Plasma NGAL as a biomarker of kidney injury and potential predictor of hypoxic brain injury in severe Plasmodium falciparum malaria: insights from India

Abstract

Severe malaria is associated with kidney and brain injury, yet potential mechanisms linking both complications remain understudied. We investigated the associations between kidney and brain injury in a cohort of Indian adults and children with severe Plasmodium falciparum malaria. We found that acute kidney injury (AKI) was prevalent in both adults (64.4%) and children (71.4%). We also found that plasma levels of the structural kidney injury biomarker NGAL strongly associated with AKI severity (p<0.0001) and negatively correlated with whole brain magnetic resonance imaging (MRI) apparent diffusion coefficient (ADC) values in cerebral malaria (r = −0.6, 95% CI: −0.8 to −0.3). Low ADC values indicate cytotoxic edema, a form of hypoxic brain injury mediated by parasite sequestration and inflammation. Severe cytotoxic edema has been shown to be associated with increased mortality in severe malaria. In our cohort there was a 5.5-fold greater risk of this form of brain injury (Prevalence Risk Ratio (PRR) 5.5, 95% CI 2.3 – 13.2) in patients with high NGAL levels (>300 ng/ml). These results suggest that plasma NGAL may play a critical role in structural kidney injury and could serve as a predictive marker for hypoxic brain injury in the context of severe malaria.

INTRODUCTION

Malaria continues to be a major public health threat with an estimated 263 million cases and 597,000 deaths occurring globally in 2023 (1). Outside of sub-Saharan Africa, India has one of the highest burdens of malaria, with Plasmodium falciparum (Pf) and Plasmodium vivax as the dominant species (2), responsible for an estimated 2 million cases and 3500 deaths in 2023 (1).

Malaria infections cause a spectrum of disease ranging from clinically asymptomatic to multi-organ failure and death, with Pf primarily driving the more severe manifestations. The pathophysiology of severe malaria is complex and involves a combination of processes such as sequestration of Pf-infected erythrocytes in the microvasculature, excessive inflammation, localized tissue hypoxia, and increased blood-brain barrier permeability (3). Multiple organ dysfunctions, including acute kidney injury (AKI) and cerebral malaria (CM), are life-threatening complications of severe falciparum malaria (4).

In settings with low and unstable transmission, such as India, severity and mortality rates increase with age due to limited acquired immunity (5). In a prospective observational study across Asia, 36.5% of severe malaria patients aged 50 years and above died, compared to only 6.1% of children under 10 years old (5). This higher mortality is attributed to a greater incidence of multi-organ dysfunction, particularly AKI, in adults (5).

There is increasing evidence that AKI is more prevalent in Pf infection than previously thought (6,7). This suspected underestimation is partly due to the World Health Organization (WHO) severe malaria criteria, which defines ‘renal impairment’ by a serum creatinine threshold of ≥3.0 mg/dL, capturing only advanced stages of AKI (4). The underreporting of kidney injury in children with malaria is supported by studies in Uganda using the Kidney Disease: Improving Global Outcomes (KDIGO) definition for AKI, as well as novel biomarkers like urinary and plasma neutrophil gelatinase-associated lipocalin (NGAL) (8,9). NGAL, a biomarker for structural AKI (10), has shown predictive value in non-malarial settings, including intensive care units (11,12) and chronic kidney disease progression in hypertensive adults (13).

In a meta-analysis of African studies on severe malaria mortality in children, WHO-defined kidney failure and coma were found to be the strongest predictors of death (14). The high prevalence of AKI in severe malaria and the common pathophysiological processes that result in kidney injury and CM give rise to the potential for kidney-brain crosstalk pathogenic pathways (15,16). The innate immune response and resultant cytokine storms, uremic toxin accumulation in the context of impaired renal clearance, endothelial activation, and blood-brain barrier disruption may collectively contribute to increased brain inflammation and swelling (17-20).

