Spectrum of thrombotic complications and their outcomes in Chinese children with primary nephrotic syndrome

This is a retrospective study performed in the Department of Paediatrics of Peking University First Hospital of China. The medical records of subjects with NS treated from January 1993 to March 2019 were retrieved from the medical record system and screened according to the discharge diagnosis codes. Subjects aged 1 to 18 years with NS complicated with thrombosis were included. NS was defined by the standard definition (24-h urinary protein  ≥50 mg/kg body weight and serum albumin < 25 g/L) stated in the Chinese Medical Association guidelines. Only subjects with NS who were diagnosed with thrombosis by imaging examinations were included. Subjects with congenital or genetic NS, secondary NS, or IgA nephropathy were excluded. This study was approved by the Ethics Committee of Peking University First Hospital.

The data collected included demographic information, the possible risk factors for thrombosis, such as a recent history (within 1 week before the diagnosis of thrombosis) of vomiting or diarrhoea, infection, renal insufficiency, central venous catheterization, a past history of thrombosis and a family history of thrombosis, and the status of NS at the onset of thrombosis, such as first episode or relapse, steroid sensitive NS (SSNS) or steroid resistant NS (SRNS), and immunosuppressive therapy. A flow chart showing the diagnostic and treatment processes for NS is shown in Fig. 1. SSNS was defined as the absence of oedema and proteinuria on 3 consecutive days within 4 weeks of oral prednisone/prednisolone therapy at dosages of 2 mg/kg/d or 60 mg/m²/d, as stated in the Chinese Medical Association guidelines. If the patient did not achieve complete remission within 4 weeks of prednisone/prednisolone therapy at dosages of 2 mg/kg/d or 60 mg/m²/d, the patient was considered to have SRNS. Complete remission and relapse were defined according to the guidelines [18]. Heavy proteinuria was defined as a urinary protein level ≥ 3+ or 24-h urinary protein level  ≥50 mg/kg, and mild or moderate proteinuria was defined as a urinary protein level of 1+ to 2+ by dipstick or a 24-h urinary protein level of > 150 mg/d and < 50 mg/kg. The information retrieved on thrombosis included the site, symptoms, signs, platelet count, fibrinogen level, plasma D-dimer findings, anti-thrombin-III level, protein C level, protein S level, antiphospholipid antibody level, lupus anticoagulant level, homocysteine level, imaging findings on DUS, CT or MRI and treatment (heparin or low molecular weight heparin with duration, warfarin), thrombosis-associated complications and outcome of thrombosis. The thrombosis-associated complications were defined as death or dysfunction of the corresponding site caused by thrombosis within 3 months from the onset of thrombosis that required a special diagnostic modality to confirm or a specific treatment, such as surgical intervention. The outcome of thrombosis was defined as the status of thrombosis, as determined by imaging methods and the functional status related to the anatomic sites of thrombosis at the time of the last follow-up. The permanent dysfunction of an organ or limb related to thrombosis was defined as a sequela.

SPSS 20.0 software (SPSS, USA) was used for analysis. A descriptive method was used. The data are expressed as the means ± standard derivations or medians (min, max), depending on the normality of the data.

Among the 27 subjects, 14 (51.9%) had SRNS, and 9 had SSNS (Table 1). At the onset of thrombosis, 2 subjects did not undergo therapy with corticosteroid and immunosuppressive agents, 25 subjects underwent therapy with corticosteroids, and 7 subjects underwent additional therapy with immunosuppressants (4 used calcinurin inhibitors, 2 used mycophenolate mofetil, and 1 used cyclophosphamide). Before the onset of thrombosis, 25 subjects had persistent proteinuria lasting 67 ± 56 days [median 49 (4, 200)]. Twenty-three of 27 (85.2%) subjects had heavy proteinuria, 2 had mild or moderate proteinuria, and 2 had trace or negative urinary protein levels. Two subjects developed thrombosis between 6 and 7 days after the remission of NS. The associated risk factors are shown in Table 1. Renal biopsy was performed before the onset of thrombosis in 10 subjects and after the onset of thrombosis in 7 subjects. The renal pathological changes included minimal change disease (MCD) in 7 (41.2%) subjects, membranous nephropathy (MN) in 5 (29.4%) subjects, focal segmental glomerulosclerosis in 3 subjects and mesangial proliferative glomerulonephritis in 2 subjects. Before the diagnosis of thrombosis, 13 subjects underwent prophylactic anticoagulation therapy.

