Archives of Medical Research 40 (2009) 109e113 ORIGINAL ARTICLE Metabolic Acidosis in AIDS Patients Elizabeth F. Daher,a Lia C. Cezar,a Geraldo B. Silva Junior,a Rafael S. Lima,a Lisandra S. Damasceno,b Ericka B. Lopes,a Fernanda R. Nunes,a Rosa S. Mota,c and Alexandre B. Libórioa a Department of Internal Medicine, Division of Nephrology, Faculdade de Medicina, Hospital Universitário Walter Cantı́dio, Universidade Federal do Ceará, Fortaleza, Ceará, Brazil b Hospital Sáo José de Doenças Infecciosas, Fortaleza, Ceará, Brazil c Department of Statistics, Universidade Federal do Ceará, Fortaleza, Ceará, Brazil Received for publication August 3, 2008; accepted November 14, 2008 (ARCMED-D-08-00347). Background and Aims. Metabolic acidosis (MA) is a frequent and serious complication in HIV-infected patients. The aim of the study is to compare patients with and without MA associated with HIV. Methods. Patients were retrospectively studied involving all HIV-infected patients with blood gas analysis performed during hospital stay admitted to a single hospital between April 2004 and July 2006. Statistical analysis was performed using SPSS 10.0 for Windows. Results. Included in the study were 159 HIV patients, 72 cases (45.3%) with MA and 87 cases (54.7%) without. The comparison of both groups showed a mean arterial pH of 7.24  0.08 vs. 7.44  0.05, HCO3 12  5.7 vs. 21  5.1 mEq/L, serum urea 81  68 mg/dL vs. 39  46 mg/dL and serum creatinine 2.7  2.6 mg/dL vs. 1.2  1.9 mg/dL in MA-HIV and non-MA-HIV, respectively ( p !0.05). Antiretroviral therapy (ART) was being administered to 38 subjects (52.8%) in MA-HIV group and 45 (51.7%) in non-MA-HIV group ( p 5 0.57). There was no association between the use of ART and MA. Mortality was higher in patients with acidosis (52.7 vs. 17.2%, p !0.0001). Conclusions. In the present study, MA was associated with acute kidney injury and increased mortality. There was no association between the use of ART and MA. Ó 2009 IMSS. Published by Elsevier Inc. Key Words: Metabolic acidosis, HIV, AIDS, Renal dysfunction, Antiretroviral therapy. Introduction Metabolic acidosis (MA) has been described in human immunodeficiency virus (HIV) patients and is associated with antiretroviral therapy (ART). The majority of cases reported in the medical literature are due to lactic acidosis (LA) secondary to ART. There are several studies to investigate which factors are associated with LA in HIV-infected patients (1,2), but few studies investigated the factors associated with MA in general and its prognostic implication. In the last few years, ART has shown a dramatic improvement in the prognosis of HIV disease. As a result, other medical problems are assuming increasing relevance Address reprint requests to: Elizabeth de Francesco Daher, Rua Vicente Linhares, 1198, Fortaleza, CE, Brasil-CEP: 60135-270; E-mail addresses: ef.daher@uol.com.br or geraldobezerrajr@yahoo.com.br in the follow-up of HIV-infected patients (3). Among these medical problems are metabolic disturbances such as dyslipidemias, lipodystrophy, glucose intolerance and increased lactic acid, generally caused by ART drugs (3,4). Nucleoside reverse transcriptase inhibitors (NRTIs) induce mitochondrial toxicity by inhibiting polymerase, which lead to mitochondrial DNA depletion and respiratory chain dysfunction. Accumulation of lactate in the cytoplasm results in severe MA (5e10). The prevalence of hyperlactatemia in outpatients on ART is |9e16% (11,12). Among untreated patients, prevalence is |2% (13). It is important to consider mitochondrial abnormalities due to ART in the differential diagnosis of MA in HIV-infected patients (14). However, other causes of MA must be considered in