Urinary Biomarkers of Kidney Diseases in HIV-infected children

Abstract

A significant number of children infected with the HIV-1 virus all over the world are at risk of developing renal diseases that could have a significant impact on their treatment and quality of life. It is necessary to identify children undergoing the early stages of these renal diseases, as well as the potential renal toxicity that could be caused by antiretroviral drugs, in order to prevent the development of cardiovascular complications and chronic renal failure. This article describes the most common renal diseases seen in HIV-infected children, as well as the value and limitations of the clinical markers that are currently being used to monitor their renal function and histological damage in a non-invasive manner. In addition, we discuss the progress made during the last 10 years in the discovery and validation of new renal biomarkers for HIV-infected children and young adults. Although significant progress has been made during the early phases of the biomarkers discovery, more work remains to be done to validate the new biomarkers in a large number of patients. The future looks promising, however, the new knowledge needs to be integrated and validated in the context of the clinical environment where these children are living.

The Joint United Nations Program on HIV/Acquired Immune Deficiency Syndrome (UNAIDS) estimated that there were approximately 3.2 million children under 15 years of age living with HIV around the world at the end of 2013, and 240,000 children became newly infected in 2013 (Gap Report, July 2014). Approximately 90% of these children live in Sub-Saharan Africa, and the majority of them (~70%) are not receiving appropriate antiretroviral therapy (ART). During the early years of the AIDS epidemic, approximately 40% of all HIV infected children experienced renal complications leading to poor growth, accelerated progression of AIDS, and/or premature death [1, 2]. Therefore a large number of HIV-infected children are at high risk of developing renal diseases, and it is necessary to identify these patients during the early stages of renal injury to prevent the development of chronic renal failure [1, 2]. This article discusses why it is important to develop biomarkers of renal diseases specifically for HIV-infected children, as well as the progress made in the renal biomarker field to assess the outcome of renal injury in HIV-infected children and young adults.

Epidemiology of HIV-related kidney diseases, risk factors, and biomarkers

According to the World Health Organization (WHO), more than 39 million HIV-infected people have died of AIDS, and there were approximately 35 million people infected with HIV all over the world at the end of 2013 (HIV/AIDS Fact sheet N⁰ 360, WHO, November 2014). With the widespread introduction of effective antiretroviral therapy (ART), renal-cardiovascular diseases have surpassed opportunistic infections as the major cause of chronic morbidity in these patients [3]. As discussed in the recent clinical guideline for the management of renal diseases in HIV-infected patients [3], HIV infection is a major risk factor for the development of chronic kidney disease (CKD). Furthermore, compared with HIV-negative people, the prevalence of albuminuria (>30 g/day) has been reported to be 2- o 5-fold higher in HIV-infected individuals [4, 5]. Several risk factors are associated with an increased risk of developing CDK in HIV-infected people. These factors include, a family history of end stage renal disease (ESRD), African ancestry, genetic predisposition (e.g. APOL-1 renal risk variants), lower CD4 cell count, advanced immunosuppression, history of acute kidney injury, injection of drugs of abuse, proteinuria, hypertension, diabetes, hepatitis C infection, some antiretroviral drugs, and higher HIV RNA levels [3, 6, 7]. Taken together, these findings suggest that HIV-1 per se can worsen the outcome of many renal diseases that are frequently seen in HIV-negative people. Therefore, it is important to develop new biomarkers for HIV-infected people, in order to identify the early stages of these renal diseases and initiate the corresponding treatment as soon as possible. Alternatively, other renal diseases are only seen in HIV-infected people, including HIVAN [6], HIV-immune complex renal diseases [8, 9], renal nephrotoxicity syndromes associated with the use of antiretroviral therapy, including indinavir [10] and several nucleotide analogs [11, 12], HIV-associated Hemolytic Uremic Syndrome [13], and other thrombotic microangiopathies (TMA) [14]. It is worth mentioning here that although the diagnosis of HIV-HUS or HIV-TMA diseases appears to be obvious on clinical grounds, this is not always the case. HIV-HUS usually has an insidious clinical onset in children, without a diarrhea prodrome, oliguria, or hypertension [2, 13], and these patients may also have pre-existing hematological abnormalities, including thrombocytopenia and anemia. For example, recently we reported a case of an adolescent girl with HIV-TTP who presented with a clinical history of idiopathic thrombocytopenic purpura (ITP), masking the early diagnosis of TTP [14]. Overall, we need new biomarkers to identify the early stages of all HIV-TMA diseases, determine the severity of each case, and predict the risk of long-term clinical outcome. In summary, all HIV-renal diseases need to be approached with special clinical considerations, and the selection of the most appropriate biomarkers for each HIV-renal disease is important.

