A 38-year-old woman presents to the prison infirmary with flu-like complaints consisting of myalgias, abdominal pain, and nausea. Her medical history is significant for Class A2 HIV disease. She has had no HIV-related illnesses except for chronic, mild diarrhea. Her CD4 T-cell count is 512 (32%), and HIV RNA by bDNA is less than 50. Her antiretroviral regimen consists of stavudine (d4T), lamivudine (3TC), and indinavir/ritonavir (IDV/RTV). Her only other illness is asymptomatic chronic hepatitis C virus (HCV) infection. She first reported to the infirmary one week ago, complaining of one week of flu-like illness. Total time since onset is two weeks. On examination today, the patient appears moderately ill. Her blood pressure is 112/64, pulse 105, respirations 22, and temperature 37.1 C. Her right upper abdomen is tender to palpation. Because there has been no improvement in her symptoms, her medical care provider decides that further investigation is warranted.
If this were your patient, what would you be concerned about right now, and how would you proceed to make the diagnosis?
It was fortunate for this woman that the alert infirmary provider did not send her back to her room to wait a little longer for the "flu" to resolve. The provider in this case promptly checked a battery of laboratory tests. The patient's urinary analysis was leukocyte esterase trace-positive and positive for ketones. Serum AST (92 units), ALT (118 units), total bilirubin 1.7 (units), and amylase normal. BUN/creatinine were 27/1.6 units, and electrolytes were Na 139 units, Cl 104, K 4.3, and HCO3 12.; The anion gap (AG) was computed as follows:
AG = (Na + K) - (Cl + HCO3) or AG = Na - (Cl + HCO3),
AG normal value < +/- 12mEq/L, or
< 16 mEq/L if potassium concentration is used to calculate value
The result is 23. The correct diagnosis for this woman's illness is gap acidemia.
The medical provider held the patient's medications and sent her to the hospital for further evaluation and treatment.
In the hospital, the medical team obtained a serum venous lactic acid level, following the ACTG Guidelines Protocol (http://aactg.s-3.com/members/psmet.htm). The reading was 6.1 mmol/L (normal is 0.5 to 2.5 mmol/L at this institution). An ultrasound of her abdomen revealed hepatomegaly with fatty infiltration. A diagnosis of severe nucleoside reverse transcriptase inhibitor (NRTI)-induced lactic acidemia with hepatic steatosis was confirmed and the patient's antiretroviral medications were discontinued.
What Is the Most Likely Cause of Anion Gap Acidosis in This Patient?
An acidosis can be caused by several things. Medication-induced renal tubular acidosis and chronic diarrhea can cause an acidosis, but not a gap acidosis, as seen here. Sepsis can cause lactic with an anion gap, but this patient did not appear to have sepsis clinically. This patient's gap acidosis was caused by medication-induced lactic academia.
Lactic Academia and Mitochondrial Toxicity in HIV Infection
The mitochondria are the body's battery packs, producing and storing adenosine triphosphate (ATP) through a (non-lactate-producing) process of cellular respiration and breaking it down when energy is required. The number of mitochondria in cells of any particular tissue varies, depending on tissue energy requirements. Cells of relatively quiescent tissue may contain only a few mitochondria; the cells of tissues with higher energy requirements -- such as muscle, liver, and nerves -- may contain thousands of mitochondria. Mitochondria have their own DNA strands (mtDNA) that are replicated by the enzyme polymerase gamma (pol gamma). Pol gamma is very similar to the HIV polymerase reverse transcriptase.
NRTIs that inhibit HIV reverse transcriptase may also interfere with polymerase gamma, apparently causing mutations in the replicating mitochondria. Mitochondria do not have a mechanism for correcting replication mutations caused by the NRTIs. As the number of non-functioning mitochondria increases, the cell loses some of its ability to produce energy from the non-lactate-producing respiration process. A "back-up" (lactate-producing) anaerobic energy production system is activated. Lactate, which is a by-product of anaerobic respiration, is released from the cell into surrounding tissues and the bloodstream, where in healthy persons, it is cleared mainly by the liver. The anaerobic process is not intended by nature to be the cell's primary energy source; rather, it is for times when extra energy is required -- during exercise, for example. A person with compromised mitochondria seems to be using this reserve system for daily energy and may not be not clearing lactic acid sufficiently. Thus, NRTIs can precipitate abnormalities, dysfunction, then toxicity, especially in someone with liver damage.
