|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Acid/Base and Electrolyte Disorders |

* Department of Internal Medicine;
Metabolic Disease Center, BRI Baylor University Medical Center, Dallas, Texas
Address correspondence to: Dr. Andrew Z. Fenves, Nephrology Division, Baylor University Medical Center, 3500 Gaston, Dallas, TX 75246. Phone: 214-820-2350; Fax: 214-820-7367; E-mail: fenvesa{at}dneph.com
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
-glutamyl cycle produces glutathione, an antioxidant substance that is involved in many important biologic functions, including inactivation of free radicals, detoxification of many compounds, and amino acid membrane transport (Figure 1). Glutathione synthetase (GS) deficiency and 5-oxoprolinase (5-OPase) deficiency are two rare inherited enzyme defects that affect the
-glutamyl cycle and result in massive urinary excretion of 5-oxoproline (pyroglutamic acid). Patients with GS deficiency also have high blood and cerebrospinal fluid 5-oxoproline levels and develop severe metabolic acidosis, hemolytic anemia, and central nervous system dysfunction (1). Heterozygous patients do not usually develop metabolic acidosis or severe 5-oxoprolinuria. Moderately increased urine excretion of 5-oxoproline also has been described in patients with propionic acidemia (2). Acquired 5-oxoprolinuria has been reported in infants who were fed the low-lactose preparation Nutramigen (3) and in patients who were taking acetaminophen (4); the anticonvulsant vigabatrin (5); or several antibiotics, including flucloxacillin and netilmicin (6).
|
-glutamylcysteine synthetase activity. This will increase production of
-glutamylcysteine, which is partially converted to 5-oxoproline. Subsequently, a small number of additional cases were reported (914). All these patients were using acetaminophen. Normal assays for GS and 5-OPase enzyme assays were reported in some cases. Often, the cause of the metabolic acidosis was multifactorial. Anion gap metabolic acidosis often develops in patients with acute acetaminophen hepatotoxicity and sometimes precedes the hepatic injury (1517). Usually the metabolic acidosis is attributed to lactic acidosis and kidney failure. However, 5-oxoproline has not been measured routinely in either the plasma or the urine of these patients and may be a contributory factor.
Rats that ingest a diet that contains 1% acetaminophen develop massive 5-oxoprolinuria, excreting up to 1 mol/L (18). 5-Oxoprolinuria could be prevented by feeding the animals methionine together with acetaminophen. The authors speculated that chronic acetaminophen ingestion leads to depletion of sulfur-containing amino acids, including cysteine, with a consequent depletion of intracellular glutathione stores. This then generates 5-oxoprolinuria via the mechanism described above (see Figure 1).
We investigated four patients, three of whom presented with severe anion gap metabolic acidosis that was unexplained by lactic acid, ketoacidosis, the ingestion of methanol or ethylene glycol, or severe kidney failure. The fourth patient developed an unexplained high anion gap metabolic acidosis after hospital admission. Each patient had been regularly ingesting acetaminophen (generally on a daily basis) for relief of chronic pain, and each patient also had some degree of kidney dysfunction. In each case, markedly elevated levels of 5-oxoproline were documented in urine and/or plasma. We could not identify any other cause to account for the severity of their metabolic acidosis.
| Case 1 |
|---|
|
|
|---|
There was no history of diabetes, hypertension, or renal disease. She smoked one pack of cigarettes per day and denied use of alcohol or illicit drugs. Her outpatient medications included phenytoin 100 mg three times daily and propoxyphene/acetaminophen as needed for pain, which was used on a regular basis.
Laboratory studies on the day of transfer included a BUN of 123 mg/dl, creatinine of 5.9 mg/dl, glucose of 85 mg/dl, potassium of 3.6 mEq/L, sodium of 145 mEq/L, chloride of 90 mEq/L, CO2 of 8 mEq/L, and an anion gap of 47 mEq/L. Renal failure was thought to be secondary to bilateral hydronephrosis. Radiation nephritis and possible acute tubular necrosis related to hemodynamic instability caused by massive retroperitoneal hemorrhage also were possible causative factors.