Renal diseases in HIV-infected children

There is a wide spectrum of pediatric renal diseases that occur in the setting of HIV-infection. HIV-infected children may present either with exacerbations of common renal conditions, or with renal diseases that are specifically induced by the viral activity of HIV-1 in the kidney. A summary of the most frequent renal diseases that affect HIV-infected children is shown in Table 1. Here, we will discuss three renal diseases that represent the most frequent glomerular and tubular lesions seen in HIV-1 infected children, and that had been studied in order to generate new candidate biomarkers for these patients.

Table 1.

Renal Disease in HIV-infected children.

Glomerular diseases: HIV associated nephropathy (HIVAN)*; HIV-associated immune complex diseases (HIVCK)*; Post-infectious glomerulonephritis; IgA-nephropathy, membranous and membranous-proliferative glomerulonephritis, and lupus-like nephritis)

Thrombotic MIcroangiopathies. HIV- associated Hemolytic Uremic Syndrome (HIV-HUS)*, Thrombotic thrombocytopenic purpura (TTP), and multiple organ failure with disseminated intravascular coagulation (DIC).

Acute Kidney Injury, sepsis, infections, drugs.

Infectious: Urinary tract infections, Hepatitis, Opportunistic infections, Tuberculosis

Tubulopathies. Renal tubular acidosis, hypercalciuria, Fanconi syndrome

Nephrotoxic drugs. Tenofovir disoproxil fumarate (TDF)*, nucleotide reverse transcriptase inhibitors (NRTI)*, Protease inhibitors (PI) (e.g., indinavir, stones)*, antibiotics, other antiviral treatments.

^*Renal diseases that are seen only in HIV-infected patients, unless the nephrotoxic drugs TDF, NRTI, and indinavir are used to treat other diseases.

HIV-associated nephropathy (HIVAN)

HIVAN is a renal disease induced directly by HIV-1 almost exclusively seen in people of African ancestry [15–17]. It is the most frequent chronic renal disease seen in children of African ancestry infected with HIV-1 [1, 2, 18]. From the clinical point of view, children with HIVAN can present either with isolated proteinuria or the classic features of nephrotic syndrome, which include heavy proteinuria, edema, and hypoalbuminemia. Those who present with isolated proteinuria can sustain normal serum creatinine values for a long period of time [1]. The renal histological lesions of HIVAN are characterized by the presence of focal and segmental collapsing glomerulosclerosis (FSGS) and microcystic transformation of renal tubules (Fig. 1A) leading to renal enlargement and chronic kidney failure [17, 19–21]. The latter changes are the most consistent renal histological findings in children with HIVAN [18]. HIVAN is triggered by the infection of podocytes and renal tubular epithelial cells [22–24], yet the mechanism is unclear. A genetic predisposition associated with two genetic risk variants of apolipoprotein L-1 (APOL1), appears to play a critical role in the pathogenesis of HIVAN [25–27]. At the present time, a renal biopsy is the only way to establish a definitive diagnosis of HIVAN, and new renal biomarkers are needed to avoid this invasive procedure in children. The early institution of ART has improved the clinical outcome and survival of these patients [15, 16, 28], but the long-term consequences of ART in children are unknown. Moreover, it is unclear whether ART can cure or completely prevent the progression of HIVAN, since patients with a very low viral load, including those receiving kidney transplants, can develop HIVAN [29–31].

Figure 1.

