Objectives: To determine the incidence and to identify the clinical parameters associated with non-gastrointestinal renal tubular and high anion gap acidosis in a cohort of HIV-1-infected children.
Methods: Records of 202 HIV-1-infected children were reviewed to identify patients with metabolic acidosis. Serum and urine chemistries of those children with persistent non-gastrointestinal acidosis were then studied prospectively. Serum and urinary anion gaps (SAG and UAG) were calculated. Those with acidosis (group 1) were compared with children without acidosis (group 2). Associations were determined with Pediatric HIV classification, height, weight, antiretroviral therapy, and Pneumocystis carinii pneumonia prophylaxis.
Results: Persistent acidosis was noted in 34 out of 202 children (17%): 16 out of 34 (47%, group 1A) had elevated SAG acidosis, and 18 out of 34 (53%) had normal SAG acidosis with a positive UAG (distal renal tubular) acidosis (group 1B). Those with acidifying defects more often received P. carinii pneumonia prophylaxis (P = 0.02 and 0.01 for groups 1 and 1A, respectively) independently of HIV-1 classification. This group was shorter in height than group 2 (P = 0.007). Differences in weight were not significant (P = 0.1). However, acidotic subjects were more immunocompromised than those in group 2 (multivariate P < 0.001 for HIV classification C3).
Conclusions: Elevated SAG acidosis and renal tubular acidosis are not uncommon among HIV-infected children with advanced disease. These disorders may be associated with height growth failure and prophylaxis with sulfur/sulfone containing antibiotics. HIV infection and/or its associated therapies may cause renal tubular damage. The causes of elevated SAG acidosis require further investigation.
From the From the Department of Child Health, St. George's Hospital and Medical School, London UK, the aDepartment of Pediatrics, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey, and the bDepartment of Child Health, at Queen Mary's School of Medicine and Dentistry, Royal London Hospital, Luckes House, Stepney Way, Whitechapel, London, E1 1BB.
Requests for reprints to: Reprint requests to G. D. McSherry, Department of Pediatrics, MSB F570A, UMD-New Jersey Medical School, 185 S. Orange Ave, Newark NJ 07103, USA.
Correspondence to R. Chakraborty, Department of Child Health St George's Hospital and Medical School, Blackshaw Rd, Tooting, London SW17 0QT, UK.
Note: This work was presented at the meeting of the American Pediatric Society/Society for Pediatric Research. May 1999, San Francisco, California, USA.
Received: 21 February 2002; revised: 6 September 2002; accepted: 8 November 2002.
HIV-associated nephropathy in children represents a disease spectrum that ranges from mild to moderate persistent proteinuria, hematuria, renal tubular acidosis (RTA), and end-stage renal disease. The prevalence varies by study and is associated with advanced immunosuppression . The pathogenesis of HIV nephropathy has not been elucidated, although the identification of HIV-1 in glomerular and tubular epithelial cells in seropositive subjects provides circumstantial support for a direct viral cytopathic effect . Other possible pathogenic mechanisms include immune complex disease , multiple viral opportunistic infections  and vasculitis. Reported histopathologic abnormalities include focal segmental glomerulosclerosis, mesangial hyperplasia, minimal lesion glomerulonephritis and ductal ectasia [4,5].
The RTA syndromes represent a diverse group of tubular transport disorders characterized by defects in the reabsorption of bicarbonate and/or the excretion of hydrogen ions that is out of proportion to any reduction in glomerular filtration rate. This results in a hyperchloremic metabolic acidosis . Three primary types have been described. Distal (Type I) RTA is caused by impaired distal H+ secretion and persistent bicarbonaturia. The disorder is characterized by an inability to appropriately lower urinary pH even in the presence of systemic acidosis , and decreased urinary ammonium excretion resulting in a positive urinary anion gap (UAG) . Common clinical manifestations include failure to thrive especially height growth failure, muscle weakness, and the development of nephrocalcinosis. Distal RTA has been reported in disease states associated with hypergammaglobulinemia . A negative UAG is seen in proximal (Type II) RTA, which is caused by an impaired capacity of the proximal tubule to reabsorb filtered bicarbonate resulting in bicarbonate wasting. Hypoaldosteronism results in Type IV RTA with hyperkalemia and a positive UAG .