While much of the knowledge on CM pathophysiology derives from post-mortem studies (21), advances in magnetic resonance imaging (MRI) in malaria-endemic regions are transforming our understanding of its underlying pathogenic mechanisms, primarily in Pf infection. MRI studies have revealed distinct patterns of cytotoxic and vasogenic edema in Pf patients with severe malaria and CM, using apparent diffusion coefficient (ADC) maps (22-26). Severe malaria patients with pronounced kidney injury showed marked cytotoxic edema resulting from hypoxic injury in the basal ganglia, characterized by lower regional ADC values (23).

The complex interplay between kidney and cerebral pathophysiology in severe malaria requires in-depth investigation. The combination of advanced neuroimaging techniques and biomarkers of kidney injury, such as NGAL, offers unique opportunities to explore this connection. This study aimed to examine the associations between AKI, defined by both KDIGO creatinine-based criteria and plasma NGAL levels, and brain MRI features in a cohort of adult and pediatric patients with severe Pf malaria admitted to hospital in India.

METHODS

Participant recruitment

This study is a secondary data analysis from an ongoing project on severe malaria-associated structural and functional brain changes in adult and pediatric patients in India (27). Adult (aged ≥16 years old) and pediatric (aged <16 years old) patients were recruited at Ispat General Hospital in Rourkela, Odisha, India. Between October 2013 and December 2023, patients presenting with Pf malaria were prospectively enrolled for MRI examinations, alongside blood parameters and clinical characteristics assessments. Eligibility was confirmed by microscopic examination of Giemsa-stained thick and thin blood films and species-specific PCR. Patients with mixed Plasmodium infections or other concomitant infections, such as bacterial or viral, were excluded.

On the day of admission, a full medical history and physical examination were recorded using a standardized clinical form. Blood samples were collected for cell counts, parasite quantification, and biochemical analyses. Patients were categorized as having severe malaria based on WHO criteria for severe falciparum malaria (28). Severe malaria cases were further classified as severe non-cerebral malaria (SNCM) or cerebral malaria (CM) depending on the presence of impaired consciousness.

Clinical care

Patient management followed WHO recommendations (4) and the national drug policy of the Government of India. Severe malaria patients received intravenous artesunate (2.4 mg/kg) immediately upon admission, followed by doses at 12 and 24 hours, and then daily until transitioning to oral artemisinin-based combination therapy. All patients received at least 3 intravenous doses before transitioning.

Magnetic resonance imaging and generation of ADC maps

Brain imaging was carried out on the day of admission using a 1.5T Siemens Symphony MRI scanner (Siemens AG, Erlangen, Germany), following established protocols (26). ADC map-derived whole-brain histograms were generated to differentiate between cytotoxic and vasogenic edema (26). Normalized whole-brain ADC histograms were created, and peak histogram locations corresponding to the most common ADC values in brain tissue were analyzed (29). Similar analyses were conducted for ADC values in the basal ganglia and white matter, regions associated with CM in adults and children, respectively (26) (Supplementary Figure 1).

Evaluation of NGAL levels

Plasma NGAL levels were measured on admission for all participants. Whole blood (4.5 mL) was collected in sodium citrate tubes, centrifuged at 500 x g for 10 minutes, and plasma was stored at −80°C prior to batch analyses. NGAL was quantified using a multiplexed bead-based enzyme-linked immunosorbent assay (Luminex Corp, Austin, USA) via a commercially available customizable assay kit (R&D Systems, Minneapolis, USA, catalogue #LXSAHM). Assays were performed according to the manufacturer protocols, in duplicate with results averaged for analyses, and by personnel blinded to study endpoints.

Assessment of kidney function

AKI was classified using KDIGO criteria based solely on serum creatinine, as urine output data was unavailable. Baseline renal function was not available in this cohort, therefore, estimated baseline creatinine for each patient was back-calculated using the Full Age Spectrum (FAS) – Age equation (30). This approach normalizes serum creatinine based on age and sex and allows for continuity of calculation from pediatric to adult patients.