Panel A shows a light microscopy picture of a renal section derived from a child with HIVAN. The black arrows point to a collapsed glomerulus and a tubular microcyst (Jones methenamine silver stain). Panel B shows a renal section from a child with HIV-HUS. The picture shows dilated capillary loops with red blood cell fragments, and tubular microcysts (Hematoxylin-eosin stain). The black arrow points to a renal arteriole with luminal narrowing due to red cell fragments and intraluminal thrombosis (A and B original magnification x 200). Pictures A and B are reproduced with permission from “Kidney disease in HIV-positive children” McCulloch, M.I., Ray P.E. Seminars in Nephrology (2008) 28, 585–594. Panel C, shows representative changes in urinary EGF, MMP-2, FGF-2, and serum creatinine (SCr), seen during the different stages of childhood HIVAN [16]. Panel D shows representative changes in urinary EGF, NGAL, FGF-2, and SCr, seen during different stages of acute kidney injury in HIV-infected children [37,42, 74].

HIV-associated Hemolytic Uremic Syndrome

HIV-infected children can develop an atypical and often lethal form of hemolytic uremic syndrome (HUS) which is defined by the presence of microangiopathic hemolytic anemia, thrombocytopenia, and acute renal failure [13]. The prevalence of HIV-HUS and other HIV-TMA disease has decreased in a significant manner in children treated with ART [1]. Nonetheless, HIV-HUS is often a lethal disease that needs to be treated as soon as possible to prevent major systemic bleeding complications [1, 2, 13]. As discussed above, performing an early diagnosis of HIV-HUS based on clinical grounds is not always easy and new markers are needed to identify children undergoing the early stages of HIV-HUS, and predict the risk of development lethal systemic vascular complications or CKD [1, 13]. HIV-HUS is characterized by an insidious clinical onset, preserved urine output, and the absence of preceding diarrhea [13, 32, 33]. Rapid progression to end stage renal disease, or death from infectious or bleeding complications is common [2]. The presence of proteinuria and high viral load in HIV-infected children are risk factors for the development of HIV-HUS, which is triggered by infectious agents that induce renal endothelial injury and thrombotic microangiopathy (TMA) [13, 32–34]. From the histological point of view, HIV-HUS show TMA lesions characterized by the accumulation of fibrin, dilated capillaries, and tubular micro-cysts (Fig 1B) [13]. It should be noted however, that HIV-infected patients can develop asymptomatic TMA or TMA associated with other diseases, including Thrombotic Thrombocytopenic Purpura (HIV-TTP) [35]. Thus, it is necessary to develop new biomarkers to identify the early stages of HIV-HUS, and other diseases associated with TMA lesions in HIV-infected children.

Acute Kidney Injury (AKI)

In addition to HIV-HUS, HIV-infected children can develop AKI due to severe dehydration, poor renal perfusion, sepsis, and nephrotoxic drugs [2]. AKI is the most frequent renal disease seen in all HIV-infected children [1, 2]. AKI is defined as an abrupt decrease in GFR (“within 48’), which is manifested by an elevated serum creatinine [36]. However, changes in serum creatinine may not accurately estimate the GFR in a patient whose renal function is changing fast. Despite these limitations, a rise in serum creatinine continues to be the most widely used laboratory marker to diagnose AKI. Years ago, a consensus definition called the RIFLE classification of AKI was developed which consists of three graded levels of renal injury (Risk, Injury, and Failure) and two outcome measures (Loss and End-stage renal disease) [37]. This definition was adapted for pediatric patients (pRIFLE) using the estimated creatinine clearance (eCrCl) and the urine output [38] (Table 2). Nonetheless, since the assessment of the eCrCl is based on the serum creatinine values, newer biomarkers are needed to identify children undergoing the early stages of AKI.

Table 2.

