Most cases occur after the ingestion of carbohydrates and eventual absorption of D-lactate from the affected intestinal segment, but others occur after the consumption of dairy products or lactobacillus tablets [2, 4]. Dahlquist and colleagues [5] reproduced the syndrome by overfeeding carbohydrates to humans with previous jejunoileostomy.

Transient neurologic symptoms are the hallmark of D-lactic acidosis: Headache, weakness, delirium, visual disturbances, dysarthria, ataxia, cranial nerve palsies, changes in affect, and even transient hypothalamic dysfunction have been reported [2-14], but the mechanism explaining these manifestations remains unknown.

Treatment has been successful with carbohydrate restriction and oral antibiotics, such as vancomycin, metronidazole, clindamycin, tetracycline, neomycin, and kanamycin [2-6, 10-13]. In some patients, however, D-lactic acidosis recurred despite the use of antimicrobial agents; in other instances, their potentially causative role was not addressed [6, 13]. The D-lactic acidosis syndrome developed in a patient who had received tetracycline [3], in another who had received metronidazole [13], and in a patient who had received both agents [5].

Case Report

Top Case Report Methods Results Discussion Author & Article Info References

First Episode

A 50-year-old man with steroid-dependent chronic obstructive pulmonary disease had jejunoileal bypass in 1975 for the treatment of obesity. He presented in 1993 after having ataxia, slurred speech, and weakness for 2 days. On the day of presentation, he completed a 14-day course of oral doxycycline that was given for bacterial bronchitis. Physical examination was otherwise unremarkable. Assessment of arterial blood gases while the patient breathed room air showed a partially compensated metabolic acidosis (Table 1). Urea nitrogen levels were 6.1 µmol/L (normal, 3.0 to 6.5 µmol/L), creatinine levels were 79.6 µmol/L (50 to 110 µmol/L), and glucose levels were 4.9 µmol/L (3.9 to 6.1 µmol/L). A computed tomographic scan of the patient's head and an electrocardiogram were both normal, and alcohol was not detected in his blood. His symptoms resolved spontaneously by the next day, and he was subsequently discharged.

Second Episode

Four days later, the patient presented with similar symptoms 5 hours after he ingested a large pasta meal. No new antibiotics had been prescribed. Physical examination showed dysarthria, and assessment of arterial blood gases showed partially compensated metabolic acidosis (Table 1). Salicylates were not detected in his blood. Computed tomographic scans of his head and toxicology screens were both normal. By the next day, his symptoms resolved and the anion gap returned to normal without intervention. A magnetic resonance imaging scan of his brain, a cranial magnetic resonance angiogram, a carotid ultrasonogram, and an echocardiogram were normal. He was discharged on a lactose-free, low-carbohydrate diet.

Provocative Testing

Two months later, the patient was hospitalized electively for the purpose of reproducing the D-lactic acidosis syndrome with oral carbohydrates, after written approval by the Hospital Human Subjects Committee was obtained. The patient had not received antibiotics since his initial attack. A protocol similar to that of Dahlquist and colleagues [5] was used. He received 4000 kcal/d in divided meals, with 64% carbohydrates, 17% protein, and 17% fat. This 60-hour challenge diet did not reproduce the syndrome (Table 1).

Third Episode

Two months later, the patient presented with confusion, slurred speech, anxiety, weakness, ataxia, liquid diarrhea, and a high anion gap metabolic acidosis after 3 days of trimethoprim-sulfamethoxazole therapy that had been initiated for another respiratory infection. He had not recently consumed dairy products or a high-carbohydrate diet. Feedings were stopped and therapy with the antibiotic was discontinued, with total improvement by the next day (Table 1).

Three months later, he tolerated 2 weeks of oral ciprofloxacin for another respiratory infection without development of symptoms. Several weeks later, yet another respiratory infection developed that was treated with trimethoprim-sulfamethoxazole. Three days after the first dose, he developed dysarthria but promptly became asymptomatic after discontinuation of therapy with the antibiotic.

Methods

Top Case Report Methods Results Discussion Author & Article Info References

Levels of D-lactate were determined enzymatically using D-lactate dehydrogenase (Sigma Chemical Company, St. Louis, Missouri). Other measurements were done using standard laboratory methods [15, 16].

Results

Top Case Report Methods Results Discussion Author & Article Info References

After elective hospitalization for provocative testing Table 1, baseline D-lactate levels were increased, but stool cultures yielded normal flora. The patient subsequently developed metabolic alkalosis that was probably caused by oral furosemide given for leg edema from prednisone-induced fluid retention. No neurologic symptoms or acidosis developed. Serum and urine D-lactate levels decreased, whereas plasma L-lactate levels increased (Figure 1).