The FAS-Age was rearranged as illustrated below. Age-based eGFR reference ranges were used which included 120 ml/min/1.73m² for ages 3 to 29, 107 ml/min/1.73m² for ages 30 to 39, 99 ml/min/1.73m² for ages 50 to 59, and 85 ml/min/1.73m² for ages 60 to 69. In those aged under 20, 120 ml/min/1.73m² has been demonstrated as an appropriate estimation when estimating baseline creatinine (9,31). Adult eGFR reference ranges come from the National Kidney Foundation (32), and are also supported by similar values in other large population-based studies (33).Q-values represent age-based median serum creatinine (mg/dL) as previously determined (30).

The original FAS-Age equations (30):

$$\begin{gathered} FAS-eGFR=\frac{107.3}{(SCr÷Q)}(\text{for ages}2\text{to}40\text{years}) \\ FAS-eGFR=\frac{107.3}{(SCr÷Q)}\times {0.988}^{(Age-40)}(\text{for age above}40\text{years}) \end{gathered}$$

These equations were rearranged to back-calculate baseline serum creatinine. The eGFR adjustment ($\times {0.988}^{(Age-40)}$) which accounts for age-related eGFR decline for those aged above 40 years was not included as it was incorporated in the assumed GFR by age, therefore:

$$SCr=\frac{107.3\times Q}{eGFR}$$

AKI stage 1 was defined as a serum creatinine 1.5–1.9 baseline, stage 2 as 2.0–2.9× baseline and stage 3 as ≥3× baseline (34).

Plasma NGAL thresholds were defined as >300 ng/mL for AKI positivity, consistent with prior studies correlating NGAL with KDIGO-based AKI in malaria and other conditions (8,35). Interpretation of plasma NGAL levels and its association with structural kidney injury must be done cautiously, however, due to extra-renal sources of the biomarker including neutrophils and other non-hematopoietic tissues (36).

KDIGO-based AKI categories were combined with plasma NGAL categories to explore the presence of functional and structural kidney injury. The KDIGO − / NGAL + group was indicative of potential subclinical AKI, a phenomenon previously demonstrated in studies of malaria-induced AKI (10) and in pooled analyses of critically ill patients (37). Conversely, the KDIGO − / NGAL − group represented an absence of kidney injury, and the KDIGO + / NGAL + group indicated combined functional and structural AKI.

Statistical analyses

Patient characteristics on admission were summarized as mean ± standard deviation (SD) for normally distributed continuous data or as median and interquartile range (IQR) for non-normal distributions, while categorical variables were reported as counts and percentages (n/N, %).

Group comparisons were performed using t-tests or Wilcoxon rank-sum tests for continuous variables, and Pearson’s Chi-squared or Fisher’s exact tests for categorical variables. For comparisons involving three or more groups, one-way ANOVA or Kruskal-Wallis tests were used, with Bonferroni-corrected pairwise comparisons using t-tests or Wilcoxon rank-sum tests applied when statistical significance was identified. Correlations between continuous variables were assessed using Pearson’s correlation coefficient (r).

Prevalence risk ratios (PRR) were calculated to compare differences in the prevalence of cytotoxic oedema across malaria severity, AKI and NGAL status and Fisher’s exact test used to compare differences in groups.

All statistical analyses were performed using R (version 4.4.1) with RStudio. A two-tailed p-value of ≤0.05 was considered to reflect statistical significance.

RESULTS

Participant demographics and clinical outcomes

A total of 59 patients were included in this study (Table 1). Of these, 14 (23.7%) were children with a median age of 5 years (IQR: 3.8, range: 3–11 years). The adult cohort had a median age of 30 years (IQR: 20, range: 16–65 years). Males predominated in both the pediatric (78.6%) and adult (68.9%) cohorts.

Table 1. Cohort demographics, clinical outcomes, and AKI/NGAL classification by age category and malaria severity.

