Chronic alcohol abuse and the acute sedative and neurophysiologic effects of midazolam

The aim of the present investigation was to examine benzodiazepine sensitivity in abstinent alcoholics. For this purpose, two escalating doses of the benzodiazepine midazolam were i.v. administered to nine alcohol-dependent patients after 2-3 weeks of abstinence and 12 healthy, non-alcoholic volunteers. A variety of dependent measures were examined, including the power spectrum of the resting electroencephalogram (EEG) and evoked EEG responses, saccadic eye movements, self-reported sedation, and vigilance task performance. Analyses revealed a significant association between plasma midazolam levels and changes in EEG beta power, pattern shift visual evoked potential amplitude, heart rate, and saccade amplitude and velocity. The patient and control groups differed significantly in the onset latencies of their saccadic eye movements, and marginally in EEG beta power, both before and after midazolam. However, no differences were detected between the groups in the dose of midazolam required to produce sedation or in midazolam's neurophysiological effects.

Relationship between hepatitis C genotype and severity of recurrent hepatitis C after liver transplantation

BACKGROUND

Recurrence of hepatitis C virus (HCV) infection after liver transplantation is universal, but the relationship between hepatitis C genotype and posttransplant outcome has been controversial. The aim of this study was to assess the relationship between hepatitis C genotype on posttransplant frequency of recurrent hepatitis, histologic severity of recurrence, and progression to cirrhosis.

METHODS

We studied 42 HCV RNA positive patients who received transplants between 1985 and 1994. Sera were tested for HCV RNA and protocol liver biopsies were in obtained the posttransplant period. Biopsies were scored according to the histologic activity index (HAI) and staged in a blinded fashion.

RESULTS

The distribution of hepatitis C genotypes distribution was as follows: 1a, 19 (45%); 1b, 17 (40%); 2b, 3 (7%); and 1 each of 2a, 3a, and 4a. There was histologic evidence of hepatitis in 38 of 42 (90.4%) of patients. Hepatitis C was mild, moderate, or severe (HAI>3) in 38% of grafts and minimal (HAI 0-3) in 62%. Overall HAI scores and histologic stage were higher in the genotype 1b group. Six of 17 (35%) genotype 1b patients had cirrhosis compared with 2 of 25 (8%) in the non-1b genotype group.

CONCLUSIONS

(1) Histologic evidence of recurrent hepatitis C is seen in 90% of liver allografts; (2) Histologic hepatitis C recurs with similar frequency in genotype 1b and non-1b recipients; (3) Genotype 1b is associated with more severe histologic disease recurrence than non-1b genotypes; (4) Genotype 1b appears to be associated with a higher degree of posttransplant fibrosis and cirrhosis than non-1b genotypes.

Increased risk of hepatocellular carcinoma in patients infected with hepatitis C genotype 1b

Infection with hepatitis C virus (HCV) genotype 1b has been reported to be associated with more severe liver disease and an unfavorable outcome. Liver transplantation allows for a complete examination of the explanted liver for the detection of hepatocellular carcinoma (HCC).

OBJECTIVE

To study the prevalence of HCC in patients with liver cirrhosis secondary to chronic infection with HCV genotype 1b compared with those infected with other genotypes.

METHODS

Sera were collected from 48 consecutive patients undergoing liver transplantation for end stage liver disease secondary to HCV infection. RNA was extracted from serum using chaotropic lysis and isopropanol precipitation. Reverse transcriptase-polymerase chain reaction of the NS5 region was performed, followed by automated sequencing on desalted amplification products. Genotype assignment followed Simmonds's classification. All explanted livers were examined for the presence of HCC.

RESULTS

HCV genotypes in our patients were as follows: subtype 1a, 20 patients (42%); 1b, 18 patients (37.5%); 2a, one patient (2%); 2b, six patients (12.5%); 3a, one patient (2%); and 4a, two patients (4%). Although five of 18 patients infected with HCV genotype 1b (28%) had HCC, only one of 30 patients (3%) infected with all other genotypes (1a, 2a, 2b, 3a, and 4a) had HCC (p = 0.02).

