Intrathecal neostigmine injection produces analgesia in volunteers and reduces hypotension from intrathecal bupivacaine in animals. Initial clinical trials with neostigmine studied doses of more than 100 microg, but animal studies suggest that smaller doses may be effective. In addition, all controlled clinical trials of neostigmine have come from one Brazilian university. This multicenter, placebo-controlled trial investigated the effects of 25-75 microg intrathecal neostigmine on analgesia and blood pressure in women undergoing vaginal hysterectomy.
After institutional review board approval was obtained at the three university centers, and after patients gave informed consent, 92 women scheduled for vaginal hysterectomy were randomized to receive an intrathecal injection of 2 ml bupivacaine, 0.75%, in dextrose plus either 1 ml saline or 25, 50, or 75 microg neostigmine. Blood pressure, heart rate, pain and nausea (both assessed by visual analog scale), and intravenous morphine use were recorded during surgery and at specified intervals afterward.
Morphine use was reduced similarly by all doses of neostigmine. Only the 75-microg dose of neostigmine increased the nausea score in the recovery room. The incidence of treatment for nausea was greater in patients receiving neostigmine (61%) than in those receiving saline placebo (29%) and was unaffected by neostigmine dose. Neostigmine did not reduce the incidence of hypotension from bupivacaine.
These data in patients after vaginal hysterectomy suggest that analgesia from intrathecal neostigmine may occur at doses less than 50 microg. In these doses, neostigmine does not reduce spinal bupivacaine-induced hypotension but may increase the need for treatment of nausea.
INTRATHECAL injection of neostigmine produces analgesia in animals, human volunteers, and patients with acute postoperative and chronic pain. Human volunteers achieve dose-dependent analgesia from doses more than 100 [micro sign]g. Intrathecal neostigmine probably produces analgesia by inhibiting the breakdown of spinally released acetylcholine, and spinal acetylcholine release is increased in the presence of pain. As a result, it might be expected that the potency of intrathecal neostigmine will increase in the postoperative period, and this was observed in animals. 
Few dose-response data exist for intrathecal neostigmine for postoperative analgesia. Intrathecal neostigmine, at a dose of 50 [micro sign]g, produced no postoperative analgesia compared with placebo in women undergoing vaginoplasty in one study, whereas 100 [micro sign]g was more effective than placebo in another study. A dose-response study in women undergoing vaginoplasty showed prolongation of time-to-first-rescue medication after 200 [micro sign]g, but not after 100 or 50 [micro sign]g neostigmine. That study, however, included only six patients per group, and the resultant low statistical power may explain the lack of a statistically significant effect of 100 and 50 [micro sign]g neostigmine, despite the two- and fivefold increases in the mean time-to-first- rescue medication observed. A larger study with 15 patients per group showed a dose-independent reduction in morphine use after administration of 25, 50, and 100 [micro sign]g intrathecal neostigmine compared with a control after lower-extremity orthopedic surgery. The primary purpose of the current study was to determine, using a multiple-institution study design with more than 20 patients per group, whether small doses of intrathecal neostigmine (25, 50, and 75 [micro sign]g) produce dose-dependent analgesia in patients after surgery.
The study protocol was approved by the human investigation committees of Wake Forest University School of Medicine (Coordinating Center), the University of Sao Paulo, and the University of Kentucky. After giving informed consent, 92 women classified as physical status 1 or 2 by the American Society of Anesthesiologists who were scheduled for total vaginal hysterectomy during spinal anesthesia were studied at the three university centers. Women older than 70 yr; those with known hypersensitivity to bupivacaine, neostigmine, propofol, or morphine; those with bradyarrhythmia; and those unable to understand the consent process or the use of the patient-controlled analgesia device were excluded.
The study was performed in a prospective, randomized, double-blind, placebo-controlled manner. Investigators at the coordinating center prospectively assigned neostigmine group quotas based on a computer-generated randomized balanced design and the anticipated availability of potential study patients. The number of patients observed was determined previously to be adequate to show a 25% linear dose-dependent reduction in morphine use (highest dose compared with placebo), an increase in the percentage incidence of nausea after surgery from 10% for placebo to 40% at the highest dose, a reduction in visual analog pain score (VAS) of 1.8 cm, and an increase in the VAS nausea score by 2 cm.
