Delayed Hematological Remission Predicts Poor Renal Outcome in Children with Atypical Hemolytic Uremic Syndrome

Introduction

Hemolytic uremic syndrome (HUS) is an important cause of acute kidney injury (AKI) in children. It is defined by the presence of microangiopathic hemolytic anemia (MAHA) (hemoglobin <10 g/dl, schistocytes ≥2% on peripheral smear, with either elevated lactate dehydrogenase [LDH] >450 IU/l or undetectable haptoglobin), thrombocytopenia (platelet count <150,000/mm³), and AKI. Patients are diagnosed to have atypical hemolytic uremic syndrome (aHUS) when there is no coexisting disease or specific infection identified. aHUS occurs mostly due to either abnormalities of the complement system or due to antibodies to complement factor H (CFH).[1] Overactivity of the alternative pathway can also occur because of gain-of-function mutations of factor B or C3 leading to aHUS. CFH is the most important protein for regulation of the alternate complement pathway, and presence of antibodies against this protein allows uninhibited activity of the complement cascade to occur on the surface of endothelial cells. Atypical HUS secondary to antibodies to CFH constitutes the majority of pediatric aHUS cases reported from India, while CFH mutation was responsible for 24%–27% aHUS cases in USA, Italy, and France.[2-5] Though eculizumab is the drug of choice for aHUS associated with inherited complement deficiencies and anti-factor H antibody–associated HUS, prompt treatment with plasma exchange (PLEX) combined with immunosuppression in the latter has been shown to have a favourable outcome in children when eculizumab is not available.[2,6]

There are very few reports from India regarding children with aHUS. We performed this study to record the clinical features and investigate the factors associated with short-term unfavourable outcomes in children with aHUS at a tertiary care center in India.

Materials and Methods

We extracted data from the case record forms of children with aHUS managed at our center from November 2013 to December 2019 with at least 3 months’ follow-up after discharge from the hospital. We excluded infants and patients who had received the initial treatment after diagnosis at another health facility. We noted the demographic details, presenting features, course in hospital, and details on follow-up after discharge. Laboratory investigations noted at presentation and at the last visit included complete hemogram, blood urea, serum creatinine, LDH, electrolytes, complement level C3, and anti-CFH antibody levels, urinalysis, and urine protein to creatinine ratio (uPCr). Renal biopsy and genetic study reports were noted where available.

Details of treatment with respect to number of plasma infusions (PI), PLEX sessions, use of immunosuppressive agents, and requirement of renal replacement therapy (peritoneal dialysis [PD], hemodialysis [HD]) were noted. The protocol for treatment of aHUS included daily PLEX with 60–75 ml/kg fresh frozen plasma (FFP) for 5–7 days, followed by three sessions per week for 1–2 weeks depending upon the course. Plasma exchange @ 15 ml/kg were instituted for children who were non-oliguric and in whom obtaining a vascular access for catheter insertion was difficult. The infusions were given daily till hematological remission and on alternate days thereafter up to attainment of renal recovery. After the guidelines for management of children with aHUS were published, cyclophosphamide/rituximab infusion and maintenance immunosuppression with corticosteroids, mycophenolate mofetil (MMF), or azathioprine were added to the management of our patients with positive anti-CFH antibodies.[6] The maintenance therapy was continued for a duration of 24 months while monitoring clinical, hematological, and biochemical renal profile. Following discharge from the hospital, these children were called for review every 1–2 months for the first year and every 2–3 months thereafter.

Hypertension was defined as blood pressure (BP) higher than 95^th percentile for age, gender, and height.[7] Standard definitions were used for defining nephrotic range proteinuria. Hematological remission was defined as platelets >150,000/μl, schistocytes <2%, and LDH less than the upper limit of normal on two consecutive days. Relapse was defined as recurrence of anemia with schistocytes ≥2%, elevated LDH, and/or thrombocytopenia (platelets <150,000/μl) with or without AKI, following remission for >2 weeks. Children with normal estimated glomerular filtration rate (eGFR) on the last follow-up visit were compared to those with chronic kidney disease (CKD) stage 2 or more.

Results

Thirty-four children with aHUS were managed during the study period. After excluding three infants and five children who were managed elsewhere, 26 children formed our study cohort.

Discussion

We describe the clinical profile of 26 children with aHUS who were managed with plasma therapy, immunosuppression when anti-factor H antibodies were present, and RRT when indicated. We compared the clinical and laboratory parameters of those whose kidney function returned to normal to those whose eGFR remained low.