CONCLUSION

Infection with HCV genotype 1b may carry a higher risk for the development of HCC than infection with other HCV genotypes.

Effect of flumazenil on ventilatory drive during sedation with midazolam and alfentanil

BACKGROUND

Patients who receive a combination of a benzodiazepine and an opioid for conscious sedation are at risk for developing respiratory depression. While flumazenil effectively antagonizes the respiratory depression associated with a benzodiazepine alone, its efficacy in the presence of both a benzodiazepine and an opioid has not been established. This study was designed to determine whether flumazenil can reverse benzodiazepine-induced depression of ventilatory drive in the presence of an opioid.

METHODS

Twelve healthy volunteers completed this randomized, double-blind, crossover study. Ventilatory responses to carbon dioxide and to isocapnic hypoxia were determined during four treatment phases: (1) baseline, (2) alfentanil infusion; (3) combined midazolam and alfentanil infusions, and (4) combined alfentanil, midazolam, and "study drug" (consisting of either flumazenil or flumazenil vehicle) infusions. Subjects returned 2-6 weeks later to receive the alternate study drug.

RESULTS

Alfentanil decreased the slope of the carbon dioxide response curve from 2.14 +/- 0.40 to 1.43 +/- 0.19 l.min-1.mmHg-1 (x +/- SE, P < 0.05), and decreased the minute ventilation at P(ET)CO2 = 50 mmHg (VE50) from 19.7 +/- 1.2 to 14.8 +/- 0.9l.min-1 (P < 0.05). Midazolam further reduced these variables to 0.87 +/- 0.17 l.min-1.mmHg-1 (P < 0.05) and 11.7 +/- 0.8 l.min-1 (P < 0.05), respectively. With addition of flumazenil, slope and VE50 increased to 1.47 +/- 0.37 l.min-1.mmHg-1 (P < 0.05) and 16.4 +/- 2.0l.min-1 (P < 0.05); after placebo, the respective values of 1.02 +/- 0.19 l.min-1.mmHg-1 and 12.5 +/- 1.2 l.min-1 did not differe significantly from their values during combined alfentanil and midazolam administration. The effect of flumazenil differed significantly from that of placebo (P < 0.05). Both the slope and the displacement of the hypoxic ventilatory response, measured at P(ET)CO2 = 46 +/- 1 mmHG, were affected similarly, with flumazenil showing a significant improvement compared to placebo.

CONCLUSIONS

Flumazenil effectively reverses the benzodiazepine component of ventilatory depression during combined administration of a benzodiazepine and an opioid.

The role of ultrasonography and automatic-needle biopsy in outpatient percutaneous liver biopsy

The risk of complications from percutaneous liver biopsy is low, but discomfort is common and complications require hospitalization in approximately 4% of patients. The optimal method of performing these biopsies is unknown. The goal of our study was to determine whether the use of ultrasonography in the biopsy room immediately prior to or during the procedure would lessen the risk of complications and to compare the safety and efficacy in obtaining tissue by use of a Trucut needle versus an automatic biopsy needle. Between 1992 and 1994, 836 patients were entered into a randomized study (489 in Rochester, MN; 347 in Barcelona, Spain). Patients were randomized immediately prior to liver biopsy into four groups: Trucut needle, or automatic biopsy needle, and with or without ultrasonography. Fisher's Exact Test and a logistic regression model were also used to assess the effect of needle and ultrasonography on the odds for complications. The four biopsy groups were well-matched at entry with respect to age, sex, underlying liver disease, hemoglobin, prothrombin time, and platelet count. The use of ultrasound was associated with a decreased rate of hospitalization for pain, hypotension, or bleeding (2 vs. 9, P < .05). No difference in safety was found between the two types of needles. The number of passes needed to obtain specimens was similar for all four groups. The average length of the specimen was slightly greater with ultrasonographic-guided biopsies (1.7 mm vs. 1.6 mm, P < .05) and with biopsies obtained using the automatic biopsy needle when compared with the Trucut needle (1.7 mm vs. 1.5 mm, P < .05), but this did not seem to be clinically important. The addition of ultrasonography reduces complications in patients undergoing percutaneous liver biopsy. The type of needle appears to offer little difference in safety or yield of diagnostic tissue. The use of ultrasonography for guidance of percutaneous liver biopsy will lead to a lower rate of complications. The value of this benefit must be weighed against the added cost of ultrasonographic guidance.