this population, especially in hospitalized patients. Some studies show a prevalence of MA at |20% in HIV (2). There are multiple factors that could lead to MA in HIV 0188-4409/09 $esee front matter. Copyright Ó 2009 IMSS. Published by Elsevier Inc. doi: 10.1016/j.arcmed.2008.12.004 110 Daher et al./ Archives of Medical Research 40 (2009) 109e113 including opportunistic infections, comorbidities such as diabetes, hepatitis, renal diseases, alcoholism and ART. The aim of this study is to describe the clinical and laboratory features of HIV-associated MA, identifying its risk factors and prognostic implications. Materials and Methods The study was conducted at the Hospital S~ao José de Doenças Infecciosas in Fortaleza in the northeastern region of Brazil. Clinical files of all HIV patients admitted from April 2004 to July 2006 were reviewed, and patients with blood gas analyses performed during their hospital stay were included. Clinical and laboratory data at admission and during the hospital stay were analyzed. Clinical data evaluated were gender, age, signs and symptoms presented at admission, comorbidities, drugs and duration of ART and length of hospital stay. Laboratory data evaluated were hemoglobin, hematocrit, white blood count, lymphocytes, platelets, serum sodium, potassium, lactate dehydrogenase levels (LDH), alkaline phosphatase, aspartate amino transaminase, alanine amino transaminase, CD4 and viral loading and blood gas analysis. Renal function was analyzed by serum urea and creatinine and acute kidney injury (AKI) was classified according to RIFLE (risk, injury, failure, loss, and end stage renal disease) criteria (15). MA was defined as arterial pH !7.35 and arterial bicarbonate !22 mEq/L. Patients with HIV MA (MA-HIV) and non-MA (non-MA-HIV) were analyzed. We compared these two groups in order to investigate the clinical manifestations, laboratory features and outcome. The study was approved by the Ethical Committee of the Institution. Statistical Analysis Statistical analysis of clinical and laboratory data was performed using SPSS v. 10.0 (SPSS Inc., Chicago, IL) and Epi Info, 6.04b, 2001 (Centers for Disease Control and Prevention, Atlanta, GA). Between-group comparison were done using Student’s t-test, Mann-Whitney test and Fisher’s exact test. Results were expressed using tables and summary measures (mean  SD) in the cases of quantitative variables. A logistic regression model was used for quantitative variables. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. The factors included in the multivariate model were those that showed a significance level !20% in the univariate analysis (Mann-Whitney test and c2 test); values of p !0.05 were considered statistically significant. Results Included in the study were 159 HIV patients with a mean age of 36  10 years. There were 110 (69.1%) male patients. MA was found in 72 cases (45.2%). No difference in age or gender was observed between MA and non-MA groups. Hospital stay was longer among patients without MA ( p !0.0001). The main cause of hospital admission was opportunistic infections including pulmonary tuberculosis (17.6%), disseminated histoplasmosis (10.0%), Pneumocystis jiroveci pneumonia (6.9%), neurotoxoplasmosis (6.9%), bacterial pneumonia (5.0%), and others (gastrointestinal infections, visceral leishmaniasis, herpes zoster, varicella, candidiasis). There was no association of any particular disease with mortality. Clinical characteristics of patients with and without MA are summarized in Table 1. Serum urea and creatinine were significantly higher in MA-HIV than in non-MA-HIV groups ( p !0.05). CD4 cell count was not significantly different between MAHIV and non-MA-HIV