		Age Category		Malaria Severity
		Children (Age < 16 years)	Adults (Age ≥ 16 years)	SNCM	CM
Total number of patients – n (%)	59 (100.0%)	14 (23.7%)	45 (76.3%)	23 (100.0%)	36 (100.0%)
Age (years) – median (IQR)	28 (22.5)	5 (3.8)	30 (20)	37 (20.5)	25 (26)
Sex – n (%)
Male	42 (71.2%)	11 (78.6%)	31 (68.9%)	19 (82.6%)	23 (63.9%)
Female	17 (28.8%)	3 (21.4%)	14 (31.1%)	4 (17.4%)	13 (36.1%)
Malaria severity – n (%)
SNCM –	23 (39.0%)	0	23 (51.1%)	-	-
CM	36 (61.0%)	14 (100.0%)	22 (48.9%)	-	-
Deaths – n (%)	6 (10.2%)	2 (14.3%)	4 (8.9%)	1 (4.3%)	5 (13.9%)
No AKI	0	0	0	0	0
AKI Stage 1	0	0	0	0	0
AKI Stage 2	1 (14.3%)	1 (33.3%)	0	0	1 (25.0%)
AKI Stage 3	5 (21.7%)	1 (50.0%)	4 (19.0%)	1 (16.7%)	5 (29.4%)
AKI Stages – n (%)
No AKI	20 (33.9%)	4 (28.6%)	16 (35.5%)	12 (52.2%)	8 (22.2%)
1	9 (15.2%)	5 (35.7%)	4 (8.9%)	2 (8.7%)	7 (19.4%)
2	7 (11.9%)	3 (21.4%)	4 (8.9%)	3 (13.0%)	4 (11.1%)
3	23 (39.0%)	2 (14.3%)	21 (46.7%)	6 (26.1%)	17 (47.2%)
NGAL >300 ng/mL – n (%)	18 (30.5%)	1 (7.1%)	17 (3.8%)	6 (26.1%)	12 (33.3%)
NGAL (ng/mL) by AKI Stage – median (IQR)
NGAL (No AKI)	85.2 (79.2)	92.7 (62.5)	82.4 (85.0)	70.5 (47.5)	119.4 (85.5)
NGAL (AKI 1)	102.1 (80.5)	86.0 (88.9)	134.3 (75.7)	128.1 (38.4)	102.1 (101.0)
NGAL (AKI 2)	108.7 (40.5)	102.3 (20.3)	131.0 (59.0)	108.7 (34.2)	108.6 (37.6)
NGAL (AKI 3)	393.4 (266.6)	327.1 (140.6)	393.4 (266.4)	440.3 (159.5)	386.3 (333)

All children were clinically diagnosed with CM. Among adults, there was a relatively even distribution of patients with 51.1% diagnosed with SNCM and 48.9% with CM. 6 patients (10.2%) died during hospital admission, including 2 children and 4 adults. All patients who died had severe AKI (stage 2 or 3) on admission equating to a 20% mortality.

Prevalence of structural and functional AKI

High plasma NGAL levels are strongly associated with severe AKI in adults

Based on KGIDO classification, 66.1% of patients with severe Pf malaria had acute kidney injury (AKI) at admission, with 50.8% classified as severe AKI (Stage 2 or 3). A larger proportion of patients with CM had severe AKI on admission (47.2%) compared to SNCM (26.1%).

Plasma NGAL >300 ng/mL was observed in 18 (30.5%) patients, including only 1 pediatric case. 9 out of 25 (36.0%) adults and 4 out of 5 (80.0%) children with severe AKI (stage 2 or 3) had plasma NGAL levels below this threshold (Figure 1A and 1B). Plasma NGAL levels were significantly higher in patients with AKI Stage 3 compared to all other stages of AKI and those without AKI (p < 0.0001). No significant differences were observed between no AKI and AKI stage 1 and between AKI Stages 1 and 2 (Figure 2).

Figure 1. Distribution of plasma NGAL and creatinine levels by AKI stage in adults and children.

