Perhaps he had an inherent bias from his college days, but this was not explored .... list are a number of diagnoses of
BLADDER BEER-A NEW CLINICAL OBSERVATION JOHN H. MULHOLLAND* and (by invitation) FRANCIS J. TOWNSEND BALTIMORE
The importance of using the sense of smell in medical diagnosis was pointed out by Susruta, a founder of ancient Hindu medicine, who noted the value of the smell of peculiar perspiration in certain conditions (1). An odor is a volatile emanation which is often difficult to define precisely, the more so than with the data derived from man's four other senses. We became aware of a pleasantly unusual and quite unexpected odor noted in several patients who had yeast infections. At the time of a laparotomy on a patient with a perforated gastric ulcer a junior house officer observed that the fluid discovered in the right abdomen smelled like beer. Perhaps he had an inherent bias from his college days, but this was not explored further. The culture of the peritoneal fluid grew only yeasts, Candida albicans and Torulopsis glabrata. The story was remarkable enough so that it travelled from the surgical to the medical house staff at a jointly held conference. Later that same month a second patient was seen, an elderly female with poorly controlled diabetes who was presumed to have septicemia. It was noted that the Foley bag contained extremely turbid urine and a urinary tract infection was suspected. The attending physician suggested prompt examination to define the cause for the turbidity, suspecting crystalluria. As the house officer collected the sample he said that it must be a yeast infection. When asked how he knew this without a microscopic examination, the response was that the sample smelled like brewing beer. The specimen was passed from hand to hand and all at the bedside were in agreement. When the patient was examined more closely it was observed that in addition to the catheterized urine, sputum, mouth scrapings, perineal fistulous fluid and the patient herself had the odor of beer. Cultures confirmed yeast in all of these areas and the urine was positive for ethanol. After the initial two patients were observed, special efforts were made to locate other patients in the hospital with yeast infections so to apply the newly defined "sniff" diagnostic test. To date we have seen five such patients. All of them have been elderly, seriously ill, four with associated surgery, four with prior antibiotics and all with diabetes. In all instances the specimen infected with yeast has had a distinctly alcoholic odor. We should note that when we refer to the odor of "alcohol" that alcohol itself * Department of Medicine, The Union Memorial Hospital, Baltimore, Maryland. 34
BLADDER BEER
35
is odorless and that the congeners of fermentation are the source of the 'alcoholic odor", even on the breath of the intoxicated patient. The index case is listed at the top of Table 1, having presented with a perforated ulcer. The second patient had undergone pelvic exenteration for cancer and had many odorous excretions including breath which had the odor of fresh bread. Alcohol was detected in the urine. Another, a 72 year old with a ruptured gallbladder had Torulopsis glabrata, a yeast similar to Candida, found in blood and urine. The latter demonstrated the characteristic smell and alcohol was present in good quantity. The next, a 53 year old diabetic with anaerobic cellulitis of the upper thigh grew yeast in the urine and "trace" alcohol was present. A most unusual and surprising patient, an 85 year old black male with an antecedent history of receiving long-term oral antibiotics for inflammatory lesions of the leg, was found to have diarrhea which had, to one of us, an odor reminiscent of a fraternity house on the morning after a party. When presented to uninformed and blindfolded fellow house officers the stool was described as smelling like wine, alcohol or beer. Gram stained smears of the stool showed sheets of yeast, many demonstrating pseudohyphae, and surprisingly few bacteria. Cultures revealed a heavy growth of Candida albicans. In this particular instance, because of the large amount and frequency of the diarrhea, the entire end of the ward reeked of alcohol. Unfortunately the stool sent for alcohol content to confirm this new scientific observation came back promptly, but was caught in modern computer technology, being reported as negative for blood. We should add that all these patients were critically ill with no access to oral or parenteral alcohol and each had poorly controlled diabetes. Four were treated with Amphotericin B and only one survived and was discharged. The measurement of alcohol levels in the urine of three patients showed all to be positive by gas chromatography. Quantitative levels were obtained on two, being 43 and 121 mg per dl. The presence of alcohol in human specimens containing glucose and yeast should come as no surprise. Several have made this observation. Under normal circumstances trace amounts of alcohol may be found in the blood; the alcohol is then channeled into an energy pathway by hepatic alcohol dehydrogenase (2, 3). Blume and Lakatua reported that a postmortem blood sample was found to contain increasing ethanol concentration upon standing (2). Ball and Lichtenwalner observed that alcohol was generated in the urine of diabetic patients when glycosuria and genitourinary candidiasis are present (4). These authors also showed that Candida will consume glucose and produce alcohol in a urine specimen kept in the laboratory for several days at room temperature.