groups ( p 5 0.09). Viral load was significantly higher in MA-HIV than non-MA-HIV group ( p 5 0.02). A comparison of laboratory data is shown in Table 2. ART was administered in 83 patients (52.2%), 38 in MA-HIV and 45 in non-MA-HIV ( p 5 0.57). The mean time of ART use was 2.7  1.2 years. There was no difference in the mean time of ART between MA-HIV and non-MA-HIV groups ( p 5 0.25; Table 1). There was no association between the use of ART and MA. Table 1. Clinical characteristics of HIV patients with and without MA MA-HIV (n 5 72) Non-MA-HIV (n 5 87) Age (years) Gender Male, n (%) Female, n (%) Length of hospital stay (days) Signs/symptoms Fever Chills Cough Dyspnea Diarrhea Vomiting Weight loss Oliguria Convulsions Headache Thoracic pain Abdominal pain ART use ART administration !1 year 1e5 years O5 years ART use (years) Death 37  8.4 36  12 49 (68.1%) 23 (31.9%) 18  16 61 (70.1%) 26 (29.9%) 24  22 49 (68.1%) 9 (12.5%) 35 (48.6%) 27(37.5%) 47 (65.3%) 38 (52.8%) 25 (34.7%) 2 (2.8%) 3 (4.2%) 13 (18.1%) 3 (4.2%) 20 (27.8%) 38 (52.8%) 55 16 53 27 44 29 40 3 3 13 10 14 45 12 (30.8%) 20 (51.3%) 7 (17.9%) 2.6  1.1 38 (52.7%) 6 (40%) 8 (53.3%) 1 (6.7%) 2.7  1.2 15 (17.2%) (63.2%) (18.4%) (60.9%) (31%) (50.6%) (33.3%) (46%) (3.4%) (3.4%) (14.9%) (11.5%) (16.3%) (51.7%) p 0.21 0.86 !0.0001 0.61 0.38 0.14 0.40 0.07 0.01 0.19 1.0 1.0 0.66 0.14 0.08 1.0 0.52 0.25 !0.0001 MA, metabolic acidosis; MA-HIV, HIV patients with MA; non-MA-HIV, HIV patients without MA; ART, antiretroviral therapy. Fisher’s exact test and Student’s t-test. Values expressed as mean  SD and %. p !0.05 was considered significant. 111 Metabolic Acidosis in AIDS Patients Table 2. Laboratory data in HIV patients with and without MA MA-HIV (n 5 72) Arterial pH 7.24 28 pCO2 (mm Hg) 12 HCO3 (mEq/L) 111 pO2 (mm Hg) 93 O2 saturation (%) Base excess (mEq/L) 13 Serum sodium (mEq/L) 133 Serum potassium 3.8 (mEq/L) Hemoglobin (g/dL) 10 Hematocrit (%) 30 7140 White blood count (mm3) 1165 Lymphocytes (mm3) 172 Platelets (103/mm3) LDH (U/L) 614 Urea (mg/dL) 81 Creatinine (mg/dL) 2.7 AP (U/L) 281 AST (IU/L) 162 ALT (IU/L) 106 228 CD4 (cells/mm3) 147,313 Viral loading (copies/mm3)         0.08 15 5.7 51 7.2 5.4 8.9 1.1  2.5  7.3  5407           Non-MA-HIV (n 5 87) 7.44 31 21 93 95 1.5 130 3.7         0.05 7.8 5.1 42 6.8 4.7 6.1 0.7 9.9  2.7 28  8.7 5478  3279 140 566 129 204 709 881 68 38 2.6 1.2 354 377 171 132 165 107 239 185 32,106 60,651           Table 4. Risk factors for death in patients in HIV with and without MA by univariate logistic regression p !0.001 0.007 !0.001 0.02 0.16 !0.001 0.03 0.67 0.35 0.16 0.15 132 !0.0001 128 0.10 1381 0.45 45 !0.0001 1.9 !0.0001 292 0.40 263 0.16 318 0.21 283 0.09 39,107 0.02 MA-HIV, HIV patients with MA; non-MA-HIV, HIV patients without MA; AP, alkaline phosphatase; AST, aspartate amino transaminase; ALT, alanine amino transaminase; LDH, lactate dehydrogenase. Fisher exact test, Student t-test and Mann-Whitney test. Values expressed as mean  SD. Significant at p !0.05. AKI was found in 53 patients (33.3%). The majority of patients with AKI had MA (66%). Patients were classified according to Risk (26.5%), Injury (24.5%) and Failure (49%) according to RIFLE classification. Failure was significantly high in MA-HIV group ( p !0.0001) (Table 3). Death occurred in 53 cases (33.3%). Mortality was higher in patients with MA (52.7% vs. 17.2%, p !0.0001). Univariate logistic regression showed that the predictors for death were high serum level of potassium and urea and low level of bicarbonate and platelets. These data are shown in Table 4. Multivariate regression model was built with the variables that showed a significance level !20% in the univariate analysis (ART