Although these disorders may not be uncommon and recommendations have been made to correct RTA in HIV-infected children , there is a paucity of data and literature in such subjects. We therefore investigated the incidence and underlying causes of persistent acidosis in a cohort of HIV-infected children.
After this study was completed, the association of HIV infection with a marked disturbance of metabolism began to be reported including wasting and altered body fat distribution . Furthermore, high anion gap lactic acidosis and evidence of mitochondrial dysfunction have been described in children and adults treated with nucleoside reverse transcriptase inhibitors (NRTI) [13,14].
Population and sampling
Records of 202 HIV-1-infected children followed prospectively at the Francois Xavier Bagnoud Center for Children in Newark, New Jersey, USA were reviewed to identify those with metabolic acidosis. The majority of children (98%) were infected with HIV-1 vertically. Bicarbonate values screened were included in serum electrolyte profiles obtained every 1–3 months as part of routine clinical care. Ethical approval for the study was obtained from the Institution Review Board (IRB) of the New Jersey Medical School and informed consent was obtained from the legal guardians of all children.
Serum and urine chemistries of children with persistent non-gastrointestinal acidosis were studied prospectively over 15 months (between September 1997 and November 1998). Acidosis was defined as serum HCO3− < 18 mole equivalents (mEq)/l (< 2 years of age) or < 21 mEq/l (> 2 years of age), on three or more visits during the study period (consecutively or intermittently) in the absence of diarrhea (group 1). Serum and urinary anion gaps (SAG and UAG) were calculated. Subjects with acidosis from persistent gastrointestinal disease and those children attending the Pediatric Renal clinic for other causes of HIV nephropathy were excluded from the study. During each visit legal guardians completed a questionnaire about the presence of diarrhea and vomiting. Other observations included age-specific height and weight percentile values, HIV classification , current medications and past medical history.
Classification of RTA and high SAG acidosis was calculated by the formula Na+ − [Cl− + HCO3−] and UAG by [Na+ + K+ − Cl−]. A presumptive diagnosis of distal RTA was made in the normal SAG group in the presence of a positive UAG and urinary Na > 30 mEq/l . A negative UAG suggested Type II RTA . Urinary pH and serum K+ measurements were used to distinguish between Types I and IV RTA. Type I (classical non-voltage dependent) RTA is associated with persistently low/normal serum K+ whereas Type I (voltage dependent) RTA is characterized by high serum K +. Both have urinary pH > 5.3. Hyperkalemic metabolic acidosis and a urinary pH < 5.3 occur in hypoaldosteronism (Type IV RTA) .
Children with persistent acidosis (group 1) were compared to non-acidotic subjects (group 2). Univariate differences between these groups were calculated using the two-tailed Fisher's Exact test on contingency tables and unconditional logistic regression. Collinearities between variables were inferred from the variance–covariance matrix. Potential confounders and effect modifiers were then controlled for within the framework of multivariate logistic regression analyses. In forward-step model selection variables were retained with a P value of < 0.2 . Variables included the Pediatric HIV classification  and mean height and weight centile measurements during the study period. A mean percentile of < 5% was used as a cutoff to compare both groups. Height, weight, antiretroviral therapy (ART) and the use of sulfa/sulfone containing antibiotics were also used to identify differences between the two groups.
Thirty-four out of 202 HIV-1-infected children (17%) had persistent acidosis in the absence of acute or chronic gastroenteritis. The mean age of group 1 was 8 years 9 months (range, 2–21 years). Eighteen out of 34 children (53%) had normal SAG acidosis with a positive UAG and urinary Na > 30 mEq/l. All of these children had low/normal serum K+ levels and a urinary pH > 5.3. They were classified as having distal non-voltage dependent RTA (group 1A). High SAG acidosis was noted in 16 out of 34 children (47%, group 1B).
