| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
1
Biochemistry,
2 Infectious Diseases and Tropical Medicine,
3 Medical Information, and
4 Virology, Faculty of Medicine, University Hospital of Clermont-Ferrand, 63000 Clermont-Ferrand, France.
aAddress correspondence to this author at: Laboratoire de Biochimie Médicale, Faculté de Médecine, 28, Place Henri Dunant, BP. 38, 63000 Clermont-Ferrand; France. Fax 33-4-7327-6132; e-mail visapin{at}courrier.u-clermont1.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We conducted a prospective cross-sectional study of venous lactate determinations with 282 consecutive HIV-positive persons who, in addition to a physical examination, had blood samples taken every 3–4 months for routine biochemical, immunologic, and viral assessment. The frequencies of hyperlactatemia and lactic acidosis were determined, and the risk factors were analyzed by a multivariate logistic regression model. The effect of modification of antiretroviral therapy in patients with moderate hyperlactatemia was also assessed.
Results: From 782 blood lactate determinations, we identified 65 (23%) patients with moderate hyperlactatemia and 5 (1.8%) with lactate concentrations >5 mmol/L (2 with severe lactic acidosis; 0.7%). Older age, drug regimens containing stavudine [adjusted odds ratio (OR) = 2.5] or a combination of stavudine-didanosine (adjusted OR = 3.1), and the use of buprenorphine (adjusted OR = 14.7) were independent predictors of hyperlactatemia. Among 65 patients with moderate hyperlactatemia, 39 did not have their treatments changed, and 26 had a new combination therapy that was associated with a clinical improvement and a more pronounced decrease in lactate (-1.66 vs -0.99 mmol/L; P <0.05).
Conclusions: Chronic compensated and moderate hyperlactatemia was common in our population study. Measurement of lactate, under standardized conditions, may be useful in optimizing management of HIV-positive persons on antiretroviral therapy.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(2)(3)(4). The active triphosphorylated forms of these nucleoside analogs are thought to act as a substrate of cellular polymerase , thus inducing mitochondrial toxicity by inhibiting polymerase , which subsequently leads to mitochondrial DNA depletion and dysfunction of the respiratory chain. A clinical syndrome related to NRTI toxicity, called type B lactic acidosis, includes myopathy, hepatic macrovesicular and microvesicular steatosis, liver dysfunction, and/or fulminant liver failure and pancreatitis (5)(6).
From a clinical point of view, clinically relevant hyperlactatemia may be defined as a lactate concentration of 2.25–5 mmol/L, and severe lactic acidosis (SLA) is defined as a lactate concentration >5 mmol/L with a pH <7.3 or bicarbonate <20 mmol/L and clinical signs of multiorgan failure (7). Previous reports showed that the incidence of increased serum lactate is 15–35% in NRTI-treated patients (7), with a clinically relevant hyperlactatemia prevalence of 1–3% (8) and a SLA prevalence of 
3.9 per 1000 persons; 40–60% of the SLA patients with multiorgan failure have died (9).
Recent studies, however, have found that SLA can occur rapidly, without previous prolonged a- or paucisymptomatic hyperlactatemia (9)(10). It has therefore been suggested that routine lactate measurements in the clinical monitoring of HIV-positive persons during ARV therapy may not be useful in preventing lactic acidosis but may be useful for recognizing paucisymptomatic hyperlactatemia (7)(8)(9)(10)(11).
We therefore conducted a prospective study to estimate the prevalence of hyperlactatemia in HIV-positive persons on NRTI-containing ARV therapy, to describe laboratory and clinical disorders associated with increased blood lactate, to identify risk factors associated with hyperlactatemia and/or lactic acidosis, and to assess the outcome of clinicians’ decisions after the recognition of clinically relevant hyperlactatemia or SLA to assess the potential benefits of routine screening of blood lactate concentrations.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
study design
This prospective cross-sectional study enrolled HIV-positive persons ≥18 years of age during a 12-month period from July 1, 2000, to June 30, 2001, with an additional 9-month follow-up of hyperlactatemic patients until March 2002. After providing written consent, patients routinely enrolled in the French National HIV cohort and followed at the University Hospital of Clermont-Ferrand had lactate determinations in addition to their routine biological (creatinine, urate, transaminases, alkaline phosphatases, bilirubin, pancreatic amylase, creatinine phosphokinase, chloride, bicarbonate, glucose, triglycerides, and cholesterol) and immunologic/virologic (CD4 cell count and HIV virus load) assessment.