A diagnosis of aHUS is reserved for patients with HUS without a coexisting disease or specific infection. It occurs due to overactivation of the alternative complement pathway due to either hereditary mutations in complement genes or autoantibodies against CFH.[9] The presence of autoantibodies directed against CFH resulting in its functional deficiency was first published in 2005.[10] The mean age of presentation in our study cohort was 6.8 ± 1.9 years, majority of whom were anti-CFH antibody positive. Sinha et al.[2] reported data on children with aHUS from multiple centers in India. The mean age of presentation was 8.4 ± 4.1 years in those with anti-CFH antibody–associated aHUS and 5.1 ± 4.9 years in those without the antibodies. Fremeaux-Bacchi et al.[4] described 89 children with aHUS; of these, 56% who had inherited defects of complement experienced disease onset before 2 years of age, while only 11% cases were due to anti-CFH antibodies and developed symptoms between 2 and 12 years of age. Dragon Durey et al.[11] reported that the median age of presentation was 8.5 years and male to female ratio was 2.5:1 in children with anti-CFH antibody–associated aHUS, similar to our finding.

In a small proportion of patients, the complete triad of HUS may be absent at presentation and manifests only during the course of illness. Three of our patients did not have thrombocytopenia at onset. One of them presented with features of chronic glomerulonephritis with intermittent episodes of gross hematuria and periorbital puffiness for 1 month. Another behaved like an SRNS and the third presented with features of rapidly progressing glomerulonephritis (gross hematuria, edema with serum creatinine of 10.5 mg/dl). All three had histopathologic features of microangiopathy on renal biopsy. Approximately 15% children may even have a normal serum creatinine at presentation.[12] The mean serum creatinine at presentation in our cohort was 4.4 ± 3.44 mg/dl, and 69% required RRT. Peritoneal Dialysis was performed in one child due to technical difficulty, and it was the initial modality in two other children who were hemodynamically unstable. Puraswani et al. similarly reported a mean serum creatinine of 5.56 ± 2.98 mg/dl, and 77.3% of children needed some form of RRT for a mean duration of 15 days.[13] The requirement of RRT has been variously reported to be 44%–95% in children with aHUS.[4,10,14,15]

Arterial hypertension is frequent and often severe in patients with HUS due to both oliguria/anuria leading to volume overload and hyperreninemia secondary to renal TMA. The prevalence of hypertension at presentation has been reported to range from 54% to 88%.[5,11,13-17] All our patients were hypertensive at presentation or during the early phase of illness. Geerdink et al. found the overall prevalence of hypertension to be 71% in children with aHUS and 100% in those with positive anti-CFH antibodies.[14] We found 83% (20/24) of children to have nephrotic range proteinuria in contrast to 56.7% as noted in another Indian cohort.[13] Hematuria, either gross/microscopic, was present in 72% of children, similar to that reported by Brockelbank.[18]

The prevalence of anti-CFH antibody–associated HUS has been reported to be 4% and 7% in two studies evaluating both adults and children with aHUS, respectively.[5,10] Studies from outside India which evaluated only pediatric patients with aHUS found a varied prevalence ranging from 9% to 26%. Lee et al. studied a Korean pediatric cohort with aHUS and found 29% of children to be anti-CFH antibody positive.[19] The prevalence of anti-CFH antibody–associated HUS in our patients was 95%, while Puraswani et al. reported that among 781 Indian children with aHUS, 56% had anti-CFH antibodies.[13-15,18,20,21]

Low C3 levels have been reported in 30% of children with aHUS.[4,5,9,14,22] Hypocomplementemia was present in 86% of our patients at presentation, similar to that reported by Uthup et al. who found it to be 88% in children with “diarrhea-negative HUS,” though measurement of anti-CFH antibody levels or mutational analysis to detect defects in complement regulatory genes was not performed by them.[17] Sinha et al. reported that 62% of children with anti-CFH antibodies had low complement levels and there was a negative correlation between the two. In addition, levels of anti-factor H antibodies greater than 8000 AU/ml correlated with poor outcome.[1] We did not attempt to find any correlation between anti-CFH antibody titres and renal outcomes as the assay for antibody was restandardized during the study period (upper limit of normal 100 AU/ml in 2017 and 20 AU/ml before that).

Renal biopsy is indicated in children with aHUS in case of unclear diagnosis or unsatisfactory clinical response to plasma therapy.[23] The former was the reason for biopsy in three of our patients who had normal platelet counts at presentation. Six children underwent a biopsy because of inadequate response to therapy. Three demonstrated features of chronic TMA and six revealed acute activity, and hence, treatment was continued.