use, CD4 count, pH, HCO3, hyperkalemia, hemoglobin level, platelet count and urea Table 3. Classification of HIV patients with and without MAassociated AKI according to RIFLE ‘‘Risk’’, n 5 14 (26.5%) ‘‘Injury’’, n 5 13 (24.5%) ‘‘Failure’’, n 5 26 (49%) MA-HIV (n 5 35) Non-MA-HIV (n 5 18) p 8 (57.1%) 7 (53.8%) 20 (76.9%) 6 (42.9%) 6 (46.2%) 6 (23.1%) 0.44 0.69 0.0001 AKI, acute kidney injury. MA-HIV, HIV patients with MA; non-MA-HIV, HIV patients without MA; Fischer exact test. Significant at p !0.05. HCO3 High urea Platelets Hyperkalemia CD4 level OR 95% CI p 0.95 1.008 0.999 1.54 0.998 0.903e0.999 1.002e1.015 0.998e0.999 1.025e2.321 0.996e1.001 0.046 0.013 !0.001 0.038 0.208 OR, odds ratio; CI, confidence interval. at admission). The independent risk factors for death were MA, hyperkalemia and low CD4 count (Table 5). Discussion In our data we identified that MA in HIV-infected patients is prevalent in patients with AKI, especially in those with its severe form (failure in RIFLE classification). Moreover, it was independently associated with a higher mortality in HIV hospitalized patients. Clinical manifestations related to MA are nonspecific, such as gastrointestinal symptoms present in O50% of cases (16). Dyspnea and tachypnea classically associated with MA are seen in 41% of cases in the same review (16). In the present study, the most frequent clinical manifestations found in patients with MA were diarrhea, vomiting, weight loss and dyspnea. The only manifestation that was statistically different between the two groups was vomiting, although it was classically associated with metabolic alkalosis. This symptom can be more a consequence of acidosis than its etiology. Several studies showed an association between MA and the use of ART, mainly the NRTIs (3,17e26). In the present study we were unable to determine any relationship between any ART drug and MA. In a population study, this suggests LA is not as frequent as other causes of MA. Otherwise, renal dysfunction was more frequent in patients with acidosis, especially in its severe form, suggesting this is greatly responsible for MA in this population causing MA directly due non-measurable anion accumulation or through higher serum levels of ART drugs. Lactate levels were unavailable because lactase measurement was not performed by the hospital where the study was performed. In the absence of lactate monitoring, periodic aminotransferase measurements are recommended (27,28). Table 5. Independent risk factors for death in HIV patients with and without MA by multivariate logistic regression MA Hyperkalemia CD4 level OR 95% CI p 10.32 5.4268 0.9916 2.09e50.92 1.63e18.00 0.9850e0.9983 0.0042 0.0057 0.0136 112 Daher et al./ Archives of Medical Research 40 (2009) 109e113 RIFLE classification is an improvement in predicting outcome of patients with AKI (29,30). AKI classified as ‘‘failure’’ was significantly high in MA-HIV group. This finding strongly supports the hypothesis that MA is mainly due to renal dysfunction. Because the number of patients in each group according to RIFLE criteria is small, association with mortality may be significant if the sample was larger. In our study, average hospital stay was higher among patients without MA. These data corroborate with severe clinical conditions of those patients with MA-HIV. Therefore, mortality was higher in patients with MA (52%). Our study is similar to previous reviews that have reported fatality ratio associated with LA in 33e60% (16,26,31). Independent risk factors for death were MA, hyperkalemia and low CD4. There are few studies to investigate the factors associated with mortality in HIV patients. Peak venous lactate level was the best predictor