Other variables evaluated for this study were age; sex; weight; likely route of HIV transmission; history of opportunistic infections (CDC/WHO classification modified in 1993); abnormal signs and symptoms, including asthenia, lipodystrophy [defined by fat loss involving the face, and/or legs, and/or arms; fat accumulation involving the neck (Buffalo hump), and/or breast, and/or abdomen; or a combination these two syndromes (12)]; gastrointestinal, muscular, and respiratory symptoms; non-HIVrelated illnesses; chronic hepatitis B and C; CD4 T-cell count and plasma HIV RNA load; and history of and current ARV treatment and other drugs taken.
Physicians who cared for patients who were identified with confirmed hyperlactatemia (two increased blood lactate values on two different samples taken at two different dates) had the opportunity either to modify or to maintain the ARV treatment. This medical decision was made after medical examination and discussion between clinicians and patients and was based on the severity of patients’ clinical signs and symptoms, on biological indices (immunology, virology), and on the lactate concentrations. Patients with confirmed hyperlactatemia during the study period were systematically followed during an additional 9-month period to assess changes in lactate after the therapeutic decision.
screening of lactate
To obtain strong reproducibility of lactate determinations during the study, we standardized preanalytical and analytical conditions. Venous blood lactate samples were collected by standard venipuncture into sodium fluoride/potassium oxalate tubes. Nurses were instructed to ensure that patients had been fasting for more than 12 h and were at rest for at least 15 min before collection without a tourniquet or fist clenching. Samples were put on ice and immediately transported to the adjacent biochemistry laboratory, where they were centrifuged (900g for 10 min at 4 °C) and analyzed. The time between blood collection and lactate analysis was <20 min. Blood lactate concentrations were determined on an automated clinical chemistry analyzer (Hitachi 912; Roche Diagnostic) based on the enzymatic conversion of L-lactate to pyruvate by a specific L-lactate oxidase (13).
Venous blood samples collected under conditions other than the previously described preanalytical and analytical conditions and/or lactate determinations for patients with current conditions (acute infections, ischemia, malignancies, or acute liver disease) known to modify lactatemia were prospectively excluded from the study.
case definitions
The reference interval for plasma lactate had been determined previously in our laboratory from 200 venous blood samples from a healthy population (sex ratio = 1; age range, 18–45 years). The range for lactate was 0.5–2.25 mmol/L [mean (SD), 1.40 (0.30) mmol/L] (14). Concentrations <2.25 mmol/L were defined as normolactatemic, concentrations between 2.25 and 5 mmol/L were considered as moderately increased (medium-lactate group), and lactate >5 mmol/L was considered highly increased (high-lactate group).
Abnormal lactate values were confirmed by a second determination. Additional blood lactate measurements were performed during patient follow-ups, when hyperlactatemia (≥2.25 mmol/L) was confirmed, and when there were clinical signs or symptoms potentially associated with hyperlactatemia, including asthenia, gastrointestinal symptoms (nausea and diarrhea), weight loss >10%, muscular symptoms (myalgia and cramps), or respiratory symptoms (cough and dyspnea).
For all patients in the high-lactate group (n = 5) and eight randomized patients in the medium-lactate group with confirmed increased lactate, biochemical evaluations of mitochondrial dysfunction were performed according to the protocol described by Touati et al. (15). Lactate, pyruvate, ß-hydroxybutyrate (ßOH), and acetoacetate (AA) were measured before and after morning meals and lunches that were glucose energy supply meals. Lactate/pyruvate and ßOH/AA blood ratios were calculated and correlated with glycemia and indices of acid–basis equilibrium.
statistical analysis
Comparisons of qualitative indicators between the groups were performed with the 
2 test, and odds ratios (ORs) and their 95% confidence interval (CIs) were determined with SAS, Ver. 8.02. Quantitative indices were analyzed with use of mean comparisons in the Student t-test. Statistical significance was defined as P <0.05. Multivariate analysis was performed with a regression logistic model.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
.
|
During the period study, 785 lactate determinations were obtained for the 282 patients, for a mean of 2.8 lactate determinations per patient (range, 1–11). The mean (SD) lactate concentration for the population study was 1.72 (1.1) mmol/L.
As a whole, 212 (75.2%) patients were normolactatemic [mean (SD), 1.31 (0.43) mmol/L], 65 (23%) had moderately increased serum lactate [2.95 (0.63) mmol/L], and 5 (1.8%) had high serum lactate [6.86 (1.67) mmol/L]. The mean lactate values of the medium- and high-lactate patient groups were statistically different from the mean values of the normolactatemic group (P <0.05) and from the reference values [mean (SD), 1.40 (0.30) mmol/L] obtained from a healthy population with no biological disorders (P <0.05).