The role of ultrasonography and automatic-needle biopsy in outpatient percutaneous liver biopsy

The risk of complications from percutaneous liver biopsy is low, but discomfort is common and complications require hospitalization in approximately 4% of patients. The optimal method of performing these biopsies is unknown. The goal of our study was to determine whether the use of ultrasonography in the biopsy room immediately prior to or during the procedure would lessen the risk of complications and to compare the safety and efficacy in obtaining tissue by use of a Trucut needle versus an automatic biopsy needle. Between 1992 and 1994, 836 patients were entered into a randomized study (489 in Rochester, MN; 347 in Barcelona, Spain). Patients were randomized immediately prior to liver biopsy into four groups: Trucut needle, or automatic biopsy needle, and with or without ultrasonography. Fisher's Exact Test and a logistic regression model were also used to assess the effect of needle and ultrasonography on the odds for complications. The four biopsy groups were well-matched at entry with respect to age, sex, underlying liver disease, hemoglobin, prothrombin time, and platelet count. The use of ultrasound was associated with a decreased rate of hospitalization for pain, hypotension, or bleeding (2 vs. 9, P < .05). No difference in safety was found between the two types of needles. The number of passes needed to obtain specimens was similar for all four groups. The average length of the specimen was slightly greater with ultrasonographic-guided biopsies (1.7 mm vs. 1.6 mm, P < .05) and with biopsies obtained using the automatic biopsy needle when compared with the Trucut needle (1.7 mm vs. 1.5 mm, P < .05), but this did not seem to be clinically important. The addition of ultrasonography reduces complications in patients undergoing percutaneous liver biopsy. The type of needle appears to offer little difference in safety or yield of diagnostic tissue. The use of ultrasonography for guidance of percutaneous liver biopsy will lead to a lower rate of complications. The value of this benefit must be weighed against the added cost of ultrasonographic guidance.

Sevoflurane versus isoflurane: induction and recovery characteristics with single-breath inhaled inductions of anesthesia

Because of its nonpungent odor and low blood-gas solubility coefficient, sevoflurane might be an ideal drug for single-breath inhaled induction of anesthesia. Fifty ASA grade I-III ambulatory surgical patients (18-76 yr old) received a single-breath induction with either 5.0% sevoflurane or 5.0% isoflurane (randomized) in a 1:1 N2O/O2 mixture. Anesthesia was maintained with the same anesthetic in 70% N2O until the end of surgery, when anesthetics were abruptly discontinued. Induction times (loss of eyelash reflex) were similar for sevoflurane (75 +/- 3 s, mean +/- se) and isoflurane (67 +/- 4 s, P = not significant). Sevoflurane patients were less likely to have complications during induction (P < 0.005); coughing occurred more frequently with isoflurane (P < 0.001). During induction, heart rate increased with both sevoflurane (from 73 +/- 3 to 90 +/- 4 bpm, P < 0.05) and isoflurane (from 70 +/- 2 to 92 +/- 2 bpm, P < 0.05); the increase with isoflurane was greater than that with sevoflurane. Times to eye opening for sevoflurane (8.1 1 +/- 1.0 min) did not differ significantly from those for isoflurane (10.6 +/- 1.3 min). Patients opened their eyes at lower end-tidal minimum alveolar anesthetic concentration (MAC)-fractions of sevoflurane (0.12 +/- 0.01 MAC) than isoflurane (0.15 +/- 0.01 MAC, P < 0.01). During recovery, patients who received sevoflurane felt less clumsy (P < 0.001) and less confused (P < 0.005) but had higher pain scores (P < 0.005) than those who received isoflurane. Sevoflurane is more suitable than isoflurane for single-breath induction, because it produces a smoother induction with a lower incidence of complications and better patient acceptance.