of mortality in HIV patients. Zidovudine is associated with higher lactate levels and higher mortality than stavudine and lamivudine (32). The presence of hypokalemia, MA, and renal failure are significantly associated with mortality in AIDS (2). This study has some limitations: 1) it is a retrospective study, being difficult to define a cause-effect relationship; 2) we do not have chloride values to perform anion gap analyses; 3) lactate levels were not measured, making it impossible to evaluate the exact importance of its accumulation in MA. In summary, MA is an important complication observed in HIV patients, which per se increases mortality. It was not directly associated with ART in the present study, although the lack of association is possible due to its low incidence compared with other causes of MA. There was a significant association between MA and AKI. It is possible that renal dysfunction plays an important role in the pathogenesis of HIV-associated MA. It is important to search for the occurrence of MA in all hospitalized HIV patients, investigate its causes and promptly correct it. Further studies, including a randomized controlled trial, are required to better establish the etiology of MA in HIV. Acknowledgments We are very grateful to the team of physicians, residents, medical students and nurses of the Hospital S~ao José de Doenças Infecciosas for the assistance provided to the patients. This research was financially supported by the Brazilian Research Council (Conselho Nacional de Desenvolvimento Cientı́fico e Tecnológico-CNPq, Brazil). References 1. Marra A, Lewi D, Lanzoni V, et al. Lactic acidosis and antiretroviral therapy: a case report and literature review. Braz J Infect Dis 2000;4: 151e155. 2. Manfro RC, Stumpf AG, Horn CL, et al. Hydroelectrolyte, acid-base, and renal function changes in patients with acquired immunodeficiency syndrome. Rev Assoc Med Bras 1992;39:43e47. 3. Roca B. Trastornos metabólicos relacionados con el VIH y el tratamiento antirretroviral. An Med Interna (Madrid) 2003;20:585e593. 4. Hirsch HH, Battegay M. Lipodystrophy syndrome by HAART in HIV infected patients: manifestation, mechanisms and management. Infection 2002;30:293e298. 5. Lai KK, Gang DL, Zawacki JK, Cooley TP. Fulminant hepatic failure associated with 2,3-dideoxyinosine (ddI). Ann Intern Med 1991;4: 115e283. 6. Johri S, Alkhuja S, Siviglia G, Soni A. Steatosis-lactic acidosis syndrome associated with stavudine and lamivudine therapy. AIDS 2000;14:1286e1287. 7. Coghlan ME, Sommadossi JP, Jhala NC, Many WJ, Saag MS, Johnson VA. Symptomatic lactic acidosis in hospitalized antiretroviral-treated patients with human immunodeficiency virus infection: a report of 12 cases. Clin Infect Dis 2001;33:1914e1921. 8. Acosta BS, Grimsley EW. Zidovudine-associated type B lactic acidosis and hepatic steatosis in an HIV-infected patient. South Med J 1999;92:421e423. 9. Miller KD, Cameron M, Wood LV, Dalakas MC, Kovacs JA. Lactic acidosis and hepatic steatosis associated with use of stavudine: report of four cases. Ann Intern Med 2000;133:192e196. 10. Honkoop P, Scholte HR, de Man RA, Schalm SW. Mitochondrial injury. Lessons from the fialuridine trial. Drug Saf 1997;17:1e7. 11. Vrouenraets S, Treskes M, Regez RM, et al. The occurrence of hyperlactatemia in HIV-infected patients on NRTI treatment. XIII International AIDS Conference: Durban 2000. 12. Datta D, Mandalia S, Morlese J, et al. Biochemical abnormalities associated with hyperlactataemia in HIV-1 positive patients. 1st IAS Conference on HIV Pathogenesis and Treatment: Buenos Aires: July 2001. 13. Moyle GJ, Datta D, Mandalia S, et al. Hyperlactataemia and lactic acidosis during antiretroviral therapy: relevance, reproducibility and possible risk factors. 