Analysis of the risk factors and clinical features of Mycoplasma pneumoniae pneumonia with embolism in children: a retrospective study

Mycoplasma pneumoniae pneumonia (MPP) is a prevalent and self-limiting disease in the community-acquired pneumonia in children [1]. However, in addition to respiratory manifestations, it may also develop extra-pulmonary complications, including the cardiovascular system, digestive organ, neurological, dermatological manifestations, etc [2]. Embolism is one of the uncommon extra-respiratory manifestations that has been reported in previous studies [3, 4]. It can present in nearly any part of the body, specifically in the lungs, then in the limbs [3]. Embolism is prone to severe sequelae and even life-threatening if not timely diagnosed and treated. In this study, we retrospectively analysed 16 children in MPP with embolism received in our hospital between January 2010 and December 2021. Additionally, we compared basic information, clinical manifestations and imaging results between 16 children in MPP with embolism as well as 32 children with MPP who had a clinical suspicion of embolism but negative radiological results. We aimed to explore the related factors predicting MPP with embolism that help clinicians in early detection and treatment.

Thrombosis has high morbidity and mortality, however, it is rare in children, specifically associated with infection. Notably, case reports and studies have confirmed the presence of embolism among patients with MPP [3, 6, 7]. MPP usually occurs in previous healthy school-age children. A few mild children may exhibit atypical clinical symptoms and are easily to misdiagnosed. In recent years, embolism caused by MPP has attracted more and more attention of pediatricians. In this study, we explore the clinical characteristics and risk factors of embolism in children with MPP by analyzing clinical information. Embolism induced by MPP may occur in any part of the body. Its location has previously been reported in the lungs, heart, brain, neck, abdomen and limb [3], as shown in our study(Table 1). In some patients, embolism can even develope in multiple parts. Herein, three patients (12%) had more than one distribution of embolism. One study [8] reported that the development of the cerebral infarction, abdominal organs and extremities takes 2 days to 3 weeks, 1 week to 1 month and 1–2 weeks respectively. In another study on cardiac embolism among children [9] caused by Mycoplasma pneumoniae infection, the average diagnostic time of 10 children was 13.3 days which is quite close to what we found in this research. Besides, the average diagnostic time for pulmonary embolism, cardiac embolism, cerebral embolism, splenic embolism and limb embolism were 13.4 days, 8.7 days, 15 days, 14 days and 10 days, respectively. Pulmonary embolism in adults is often characterized by chest pain, chest tightness and dyspnea, however, the symptoms of children are often atypical and easy to be ignored[10]. Sheng et al.[11] showed that all the children in MPP with embolism present fever and cough, whereas 5 (71.43%) children developed dyspnea. In our study, all children with pulmonary embolism had fever and cough. Two children (10%) presented shortness of breath and three children (30%) developed pain in various parts of the body. Only one child developed dyspnea, whereas none had hemoptysis. Children with cardiac embolism may have clinical manifestations, including chest pain, shortness of breath or complicated with pleural effusion, but without specificity [9]. In our study, children with cardiac embolism not only had shortness of breath, but also developed hypoxemia and neck pain. Children with cerebral embolism may manifest neurological symptoms, including disorder of consciousness and unilateral limb weakness [3]. In our study, limb weakness occurred in 3 children (100%) with cerebral embolism, whereas one of them had positive pathological signs and dyspnea respectively. Children with splenic embolism had left upper abdomen and periumbilical abdominal pain with vomiting (Table 1). The mechanism of embolism development in patients with MPP remains unclear. So far, relevant animal studies have not been reported. The mechanism may include the following aspects [4, 12‐15]: i. MP can cause microvascular endothelial injury and release cytokines, resulting in local vasculitis and thrombotic vascular occlusion without systemic hypercoagulability. ii. MP