Urinary candidate biomarkers for HIV-renal diseases

Renal Disease	Increased (> cut-off values).	Decreased (< cut-off values).	No changes (> or < cut-off values)
HIVAN	FGF-2, MMP-2, NGAL, gross proteinuria, RBP, β2-M, cystatin C	EGF
HIV-HUS	FGF-2, NGAL, iron, haptoglobin, hemopexin, lactoferrin.	EGF
Sepsis-AKI	FGF-2, NGAL, RBP, γ-GTP, β2-M,	EGF
ART-induced toxicity.	RBP, β2-M, γ-GTP N-Aβ-D-glucosamine, α-1-M.		FGF-2, EGF

Abbreviations. Fibroblast Growth Factor-2 (FGF-2); Matrix Metalloproteinase-2 (MMP-2); Epidermal Growth Factor (EGF); Neutrophil Gelatinase Associated Lipocalin (NGAL); Retinol binding protein (RBP); β-2 Microglobulin (β-2M); γ-glutamyl-transpeptidase (γ-GTP); N-acetyl β-D-glucosamine (N-A β-D-glucosamine); α-1 microglobulin (α-1-M).

Assessment of renal injury in HIV-infected children

The following renal findings are used in the clinical setting of HIV-infected children to define CKD: (1) A random urinary protein-to creatinine ratio > 0.2 (mg/mg) [39]. Urine protein to creatinine ratios > 1.0 (mg/mg) are considered nephrotic range proteinuria (> 100 mg/m²/d or 40 mg/m²/h) [39, 40]; (2) Urine sediment abnormalities, including the presence of cast and renal epithelial cells [1]; (3) Fluids-electrolytes abnormalities due to tubular disorders; (4) abnormal renal histology; (5) Structural abnormalities detected by imaging, including enlarged echogenic kidneys; (6) Decreased glomerular filtration rate (GFR) < 60 ml/min per 1.73 m², assessed by estimating the creatinine clearance (CrCl); 7) Increase serum creatinine values according to age. In younger children (< 2 years of age), the urinary protein excretion and GFR values are adjusted according to the normal values for the corresponding age. Although the serum creatinine concentration performs well for assessing kidney function in patients with stable CKD [41, 42], it has limitations in HIV-infected children, because these values are influenced by the total protein intake, body weight, nutritional status, muscle mass or breakdown, and the tubular secretion of creatinine in the context of ongoing renal injury and multiple medications. Unfortunately, all these factors are usually affected in HIV-infected children.

Previous studies have suggested that cystatin C could become a valuable renal biomarker in HIV-infected patients [43–46]. Cystatin C is a nonglycosilated low molecular weight protein that is synthesized and secreted at a nearly constant rate by almost all nucleated cells, and is freely filtered by the glomeruli [36, 47]. In contrast to creatinine, cystatin C is not excreted in the urine, and it is reabsorbed and metabolized by proximal tubular cells [47]. Several studies done in HIV-negative people suggest that cystatin C may be superior to serum creatinine [47, 48]. Furthermore, cystatin C has been proposed as a more sensitive marker of kidney disease in HIV-infected adults [43–46]. However, other studies [44, 45] have questioned this notion arguing that these studies lack gold standard measurements of GFR, or that the serum cystatin C levels are affected by changes in viral load and systemic inflammation [44, 49, 50]. Thus the value of cystatin C as marker of GFR in HIV-infected children is unclear at the present time. We have found elevated levels of cystatin C in the urine of HIV-infected children with proteinuria [51]. These findings suggest an impaired ability of the proximal tubular epithelial cells to reabsorb and metabolize cystatin C in these patients. Overall, given the inherent limitations of proteinuria assessments and GFR estimations, there is a need to develop new biomarkers to assess both the GFR and renal histological damage in HIV-infected children.