Liver resection and cyst fenestration in the treatment of severe polycystic liver disease

BACKGROUND/AIMS

There is limited information on treatment options for massive, highly symptomatic polycystic liver disease. The aim of the study was to analyze the immediate and long-term outcome of combined liver resection and fenestration.

METHODS

Information was abstracted from medical records. Follow-up was obtained by mailed questionnaire. Liver volume was quantified by computed tomography.

RESULTS

Thirty-one patients underwent liver resection and fenestration between July 1985 and June 1993. Mean liver volume was 9357 mL before and 3567 mL after surgery. There was one death from postoperative intracerebral bleed. Eighteen patients experienced complications, usually transient pleural effusions or transient ascites. Twenty-eight of 29 surviving patients with adequate follow-up have experienced immediate and sustained relief of symptoms and improvement in quality of life. After median follow-up of 2.4 years (range, 0.2 to 7.9 years), most patients have not had clinically significant enlargement of the liver. Sequential computed tomography scans before and after surgery suggest that hepatic enlargement in the age range of the patients in the study mainly resulted from the expansion of existing cysts rather than from the development of new cysts.

CONCLUSIONS

Selected patients with severe symptomatic polycystic liver disease and favorable anatomy benefit from liver resection and fenestration with acceptable morbidity and mortality. The extent of hepatic resection and fenestration is important for the long-term effectiveness of this procedure.

Hepatic venous outflow obstruction in autosomal dominant polycystic kidney disease

To discuss the clinical presentation, diagnosis, and treatment of hepatic venous outflow obstruction as a complication of polycystic liver disease, four cases diagnosed and treated at our institution have been reviewed and the information from six previously published case reports has been summarized. Eight of the 10 patients were women. All presented with severe ascites. Nine had polycystic kidneys. Three had moderate-to-advanced renal insufficiency, four were on hemodialysis, and one had a renal allograft. Possible predisposing factors were identified in seven patients; the most common was recent abdominal surgery, which, in three cases, was a bilateral nephrectomy. All patients had extrinsic compression of the hepatic veins and the inferior vena cava by hepatic cysts, and four had proven superimposed thrombosis of the inferior vena cava and/or hepatic veins. In the patients seen in this institution, magnetic resonance imaging was helpful in determining the level of obstruction in the inferior vena cava and the patency of the hepatic and portal veins. The outcome was worse in the patients with thrombosis; one recovered after a portocaval shunt, and the remaining three patients died. On the other hand, five of the six patients without thrombosis recovered after alcohol sclerosis of a large dominant cyst (one patient) or after hepatic resection and cyst fenestration (four patients). Hepatic venous outflow obstruction probably has been underrecognized as a cause of portal hypertension, ascites, and liver dysfunction in polycystic liver disease. The diagnosis can be reliably established with current imaging techniques, especially magnetic resonance imaging.(ABSTRACT TRUNCATED AT 250 WORDS)

Diphenhydramine enhances the interaction of hypercapnic and hypoxic ventilatory drive

BACKGROUND

Although diphenhydramine is frequently used to treat pruritus and nausea in patients who have received neuraxial opioids, there are no data regarding its effect on ventilatory control. We conducted the current study to evaluate the effects of diphenhydramine on hypercapnic and hypoxic ventilatory control in healthy volunteers.

METHODS

First, we measured the steady-state ventilatory response to carbon dioxide during hyperoxia with an end-tidal carbon dioxide tension of 46 or 54 mmHg (alternate subjects) in eight healthy volunteers. We then determined the hypoxic ventilatory response during isocapnic rebreathing at the same carbon dioxide tension. After a 10-min recovery period, we repeated the steady-state and hypoxic ventilatory response measurements at the other carbon dioxide tension (54 or 46 mmHg). Ten minutes after subjects received diphenhydramine 0.7 mg.kg-1 intravenously, we repeated this sequence of ventilatory measurements.