Univariate analysis revealed that children with acidifying defects were significantly shorter in height than those in group 2 (P = 0.007). This difference was not observed when weights were compared (P = 0.1) (Table 1). Other significant risk factors included Centers for Disease Control and Prevention (CDC) stage C clinical classification (P < 0.001) and immunologic category 3 (P = 0.001). Multivariate analysis (which controlled for potential confounders including age at study recruitment and treatment with ART) suggested that children within the CDC C classification were 5.3 times more likely to have acidifying defects than those from group 2 independent of other variables (P < 0.001). Differences in height and weight between group 1 and 2 were therefore not independent of differences in clinical classification (Tables 1 and 2).
Children receiving prophylaxis with sulfur/sulfone containing antibiotics were three times more likely to have acidifying defects than those in group 2 (P = 0.02) (Table 1). This association was most striking among those in group 1A with high SAG acidosis [odds ratio (OR), 15; P = 0.01] but was not significant for subjects in group 1B (P = 0.5).
Although all children in group 1 received ART (versus 146/167 in group 2) a univariate comparison for this variable was not significant (P = 0.135; OR, 4.75; Table 1).
Calculating SAG and UAG in conjunction with urinary pH and serum K+ may be helpful in screening for an array of correctable disorders causing hyperchloremic metabolic acidosis. A stepwise approach using these relatively inexpensive investigations is an appropriate screening tool that may reveal an unsuspected diagnosis with major clinical implications. In this study we demonstrated that elevated SAG acidosis and distal RTA were not infrequently observed in HIV-infected children with advanced disease, and may have contributed towards height growth failure. We therefore recommend routine use of the above screening methods as a means of evaluating for unexplained acidosis in HIV-infected children.
One previous report in the Pediatric literature cites the prevalence of distal RTA as being 50% among eight children receiving therapy for acute lymphoblastic leukaemia . The investigators from this and other studies attributed the hyperchloremic metabolic acidosis, renal bicarbonate wasting and growth failure to co-trimoxazole usage [18,19]. However, ART may improve tubular and glomerular histopathological changes of HIV-1-associated nephropathy in dialysis-dependent patients . It seems plausible that a combination of advanced HIV-1 disease and prophylaxis with sulfur/sulfone containing antibiotics could independently contribute to tubulointerstitial damage, and that ART may be protective by controlling HIV-1 replication. This may have accounted for the lack of association in children from group 1B with this variable. Nevertheless, an association with high SAG acidosis and Pneumocystis carinii pneumonia (PCP) prophylaxis, independent of ART, was noted (P = 0.01; data not shown). This has not previously been reported in the literature and requires further investigation.
ART can independently contribute to high SAG acidosis . In this study a univariate comparison for this variable did not identify any association (P = 0.135; OR, 4.75), as only 21 children in the entire cohort were naive to ART. With further enrolment significance may have been established.
Mitochondrial toxicity with lactic acidosis and myopathy appear to be causally related to treatment with NRTI . The first pediatric case to be identified with this syndrome was an HIV-infected child treated with combination ART . In our study 16 children were identified with high anion gap acidosis in 1998—1 year before the first descriptions of mitochondrial toxicity caused by NRTI. At the time we assayed urine specimens empirically by chromatography in order to identify urinary metabolites that could account for the high anion gap acidosis in these children. These included organic acids, branched chain amino acids and qualitative lactate. However none were detected. Hyperlactatemia is a known feature of many mitochondrial disorders , and has been observed among many subjects receiving NRTI [22–24]. Obtaining blood lactate and pyruvate levels may be diagnostic in establishing the cause of unexplained high SAG acidosis in HIV-infected subjects receiving ART  and PCP prophylaxis.
Acidifying defects may have contributed to height growth failure. However the influence of acidosis on height could not be separated from the effects of disease progression as children in group 1 were more immunocompromised than those in the non-acidotic group. Furthermore, we could not differentiate between truly persistent from intermittent acidosis. We believe height growth failure may be related to the former. In addition to ART, it would seem prudent to attempt to manage distal RTA with alkalinizing agents and minerals . However, it remains unclear whether this approach would have a beneficial effect on height growth.
There is a need for similar studies. If this data is confirmed by investigators elsewhere, then screening immunocompromised children with unexplained persistent acidosis for the disorders described might be warranted. Improved understanding of the pathogenesis of NRTI-associated toxicity will aid the development of appropriate agents to minimize such manifestations among children who may have to receive these medications indefinitely.
