Long Shadows

AKI affects a growing number of children and neonates and is associated with high morbidity, mortality, and health care expenditure.¹ Implementation of consensus criteria to diagnose and define the severity of pediatric AKI has enabled quantification of its incidence and early outcomes across multiple clinical contexts, including critical illness, cardiac surgery, and neonates. Although previously considered to be a self-limiting and reversible process, the long-term sequelae from AKI are now well recognized. In adults, significantly higher risks of CKD, kidney failure, cardiovascular disease, and death have been reported after an episode of AKI.^2,3 Similar risks have been identified in high-risk pediatric cohorts, such as critically ill children and children requiring acute KRT.^4,5

In this issue of JASN, Robinson et al. present the findings of a large cohort study on the long-term outcomes of AKI in hospitalized children who did not require acute KRT.⁶ They followed 4173 pediatric AKI survivors (without a history of CKD, AKI, or KRT) and 16,337 matched non-AKI comparators for a median of 9.7 years to determine their subsequent incidence of major adverse kidney events (incident CKD, KRT, and all-cause mortality), hypertension, or recurrent AKI. Compared with non-AKI controls, children with AKI had a four-fold higher risk of developing a major adverse kidney event, which was predominantly driven by a higher risk of CKD (eight-fold higher). They were also at higher risk of kidney failure, hypertension, and subsequent AKI, but not all-cause mortality.

This work fills an important gap in the literature because previous studies on the long-term outcomes of pediatric AKI have largely focused on high-risk groups or followed patients for insufficiently long periods. Although Robinson et al. used administrative coding data rather than consensus definitions for AKI and CKD, their large province-wide cohort focuses on children who did not require acute KRT, and a 10-year follow-up period with minimal attrition contributes important data to the field. Importantly, these data also complement their previous analyses of the outcomes of children who required acute KRT.⁵ It is concerning to note that the incidence of adverse outcomes in children who did not require acute KRT was almost identical to those who did.

Although logistically challenging, ensuring a sufficiently long period of follow-up is critical when examining the incidence of post-AKI outcomes such as CKD. After an episode of AKI, hyperfiltration of residual nephrons may mask significant nephron loss, resulting in return of serum creatinine and eGFR to pre-AKI levels. Because survivors of childhood AKI are more likely than their adult counterparts to have higher kidney functional reserve, they often have greater capacity to augment filtration capacity in response to physiological stress. For AKI survivors, this means that CKD may not manifest for years or even decades after the initial insult. Indeed, one third of children in the study by Robinson et al. who eventually developed CKD did so between 1 and 5 years after AKI. Whether distant CKD is a direct or indirect effect of AKI is uncertain, as is the contribution of shared risk factors to the subsequent development of CKD.

Comparison of this work with the adult literature highlights several key differences in the long-term outcomes of AKI between children and adults. First, major adverse kidney events after childhood AKI are predominantly driven by incident CKD rather than death, which contrasts with the adult literature.² This has important therapeutic implications because early detection of CKD could allow timely implementation of management strategies to prevent subsequent disease progression. Second, although the absolute incidence of CKD after AKI is considerably lower in children than in adults (17.4 versus 177.6 per 1000 person-years), the relative risk is much higher (hazard ratio [HR] 7.90 versus HR 2.67). This is likely explained by the lower overall incidence of CKD in the general pediatric population, as well as by the irreversible nephron loss and longitudinal exposure to CKD risk factors (e.g., hypertension, recurrent AKI) experienced by many children with AKI. Finally, the cohorts with the strongest association between AKI and adverse outcomes in adult studies (e.g., cardiac surgery, critically ill) demonstrated the weakest association in the study by Robinson et al. This may reflect differences in the nature of the inciting events, illness severities, or the relative contribution to creatinine elevation by functional and structural changes between pediatric and adult cohorts.

This study is an important step forward in the pediatric AKI literature, but more work is needed to generate robust estimates of the long-term outcomes of childhood AKI to guide clinical care. Large-scale studies using consensus definitions for AKI and CKD are likely to identify a far greater population at risk. Low frequency of serum creatinine testing after childhood AKI is a major weakness of retrospective studies because it may bias the ascertainment of CKD, leading to underestimation of its true incidence. Moreover, phenotyping of patients with AKI according to shared characteristics (e.g., AKI duration or severity) has been reported to have important prognostic implications in adult cohorts, and the same is likely in children. Next, identification of prognostic biomarkers and the development of pediatric-specific risk prediction scores could be used for patient stratification. Such risk scores should consider prenatal indicators of nephron endowment, including gestational age, birth weight, intrauterine growth restriction, and preeclampsia.⁷ Postnatal exposure to nephrotoxins, sepsis, surgical interventions, and previous episodes of AKI are also relevant. In the future, novel genetic risk factors of both AKI and the AKI to CKD transition may inform the identification of AKI endotypes, which share pathophysiological mechanisms, clinical trajectories, long-term outcomes, and, potentially, response to therapies.

Accepting that children affected by AKI have a higher lifetime risk of CKD necessitates a plan for follow-up. The optimal timing, duration, and provider of follow-up is controversial and will likely differ between health systems and health services. The low workforce density of pediatric nephrologists makes it impractical for all children to be followed by specialist services.⁸ Rather, such services are likely best reserved for children at highest risk of progressive CKD on the basis of premorbid kidney function, severity of AKI, and degree of recovery. Conversely, AKI survivors who identified as being at low risk of adverse sequelae may be cared for by their pediatrician or family doctor. To ensure comprehensive care for all children regardless of practitioner, it is imperative that we develop detailed recommendations specifying appropriate intervals for review and outlining the key components of their clinical assessment such as height, weight, BP, urinalysis for proteinuria, and kidney function assessment. A holistic approach that captures developmental, social, and educational factors is especially important in the pediatric setting.

Eventually, it is hoped that post-AKI follow-up care will serve the additional purpose of implementing evidence-based therapies to mitigate the AKI to CKD transition. Currently, no such therapies have shown efficacy in the adult or pediatric literature. However, potential candidate therapies include established treatments such as renin-angiotensin system inhibitors; newer therapies such as sodium-glucose cotransporter-2 inhibitors and finerenone; and management strategies that focus on diet, exercise, and overall quality of life. The inclusion of children in trials testing novel interventions to slow kidney disease progression after AKI is essential to generate evidence-based recommendations. Children are too often excluded from clinical trials because of ethical concerns, a perception that their inclusion will affect trial design, or concerns about drug formulation, resulting in a significant evidence gap in this population. Trials should be codesigned with patients and families, ensuring the development of patient-centered models of care tailored to children who may be at risk of CKD but who are not currently unwell.⁹

In conclusion, these findings confirm the poor long-term outcomes of AKI in hospitalized children who do not require acute KRT. Given the significant health and economic effect of AKI, and the decades-long shadow cast by childhood AKI, there is the potential to significantly improve the health of the population by identifying those at risk early in life and generating an evidence base for their follow-up and management.¹⁰

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E845.

Funding

None.

Author Contributions

Conceptualization: Catherine Quinlan, Emily J. See.

Writing – original draft: Catherine Quinlan, Emily J. See.

Writing – review & editing: Catherine Quinlan, Emily J. See.