RESULTS

Under hyperoxic conditions (inspired oxygen fraction > 0.5) diphenhydramine did not affect the ventilatory response to hypercapnia. Similarly, at an end-tidal carbon dioxide tension of 46 mmHg, neither the slope nor the position of the hypoxic ventilatory response curve changed significantly after diphenhydramine. However, at an end-tidal carbon dioxide tension of 54 mmHg, the slope of the hypoxic ventilatory response increased from 1.28 +/- 0.33 to 2.13 +/- 0.61 l.min-1.%SpO2(-1) (mean +/- standard error), and VE at an arterial hemoglobin oxygen saturation of 90% increased from 31.2 +/- 3.1 to 43.1 +/- 5.4 l.min-1).

CONCLUSIONS

We conclude that although it did not affect the ventilatory response to carbon dioxide during hyperoxia or the ventilatory response to hypoxia at an end-tidal carbon dioxide tension of 46 mmHg diphenhydramine augmented the hypoxic response under conditions of hypercapnia in our young healthy volunteers. Although these findings may help to explain the apparent safety of diphenhydramine, they may not be applicable to debilitated patients or those who have received systemic or neuraxial ventilatory depressants.

Propofol depresses the hypoxic ventilatory response during conscious sedation and isohypercapnia

BACKGROUND

Propofol infusion at subanesthetic doses provides reliable conscious sedation. However, the ventilatory effects of sedative doses of propofol have not been established. The current study was conducted to determine the effects of propofol sedation on the hypoxic ventilatory response.

METHODS

Eight healthy, male volunteers received 1 mg.kg-1 propofol followed by a propofol infusion adjusted to maintain a constant, subanesthetic level of sedation. Hypoxic ventilatory response was measured using an isocapnic rebreathing technique: while keeping PETCO2 constant (approximately 6 mmHg above prestudy baseline), the authors continuously recorded minute ventilation and tidal volume, as oxygen saturation (SpO2) decreased from 98 to 70%. Hypoxic response determinations were performed before and during propofol infusion, as well as 30 and 60 min after termination of the propofol infusion.

RESULTS

The slope of the hypoxic ventilatory response curve (VE vs. SpO2) decreased from 0.88 +/- 0.15 to 0.17 +/- 0.03 l.min-1.%SpO2 -1 during propofol sedation (mean +/- SE). Thirty minutes after discontinuation of the propofol infusion, slope returned to its prepropofol value. In addition, minute ventilation at SpO2 = 90% decreased during propofol sedation, from 16.1 +/- 0.8 to 8.7 +/- 0.4 l.min-1, accompanied by a similar decrease in tidal volume at SpO2 = 90%, from 1,099 +/- 87 to 523 +/- 21 ml. Thirty minutes after discontinuation of the propofol infusion, these variables also returned to their prepropofol values.

CONCLUSIONS

The authors concluded that propofol infusion for conscious sedation significantly decreases the slope and causes a downward shift of the hypoxic ventilatory response curve measured during isohypercapnia.

Prevalence of peritonitis and the ascitic fluid protein concentration among chronic liver disease patients