The prevalence of hyperlactatemia (lactate ≥2.25 mmol/L) in the whole population was 24.8% (70 of 282) during the study period. At inclusion in the study, patients treated with ARV drugs, including NRTIs (87%; 245 of 282), tended to have higher lactate concentrations [mean (SD), 1.75 (1.14) mmol/L] than untreated (7.5%; 21 of 282) patients [mean (SD), 1.53 (0.97) mmol/L] as well as ARV-naive (5.5%, 16 of 282) patients [1.44 (0.97) mmol/L], but the differences were not statistically significant. The 1-year cumulative incidence of hyperlactatemia (lactate ≥2.25 mmol/L), i.e., the number of patients who became hyperlactatemic during the cross-sectional study, was 14% among HIV-positive persons undergoing treatment.
The proportion of patients whose lactate/pyruvate and ßOH/AA blood ratios disclosed alterations of mitochondrial function (defined by lactate/pyruvate ratio >30 and ßOH/AA ratio >2) was significantly (P <0.05) higher in the high-lactate group (40%) than in the medium-lactate group (18%).
We identified an association between hyperlactatemia (lactate ≥2.25 mmol/L) and lipodystrophy, weight loss from the initiation of the current regimen (>10%), gastrointestinal symptoms (vomiting, nausea, and diarrhea), respiratory symptoms (cough and dyspnea), and muscular symptoms (myalgia and cramps; Table 2
). By contrast, increased lactate was not associated with fever, rash, asthenia, neurologic symptoms (dizziness and neuropathy), hepatitis C, and sleep disorders. The presence of diabetes (using WHO recommendations based on blood glucose determinations) was also statistically associated with a higher risk of hyperlactatemia (lactate ≥2.25 mmol/L). Several analytes, such as glucose, chloride, bilirubin, triglycerides, alanine aminotransferase, alkaline phosphatase, and bicarbonates, were significantly different between the normolactatemic group and the group of patients with increased lactate (>2.25 mmol/L; Table 2
). By contrast, increased lactate was not associated with differences in blood cholesterol, amylase, creatine kinase, urea, and creatinine.
|
Four patients with hyperlactatemia (lactate >5 mmol/L) were admitted to an intensive care unit to manage disorders associated with the hyperlactatemia. Over the study period, two patients developed SLA with multiple organ failure. The 1-year cumulative incidence of lactic acidosis was 0.7%. Three patients with lactate >5 mmol/L (60%) had chronic hepatitis C. One of them was receiving interferon–ribavirin therapy at the time of admission, one had been treated with interferon–ribavirin a few weeks previously, and the third patient had never been treated. No deaths were related to SLA. Three deaths (1.1%) were reported during the study period. These were attributable to liver disease, opportunistic infection, and a non-Hodgkin lymphoma.
Univariate analysis also showed that patients with increased lactate (≥2.25 mmol/L) were significantly older (mean age, 47 vs 43 years), were more likely (P <0.05) to receive a stavudine- or stavudine–didanosine-containing regimen, or were more likely to receive buprenorphine than patients with a lactate value within the reference interval (Table 2
). By contrast, increased lactate was not associated with treatment with lamivudine, zidovudine, abacavir, zalcitabine, efavirenz, nevirapine, and interferon and/or ribavirin. Multivariate analysis using a regression logistic model included significant variables in the univariate analysis. Two models were used. The first model, which included all statistically significant variables except the combination therapy stavudine–didanosine, identified the following as independent predictors of hyperlactatemia: age, stavudine-containing regimen, and the use of buprenorphine (Table 3
). The second model, which included all statistically significant variables and the combination stavudine–didanosine, identified only the combination stavudine–didanosine as a predictor of hyperlactatemia (Table 3
).
|
All patients with lactate >5 mmol/L had all ARV therapy interrupted and were prescribed, in addition to supportive care, a combination of L-carnitine, vitamins B1 and B2, folic acid, biotin, and enzymatic cofactor Q. All had progressive decreases in their lactate concentrations, which were associated with clinical, biological, and mitochondrial function improvements (defined by normalization of mitochondrial tests). After the normalization of mitochondrial tests (defined by lactate/pyruvate ratio <30 and ßOH/AA ratio <2), an effective combination therapy (one antiprotease and two NRTIs, or three NRTIs excluding either stavudine or didanosine) was reintroduced several weeks to months later, and no lactic acidosis occurred (median follow-up, 6 months).