infection promotes the body to produce autoantibodies such as anticardiolipin antibodies (ACA), β2-glycoprotein antibodies or lupus anticoagulant antibodies, and then form immune complexes, resulting in the injury of respiratory tract and other organs outside the lungs, or MP activates complement and then activates the coagulation system which may result in thrombotic vessel occlusion. Antinuclear antibody test was performed in 6 children and two children were positive (both titers were 1:100). Two children were examined for ACA and one of them had increased ACA-IgA. Protein C activity and protein S activity were performed in 6 children. Protein S activity decreased in both children, whereas protein C activity increased in one child (Table 1). After treatment, all the children returned to normal. Elsewhere, Liu et.al found that protein S activity was low in 5.1% (2/39) of patients and protein C activity was normal. ANA was positive in 51.2% (22/43) of patients [3]. In our study, half of the pulmonary thrombosis patients occurred on the same side as the pulmonary consolidation, further confirming the hypothesis that isolated arterial thrombosis may be in situ thrombosis [3]. D-dimer is a specific degradation product of cross-linked fibrin that primarily reflects blood hypercoagulability, intravascular thrombosis and secondary fibrinolysis [16]. One study has shown that D dimer can be recognized as an indicator to evaluate the severity of MPP and the incidence of extrapulmonary complications increased with the increase of D-dimer level [17]. D dimer significantly increased in the past reported cases of MPP with embolism [11, 18‐20]. According to Liu et al., the average D-dimer in 10 children with cardiac embolism was 12.685 mg/L [9]. Sheng et al. found that the average D-dimer in 7 children with pulmonary embolism was 3.626 mg/L [11]. Furthermore, Liu et al.[3] found that D-dimer > 11.1 mg/L (even > 5.0 mg/L) would help in the early diagnosis of thrombosis in MPP children. However, D dimer has a high sensitivity but poor specificity, which can be increased in several diseases [4]. The results of our case–control study showed that D-dimer (closest to CTA/MRA) > 3.55 mg/L, pulmonary consolidation (⩾ 2/3 lobe) and pleural effusion were independent risk factors for embolism in children with MPP. This further confirms that the severity of MPP is the most strongly associated risk factor for MP-associated thrombosis [3]. The American Society of Hematology guideline panel developed detailed guidelines for the treatment of pediatric venous thromboembolism in 2020 [21]. The guidelines suggested ① using direct oral anticoagulants (DOACs) over vitamin K antagonists (VKAs) for patients with DVT and/or pulmonary embolism (PE), however, it may not apply to certain patients with such as renal insufficiency (creatinine clearance < 30 mL/min), moderate to severe liver disease and orantiphospholipid syndrome et.al; ② using thrombolytic therapy followed by anticoagulation over anticoagulation alone in patients with PE and hemodynamic compromise; ③ for patients with PE with echocardiography and/or biomarkers compatible with right ventricular dysfunction but without hemodynamic compromise (submassive PE), using anticoagulation alone over the routine use of thrombolysis in addition to anticoagulation (using systemic thrombolysis over catheter-directed thrombolysis); ④ using a shorter course of anticoagulation for primary treatment (3–6 months) over a longer course of anticoagulation for primary treatment (6–12 months) for primary treatment of patients with DVT and/or PE. In our hospital, all the children were applied LMWH to anticoagulate for 2 weeks. Two children were used urokinase for thrombolysis for 3 to 5 days. This study also has some limitations. Firstly, it was a retrospective study and some data were missing. Secondly, the sample size in our study is small and we need to further carry out large sample prospective studies. In addition, relevant animal studies need to be established to study the mechanism of embolism development in patients with MPP.

In conclusion, children with embolism caused by MPP have higher D-dimer level and severe imaging findings. D-dimer (closest to CTA/MRA) > 3.55 mg/L, pulmonary consolidation (⩾ 2/3 lobe) and pleural effusion were independent risk factors for embolism in children with MPP. Early diagnosis and anticoagulant therapy can effectively reduce complications and mortality.