Discussion:
Our study is one of the few trying to elaborate various complement abnormalities, presentation and outcome in pediatric aHUS, albeit in the absence of targeted therapy. While a myriad of mutations have been documented, yet the whole picture is still not as vivid as we want it to be. Besides, the phenotype association in different mutations is something that we need to understand further.
In our cohort, the median age of presentation was 49 months which was different from other studies where it is around 24 months [7-10].The reason for this difference could be that previous studies included more of infection associated HUS, especially STEC HUS, while our study population had included only aHUS. However, the possibility remains that in India, HUS is more common in older age group. A retrospective study by Aditi et al. in children with anti-FH antibody HUS observed that mean age was 8.4 ± 4.1 years [11].
It was a male predominant study similar to that done by Srivastava et al. [7]. However, a lower sex ratio was noted in western countries with a ratio of 0.9 [12]. This difference could be due to the male predominant society, though the possibility that boys have a more propensity for these abnormalities remains.
Oliguria was the most common presentation (71.4%) followed by progressive pallor (48.6%) as HUS itself is a triad of hemolytic anemia, thrombocytopenia and AKI. One-third children had loose stools (1 with bloody diarrhea) as prodromal feature suggesting, loose stools as a prodromal feature can also be noted in aHUS as shown in previous studies [11,13]. Seizure was the commonest extra-renal presentation similar to that observed by Hughes et al. [8]. Neurological involvement might be due to disease as such or can be secondary due to sequel of disease like azotemia, electrolyte imbalances and posterior reversible encephalopathy due to hypertension.
Prevalence of acute hypertension in our cohort was similar to that observed by Hughes et al. They documented that 61% had hypertension. However, other studies showed an incidence of 15–30% [7, 14]. This may be due to the fact that STEC HUS is known to have less severe manifestations. Hypertension in HUS is mainly due to pre-glomerular vessel involvement which causes a break in renin angiotensin axis and result in hyperreninemic state aggravating hypertension [15].
Severe anemia was noted initially in half, however, because of on-going hemolysis, 88% subsequently had severe anemia. Similar haemoglobin levels were noted in other studies where children were assessed retrospectively after being diagnosed as HUS [8,9]. Anemia in HUS is mainly due to intravascular hemolysis caused by stress of constricted small vessels and microthrombi formed due to endothelial damage. In our study, there was no correlation between levels of haemoglobin and outcome, though it was not powered for the same. Thus, further larger studies are required. Two-third children had moderate to many schistocytes in peripheral smear suggesting severe hemolysis. Studies have been published for standardization of schistocyte counting and reporting, that may help in decreasing the observer bias [16].
Severe thrombocytopenia was noted in one-third of children at admission but there was a two-fold decrease during the course. Similar results were noted in other studies [7-9]. This is due to excessive consumption of platelets in endothelial damage induced fibrin and platelet thrombi formation.
In our study, 80% children presented with AKI stage 3. Staging of AKI helps in prognostication and assessment of outcome. Although 93% children had AKI stage 3 during hospital course, 14 children showed recovery.
Proteinuria was found consistently in studies where urine analysis was performed [16]. It can result due to podocyte dysfunction caused by activated alternate complement pathway. Interestingly, nephrotic range proteinuria was noted in a higher proportion (34.3%) in our cohort compared to other studies [7]. Hematuria was noted in 60% children which was consistent with other studies [9]. It is due to complement mediated damage to endothelium of glomerular arterioles. Hematuria and proteinuria suggests glomerular involvement. Doorn et al, reported a child who was initially treated as MPGN and was later diagnosed to have HUS because of recurrent relapses [17].
We observed that C3 was low in 57.1% and C4 in 25.7 % of our study cases. Kaplan et al., and Stuthlinger et al., [18, 19] have described hypo-complementemia in HUS. They explained that low C3 is due to intravascular activation of classical complement pathway. This might also explain low C4 levels, as it is part of classical pathway. Similar data of low C4 levels were noted in a study by Song et al., where he assessed C4 levels in anti-FH antibody associated HUS children [20]. Both C3 and C4 can be low in SLE which is a disease affecting classical pathway. All our 6 children who had low C3 and C4 were screened with ANA which was negative. Therefore, it maybe plausible that the classical pathway is activated in aHUS.
Further, three-fourth of the study population had abnormalities in at least one component of complement system. Anti-FH antibody HUS was found in 15 children (44%). This is different to a European study where only 22% children had anti-FH associated HUS [21]. However, it is similar from studies from India where it forms the predominant group [10].In our study, low CD46 expression which constituted 26.7% of study population was more compared to that by Khandelwal et al where 12.7% had CD46 defect [22]. However, this has to be confirmed with genetic mutation studies. Interestingly, three children in this sub-group also had elevated anti-FH levels which had been previously described in literature [22].
Treatment of HUS is based on etiology. In our study, children with suspicion of aHUS based on clinical presentation and investigations, especially antibody associated HUS, underwent plasma therapy if amenable. In India, as per literature, anti- FH antibody HUS is more common than other forms of atypical as well as STEC associated HUS [10]. Hence empirical treatment with PEX and immunotherapy was offered.
Almost all children who underwent PEX were initiated with immunosuppressive therapy which included methylprednisolone followed by oral steroids and cyclophosphamide. It has been found that PEX followed by immunosuppression had a better renal outcome than immunosuppression alone in case of rapidly progressive glomerulonephritis [23]. Sana et al., observed that in anti-FH HUS too, early immunosuppression using cyclophosphamide pulse therapy along with PEX caused sustained remission, decreasing need of maintenance therapy [24].
Previous literature described a lower mortality rate of 25% after increased utility of plasma therapy in HUS [25, 26] and a French cohort reported only in 8.6% [27]. However, overall outcome of our cohort was worse. This could be due to many factors, but etiology, poor socio-economic status, affordability for PEX and non-availability of eculizumab could be important reasons. However, further larger studies are needed to assess prognostic factors.
Age of onset of illness is an important predictor for outcome of disease. Previous studies suggest that poor outcome is noted in children of younger age when compared to toddlers and school going age group [11]. Similarly, our study also showed a significant poorer outcome in younger children, especially in infants (p value <0.01). This age dependent outcome might be due to multiple factors like greater prevalence of familial HUS, difficult in amenability to treatment modalities like hemodialysis, PEX. Hence children with HUS presenting at less than 2 years needs to be worked up and diagnosed early. Further, there is a need for improving targeted therapy.
Conclusion: We observed that aHUS was more common in boys, with hypertension as a common presentation. While various complement abnormalities were detected, anti-FH antibody and low MCP expression seems to be common. Remission was seen only in half of the patients. We recommend a comprehensive complement analysis in pediatric aHUS as it has therapeutic implications.