Evaluation of the high anion gap metabolic acidosis included two separate measurements of plasma lactic acid, both normal, and no detectable d-lactate acid. There was no evidence of ethanol, methanol, or ethylene glycol ingestion. Serum ketone assay was repeatedly negative. Although she had severe kidney failure, it was not thought that this could account for the severity of her high anion gap metabolic acidosis. The anion gap remained between 35 and 40 mEq/L.
A right percutaneous nephrostomy tube, placed because of persistent severe right-sided hydronephrosis, drained approximately 400 ml/d urine. Analysis of her urine for organic anions with gas chromatographymass spectroscopy (GC-MS) revealed 0.7 mmol 5-oxoproline/mmol creatinine. The patients medications during the hospital included vancomycin and gentamicin to treat blood culturenegative fevers, which resolved after the ureteral stents were removed.
While receiving oral sodium bicarbonate (Bicitra 30 ml orally twice daily) and regular hemodialysis, her anion gap remained in the 32- to 41-mEq/L range. At this time, she continued to ingest large daily doses of propoxyphene/acetaminophen for pain. The patients invasive squamous cell carcinoma could not be controlled, and she died in January 1991.
| Case 2 |
|---|
|
|
|---|
Her medical history was remarkable for migraine headaches for >30 yr. She also had a 6-yr history of hypertension. She had had an abdominal hysterectomy many years before and laparoscopic cholecystectomy 1 mo before admission. Home medications included enalapril 10 mg/d, Premarin 0.625 mg/d, and acetaminophen either alone or in combination with Propoxyphene as needed for pain.
On admission to BUMC, the patient was in respiratory distress and had orthostatic hypotension. Admission laboratory tests included arterial blood gas on room air: pH 6.88, Pco2 28 mmHg, and Po2 145 mmHg. White blood cell count was 20,000 with 15% bands and 93% total polysegmented neutrophils. Hemoglobin was 10.1 g/dl, and hematocrit was 32.2%. Glucose was 534 mg/dl, sodium was 131 mEq/L, potassium was 4.0 mEq/L, chloride was 90 mEq/L, and bicarbonate was 8 mEq/L, with an anion gap of 33 mEq/L. Serum ketone assay was negative. Serum creatinine was 2.0 mg/dl.
-Glutamyl transferase was 1100 U/L, serum glutamic oxaloacetic transaminase was 2200 U/L, serum glutamate pyruvate transaminase was 900 U/L, lactate dehydrogenase was 2000 U/L, and creatinine phosphokinase was normal at 159 U/L. The ammonia level was 48 µmol/L. Lactic acid level was 15.2 mEq/L. Erythrocyte sedimentation rate was 1 mm/h. Admission drug screen was positive for opioids but negative for amphetamines, barbiturates, cocaine, methadone, or PCP. Salicylate level on admission was <3 mg/dl, and acetaminophen level was 6.8 µg/ml, confirming that the patient was taking acetaminophen-containing medications. Alcohol level was undetectable. Lumbar puncture was unremarkable. Urine, blood, and cerebrospinal fluid cultures all were negative.
During the next 48 h, treatment with intravenous fluids and supportive therapy produced general overall improvement, and her lactate level fell to the normal range. However, the anion gap remained very high, >32 mEq/L over the next 3 d, and her metabolic acidosis persisted. Plasma 5-oxoproline level by GC-MS was 6.4 mmol/L and was characterized further as the L-isoform, which probably excludes a major contribution of bacterial origin.