Biomarkers for HIV-associated renal diseases

Iron binding proteins as biomarkers for childhood HIV-associated renal diseases

As discussed above, the identification of new biomarkers to facilitate the early diagnosis of HIVAN, HIV-HUS, AKI, and other HIV-associated renal diseases, could have significant impact in the clinical care of HIV-infected children. To explore this issue, we assessed the value of eleven urinary proteins that are considered excellent biomarkers of glomerular diseases in HIV-negative adult patients [52] in children with HIVAN and HIV-HUS [53]. These proteins are: orosomucoid, transferrin, α-1 microglobulin, zinc-α2-glycoprotein, α1-antitrypsin, complement factor B, haptoglobin, transythyretin, retinol binding protein, albumin, and hemopexin [52]. Using two dimensional Gel Electrophoresis (2-DE) and mass spectrometry analysis we found that all children with HIVAN and HIV-HUS showed high urinary levels of all these proteins, when compared to samples collected from HIV-infected children without renal disease [53]. However, we were unable to identify a single urine protein biomarker to distinguish children with HIVAN vs. HIV-HUS. The more relevant changes were noted in the urinary excretion of the iron related proteins haptoglobin, hemopexin, and transferrin. Haptoglobin, is a plasma protein that participates in the recycling of heme iron [54]. Hemopexin, is a heme binding plasma glycoprotein, that protects against hemoglobin-mediated oxidative damage during intravascular hemolysis [55], and transferrin is a key protein involved in the transport of iron [56]. Moreover, we found elevated levels of iron, lactoferrin, and neutrophil gelatinase-associated lipocalin (NGAL) in the urine of children with HIVAN and HIV-HUS [53]. Lactoferrin is an antioxidant iron-binding protein [57], and NGAL is an iron trafficking protein [58], which is also an established biomarker of AKI [59, 60]. In agreement with these findings, HIV-Tg rats with renal disease and TMA lesions also showed elevated urinary levels of iron and NGAL [53]. In addition, since endothelial injury plays a critical role in the pathogenesis of TMA lesions, and injured endothelial cells release FGF-2 [61], we measured and detected high levels of FGF-2 in the circulation and urine of children with Stx-HUS and HIV-HUS [33, 34]. In summary, these findings identified FGF-2, iron, NGAL, and other iron-related proteins as potential candidate urinary biomarkers for children with HIV-HUS. Considering that iron can induce renal cytotoxic effects [62], it is tempting to speculate that the accumulation of iron may play a role in the pathogenesis of these diseases as well.

NGAL and HIV-infection

NGAL is produced by distal renal tubular cells, but is also produced by other cell types including neutrophils [63], and is a critical component of the immunity to bacterial infections [64]. In agreement with this notion, a previous study in HIV-infected patients showed decreased NGAL serum levels, potentially reflecting decreased number and function of neutrophils as well as impaired bone marrow myelopoiesis, and these changes were reversed with ART [65]. Therefore, in patients undergoing systemic infections or multi-organ failure, the serum levels of NGAL are a less specific biomarker of renal injury [64, 66]. Nevertheless, in support of the notion that NGAL could become a reliable biomarker for children with HIVAN, studies in adults demonstrated that the expression of NGAL was much higher in patients with biopsy proven HIVAN than in HIV-positive and HIV-negative patients with other forms of chronic kidney disease [67]. Furthermore, in an HIV-transgenic mouse model, NGAL mRNA was increased in dilated microcystic segments of the nephron [67]. In contrast urinary NGAL did not correlate with proteinuria in human or mouse models. Based on these findings, the authors concluded that urinary NGAL could become an excellent biomarker to screen for HIVAN-related tubular damaged [67]. A follow up study comprising 25 HIV-infected adults, including 18 patients with HIVAN and 7 with other glomerular diseases, demonstrated that NGAL had 94% sensitivity and 71% specificity for the diagnosis of HIVAN with an area under the receiver operator characteristic (ROC) curve of 0.88 [68]. Thus, further studies are warranted in a larger cohort to determine the value of NGAL for the noninvasive diagnosis of HIVAN.