View larger version (15K): [in this window] [in a new window]   Figure 1. D-Lactate levels after the intake of oral carbohydrates was increased. The patient had the short-bowel syndrome and received a challenge diet of 4000 kcal/d (64% carbohydrates, 17% protein, and 17% fat). Normal plasma and urine D-lactate levels are 0 to 0.25 mmol/L.

Stool cultures collected during the third episode were negative for normal coliforms and enteric pathogens, yielding only Lactobacillus acidophilus. This organism was resistant to trimethoprim-sulfamethoxazole (minimum inhibitory concentrations > 100 µg/mL and 20 µg/mL, respectively), doxycycline (20 µg/mL), and kanamycin (>20 µg/mL); sensitive to vancomycin (2 µg/mL) and ciprofloxacin (0.5 µg/mL); and moderately resistant to metronidazole (50 µg/mL). After 24 hours of incubation, L. acidophilus produced 58.3 mmol/L of D-lactate from glucose at a bacterial suspension of 1 x 10¹¹ colony-forming units/L.

Discussion

Top Case Report Methods Results Discussion Author & Article Info References

This case documents D-lactic acidosis in a patient with the short-bowel syndrome receiving oral antibiotics and suggests that these agents induced the selective overgrowth of D-lactate-producing L. acidophilus.

In humans, L-lactate is formed directly from pyruvic acid as a product of anaerobic carbohydrate metabolism. Some intestinal bacteria metabolize carbohydrate into D-lactate by the action of isomer-specific D-lactate dehydrogenase. Even though mammals may generate small amounts using the methylglyoxal pathway [17], plasma D-lactate is derived mostly from intestinal absorption and requires special techniques for its detection that are unavailable in most clinical laboratories.

D-Lactic acidosis is usually self-limited, and its duration, from hours to days, seems to depend on a balance among the rates of D-lactate production, absorption, utilization, and renal excretion. Interestingly, a similar syndrome in cattle was described before the first report in humans [18]. Ruminants with increased production of D-lactate in their rumens developed blindness and ataxia [19, 20].

A delay of months to years usually occurs between the time of bowel surgery and the development of the D-lactic acidosis syndrome. This patient received multiple antibiotic regimens through the years, and his colonic Lactobacillus population probably increased until the amount of D-lactate produced exceeded his metabolic and renal excretory capacity. Once D-lactate levels reached a critical point, the full-blown syndrome became apparent.

This patient had increased baseline plasma and urine D-lactate levels when hospitalized for provocative testing. Although symptoms are known to occur at high D-lactate levels, the clinical threshold may vary from patient to patient [8]. Subclinical D-lactic acidosis with chronic or intermittent neuropsychiatric impairment may develop in some patients. In fact, in a study by Thurn and colleagues [8], 16 of 33 patients with jejunoileal bypass reported symptoms consistent with D-lactate encephalopathy; of these, 9 had random serum D-lactate levels of more than 0.5 mmol/L [8].

The decrease in urine and plasma D-lactate levels and the absence of symptoms during our test indicate that selective bacterial overgrowth by antibiotics may have been the determining factor in the development of the syndrome and that oral carbohydrates played only an additive role. In fact, this patient developed the syndrome twice without previous dietary indiscretion. The increase in L-lactate levels with dietary challenge was probably a result of the metabolic activity of L-lactate-producing bacteria, with the inhibition of other species that preferentially produce D-lactate. A more prolonged test would allow the determination of the changing bacterial populations and the amounts of D- and L-lactate ultimately produced.

The antibiotic resistance patterns of the stool isolates also support the idea that selective overgrowth of lactobacilli occurred before the onset of symptoms. The uneventful course of oral ciprofloxacin is consistent with the sensitivity of L. acidophilus. Because ciprofloxacin probably eradicated colonic L. acidophilus, it did not induce the syndrome.

The pathophysiologic mechanism explaining the neurologic manifestations is still not understood. Acidemia does not by itself account for the symptoms because even more profound acidosis from other causes is not associated with these manifestations. D-Lactate has been thought to be directly toxic to the central nervous system, but intravenous infusions of D-lactate in healthy humans have not reproduced the syndrome [14]. Plasma levels achieved in these studies may have not been high enough, and perhaps patients with the short-bowel syndrome develop symptoms at lower concentrations. Certain nutritional deficiencies may predispose these patients to neurologic dysfunction when they are exposed to D-lactate. Infusing D-lactate into these patients would clarify this issue, but ethical issues must be considered.

The judicious use of antibiotics in patients with the short-bowel syndrome along with carbohydrate restriction and avoidance of lactobacillus-containing preparations seem justified. Fecal cultures with antimicrobial susceptibility testing may aid in the evaluation and management of these patients. Such practices may also help in the selection of specific antibiotics, thereby allowing therapy to be tailored to the microbial population found in each patient.