Scatter plots showing individual plasma NGAL and creatinine levels (y-axis) against age (x-axis) for adults and children, categorized by AKI stage according to the KDIGO guidelines. Estimated baseline creatinine levels were back-calculated using the FAS-Age equation. The black dashed line in 1A and 1B represents the NGAL > 300 ng/ml threshold. In 1C and 1D the black dashed line represents the WHO severe malaria guideline criterion for renal impairment (single serum creatinine > 3.0 mg/dL). AKI: Acute Kidney Injury. NGAL: Neutrophil Gelatinase-Associated Lipocalin.

Figure 2. Distribution of plasma NGAL over AKI stages.

Boxplot comparing plasma NGAL levels in patients grouped by AKI status: no AKI, AKI stage 1, stage 2, and stage 3, categorized according to the KDIGO guidelines. P-values were calculated using the Wilcoxon Rank Sum Test and adjusted for multiple pairwise comparisons with the Bonferroni correction method. AKI: Acute Kidney Injury. NGAL: Neutrophil Gelatinase-Associated Lipocalin.

Prevalence of potential sub-clinical AKI using NGAL classification

Combining KDIGO-based AKI and NGAL criteria, 19 patients (32.2%) exhibited no structural or functional kidney injury (KDIGO − / NGAL −). Functional kidney injury (KDIGO + / NGAL −) was present in 22 patients (37.3%). Only 1 patient (1.7%) had potential subclinical AKI (KDIGO − / NGAL +). Nearly a quarter of patients (28.8%) had combined structural and functional kidney injury (KDIGO + / NGAL +) (Figure 3).

Figure 3. Categorization of patients according to AKI and NGAL status.

Patients were classified as KDIGO-positive (+) if admission serum creatinine exceeded 1.5× the estimated baseline. Plasma NGAL was considered positive (+) if levels exceeded 300 ng/ml. AKI: Acute Kidney Injury. NGAL: Neutrophil Gelatinase-Associated Lipocalin.

WHO definition of renal impairment and pediatric AKI

None of the children in our cohort with AKI met the 2014 WHO severe malaria guideline criteria for kidney injury (defined as serum creatinine ≥3.0 mg/dL). Of significance, all 5 children with severe AKI (Stages 2 or 3) as defined by KDIGO standards did not meet this threshold. Similarly, 10 out of 25 (40.0%) adults with severe AKI would not have been classified as having kidney injury under these guidelines (Figure 1C and 1D).

Whole brain ADC categorization and associations with AKI and NGAL

Whole-brain ADC values obtained via MRI on admission were categorized into normal ADC, vasogenic edema (high ADC values), or cytotoxic edema (low ADC values). After excluding two cases due to motion artifacts, 56 patients were analyzed. Normal ADC values were observed in 13 patients (22.4%%), while 28 (48.3%) exhibited features of vasogenic edema and 17 (29.3%) showed cytotoxic edema.

Correlation analyses between whole brain ADC values and creatinine revealed a moderate negative association in adults and children with SNCM (r = −0.5, 95% CI −0.8 to −0.2, p = 0.01) and CM (r = −0.5, 95% CI: −0.7 to −0.3; p < 0.001) (Figure 4A, Supplementary Table 1A).

Figure 4. Quantitative ADC changes and correlation with plasma creatinine and NGAL levels in adults and children by whole brain and most affected region.

Plasma NGAL and creatinine values plotted against whole brain ADC values of adults and children, for basal ganglia ADC in adults and white matter ADC in children. Points plotted represent patients and their malaria severity diagnosis. Area shaded in red represents ADC range in keeping with cytotoxic edema, green area represents normal ADC range and blue zone represents vasogenic edema. Black lines represent best-fit linear regression line calculated using Ordinary Least Squares. ADC: Apparent Diffusion Coefficient. NGAL: Neutrophil Gelatinase-Associated Lipocalin.

Plasma NGAL also negatively correlated with whole-brain ADC values in adults and children with SNCM (r = −0.5, 95% CI −0.8 to −0.08, p = 0.02) and CM (r = −0.6, 95% CI: −0.8 to −0.3; p < 0.001) (Figure 4B, Supplementary Table 1B).