No patients with lactate between 2.25 and 5 mmol/L (medium-lactate group) developed SLA during the study period. Among the 65 patients in the medium-lactate group, 39 (60%; mean lactate concentration, 2.9 mmol/L; range, 2.3–4.3 mmol/L) did not have their treatment modified during the clinical follow-up. Seven patients had signs and symptoms associated with hyperlactatemia: lipodystrophy (5), neuromuscular signs (1), and diabetes mellitus (1). Seven patients were receiving a stavudine-containing regimen, and one was receiving a combination of stavudine and didanosine. The remaining 26 patients (mean lactate, 3.6 mmol/L; range, 2.3–4.9 mmol/L) were more frequently symptomatic (19 of 26) and had their treatment modified during their clinical follow-up. Twenty-two patients were on stavudine; among these, 15 were on the combination stavudine–didanosine. The stavudine and combination stavudine–didanosine therapies were interrupted, and patients were offered a combination of one or two other reverse transcriptase inhibitor(s) and a single- or double-protease inhibitor regimen.
The decrease in lactate concentrations among the 26 patients who had their treatment changed was statistically (P <0.05) greater than that of the 39 patients for whom treatment was unchanged [-1.66 (1.63) vs -0.99 (0.80) mmol/L]. In addition, the magnitude of the decrease in the lactate concentrations was correlated with a clinical improvement (defined by the disappearance of at least one of the three types of symptoms: gastrointestinal, muscular, or respiratory). Patients who had their ARV treatment modified reported a significantly higher rate of clinical improvement 9 months later than patients who did not have their treatment changed (88% vs 66%; P <0.05).
Finally, available lactate concentrations from our population study (282 patients) on March 31, 2002, were within the reference values (<2.25 mmol/L) in 276 (97%) patients and moderately increased in 6 (3%) patients. No patients had a lactate >5 mmol/L. The mean (SD) lactate concentrations were 1.26 (0.43) mmol/L in the normolactatemic group and 2.63 (0.60) mmol/L in the medium-lactate group.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The general application of our results will rely on several criteria. The characteristics of our population study do not differ from those of previously published cohorts in terms of sex ratio, weight, likely route of HIV transmission, CDC/WHO classification, CD4 cell count, HIV RNA viral load, co-infections with hepatitis C and or B virus, and ARV and anti-hepatitis C treatment (9)(16)(17)(18)(19). Moreover, for blood lactate determinations, preanalytical and analytical conditions were standardized to provide consistent results. Blood samples were put on ice and transferred immediately to the adjacent laboratory. Although lactate values remain stable for at least 8 h at room temperature in tubes containing sodium fluoride, blood lactate concentrations may be seriously increased by the ischemia caused by a tourniquet or decreased by the storage of blood without sodium fluoride (20). Some authors have even proposed that measurement of the lactate/pyruvate ratio during functional respiratory or energy supply tests seem to be more sensitive than simple blood lactate measurements for evaluating mitochondrial dysfunction (21). In our study, mitochondrial tests were performed in patients with severe hyperlactatemia. After interruption of the ARV therapy in those patients, decreases in lactate concentrations were associated with normalization of the molar lactate/pyruvate ratio under basal conditions and after energy supply.
Our study was able to identify the age of the patients as an independent predictor of hyperlactatemia, with older patients being more likely to develop hyperlactatemia. To our knowledge, this has been reported only once (10). Hyperlactatemia is a marker of altered mitochondrial function. Because NRTIs are known to be potential substrates of polymerase , inhibition of this polymerase decreases mitochondrial DNA concentrations and, thus, the synthesis of mitochondrial enzyme protein subunits encoded by mitochondrial DNA (22)(23). Alterations in mitochondrial function have been described during aging (24). This may explain why older patients with physiologic alterations in their mitochondrial capacities are more likely to have higher lactate concentrations with NRTIs than younger patients.
Stavudine-containing regimens and the association of stavudine and didanosine in a combination therapy were also identified as strong predictors of hyperlactatemia. In addition, these drugs were involved in all episodes of lactate >5 mmol/L and all episodes of lactic acidosis reported in our study. We consistently observed a dramatic reduction in the prevalence of increased lactate after a switch in therapy strategies: the frequency decreased from 24% to 3% at the end of the follow-up period, as reported previously (21)(25). The mitochondrial toxicity of the combination stavudine–didanosine has a major impact in terms of therapeutic strategies in naive and pretreated patients with therapeutic failure (26). Thus, this combination is now not recommended for any patients in France, particularly for pregnant women at increased risk for lactic acidosis or for postexposure prophylaxis (12).