JOHN H. MULHOLLAND
36
4i
0
"
0 CD
4)
0
4
4)
4
4-)0
*
"0)
)
4)
C)
() tD
16. 44J E
r--4
0
w
04 m
!t E-4
*
@ rn
4)
O
$
_
w
t
4)e CD2
4)C
22hX E 0
4)
"_
C0 4)
4)
4)
4
4)
-4)4
)
4)
4)
PU
4)
6-
4)
4)
O
C)
L
10 r. (L)
co0 to
co
a LO
6q
0
z
BLADDER BEER
37
We have been unable to find in the English or American literature a note of alcohol or beer odor in specimens from patients with yeast infections, even in the urine, much less the gastrointestinal tract. However, the Japanese report the "auto-brewery syndrome" in which they have seen middle aged patients with bowel abnormalities, most often surgery, who have yeast overgrowth, usually Candida, in the GI tract and who ferment ingested carbohydrates, producing enough alcohol to result in drunkenness (5). This observation has not been reported in other countries. We have all learned to identify anaerobic infections by the presence of a foetid odor emanating from biologic specimens. In the hopes of improving the olfactory skills of fellow physicians Smith, Smith and Levinson provided a recent review of "Smells and Possible Diagnosis" which appeared in the Lancet one year ago (6). Included in their comprehensive list are a number of diagnoses of particular relevance to some of the assembled audience-the stale beer smell of scrofula, the freshly baked brown bread odor of typhoid. Recent interest has been paid to phenylketonuria, an inborn error of metabolism which produces a musty, stale locker room towel odor in urine. Also noted in this report is another rare congenital enzymatic block called oast-house urine disease or "beery baby syndrome" which is associated with a urine odor of dried malt or hops. The issue of the relationship of the appearance of alcohol in the cerebrospinal fluid of patients with cryptococcal meningitis has been seriously questioned (7), although, alcohol smell of CSF in infected patients is listed by Smith et al. (6). A prior member of this group was reported to diagnose diphtheria by the sweetish odor detected in the corridor outside the room of an infected patient even before the resident had presented the case history. To the comprehensive list of diagnoses suggested by the physician's senses we would like to add as a new clinical observation the beery smell of yeast infections in urine, stool and peritoneal fluid. As clinicians at the bedside, we should interpose our noses, directing the use of one of our given senses to complement the multiple diagnostic tests which are inevitably performed in patients requiring intensive medical care. 1. 2. 3. 4.
BIBLIOGRAPHY Cone TE, Jr. Diagnosis and treatment: Some diseases, syndromes, and conditions associated with an unusual odor. Ped 1968; 41: 993. Blume P, Lakatua DJ. The effect of microbial contamination of the blood sample on the determination of ethanol levels in serum. Am J Clin Pathol 1973; 60: 700. Krebs HA, Perkins JR. The physiological role of liver alcohol dehydrogenase. Biochem J 1970; 118: 635. Ball W, Lichtenwalner M. Ethanol production in infected urine. New Engl J Med 1979; 301:614.
38
JOHN H. MULHOLLAND
5. Kaji H, Asanuma Y, Ide H, et al. The auto-brewery syndrome-the repeated attacks of alcoholic intoxication due to the overgrowth of Candida (albicans) in the gastrointestinal tract. Materia Medica Polona 1976; Fasc 4(29): 429. 6. Smith M, Smith LG, Levinson B. The use of smell in differential diagnosis. Lancet 1982; 2: 1452. 7. Wilson DE, Williams TW Jr, Bennett JE. Further experience with the alcohol test for cryptococcal meningitis. Am J Med Sci 1966; 62: 532.