The prevalence of spontaneous bacterial peritonitis (SBP) or its variants, bacterascites (BA), and culture-negative neutrocytic ascites (CNNA), may vary depending on the underlying liver disease and protein content of ascites. In this study, we compared the frequency of peritonitis (SBP, BA, CNNA) upon admission in alcoholic (ALD), cholestatic (CLD), and hepatocellular liver disease (HLD); determined the relationship between Child's class and prevalence of peritonitis; and assessed ascitic total protein as a risk factor for peritonitis. Between January 1989 and April 1991, 113 consecutive patients were admitted with chronic liver disease and ascites (49, ALD; 22, CLD; and 42, HLD). All had admission paracentesis. SBP was defined as polymorphonuclear cell count (PMN) > or = 250 mm3 with a positive culture, BA as PMN < 250/mm3 and positive culture, and CNNA as PMN > or = 250/mm3 with negative culture. No patients with obvious intraabdominal source for infection (i.e., secondary peritonitis) were included in the analysis. The prevalence of peritonitis was 8/113 (7%); four patients had SBP, one BA, and three CNNA. The occurrence of peritonitis was independent of the type of liver disease (ALD, 8%; CLD, 9%; HDL, 5%). Neither ascitic fluid total protein nor the severity of liver disease (Child's class) predicted the occurrence of peritonitis. We conclude that the occurrence of peritonitis is unrelated to the type of liver disease, and severity of liver disease did not predict the presence of peritonitis. Also, ascitic fluid total protein < 1.0 g/dl may not be a sensitive predictor of risk of peritonitis.

Nerve stimulator polarity and brachial plexus block

To determine whether needle polarity significantly affects nerve stimulation during peripheral nerve block, we performed a randomized double-blinded study of 10 patients undergoing axillary block for upper extremity surgery. Using an insulated needle, we determined the minimum current necessary to elicit muscle contraction with positive and negative needle polarity at two needle placements: (A) where stimulation was first observed and (B) where stimulation was maximal. At Position A, stimulation required significantly more current when the needle was positive (2.32 +/- 0.45 mA, mean +/- SEM) than when it was negative (1.05 +/- 0.23 mA, P < 0.001). Similarly, at Position B, stimulation required more current when the needle was positive (1.49 +/- 0.49 mA) than when it was negative (0.47 +/- 0.15 mA, P < 0.001). The mean ratio of positive to negative threshold stimulation current at Position B (3.11 +/- 0.20) was significantly greater than that at Position A (2.37 +/- 0.19, P < 0.05). Our results emphasize the importance of attaching the negative terminal of the nerve stimulator to the stimulating electrode. Use of the positive terminal could lead to abandoning a block if stimulation were not obtained at a low enough current; alternatively, motor contraction might not be observed before neural contact or vascular puncture.

High-volume postobstructive choleresis after transhepatic external biliary drainage resolves with conversion to internal drainage

We report high-volume postobstructive choleresis in two patients who underwent transhepatic external drainage for malignant biliary obstruction. Excessive loss of bicarbonate-rich biliary fluid (up to 6.5 L/day) caused orthostatic hypotension, prerenal insufficiency, hyponatremia, and a decrease in serum bicarbonate. Therapy with isotonic fluids containing sodium, chloride, lactate, bicarbonate, and potassium was based on measurement of biliary fluid volume and electrolyte concentrations. Biliary fluid loss was terminated by conversion to internal biliary drainage. The reason for this rare complication of external drainage of biliary obstruction is unknown, but such patients must be closely monitored for volume loss. When high-volume choleresis occurs, biliary fluid and electrolyte losses should be precisely measured and replaced, and external biliary drainage converted to internal drainage.

Sedative doses of propofol increase beta activity of the processed electroencephalogram

The effects of sedative infusions of propofol on the processed electroencephalograms (EEG) of eight healthy male volunteers were studied. EEG data for aperiodic analysis were collected during three 5-min periods: before propofol, during propofol infusion, and 30 min after termination of the infusion. After an initial dose of 1 mg/kg, subjects received a propofol infusion titrated to produce a standard level of conscious sedation. The infusion rate was 84 +/- 27 micrograms.kg-1 x min-1 (mean +/- SE) and plasma propofol levels were 2180 +/- 43 ng/mL. Total EEG power, defined as the sum of the squares of peak-to-peak amplitudes during each 5-s epoch, increased from 1350 +/- 295 microV2 x epoch-1 to 9675 +/- 2390 microV2 x epoch-1 during the propofol infusion (P < 0.05); it returned to 1445 +/- 145 microV2 x epoch-1 30 min after the infusion was discontinued (P < 0.05 vs the result during propofol). The change in total power was accompanied by a change in the distribution of power within the EEG spectrum, as the fraction of activity in the beta-band (12-35 Hz) increased during the infusion from 23% +/- 3% to 44% +/- 5% (P < 0.05). Thirty minutes after the infusion was terminated, the distribution of activity within the EEG spectrum had reverted to pre-propofol patterns. The similarity of EEG effects seen with sedative doses of propofol and benzodiazepines suggests that these drugs may share some neurochemical effects.