Biomarkers for childhood HIVAN

Using the proteomic techniques discussed above, we were unable to identify a biomarker profile that could distinguish children with HIVAN from those with other glomerular diseases. Therefore, we used an alternative approach based on our knowledge of the pathogenesis of childhood HIVAN. Briefly, previous studies documented a progressive renal accumulation of FGF-2 in HIVAN [34, 69–71] in correlation with the progression of the microcystic tubular lesions. As discussed above, these lesions are the most consistent renal histological findings in children with HIVAN [72]. Therefore, we measured the urinary levels of FGF-2 in HIV-infected children with and without renal diseases, in combination with Epidermal Growth Factor (EGF) and Matrix Metalloproteinase-2 (MMP-2) [51]. FGF-2 plays an important role in the regeneration of renal tubules [73, 74] and can induce the formation of renal cysts [75, 76]. Moreover, FGF-2 is accumulated in the kidney of children with HIVAN bound to an increased number of heparan sulfate proteoglycans (HSGP) [70, 71]. The activity of MMP-2 is enhanced during the process of renal tubular regeneration [77, 78], and MMP-2 facilitates the release and activation of FGF-2 bound to HSPG [79]. EGF is a powerful mitogen which is predominately synthesized by renal tubular epithelial cells [80, 81]. However, its synthesis and release into the tubular lumen is decreased in injured renal tubular epithelial cells [80–83]. In agreement with these observations, we found elevated urine levels of FGF-2 and MMP-2, in association with reduced urine levels of EGF in children with HIVAN (Fig 1C). In summary, these findings suggest that the urinary biomarker profile comprised of high levels of FGF-2 and MMP-2, and low levels of EGF, could become a useful biomarker tool to identify children with HIVAN [51, 84].

A new biomarker profile for AKI in critically ill children

Based on the findings described above, and considering that HIV-infected children frequently develop AKI due to changes in renal perfusion or sepsis, we tested the value of a urinary biomarker profile comprised of elevated urinary levels of NGAL, FGF-2, and decreased urinary levels of EGF, in critically ill children [85]. NGAL was selected as a marker of renal tubular injury because it is considered an established biomarker of AKI [59, 60]. FGF-2 was selected as marker of renal epithelial and endothelial injury and regeneration, because these processes play a critical role in the pathogenesis of AKI induced by poor renal perfusion and septic shock, and because injured endothelial cell release FGF-2 [34, 85, 86]. Finally, EGF was selected because previous studies show that the urinary levels of EGF could be used to monitor the outcome of AKI in children [82, 87, 88] (Fig. 1D). Using selected cutoff points for NGAL, FGF-2 and EGF, this biomarker profile demonstrated a 89% sensitivity and 89% specificity for the diagnosis of AKI in critically ill children admitted to a Pediatric Intensive Care Unit, with an area under the ROC curve of 0.94 [85]. The positive predictive value of this test was 0.85 (95% CI 0.65 to 0.98) and the negative predictive value was 0.93 (95% CI 0.71–0.97) [85]. This biomarker profile can be used in both HIV-positive and HIV-negative children, although further studies in larger cohort of children are warranted to validate these data. Several other AKI biomarkers have been identified in HIV-negative patients, as described in detail in previous publications [36, 89]. More studies are needed to define their role in HIV-infected children.

Biomarkers of renal tubular dysfunction and injury

Previous reports described the performance of well established biochemical markers to assess the tubular function of HIV-infected patients [90]. These markers included hyperaminoaciduria, phosphaturia, glucosuria in patients without diabetes, and increased fractional excretion of uric acid. In addition, hypercalciuria and acidosis were the most relevant complications seen in HIV-infected children followed in Venezuela [91]. Similar renal tubular features were seen in HIV-infected children treated in the US during the early years of the AIDS epidemic [1].

