SUMMARY

Hypoxic hepatitis, also called ischemic hepatitis or shock liver, represents damage to the liver as a result of insufficient oxygen supplied by the blood. Most often, this hypoxemia results from insufficient blood supply, as in congestive heart failure, though other causes exist, among other possibilities low arterial oxygen levels, though this has yet to gain wide acceptance. During a period of 10 years, from 1983 to 1993, the authors collected 142 cases of hypoxic hepatitis from their intensive care units. Two of these had normal cardiac function, but extremely low arterial oxygen; both were morbidly obese women, leading to suspicion of obstructive sleep apnea. The first patient was a 64-year-old woman admitted comatose, in respiratory distress. She had a history of depression, mild hypertension, and obesity (with a Body Mass Index (see Glossary) of 47.4. She had been taking prothipendyl chloralhydrate (presumably a sleeping medication), chlorazepam (unlisted under this spelling in my Physicians Desk Reference (1997 version), but probably the same as either clonazepam/Klonopin or clorazepate/Tranxene, both benzodiazepines presumably used for anxiety or insomnia), triazolam (Halcion, a short-acting benzodiazepine sleeping medication), sulpiride (I believe this may be a European antipsychotic drug), and propranolol (Inderal, a beta-blocker most comonly used as an antihypertensive, also occasionally used for other symptoms including anxiety). Five days before admission she had fallen ill with an apparent respiratory infection for which she was given lincomycin, an antibiotic. She became more and more sleepy, unable to speak, and finally unconscious. On admission her systolic blood pressure was slightly high, at 150/70, with a normal heart rate of 74 beats per minutes, and no physical signs of congestive heart failure. She did appear to have a lung infection for which she was treated with new antibiotics. Blood tests on admission showed remarkably low arterial oxygen (22, where normal is 80-98) and oxygen saturation (34, where normal is 94-96) and high carbon dioxide (64, where normal is 35-45). These were corrected by artificial ventilation within hours of admission. However, liver enzymes, much elevated on admission, continued to increase through the second day, then declined but had not yet normalized by the time of discharge from the ICU four days later. A liver biopsy demonstrated hypoxic hepatitis. Two nocturnal oximetry records showed repeated episodes of oxygen desaturation, mainly at the end of the night presumably in association with REM. The patient, who had been uncooperative throughout, refused polysomnography. It may be relevant that, while tests for other drugs were negative, her blood level of benzodiazpines was quite high (3910 ng/ml, where therapeutic range is 100-1500 ng/ml). Five years later the patient was still alive but had suffered several similar episodes. The second patient was a 70-year-old woman admitted with respiratory distress. Four months earlier she had been admitted to the cardiology ward with fatigue, hypertension, and rapid weight increase. At that time cardiac and pulmonary exams were normal except for moderate hypertension (170/100), mild ankle edema, and obesity (BMI=39). She was discharged on a triad of antihypertensives: lisinopril (an ACE inhibitor), furosemide (a diuretic), and guanfacine, an alpha-2-adrenoreceptor blocking agent. When readmitted, she was half-asleep, with mild difficulty breathing and cyanosis (blue tinge to the blood from lack of oxygen). Like the first patient, she had a very low blood oxygen level of 26, a very low oxygen saturation of 31, and a much elevated carbon dioxide of 83. She also had elevations of blood liver enzymes. She was immediately intubated and mechanically ventilated, with progressive improvement in blood gases over her four-day stay in the ICU. The liver tests continued to rise the day after admission, then subsided but not into the normal range at the time of discharge, much like patient #1. However, this patient allowed polysomnography, which showed numerous episodes of oxygen desaturation, apneas, and hypopneas. She went on to accept BiPAP treatment and was still alive 4 years later. The authors note that ischemia (cell damage or death due to lack of blood flow) has been considered the major, if not the sole, mechanism of hypoxic hepatitis. The mechanism represented here was thought rare to nonexistent because of animal studies which showed that liver damage would result only from extremes of hypoxia below that ordinarily seen in humans, and because liver function of patients with asthma or chronic obstructive pulmonary disease were generally normal. In these two cases, an ischemic (lack of blood) cause can be dismissed; neither patient had cardiac disease or signs of circulatory failure. On the other hands, these two patients showed extremely low blood oxygen seen in none of the other 140 cases the authors studied—even in those patients with severe cardiorespiratory failure. In the first case, the severe hypoxemia was thought due to probably sleep apnea syndrome combined with benzodiazepine abuse and bronchial infection. In the second case, hypoxemia was probably due more simply to sleep apnea.

COMMENTS

The authors are using two cases out of a series of 142 to prove that a certain mechanism of hypoxic hepatitis, doubted by many, did occur. In other words, obstructive sleep apnea can bring blood oxygen down to the very low levels necessary for liver damage. In doing so, it exceeds the extremes of hypoxia seen in several other, more common cardiorespiratory diseases. The liver may require unusually low levels of oxygen to show this kind of hypoxic damage, but other organs of the body, such as the brain, do not. In clinical case reports like this, one often is left wanting to know more than the authors had space to say. I would be especially interested to know how the pattern of uncooperativeness and noncompliance arose in the first patient, and persisted despite several later episodes of what amounts to coma. It is typical of drug abusers that they fear the doctor gaining too much knowledge about their condition and eventually depriving them of their drug, or that they sign out of the hospital prematurely because the amount of drug they receive in the hospital is not comparable to what they were used to taking outside, so they experience withdrawal or rebound symptoms. On the other hand, I note that this patient was apparently prescribed two different benzodiazepines—a dubious practice even in a patient without OSA and all the more so in its presence. If my reading of her pre-hospital medications is correct, despite her diagnosis of depression she was not on any antidepressants. We are given enough information only to suspect the possibility of benzodiazepines abuse, but the high blood level might also have resulted from the combination of two benzodiazepines, dosage unspecified, and perhaps their interaction with another drug like sulpiride. Looked at from this perspective, her decline could have resulted from oversedation from unintentional effects of a medication complex in a condition (OSA) known to respond to sedatives with worsened respiratory problems. Likewise, her bronchitis might have aggravated the same problem. An angry, resistant patient is all too likely to end up repeatedly on combinations of high doses of sedatives. We might also wonder about the cumulative sedative effects of the three antihypertensive medications administered to the first patient, but excepting an idiosyncratic reaction to one of them, none seems sedative enough to account for her decline. The interesting point of recent severe weight gain seems to tie into this decline, but whether as cause or effect we cannot know. So in uncontrolled clinical observations like these many interesting questions can be raised but rarely answered. They central point of the authors—that hypoxia from OSA can become severe enough to affect the liver—seems well documented.





This page hosted by Get your own Free Home Page