Randomized controlled trial of recombinant alpha-2a-interferon for chronic hepatitis C. Comparison of alanine aminotransferase normalization versus loss of HCV RNA and anti-HCV IgM

We enrolled 32 patients with chronic hepatitis C into a randomized, controlled trial to evaluate the efficacy of recombinant alpha-2a-interferon treatment. Sixteen patients were randomized to receive 1.5 million units of recombinant alpha-2a-interferon subcutaneously, thrice weekly, for six months while the remaining 16 patients were randomized to a control group that received no treatment. The mean serum alanine aminotransferase (ALT) level during the six-month study period, expressed as a percentage of the prestudy baseline value, was 82% for the control group compared to 56% for the treatment group (P = 0.014). One fourth of the treatment group normalized their serum ALT level compared to only 6% of the controls (P = 0.05). During posttherapy follow-up, 86% of responders clinically relapsed. Loss of anti-HCV IgM and HCV RNA occurred exclusively in interferon-treated responders. Anti-interferon antibodies developed in 32% of all treated patients. Forty percent of nonresponders developed anti-interferon antibodies compared to only 14% of responders (P = NS). We conclude that recombinant alpha-2a-interferon is clinically effective in patients with chronic hepatitis C. However, most responders in this trial of low-dose interferon relapsed upon cessation of treatment.

The effect of flumazenil on midazolam-induced depression of the ventilatory response to hypoxia during isohypercarbia

BACKGROUND

While flumazenil reverses benzodiazepine-induced sedation, its ability to antagonize the ventilatory depressant effects of benzodiazepines has not been fully established. A randomized, double-blind study was conducted to determine whether flumazenil effectively reverses midazolam-induced depression of the hypoxic ventilatory response.

METHODS

Twelve healthy male volunteers received intravenous midazolam 0.12 +/- 0.01 mg.kg-1 followed by either flumazenil 1.0 mg or placebo. Hypoxic ventilatory response was measured using an isocapnic rebreathing technique: as Spo2 decreased to 70% VE and tidal volume were continuously recorded. Hypoxic response determinations were performed before and after midazolam, as well as 3, 30, 60, 120, and 180 min after flumazenil or placebo.

RESULTS

After midazolam, the slope of the hypoxic ventilatory response curve (VE vs. SpO2) decreased to 0.59 +/- 0.05 (means +/- SE) times its premidazolam baseline; likewise, at Spo2 = 90%, minute ventilation (VE90) and tidal volume (TV90) decreased to 0.70 +/- 0.04 and 0.62 +/- 0.03 times baseline, respectively. Three minutes after flumazenil, the slope increased to 1.10 +/- 0.13 times baseline (P < 0.05 vs. postmidazolam), while following placebo, it was only 0.81 +/- 0.09 times baseline (P = NS vs. postmidazolam, P < 0.05 between treatments). VE90 and TV90, after flumazenil, increased to 1.45 +/- 0.15 and 1.27 +/- 0.09 times baseline, respectively (P < 0.05 vs. postmidazolam); these increases were significantly greater than the corresponding changes observed after placebo (P < 0.05 between treatments).

CONCLUSIONS

It was concluded that, after sedation with midazolam, flumazenil causes a greater increase in hypoxic ventilatory response during isohypercarbic conditions than does placebo, and may, therefore, be useful in the treatment of midazolam-induced ventilatory depression.