We thank S. Timmapuri, D. Diawatan, L. Y. Shih and R. Booy for their assistance.
1.Abuzaitoun OR, Hanson IC. Organ-specific manifestations of HIV disease in children. Pediatr Clin North Am
2.Wali R, Drachenberg CI, Papadimitrou JC, Keay S, Ramos E. HIV-1-associated nephropathy and response to highly-active antiretroviral therapy. Lancet
3.Connor E, Gupta S, Joshi V, DiCarlo F, Offenberger J, Minnefor A, et al
. Acquired immunodeficiency syndrome-associated renal disease in children. J Pediatr
4.Foster S, Hawkins E, Hanson CG, Shearer W. Pathology of the kidney in childhood immunodeficiency: AIDS-related nephropathy is not unique
. Pediatr Pathol
5.Ingulli E, Tejani A, Senih F, Nicastri A, Chen CK, Pomrantz A. Nephrotic syndrome associated with AIDS in children. J Pediatr
6.Batlle D, Flores G. Underlying defects in distal RTA: New understandings. Am J Kidney Dis
7.Rodriguez-Soriano J, Vallo A. RTA. Pediatr Nephrol
8.Batlle DC, Hizon M, Cohen E, Gutterman C, Gupta R. The use of urinary anion gap in the diagnosis of hyperchloremic metabolic acidosis. N Engl J Med
9.Morris RC, Fudenberg HH. Impaired renal acidification in patients with hypergammaglobulinemia. Medicine
10.Zelikovic I. RTA
. Pediatr Ann
11.Laufer M, Scott GB. Medical management of HIV disease in children. Pediatr Clin North Am
12.John M, Nolan D, Mallal SA. Antiretroviral therapy and the lipodystrophy syndrome. Antiviral Ther
13.Church JA, Mitchell WG, Gonzalez-Gomez I, Christensen J, Vu TH, Dimauro S, et al
. Mitochondrial DNA depletion, near-fatal metabolic acidosis, and liver failure in an HIV-infected child treated with combination antiretroviral therapy. J Pediatr
14.Roge BT, Calbet JA, Moller K, Ullum H, Hendel HW, Gerstoft J, et al
. Skeletal muscle mitochondrial function and exercise capacity in HIV-infected patients with lipodystrophy and elevated p-lactate levels. AIDS
15.Hosmer D, Lemeshow S. Applied Logistic Regression. New York: Wiley Series in Probability and Mathematical Statistics.
16.Centers for Disease Control and Prevention. 1994 Revised classification system for human immunodeficiency virus infection in children less than 13 years of age. MMWR
17.Murphy JL. RTA in children treated with TMP-SMZ during therapy for acute lymphoid leukemia. Pediatrics
18.Kaufman AM, Hellman G, Abramson RG. Renal salt wasting and metabolic acidosis with TMP-SMZ therapy. Mt Sinai J Med (NY
) 1983, 50
19.Murphy JM, Griswold WR, Reznik VM, Mendoza SA. TMP-SMZ-induced RTA
. Child Nephrol Urol
20.White AJ. Mitochondrial toxicity and HIV therapy. Sex Trans Infect
21.Stacpoole PW. Lactic acidosis and other mitochondrial disorders. Metabolism: Clin Exp
22.Harris M, Tesiorowski A, Chan KJ, Montaner JS. Lactic acidosis complicating antiretroviral therapy: frequency and correlates. Antiviral Ther
23.John M, Moore C, James IR, Nolan D, Upton RP, McKinnon EJ, et al
. Chronic hyperlactataemia in HIV-infected patients taking antiretroviral therapy. AIDS
24.Carr A, Miller J, Law M, Cooper DA. A syndrome of lipoatrophy, lactic acidaemia and liver dysfunction associated with HIV nucleoside analogue therapy: contribution to protease inhibitor-related lipodystrophy syndrome. AIDS
25.Chariot P, Ratiney R, Ammi-Said M, Herigault R, Adnot S, Gherardi R. Optimal handling of blood samples for routine measurement of pyruvate and lactate. Arch Pathol Lab Med