Alternatively, other studies have explored the role of low molecular weight proteins (LMWPs) as tubular biomarkers in HIV-infected patients [44]. These proteins are freely filtered through the glomerulus, and are subsequently reabsorbed and catabolized by proximal tubular cells. Therefore, in patients with normal tubular function the urine levels of LMWPs are very low. As reported before [53], using 2-DE we identified three LMWPs that are established biomarkers of proximal tubular dysfunction in HIV-infected children with proteinuria. These markers are, α–1 microglobulin (α–1M), β2- microglobulin (β-2M) [92] and retinol binding protein (RBP) [53]. α–1M is glycosylated protein of ~ 26–33 kDa produced mainly in the liver [93]. β–M2 is a 12 kDa protein that is a component of the major histocompatibility complex (MHC) class I molecules, and is present in all nucleated cells [94]. RBP is a 21 kDa protein that circulates in plasma bound to transthyretin, although its unbound fraction is freely filtered and reabsorbed by the proximal tubule via an ATP dependent endocytic mechanism [94]. Therefore, an increased RBP-creatinine ratio can be used to identify a tubular transport defect [95]. Taken together, high urinary levels of α–1M, β-M2, and RBP indicate the presence of tubular dysfunction or tubular injury, while other biomarkers discussed above (e.g. NGAL or EGF), could further confirm the presence of tubular injury. The elevated urinary levels of β-M2, and RBP in our cohort of HIV-infected children, were also associated with elevated urinary cystatin C levels [51], further documenting the failure of proximal tubular cells to reabsorbed and/or metabolized cystatin C in these patients. HIV-infected children with trace proteinuria measured by dipstick, also show elevated levels of β-M2, and RBP, and these findings suggest that proximal tubular injury is an early event in the pathogenesis of many HIV-associated glomerular diseases, including HIVAN [53]. These findings are in agreement with the results reported in HIV-Transgenic mice [71]. Unfortunately, it is not always possible to determine whether the elevated urinary levels of LMWPs are due to the presence of isolated tubular injury or the early stages of HIV-glomerular diseases. In agreement with this notion, we found elevated levels of γ–glutamil transpeptidase (γ-GT) and lysozyme, two enzymes primarily located in the brush border of proximal tubular cells [96], in a group of HIV-infected children with proteinuria. The urinary lysozyme levels were elevated in HIV-infected children with trace proteinuria, suggesting an early involvement of the brush border of proximal tubular cells in several glomerular renal diseases [51]. Taken together, these studies suggest that LMWPs are sensitive biomarkers to detect renal tubular dysfunction in HIV-1 infected children, but that they should be used in combination with other biomarkers to rule out the presence/absence of glomerular and tubular injury.

Biomarkers of renal toxicity induced by ART

Although ART has dramatically reduced the mortality and improved the renal outcome of all HIV-infected children, most ART drugs have some toxicity. Several antiretroviral drugs have been linked to renal toxicity, including tenofovir disoproxil fumarate (TDF) and nucleotide reverse transcriptase inhibitors (NRTI). TDF is a nucleotide analogue of adenosine 5′-monphosphate that is eliminated unchanged in the urine by a combination of glomerular filtration and proximal tubular secretion [97, 98]. Two other nucleotide analogues with structural similarities to TDF, cidofovir and adefovir, can also cause nephrotoxicity [99]. These drugs are actively transported into renal proximal tubular cells and secreted into the tubular lumen. A number of mechanisms and drugs given to HIV-infected patients can interact with the transport mechanisms and facilitate the accumulation of TDF inside renal tubular cells, causing cytotoxicity through mitochondrial disruption [100].

For this reason, current clinical practice guidelines from the Infectious Diseases Society of America, recommend avoiding TDF HIV-infected patients who have a GFR < 60 ml/minute/1.73 m², whenever feasible [3]. These guidelines also suggest avoiding TDF as part of first-line therapy in prepuberal children (Tanner stages 1–3), because TDF is associated with increased renal tubular abnormalities and bone mineral density loss in this age group. Fanconi’s syndrome is the most important clinical consequence of tubular toxicity in HIV-infected children, because it affects the regulation of calcium and phosphorus balance, inducing growth retardation and bone problems [101]. TDF has been the drug most often associated with the development of Fanconi’s, syndrome, albeit at a very low incidence (0.2–2%) [100]. Alternatively, TDF can cause isolated hypophosphatemia, normoglycemic glycosuria, and/or phosphaturia [90, 102]. In a large pediatric study in older children, the prolonged used of TDF was associated with low incidence of hypophosphatemia (4%), which was reversible following the TDF withdrawal [98]. The urinary levels of β2M and RBP are considered sensitive biomarkers to detect the TDF-induced tubular injury in children and adults [92, 100, 103, 104]. However, as discussed above, these biomarkers are also elevated in children with proteinuria who are not treated with TDF [51]. Finally, one study conducted in HIV-infected adults without proteinuria or impaired glomerular function who were treated with highly active antiretroviral therapy (HAART), found that ~ 25 % of these patients developed subclinical tubular damage, which was diagnosed when at least three of the following urinary biomarkers were elevated: N-acetyl-β-D-glucosaminidase, γ-glutamyl transpeptidase, β-2M, and α-1 microglubulin [105]. Overall, these studies suggest that many biomarkers are available to assess the magnitude of ART-induced tubular dysfunction and injury in HIV-1 infected children. However, these biomarkers are also affected in many glomerular diseases seen in HIV-infected children.