Mitochondrial dysfunction leads to varied pathology and is not easy to predict. HIV disease alone can cause a variety of abnormalities. This is a list of selected in vivo manifestations of NRTI-associated mitochondrial toxicity:
Neuromuscular -- Myopathy: zidovudine (AZT); Polyneurophathy: zalcitabine (ddC), didanosine (ddI), stavudine (d4T)
Hepatic/GI -- Steatosis, lactic acidosis: zidovudine (AZT) didanosine (ddI), stavudine (d4T), zalcitabine (ddC); Pancreatitis: didanosine (ddI), stavudine (d4T)
Hematologic -- Pancytopenias: zidovudine (AZT)
Nephrologic -- Proximal renal tubular dysfunction: adefovir (Hepsera)
Metabolic -- Lipodystrophy (new theory): stavudine (d4T)
Why Are So Many Different Tissues Affected?
Some thoughts are that 1) each tissue may have different NRTI kinetics (tissue levels of drugs may differ); 2) each tissue may have different activation enzymes (levels of active drug may differ); 3) each tissue may have different underlying proportion of mutant mtDNA; 4) it is known that NRTIs vary in their ability to inhibit mtDNA polymerase; 5) the role of mitochondria may be more or less important in certain tissues.
The Food and Drug Administration (FDA) reported 106 cases of lactic acidemia through June 1998. There were 46 cases associated with the use of a single NRTI (mainly AZT) and 61 cases associated with a combination of NRTIs including d4T, ddI, or AZT. In 69% of these cases, hepatic steatosis was present. There were 20 fatalities (85% female; 65% females obese). The French reported 11 cases in 867 (0.84%) patients over 18 months: d4T/ddI (n=7); d4T/3TC (n=2); d4T (n=2). Hepatic steatosis was present in four of five biopsies. There was one fatality. Four of the patients had HCV co-infection. They report treating lactic acidemia with riboflavin and carnitine.
Treating Mitochondrial Toxicities
Effective treatments still need to be evaluated. Current theories of treatment focus on assisting the respiratory chain function with 1) Coenzyme Q -- electron transfer with complex III (portions of the ATP electron transport chain in the mitochondria), 2) riboflavin -- a cofactor for electron transport complexes I and II, or 3) L-carnitine, a shuttle mechanism for fatty acid transport across the mitochondria. These have shown varying efficacy in uncontrolled trials to date. Challenges in studying mitochondrial toxicity (MT) include the lack of cell line or animal model to reliably predict MT. In vitro models may provide important information but experience with fialuridine (FIAU) and recently with lodenosine (FddA) shows that clinical experience is the most reliable indictor.
Hepatic steatosis is a frequent finding at biopsy, most often attributed to alcohol, obesity, diabetes, or drugs. If fatty acid oxidation in hepatic mitochondria is impaired, triglycerides may accumulate as small lipid vesicles in hepatocytes. Acute microvesicular steatosis can be very serious, leading to liver failure and death.
Resuming Antiretroviral Therapy
This patient improved after two days in the medicine intensive care unit, and an additional four days in the acute care ward. She returned to the facility and was seen in the infirmary for follow up. At one month, her labs were rechecked and results showed Na 137, Cl 106, and CO2 21. At two months, post-discharge her electrolytes were normal, and she was no longer acidemic. However, the patient's HIV RNA increased to 16,000 units (it had been undetectable) and her CD4 count was 392 units (down from 512 units). The patient was hesitant to restart antiretroviral therapy, and since her viral load and CD4 are at acceptable levels for now, the bottom line for this patient is "wait and see." When she does require therapy, NRTIs previously used in her treatment regimen will not be used and an NRT-sparing regimen will be considered.
If lactic acidemia is not specifically looked for, it may not be diagnosed in time. The disease varies in presentation, and its common symptoms -- nausea, vomiting, fatigue, myopathy, abdominal pain, and recent weight loss -- are also symptomatic of HIV disease itself. There are currently limited diagnostic tests for toxicity and a serum specimen to measure for lactic acid must be drawn carefully according to protocol.
Stephen Tabet, M.D., M.P.H. is Assistant Professor of Medicine at the University of Washington, and Director of the Northwest Correctional Medicine Education Program. This case presentation is a collaboration with the Northwest AIDS Education and Training Center, with Stephen Tabet, M.D., and Kate Willner, trainer.
AACTG Metabolic Guidelines for Hyperlactatemia and Lactic Acidosis, http://aactg.s-3.com/metabolic/lactic.pdf.
For a list of references please email Kate Willner at firstname.lastname@example.org.
Back to the HEPP Report June 2003 contents page.