There were 24 subjects with venous thrombosis and 3 with arterial thrombosis. Twenty subjects had thrombosis at a single site, whereas 7 subjects had thrombosis at multiple sites.

Among the sites of venous thrombosis (Table 2), the RV (9/27, 33.3%) was the most common site, followed by the lungs (7/27, 25.9%), DVL (6/27, 22.2%), cerebral venous sinus (CVS) (6/27, 22.2%), inferior vena cava (IVC) (3/27, 11.1%), portal vein (PV) (2/27, 7.4%), superior mesenteric vein (SMV) (2/27, 7.4%) and internal jugular vein (IJV) (1/27, 3.7%). The various manifestations are shown in Table 2 according to the site of thrombosis. The laboratory findings are shown in Table 3.

From the onset of thrombosis until the disappearance of thrombosis, no subjects died of thrombosis, and complications were detected in 4 (4/27, 14.8%) subjects. Intestinal necrosis and intestinal perforation occurred in 1 subject with thrombosis in the PV, SMV and splenic vein. Persistent intracranial hypertension occurred in 1 subject with CVS thrombosis. One subject with PE had an arterial blood oxygen saturation level < 95%; he needed oxygen at 1 L/min from a nasal catheter to maintain an oxygen saturation level > 95%. He had pulmonary hypertension [mean pulmonary arterial pressure of 48.7 mmHg]. Twenty days later, his mean pulmonary arterial pressure decreased to 31.6 mmHg. Lameness occurred in 1 subject with popliteal arterial thrombosis.

When patients had the manifestations shown in Table 2, such as pain or swelling in a specific location or an increased D-dimer level, thrombosis was suspected, and imaging methods such as DUS, MRI and CT were used for the diagnosis of thrombosis. The imaging methods used to diagnose thrombosis and their findings are shown in Table 4, and DUS was the most commonly used method.

After the diagnosis of thrombosis, 2 subjects left the hospital voluntarily, and 25 subjects received standard anticoagulation therapy with heparin or low-molecular-weight heparin first, followed by warfarin for a planned duration of 3–6 months, depending on the sites of thrombosis and the NS status of the subject [17]. Heparin was given in a bolus dose of 75–100 U/kg, the maintenance doses were 10–20 U/kg per hour, and the activated partial thromboplastin time was monitored (target 60–85 s). Low-molecular-weight heparin was given in doses of 100 U/kg every 12 h subcutaneously, and anti-Xa level (target 0.5–1.0 U/ml) was monitored. Warfarin was initially given in a dose of 0.1–0.2 mg/kg (< 5 mg), and the dosage was adjusted according to the international normalized ratio (target 2.0–3.0). Within the follow-up period, the subjects received anticoagulation therapy for 4.4 ± 3.9 months. Seven subjects underwent thrombolysis.

In addition, 1 patient with intestinal perforation underwent laparotomy, end-to-end anastomosis, and an abdominal operation and fully recovered. One patient with persistent intracranial hypertension underwent a lumbar cistern-peritoneal shunt operation, and his condition improved. One patient with popliteal arterial thrombosis failed to show improvement after anticoagulation therapy and thrombolysis. She underwent embolectomy successfully.

The outcomes of thrombosis at different sites are shown in Table 5. In nineteen (70.4%) subjects, the symptoms resolved completely or improved without complications or sequelae. In three subjects who experienced severe complications, including 1 with intestinal perforation, 1 with persistent intracranial hypertension and 1 with pulmonary hypertension, the symptoms resolved completely without sequelae. Three (3/27, 11.1%) subjects developed sequelae; among them, 1 subject with intracranial arterial thrombosis had epilepsy, 1 subject with intracranial arterial thrombosis had an askew mouth (due to facial paralysis), and 1 subject with popliteal arterial thrombosis had lameness. Two subjects were lost to follow-up. Recurrent thrombosis was observed in 1 subject who had CVS thrombosis associated with the relapse of NS.