Genetic evaluation is not required for urgent therapeutic decisions, but is necessary to establish the etiology, to predict the risk of relapses and risk of progression to end-stage kidney disease (ESKD), for genetic counseling, and for making decisions regarding duration of complement blockade treatment when available and prior to kidney transplantation. The strong correlation between the occurrence of CFH autoantibodies and absence or reduction of CFHR1/CFHR3 in plasma was first reported by Joszi et al.[20] Multiplex ligation-dependent probe amplification in five of our patients with positive CFH antibodies revealed homozygous deletion in both CFHR1 and CFHR3 in four patients. One patient who was CFH antibody positive had a homozygous deletion in CFH gene on NGS. In cohorts reported by Hofer et al., Jozsi et al., and Moore et al., homozygous deletion of CFHR1 had been noted in 80% patients.[15,20,21] As we could not perform MLPA for all our anti-CFH antibody–associated aHUS patients, this does not reflect the true prevalence of genetic mutations in our cohort. Mutations in other complement genes like CFH, CFI, CD46, and C3 variants have been reported in addition to presence of anti-CFH antibodies, thereby suggesting multiple susceptibility factors in some patients to manifest the disease.[21] In addition, up to 12% of patients with aHUS have been found to have various combinations of two or more mutations of CFH, CFI, MCP, C3, CFB, or thrombomodulin.[3,5,22,24]

Eculizumab has been proposed as the first-line treatment, and if it is not available, it is recommended to initiate early PLEX or PI (if PLEX is not possible).[6] We initiated plasma therapy in 25 patients within a median of 6 days from the onset of symptoms (IQR: 3–9 days). The delay was caused because these patients were referred to us after initial evaluation and symptomatic treatment at centers of primary care. Puraswani et al. reported patients from different centers from India; median time to initiation of therapy was 12 days (IQR: 6–24).[13] Sinha et al. showed that delayed initiation of PLEX, 2–3 weeks following onset, predicted adverse outcomes.[2] We too found that children who had CKD stage 2 or more had delayed initiation (4 vs. 14 days) of plasma therapy compared to the group with normal eGFR. The time to achieve hematological remission in our patients was a median of 17 days in contrast to 27 days reported by Puraswani et al.[13]

A relapse proportion of 40%–50% has been reported in children with aHUS.[4,14] Of those with anti-CFH antibodies, 11%–43% showed relapses, with majority showing them in the first 2 years after the first episode.[2,11,13,18,21] Dragon-Durey et al.[10] summarized treatment outcomes in 243 patients with anti-CFH antibody–associated HUS and found that 10.7% patients treated with PLEX and immunosuppression relapsed, compared to 28.7% of those treated with PLEX alone in previously published studies. Sixteen percent (4/26) of our patients relapsed. All four had anti-CHF antibodies; two of these children had been treated with PLEX and steroids, one received PLEX alone (managed before the publication of the guidelines), and one had presented as SRNS (described above). The relapses were treated with PLEX and immunosuppression.

Twenty-two children were followed up for a median of 27 months. An equal number had CKD stage 2 or more and eGFR >90 ml/min/1.73 m² [Table 2]. The mean age at presentation and median duration of follow-up were similar in these two groups, but the mean eGFR at the last visit was significantly different, justifying our objective of comparing these two groups with respect to factors contributing to lower residual eGFR. The time to initiate plasma therapy and the mean number of days to achieve hematological remission were significantly higher in children with CKD stage 2 or more. More children in the latter group needed RRT during the acute illness, but the difference was not statistically significant. Delayed hematological remission and need for prolonged dialysis were associated with adverse outcome, defined as CKD stage 4/5 or death in children with anti-CFH antibody–associated HUS.[2] Anti-CFH antibody–associated aHUS has been associated with a high rate of ESKD (27%–63%).[2,5,10,25] The low proportion of ESKD (5.2%) in our patients is explained by the short duration of follow-up.

Geerdink et al.[14] found chronic hypertension in 54% of their patients with aHUS and 47% had sustained proteinuria. Puraswani et al. reported prevalence of stage 2 hypertension or proteinuria in 42.7% of children with anti-CFH antibody–associated HUS at 3 months of follow-up.[13] The prevalence of hypertension and proteinuria in our cohort was 68% and 31.8%, respectively.

The mortality in our patients was 7.6% (median follow-up: 27 months); one child presented late and the other was withdrawn from treatment by caretakers. The French cohort included 89 children with 11% anti-CFH–associated disease; mortality was 7.8% during a median follow-up period of 45 months.[4] Geerdink et al.[14] managed 45 children with aHUS with plasma therapy (13% anti-CFH antibody associated) and reported a mortality of 9% during a follow-up period of 7.5 years. Puraswani et al.[13] recorded a mortality of 6.5% in children with anti-CHF–associated HUS managed with plasma therapy and immunosuppression. The high mortality in our patients could be attributed to the delay in reaching our center and noncompliance with therapy.

The limitation of our study is its retrospective observational design and small sample size. Since new literature on management of aHUS was emerging, there was heterogeneity in treatment of our patients. Secondly, we did not look for other genetic defects in those in whom we had found one (e.g., we did not perform NGS in those where a copy number variation had been identified by MLPA).

Conclusion

This single-center study showed better outcome in children with anti-factor H antibodies when they were given early therapy with PLEX and immunosuppression, with lower morbidity in terms of developing chronic kidney disease stage 2 or more. Delayed hematological remission identifies those likely to progress to CKD 2 or more. Children with aHUS have persistent hypertension and proteinuria despite attaining hematological remission and irrespective of their CKD stage.