Acetaminophen was discontinued. Renal function initially worsened, and her serum creatinine increased to 3.7 mg/dl but then stabilized and improved. At that time, her anion gap metabolic acidosis also resolved. Two weeks after this episode, the patient consented to a skin biopsy. GS activity in cultured fibroblasts was in the supranormal range.
| Case 3 |
|---|
|
|
|---|
Her medical history was significant for severe chronic obstructive pulmonary disease, peripheral vascular disease, chronic pain, breast cancer, atrial fibrillation, and osteoporosis. Medications included acetaminophen/hydrocodone as needed for pain, diltiazem 180 mg/d, digoxin 0.125 mg/d, Trental 400 mg/d, furosemide 20 mg twice daily, albuterol inhaler, and ipratropium bromide inhaler. She also had a history of ongoing alcohol abuse and previous tobacco use. On admission, the patient was in respiratory distress with markedly labored breathing. Admission laboratory was remarkable for an arterial blood gas on room air: pH 7.16, Pco2 14 mmHg, and Po2 111 mmHg. White blood cell count was 14,500/UL with 84% polysegmented neutrophils. Hemoglobin was 16.1 g/dl, hematocrit was 46.2%, and platelets were 546,000/UL.
Glucose was 166 mg/dl, sodium was 143 mEq/L, potassium was 4.7 mEq/L, chloride was 114 mEq/L, and bicarbonate was <5 mEq/L, yielding an anion gap of >24 mEq/L. Albumin was 3.0 g/dl, total bilirubin was 1.0 mg/dl, alkaline phosphatase was 138 U/L, ALT was 74 U/L, AST was 195 U/L, prothrombin time was 15.9 s, and partial thromboplastin time was 42.2 s. Serum osmolality measured at 318 mOsm/kg (calculated 300). Lactic acid was 2.5 mmol/L, semiquantitative serum ketone testing was negative, salicylate was <25 mg/L, and ethanol and methanol were negative. Plasma 5-oxoproline level was 2.8 mmol/L, and urine contained 1 mmol/mmol creatinine (both by GC-MS).
During the next 24 h, the patients status worsened. She became progressively more dyspneic and obtunded. Creatinine increased, respiratory acidosis complicated her metabolic acidosis, and she died secondary to ventricular asystole approximately 24 h after her admission.
| Case 4 |
|---|
|
|
|---|
| Discussion |
|---|
|
|
|---|
-glutamyl cycle, which is shown in Figure 1. Several inherited enzyme defects that have been characterized are extremely rare and usually present at a very early age with neurologic and hematologic abnormalities. GS deficiency is an autosomal recessive disorder and is characterized by mental retardation, ataxia, hemolytic anemia, and chronic metabolic acidosis in the homozygous patients (1). With GS deficiency, reduced glutathione levels increase
-glutamylcysteine synthetase activity, and the resulting high
-glutamylcysteine levels are partially converted to 5-oxoproline. 5-Oxoprolinase deficiency is a second rare autosomal recessive disorder that presents with 5-oxoprolinuria (19). It is interesting that these patients often develop kidney stones at a young age. Acquired 5-oxoprolinuria has been described in adults. Several contributory factors include malnutrition (20,21), pregnancy (22,23), and strict vegetarian diet (24). Limited glycine availability may be a common precipitating cause. Increased urinary 5-oxoproline excretion also has been described in patients with type 2 diabetes (25). Severe 5-oxoproline aciduria and acidemia has been described in patients who use several different medications. These include acetaminophen (paracetamol) (4,7,814), vigabatrin (5), and the antibiotics flucloxacillin and/or netilmicin (6). Glycine deficiency and these other clinical conditions all may result in depletion of glutathione. This is proposed as the explanatory mechanism for the development of excess 5-oxoproline generation. This critically important tripeptide, which contains glutamic acid, cysteine, and glycine, has major reducing and antioxidant effects, participates in transhydrogenation reactions and amino acid transport, modulates immune function, and detoxifies many drugs and poisons. Cysteine and glycine are required for the synthesis of glutathione (Figure 1).
The liver has one of the highest organ contents of glutathione, and it is known that hepatic glutathione stores are depleted in patients with acetaminophen toxicity (26). N-acetyl benzoquinonimine, an acetaminophen metabolite, binds irreversibly to glutathione (26,27). In vitro studies show that the intracellular glutathione content of liver cells must fall to <20% of control values before acetaminophen toxicity develops (28).
High anion gap metabolic acidosis occurs frequently in patients with acetaminophen toxicity (15,29,30). This generally is attributed to lactic acidosis and kidney failure. However, in some cases, the anion gap cannot be explained (30,31). These patients often have less severe toxicity and a history of chronic acetaminophen ingestion.