In the basal ganglia for adults with SNCM, similar moderate negative correlation was demonstrated between both creatinine (r = −0.5, 95% CI: −0.8 to −0.2; p = 0.02) and NGAL (r = −0.5, 95% CI −0.7 to −0.04, p = 0.03) (Figure 4C and 4D, Supplementary Table 1A and 1B).

No correlation was observed between creatinine or NGAL and ADC values in the white matter of children with CM (Figure 4E and Figure 4F, Supplementary Table 1A and 1B).

NGAL > 300 ng/ml is associated with greater prevalence of cytotoxic edema

In this cross-sectional study cytotoxic oedema was a common finding in our cohort of patients with severe malaria (30.4%). Further analyses compared the prevalence of low ADC values in whole brain MRI between different exposure groups. These groups were categorized by malaria severity, AKI and NGAL status and included both adults and children. Patients with a plasma NGAL >300 ng/ml were found to be 5.5 times more likely to have cytotoxic edema when compared to lower NGAL values (PRR 5.5, 95% CI 2.3 – 13.2, p < 0.001). AKI status was also found to be associated with greater prevalence of cytotoxic edema (PRR 7.0, 95% CI 1.0 – 48.5, p = 0.009). However, the lower limit of the 95% CI equaling 1.0 and broad confidence intervals means conclusions cannot be drawn from this result. Malaria severity was not significantly associated with greater prevalence of cytotoxic edema in this cohort (PRR 1.9; 95% CI 0.7 - 5.2) (Table 2).

Table 2. The prevalence risk of cytotoxic edema on MRI measured by malaria severity, AKI and NGAL status.

Prevalence risk ratio (PRR) represents the ratio of the probability of cytotoxic edema (determined by whole brain MRI ADC values) occurring in different exposure groups (malaria severity, AKI status and NGAL status). PRR value of greater than 1 indicates a higher prevalence of cytotoxic edema in the exposed group relative to reference (Ref) group. PRR of 1 indicates no difference between the groups. PRR is calculated by dividing the prevalence in the exposed group by the prevalence in the unexposed group and p-values calculated using Fisher’s exact test.

	Cytotoxic edema N (%)	No cytotoxic edema N (%)	PRR (95% CI)	p-value
Malaria Severity
SNCM	4 (19.0%)	17 (81.0%)	Ref	Ref
CM	13 (37.1%)	22 (62.9%)	1.95 (0.73 – 5.20)	0.23
AKI status
No AKI	1 (5.9%)	16 (94.1%)	Ref	Ref
AKI	16 (41.0%)	23 (59.0%)	7.0 (1.0 – 48.5)	0.011
NGAL status
NGAL ≤ 300 ng/ml	5 (12.8%)	34 (87.2%)	Ref	Ref
NGAL > 300 ng/ml	12 (70.6%)	5 (29.4%)	5.5 (2.3 – 13.2)	< 0.001

DISCUSSION

This study sheds light on the associations between kidney injury and neuroimaging and offers novel insights into the manifestations of AKI and plasma NGAL levels in a cohort of Indian adults and children hospitalized with severe Pf infection.

Creatinine-defined functional AKI was notably prevalent in this cohort and severe AKI had a 20.0% mortality rate. Nearly two-thirds (66.1%) of all patients with severe malaria presented with AKI, including 64.4% of adults and 71.4% of children. These figures are higher than those reported in African cohorts, where 24–59% of children with severe malaria develop AKI (7). Our cohort is limited to children with CM, however the high prevalence of AKI in Indian children may also reflect regional differences in malaria transmission dynamics, underlying immunity, and host genetic factors. Estimating baseline creatinine in this population was also limited by the lack of robust evidence for population and age-based normal eGFR values in India and reference ranges used in this study were established in a Caucasian population (32,33).