Future needs and directions

The early identification of kidney injury induced by HIV-1 or ART should be an important clinical goal in the clinical follow up of HIV-infected children. The studies discussed above have identified a significant number of new candidate biomarkers that need to be tested and validated in a large group of HIV-infected children.

Five phases of biomarker development has been described for the early detection of cancer [106], and these phases have been discussed in the context of acute kidney injury [41]. Briefly, phase 1 includes preclinical laboratory studies that identify genes and proteins that are over-or under-expressed. The availability of HIV-Tg mice and rats, that mimic the renal diseases seen in patients with HIVAN, has facilitated the discovery of NGAL, FGF-2, EGF, MMP-9, and several iron binding proteins as potential candidate renal biomarkers for HIV-infected children [18, 72]. Phase 2 studies involve the translational phase of biomarker development, and include testing new clinical assays to assess the presence or absence of renal disease in samples that are collected from patients. These studies should distinguish patients with renal diseases from those without renal diseases. A few pilot studies have been done in HIV-infected children to test new candidate biomarkers for HIVAN [51, 53]. Phase 2 studies in general do not allow determination of whether the disease can be detected earlier using the proposed biomarkers, but include the estimation of true-positive and false positive rates, and the generation of receiver operating curves (ROC) and cut-off values, to assess the performance of the selected candidate biomarkers, as described for NGAL in adult patients with HIVAN [68]. Phase 3 studies are similar to Phase 2 studies, except that the biomarkers are then evaluated for their capacity to detect pre-clinical disease at earlier time points. The identification of patients carrying the APOL1 genetic risk variants that predispose to the development of HIVAN [25–27], in combination with very sensitive and specific biomarkers, may allow the identification of these patients. However, these studies have not been done in children with HIVAN due the lack of clinical specimens collected before and after these patients develop HIVAN, and the limited number of patients available in single centers. Finally, phases 4 and 5 involve large-scale validation studies of the selected biomarkers in a prospective manner. The primary questions addressed in the phase 4 studies include the understanding of the performance of the biomarker as a screening test [41, 106]. These studies can also help understand the cost of using the new biomarkers in standard clinical care. Because the prevalence for HIVAN is low, these studies will require a large sample size, and the participation of multiple clinical centers to recruit these patients. Phase 5 studies usually explore the impact of using the selected biomarker tests in relationship with the outcome of the disease [41, 106].

In summary, significant progress has been made during the last years in the identification of renal biomarkers for HIV-infected children. Nevertheless, a great challenge remains ahead, if we intent to validate many of the biomarkers discussed above in these children. In addition, we hope that the availability of more sensitive proteomic technologies will facilitate the discovery of new biomarkers to measure the GFR in a more accurately manner, and to distinguish the different stages of CKD and AKI in HIV-infected children and young adults. Particular emphasis should be placed in using new proteomic techniques to assess the GFR and status of CKD in children < 2 years of age, since this is a unique period of kidney growth and maturation that may require a unique set of biomarkers to assess the renal function and injury.

Abbreviations

KDIGO

Kidney Disease improving global outcomes

HIV-1

Human immunodeficiency virus -1

ACR

albumin-to-creatinine ratio

GFR

Glomerular Filtration Rate

CKD

Chronic Kidney Disease

ART

Antiretroviral therapy

TMA

thrombotic microangiopathy

FGF-2

Fibroblast Growth Factor-2

EGF

Epidermal Growth Factor

Matrix

metalloproteinases-2

HIVAN

HIV-associated nephropathy

HIV-HUS

HIV-associated hemolytic uremic syndrome

AKI

Acute Kidney Injury