Buprenorphine was also found to be associated with the occurrence of hyperlactatemia. Further studies are warranted to confirm this finding. It has already been demonstrated in rat liver that buprenorphine impairs mitochondrial respiration and ATP formation and that the hepatotoxicity of high concentrations of buprenorphine is related mainly to its mitochondrial effects (27). In addition, a clinical case report suggested that high concentrations of buprenorphine can trigger hepatitis, particularly in patients whose mitochondrial function might be impaired by hepatitis C or NRTIs (28). Consistently, all of our patients on buprenorphine had chronic viral hepatitis C related to unsafe injection of drugs.
The present study has several limitations. Not all of the patients in our cohort were included during the study period because some of them did not fulfill the inclusion criteria. This might have led to a higher rate of hyperlactatemia and might have impaired external validity. Patients included had different numbers of lactate determinations. Lactate determinations were performed from venous samples and not from arterial samples. No muscle, liver, or other tissue biopsies were performed in patients with lactic acidosis to seek evidence of mitochondrial dysfunction at the tissue level. In addition, the analysis of risk factors for hyperlactatemia did not take into account previous exposure to NRTIs, the duration of this preexposure in treated patients, or the use of hydroxyurea, which has been reported to be a risk factor for symptomatic hyperlactatemia (29). However, we do not use hydroxyurea in combination with ARV drugs in our center, and no patient included in this study had ever been offered hydroxyurea.
Mild hyperlactatemia (lactate <5 mmol/L) found in a- or paucisymptomatic patients is known to have a poor positive predictive value for the development of SLA (9). This does not support systematic lactate determinations in addition to routine immunologic/virologic assessment in treated patients. This is consistent with the current French National guidelines published in June 2002, which do not recommend routine lactate determinations for the management of treated HIV-positive persons (12).
Physicians and biochemists involved in the study were more likely to consider a trend in lactate concentrations, rather than a single value in a symptomatic patient at high risk of developing lactic acidosis, to monitor side effects associated with ARV treatment. Physicians need to make a timely diagnosis of symptomatic hyperlactatemia: an increase in successive lactate measurements in combination with the presence of clinical symptoms should alert the clinician to a possible need to revise the ARV therapy. We therefore suggest that a baseline lactate measurement be performed at the time of initiating treatment. Taking into account the poor positive predictive value of mild hyperlactatemia in asymptomatic patients for the development of SLA, lactate measurements might be performed only every 6 months under standardized conditions in asymptomatic patients to detect a trend of increasing lactate concentrations. In addition, lactate should be measured as soon as clinical manifestations occur that suggest possible symptomatic hyperlactatemia. These determinations are rapid, simple, and inexpensive and may lead to modification or cessation of the ARV treatment.
Within this perspective, we propose a new algorithm, compared with the previously reported algorithm (7) that we currently use in practice (Fig. 1
), based on lactate concentrations and on the presence or absence of clinical signs and symptoms suggesting symptomatic hyperlactatemia to optimize the management of patients on NRTIs. With this new algorithm, some improvements could be noted. For example, since the end of the study and the beginning of our use of the new clinical and biological scheme to monitor HIV-positive persons treated by NRTIs, no new clinical cases of SLA have been observed and our SLA frequency has been lower than the frequencies described in the literature. In addition, the use of blood lactate as a tool for monitoring HIV-positive patients treated by NRTIs led to an increase in patients with a lactate values within the reference interval [276 (97.0%) vs 212 (75.2%) at the beginning of the study].
|
In conclusion, we have demonstrated that chronic compensated and moderate hyperlactatemia is common in HIV-infected patients treated with NRTIs. Measurement of lactate under standardized conditions may be useful to optimize the management of these HIV-infected patients on NRTI therapy.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. Prod'homme, C. Rochon, M. Balage, H. Laurichesse, I. Tauveron, C. Champredon, P. Thieblot, J. Beytout, and J. Grizard Whole body leucine flux in HIV-infected patients treated with or without protease inhibitors Am J Physiol Endocrinol Metab, April 1, 2006; 290(4): E685 - E693. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Fall and H. M. Szerlip Lactic Acidosis: From Sour Milk to Septic Shock J Intensive Care Med, September 1, 2005; 20(5): 255 - 271. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