3rd International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV. Athens: 2001 (abstract 98). 14. Morgello S, Wolfe D, Godfrey E, Feinstein R, Tagliati M, Simpson DM. Mitochondrial abnormalities in human immunodeficiency virus-associated myopathy. Acta Neuropathol (Berl) 1995;90:366e374. 15. Kellum JA, Bellomo R, Ronco C. Classification of acute kidney injury using RIFLE: what’s the purpose? Crit Care Med 2007;35:1983e1984. 16. Arenas-Pinto A, Grant AD, Edwards S, Weller IV. Lactic acidosis in HIV infected patients: a systematic review of published case. Sex Transm Infect 2003;79:340e344. 17. Marceau G, Sapin V, Jacomet C, et al. Frequency, risk factors, and outcome of hyperlactatemia in HIV-positive persons: implications for the management of treated patients. Clin Chem 2003;49:1154e1162. 18. White AJ. Mitochondrial toxicity and HIV therapy. Sex Transm Inf 2001;77:158e173. 19. Chariot P, Bourokba N, Monnet I, Gherardi R. Hyperlactatemia and human immunodeficiency virus infection: lessons from the era of antiretroviral monotherapy. Antimicrob Agents Chemother 2002;46:3683. 20. Claessens YE, Chiche JD, Mira JP, Cariou A. Bench-to-bedside review: severe lactic acidosis in HIV patients treated with nucleoside analogue reverse transcriptase inhibitors. Crit Care 2003;7:226e232. 21. Leonard EG, McComsey GA. Antiretroviral therapy in HIV-infected children: the metabolic cost of improved survival. Infect Dis Clin North Am 2005;19:713e729. 22. Mendoza C, Blanco F, Soriano V. Toxicidad mitocondrial de los antirretrovirales: diagnóstico y monitorización. Med Clin (Barc) 2003;121:310e315. 23. Geddes R, Knight S, Moosa MYS, Reddi A, Uebel K, Sunpath H. A high incidence of nucleoside reverse transcriptase inhibitor (NRTI)-induced lactic acidosis in HIV-infected patients in a South African context. S Afr Med J 2006;96:722e724. 24. Eshun-Wilson I, Soentjens P, Zeier M, Taljaard J. Symptomatic hyperlactataemia and lactic acidosis in the era of highly active antiretroviral therapy. S Afr Med J 2005;95:929e930. Metabolic Acidosis in AIDS Patients 25. Sundar K, Suarez M, Banogon PE, Shapiro JM. Zidovudineinduced fatal lactic acidosis and hepatic failure in patients with acquired immunodeficiency syndrome: report of two patients and review of the literature. Crit Care Med 1997;25:1425e 1430. 26. Mégarbane B, Brivet F, Guérin JM, et al. Acidose lactique et défaillance multi-viscérale secondaire aux thérapeutiques antirétrovirales chez les patients infectés par le VIH. Presse Med 1999;28:2257e 2264. 27. Chattha G, Arieff AI, Cummings C. Tierney LM Jr. Lactic acidosis complicating the acquired immunodeficiency syndrome. Ann Intern Med 1993;118:37e39. 28. Roy PM, Gouello JP, Pennison-Besnier I, Chennebault JM. Severe lactic acidosis induced by nucleoside analogues in an HIV-infected man. Ann Emerg Med 1999;34:282e284. 113 29. Hoste EA, Clermont G, Kersten A, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006;10:R73. 30. Abosaif NY, Tolba YA, Heap M, Russell J, El Nahas AM. The outcome of acute renal failure in the intensive care unit according to RIFLE: model application, sensitivity, and predictability. Am J Kidney Dis 2005;46:1038e1048. 31. Falcó V, Rodrı́guez D, Ribera E, et al. Severe nucleoside-associated lactic acidosis in human immunodeficiency virus-infected patients: report of 12 cases and review of the literature. Clin Infect Dis 2002; 34:838e846. 32. Tripuraneni NS, Smith PR, Weedon J, Rosa U, Sepkowitz D. Prognostic factors in lactic acidosis syndrome caused by nucleoside reverse transcriptase inhibitors: report of eight cases and review of the literature. AIDS Patient Care STDS 2004;18:379e384.