-quarters of the patients showed evidence of infection with SARS-CoV-2, either by RT-PCR or serology. In a recent systematic review, this frequency was 84.7% [22]. As about one fifth of the patients did not undergo a serology, the lowest frequency of evidence of a previous infection by SARS-CoV-2 found may be justified, since serology has a higher positivity rate than RT-PCR in PIMS-TS [30].
Our pediatric intensive care unit admission rate of 68% was slightly lower than that reported in the literature (71%), however, our patients stayed more days in the hospital (10 days) when compared to the literature (7.9 days) [22]. At the beginning of the case series descriptions, most of the patients were only using IVIG due to the overlap of clinical manifestations between KD and PIMS-TS and consequently the extrapolation of IVIG use in KD [4,5,6]. Recently, some guidelines and observational studies have advocated, especially in severe cases, the association of IVIG and glucocorticoids [31,32,33,34,35]. In our cohort, we observed a higher frequency in the combined prescription of IVIG and glucocorticoids. These patients also presented a higher frequency of admission to the pediatric intensive care unit but a similar length of hospital stay when compared to patients who received only IVIG or glucocorticoid. These results were expected since the former group presented a more severe disease with need for intensive care. At the same time, they presented a similar outcome to the other group, as there was no difference in the days of hospitalization.
A mechanism of action of IVIG in these patients was recently described. Peripheral blood samples were collected before and after IVIG infusion. It was observed that patients had a high concentration of IL-1β producing neutrophils before IVIG infusion and after infusion there was a marked reduction of these neutrophils, indicating an important anti-inflammatory effect of IVIG by reducing IL-1β production [36, 37]. Furthermore, an in vitro assay demonstrated that IVIG causes the cell death of these neutrophils [36, 37].
Five patients in our cohort received neither IVIG nor glucocorticoid. These patients had a mild clinical phenotype, with only mucocutaneous and gastrointestinal manifestations. None of them required admission to a pediatric intensive care unit. Patients with mild symptoms do not always require immunomodulatory treatment, only close monitoring [31].
This study corroborates the findings in literature on PIMS-TS to date, through a representative sample of Brazilian patients who were followed by pediatric rheumatologists, with extensive experience in inflammatory diseases. However, it presents some limitations due to the retrospective nature of the data collection, such as the non-standardization in the management and the request of complementary exams; no documentation of a history of exposure to SARS-CoV-2, besides the fact that some patients did not perform one of the laboratory tests to confirm SARS-CoV-2 infection. For this reason, it was not possible to use the CDC [10] or the WHO [11] criteria, and the RCPCH criteria were used for PIMS-TS case definition [9].