Pitt (4) measured urine 5-oxoproline excretion in patients who ingested acetaminophen compared with control subjects. Urine 5-oxoproline excretion in the acetaminophen group was >100 times higher than that in the control group, although still much lower than occurs in patients with inherited GS deficiency and most of the reported cases of acquired 5-oxoproline acidosis (3).
Chronic acetaminophen ingestion is associated with reduced plasma glutathione levels (32). Review of the literature revealed 18 adult cases of acquired, often transient, 5-oxoprolinuria and high anion gap metabolic acidosis associated with chronic "therapeutic" ingestion of acetaminophen (this excludes pediatric patients and patients with acute acetaminophen poisoning). The patient characteristics and biochemical findings of the previously reported patients and four new patients whom we report are summarized in Tables 1 and 2.
|
|
-glutamyl cycle are know to be different in men and women (33,34). These differences may increase the susceptibility of women to develop this disorder. It should be emphasized that acetaminophen ingestion alone probably does not generate clinically significant 5-oxoprolinuria or metabolic acidosis. In most cases, synergistic interaction between acetaminophen ingestion and multiple other factors exists (9). All of these patients had underlying or preceding illnesses, and most were malnourished. This probably depleted hepatic glutathione stores. This undoubtedly increases the susceptibility to toxic effects of chronic acetaminophen use. Many but not all of the patients had abnormal liver function tests. Several also had a history of chronic alcohol abuse, which also is known to reduce glutathione levels (35).
Healthy individuals who ingest diets that are depleted of cysteine, methionine, or glycine increase their excretion of 5-oxoproline (36). Patient 2 in Table 1 was a vegetarian, and patient 3 was pregnant (2123). Many of the reported patients also had infection or sepsis. The cause of high anion gap metabolic acidosis with sepsis often cannot be elucidated (3739). Glutathione synthetic rates in critically ill septic children fall by approximately 60% (40). Furthermore, some antibiotics (fludoxacillin and netilmicin) have been implicated in the development of this syndrome.
All our patients and many of those previously reported had renal insufficiency or failure. A reduction of kidney function will reduce urine excretion of 5-oxoproline and therefore may cause greater systemic accumulation.
The course of our case 4 (patient 22 in Tables 1 and 2) illustrates the critical role of acetaminophen in the development of otherwise unexplained high anion gap metabolic acidoses in some patients. She had multiple comorbid conditions that contributed to depleted glutathione stores, and she received large doses of acetaminophen for pain. This patient developed severe 5-oxoproline metabolic acidosis during her hospitalization in response to therapeutic administration of acetaminophen. Chronic renal insufficiency may have contributed to higher plasma 5-oxoproline levels. The observation that the high anion gap metabolic acidosis resolved with cessation of the drug is very important.
We believe that treatment of this unusual form of metabolic acidosis should focus on the recognition of its presence and cessation of acetaminophen intake. The importance of treatment of sepsis and renal and/or hepatic dysfunction is self-evident. N-acetylcysteine has been used with some effectiveness in patients with GS deficiency because it is thought to increase the low intracellular glutathione and cysteine concentrations (41). One adult patient with the acquired syndrome of high anion gap metabolic acidosis as a result of 5-oxoprolinemia was treated with intravenous N-acetylcysteine and improved (the only other reported patient who was treated was a 5.8-yr-old child, who also improved). In view of N-acetylcysteines low toxicity and theoretical benefit, its use seems reasonable until better studies become available.
A prospective study of 23 patients who were in the intensive care unit with unexplained high anion gap metabolic acidosis failed to identify any patient with high 5-oxoproline levels (42). This entity undoubtedly is rare but very important to recognize because it can be treated by withdrawing a very common therapeutic agent.
| Conclusion |
|---|
|
|
|---|
| Acknowledgments |
|---|
| Footnotes |
|---|
Received October 21, 2005. Accepted February 28, 2006.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. D. Navaneethan, R. Mooney, and J. Sloand Pseudo-anion gap acidosis NDT Plus, April 1, 2008; 1(2): 94 - 96. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |