Jerome B. Posner

THE TOXIC EFFECTS OF CARBON DIOXIDE AND ACETAZOLAMIDE IN HEPATIC ENCEPHALOPATHY

Striking alterations in respiratory gas and blood ammonia levels usually accompany hepatic coma. The arterial pH is elevated and the carbon dioxide tension reduced (1). The blood ammonia level (2-4) is increased in many instances and the brain arteriovenous ammonia difference is high (5, 6). The cerebral oxygen uptake is depressed (7, 8). Whether the pronounced alkalosis which accompanies many instances of hepatic comiia contributes to the encephalopathy is unknown. Severe blood alkalosis impairs oxyhemoglobin dissociation (9), lowers the arterial oxygen partial pressure and interferes with cerebral function in normal subjects (10). Alkalosis is reported to enhance ammonia toxicity (11-13). These effects led Roberts and associates (14) as well as others (13) to suggest using carbon dioxide inhalation to treat hepatic coma. Prelimiiinary clinical observations on our own services suggested that inhaled CO2 worsened rather than improved patients with the encephalopathy of liver disease. Accordingly, detailed clinical and cerebral metabolic studies were carried out to quantitate the effects on brain function of raising blood and tissue CO2 tensions. The results of the study are reported here. Cerebral blood flow and cerebral oxygen uptake were measured at rest in control and cirrhotic patients and the results were correlated with the patients mental and somatic neurological function. Following the baseline study, selected patients inhaled carbon dioxide and the clinical cerebral metabolic observations were repeated. Subsequently, to eliminate the undesirable side effects of hyperpnea and dyspnea as contributing to the observed changes, similar observations were made before and after raising tissue CO2 tensions by rapidly infusing acetazolamide (Diamox).

MRI demonstrates descending transtentorial herniation

Descending cerebral transtentorial herniation (DTH) is a serious and often fatal complication of intra-cranial mass lesions. The condition can be inferred from clinical neurologic signs, but has not been visualized during life. Using midsagittal magnetic resonance images (MRIs), we compared vertical brainstem position on 50 images from normals and 21 images from 15 clinically stable patients with large supratentorial tumors. The length of Twinings line (T), the perpendicular distance from T to the pontomesencephalic junction (T-PMJ), and from T to the apex of the midbrain aqueduct (T-A) were measured. We also measured lateral shifts of the diencephalon and midbrain on axial images. T-PMJ decreased from 2.04 ± 0.06 mm in normals to 0.94 ± 0.2 mm in patients with large cerebral tumors (p < 0.0001). Similarly, T-A decreased from 6.35 ± 0.13 mm in normals to 4.83 ± 0.35 mm in patients (p = 0.001). Lateral diencephalic-midbrain shifts often accompanied DTH but to an unpredictable degree. Either lateral or downward brainstem shift could occur alone and did not necessarily produce specific neurologic signs or an altered state of consciousness. Anatomic DTH occurs in life, it can be quantified with MRI, and in slowly developing cerebral mass lesions the process can precede the appearance of neurologic signs and symptoms that indicate lower-diencephalic or midbrain dysfunction.