Our results show significant limitations in the current WHO severe malaria guideline for AKI diagnosis, which relies on a single serum creatinine threshold (≥3.0 mg/dL). This approach would have missed all severe AKI (stage 2 and 3) in children and 40.0% of severe AKI in adults in our cohort and further supports the argument that the WHO guideline is inadequate in the assessment of kidney injury, particularly in pediatric severe malaria (6). This highlights the need for revised, age- and potentially region-specific criteria for AKI assessment in the context of malaria infection.

The use of plasma NGAL as an early predictive tool for AKI has been demonstrated in a variety of contexts including malaria (10,11,38), and its threshold-based sensitivity has been shown to vary between adults and children (39). Despite the high prevalence of functional AKI in our cohort, only 64% of adults with severe functional AKI (stage 2 or 3) exceeded the NGAL threshold of 300 ng/mL considered high-risk for structural kidney injury. This suggests that the established cut-off may be less sensitive in populations with a high AKI burden. Plasma NGAL >300 ng/mL was significantly associated with severe AKI (stage 3) in our patients, indicating its potential as a diagnostic marker in severe disease, but a lower threshold would likely improve its clinical utility as a predictive biomarker. A recent study in children identified a urine NGAL cut-off of 125ng/ml as the best predictor of severe AKI in critically ill children (40), however the ideal cut-off in malaria requires larger validation studies with longitudinal follow-up to establish its predictive power for AKI.

Importantly, plasma NGAL is influenced by systemic conditions like sepsis (41-43) and inflammation (41,44), complicating its interpretation in malaria. As an acute-phase protein, NGAL is upregulated in neutrophils by inflammatory cytokines, a process amplified in severe falciparum malaria (3). This systemic activation limits the specificity of plasma NGAL as a marker of structural kidney injury, as observed in similar contexts like intensive care (43) and cardiac surgery (44).

Despite these limitations, our study highlights a robust association between plasma NGAL and hypoxic brain injury in severe malaria, measured via MRI-derived ADC values. NGAL was shown to negatively correlate with ADC values, particularly in CM cases. Significantly, the increased risk of cerebral cytotoxic edema in patients with high levels of plasma NGAL (>300 ng/mL) (PRR 5.5, 95% 2.3 – 13.2) underscores a potential link between structural kidney injury, inflammation and hypoxia-mediated brain injury in severe malaria. NGAL has been shown to be associated with other hypoxic cerebral insults such as ischemic stroke (45) and the use of NGAL as a hypoxic brain injury biomarker warrants further investigation, particularly in malaria-endemic resource-limited settings where high-resolution brain imaging may not be available.

The association between NGAL and cerebral hypoxic lesions suggests a shared inflammatory pathway driving kidney and brain injury in malaria. The higher incidence of NGAL >300 ng/ml in our adult cohort compared to children could be explained by the strong association with severe cerebral hypoxic lesions in adults compared to children as demonstrated elsewhere (26). AKI and cerebral impairments are mediated not only by parasite sequestration but also by inflammatory cascades, including cytokine induction, complement activation, and endothelial and immune cell responses (3,15,18,19,46). The role of NGAL as an inflammatory mediator (41) likely amplifies these processes, positioning it as a central player in the pathophysiology of malaria-associated brain injury.

While associations between creatinine levels and brain injury were less robust in this study, negative correlations with ADC values were demonstrated and a larger study may be able to refine estimates of the risk of cytotoxic edema in the context of functional AKI. Such research may further illustrate potential kidney-brain crosstalk mechanisms in severe falciparum malaria, where kidney injury exacerbates cerebral injury, although other possibilities have been explored elsewhere (15).

Together, these findings emphasize the dual role of NGAL as a biomarker of severe AKI and potential predictor of hypoxic brain injury in severe malaria. While its inflammatory role complicates its use as a specific indicator of kidney injury, its strong association with neuroimaging findings suggests broader utility in assessing hypoxia-driven cerebral changes. Future research should focus on elucidating the mechanisms underlying the longitudinal effects of NGAL on brain edema and to understand its role as a marker of AKI through the measurement of urine levels of NGAL. Exploration of therapeutic strategies targeting NGAL to mitigate kidney-brain injury in severe malaria is also warranted.