DISCUSSION Glaser (Menlo Park): I wonder whether you've thought about calling this entity "The Budweiser Syndrome"? Sessions (Chapel Hill): Well, as I'm sure will be the recurring theme through the rest of the meeting, this brings to mind the Presidential Address because it will be difficult to smell anything from the conference room down the hall and it testifies to our new members following your admonishments, Dr. Glaser. It isn't surprising to me that this was described by the Japanese because they have a diminished tolerance toward alcohol and in fact it has been said with something less than a joking manner that it was the difference in the tolerance to alcohol that saved the American occupation of Japan. Therefore, accumulating even small blood levels of alcohol could be expected to produce some of the flushing of the face and giddiness of manner. I'm fascinated by being able to bring those two points together that your paper presented. Thank you for an interesting paper. Wolf (Bangor): It is interesting that plants are capable of producing alcohol. They do so when forced into anaerobic metabolism. Most land plants get their oxygen through their roots and breath out carbon dioxide through their leaves. When oxygen uptake is impaired because of too much water in the ground, they resort to anaerobic glycolysis with a prominent Pasteur effect with the production of ethanol and lactic acid. Marsh plants on the other hand are able to get oxygen through their stems and loose appendages above the water so they easily tolerate having wet roots. Certain land plants are relatively tolerant of periodic water flooding. To cope with it they have a special intermediary metabolic pathway that involves the induction of alcohol dehydrogenase that in turn mitigates the production of ethanol. Without it, the ethanol would kill the plant. You might say that plants that lack this metabolic protection against ethanol, peas for example, may, in wet weather, succumb to alcohol abuse. Furman (Indianapolis): I'd like to comment further on the observation in the Japanese. Alcoholism is not the problem among the Japanese that it is in the United States. About 10-20% of the Japanese get a very disconcerting flush when taking alcohol and it's been largely established, I think, that it is due to the fact that they lack the enzyme or certain iso-enzyme of acid aldehyde dehydroginase. Can I add one historical note? As a house officer in New Haven Hospital in the early 40s, the pediatric house staff was wont to call down to the emergency room one of the older staff nurses when they suspected they had a diptheritic pharyngitis. She would sniff and make her pronouncement and I can tell you she was more often right than wrong. Barondess (New York): Jack, did some of your patients have candidemia and do you have any observations on the blood alcohol levels in those patients? My comment is that for some time it has seemed to me that people in left ventricular failure and especially those in acute pulmonary edema have a rather characteristic odor to their breaths. It was suggested to me at one point by Dr. Stewart Bondurant that that might be the odor of surfactant, and he sent me a batch that he had prepared which, unfortunately, being from a poor year, had no odor at all. I wonder if you have any comment about that. Mulholland: Only one of our patients had septicemia and that was with Torulopsis. In the others we could not establish that, but we certainly know that Candida septicemia is a
BLADDER BEER
39
very difficult diagnosis to establish and it is more often that not, I believe, that patients are being treated now with that presumptive diagnosis after cultures are obtained because of the difficulty with finding a positive blood. The organisms just don't grow that fast. The second question was a good one and I think is part of the lack of scientific methodology, perhaps, reported earlier in our presidential address. Jerry, we haven't been tooled up well enough to get that serum specimen. We blew it on the stool, obviously, when that would have been certainly fun to have shown that colonic stool had alcohol. Actually, Dr. Sessions is probably aware of this too, that stools from small bowel aspirates in patients with sprue are found to have alcohol present as opposed to stool from normals. But we haven't gotten a blood at this point. Thier (New Haven): I wondered whether you had done any studies outside of the patient such as adding yeast to urine with varying sugar contents to see whether you could get any clues as to the lack of control and whether you knew at what sugar and yeast levels you were likely to see alcohol form. Mulholland: Well, the experiments that you've commented on have not been done very scientifically but are in the literature. Glucose and yeast in urine will produce alcohol and I think it really is a matter of how much glucose is available. Our patients, as you would well be aware, with a Foley catheter inserted in the intensive care unit, are producing varying quantities of surgar-usually quite a bit. And it only takes, I think, a small amount of sugar and yeast to do that. From the one slide shown here from a letter to the editor in the New England Journal, I would suggest that glucose content is probably the limiting aspect; and as long as the glucose remains present the yeast will thrive and ferment alcohol.