No premedication was administered on the night before surgery, although benzodiazepine was not discontinued in any patients who were routinely receiving it at night. The concept of a VAS, which consisted of a 10-cm line, with 0 signifying “no nausea” or “no pain at all” and 10 representing “the worst possible nausea or vomiting” or “the worst possible pain,” was introduced the day before surgery. On the day of surgery, an intravenous catheter was inserted, and midazolam, up to 4 mg in 0.5-mg increments, was administered intravenously in the holding room as needed for anxiolytic' purposes. Lactated Ringer's solution, 15 ml/kg, was infused over 30-60 min before intrathecal drug injection. Lumbar puncture was performed in the operating room with the patient in the lateral decubitus position, using either a 25- or 27-gauge Whitacre needle at the L3-L4 or L4-L5 interspace. After free flow of clear cerebrospinal fluid was obtained, the study solution was injected, with the needle orifice facing the ceiling. The final volume of 3 ml was injected at 1 ml per 10 s. Patients received 2 ml bupivacaine, 0.75%, in dextrose plus either 1 ml saline or an equal volume of saline containing 25, 50, or 75 [micro sign]g neostigmine. The neostigmine study solution was prepared immediately before lumbar puncture by a local anesthesiologist who was not involved in the study, and the investigator (blinded to the study solution content) added this 1 ml of solution to the bupivacaine solution. All centers used the same preparation of neostigmine (IMS Ltd, El Monte, CA). Patients were positioned supine immediately after injection and in the lithotomy position within 15 min of spinal injection. After adequate sensory blockade from spinal bupivacaine was established, all patients received propofol by constant intravenous infusion that was titrated to deep sedation during surgery.
Sensory level to pinprick was assessed every 5 min for 15 min after spinal injection, when patients arrived in the recovery room, and every 30 min until discharge. At these times in the recovery room, motor strength in the lower extremities were assessed by a four-point scale (0 = able to move normally, 1 = subjective weakness, 2 = able to move feet but not flex knees, 3 = unable to move lower extremities). Blood pressure was monitored noninvasively every 5 min during surgery, and heart rate and oxyhemoglobin saturation were continuously monitored throughout surgery. Blood pressure and heart rate were recorded before drug injection, at 5-min intervals during surgery, at 15-min intervals after surgery until 3 h from the time of intrathecal injection, hourly until recovery room discharge, then every 6 h until 24 h after injection. A decrease in mean arterial pressure of more than 20% below preanesthetic baseline level was treated by 5-mg intravenous increments of ephedrine and by intravenous fluid administration. Decreases in heart rate of less than 50 beats/min were treated with 0.25-mg increments of atropine administered intravenously. Nausea at any time during the study was treated with either 2-4 mg ondansetron or 20 mg metoclopramide administered intravenously.
Any intraoperative pain was treated with intravenous fentanyl in 25-[micro sign]g increments. Pain was assessed by the 10-cm VAS at arrival in the recovery room and every 30 min until discharge and was treated with intravenous patient-controlled morphine, with initial settings of a 2-mg dose, 10-min lockout, and an hourly limit of 12 mg. The VAS was also recorded at discharge from the recovery room and 24 h after the study solution was injected. At these same times, patients rated their nausea on a 10-cm VAS scale.
Data are presented as mean +/− SD or median and quartiles, as appropriate. Groups were compared for demographic data, and the duration of surgery and time in the recovery room were determined by one-way analysis of variance. The time-to-first-patient-controlled analgesia use was compared among groups using the Kaplan-Meier analysis and the Wilcoxon's rank sum. The incidence of adverse events and adjuvant drug use were compared among the groups by chi-squared analysis. Blood pressure, heart rate, and VAS scores were compared among the groups by two-way analysis of variance for subsequent measures. Categoric scale data and cumulative morphine use for 8 h after intrathecal injection (the expected duration of drug action) were compared among the groups using the Kruskal-Wallis test followed by the Wilcoxon's rank sum test. P values were corrected by post hoc tests, and a level of 0.05 was considered significant.
Sixty patients were observed at the University of Sao Paulo, 23 were observed at Wake Forest University School of Medicine, and 9 were observed at the University of Kentucky. Demographic characteristics, duration of surgery, duration of stay in the recovery room, or intraoperative fluid or drug administration did not differ in the neostigmine-dose groups (Table 1). Blood pressure or heart rate before intrathecal injection or after bupivacaine injection did not differ in the groups (Figure 1). Similarly, blood pressure or heart rate in the recovery room or for the subsequent 24 h (data not shown) did not differ. The incidence of treatment with ephedrine for hypotension did not differ between saline control treatment (25%) and neostigmine treatment (25%), and it did not differ with neostigmine dose (25% after placebo, 36% after 25 [micro sign]g, 27% after 50 [micro sign]g, 13% after 75 [micro sign]g; P = 0.18 by chi-squared analysis). The incidence of treatment for bradycardia did not differ between saline control treatment (0%) and neostigmine treatment (9%), and it did not differ with neostigmine dose (5% after 25 [micro sign]g, 14% after 50 [micro sign]g, and 9% after 75 [micro sign]g). Neostigmine did not affect the onset of sensory blockade from intrathecal bupivacaine (Table 2).
Only the 75-[micro sign]g dose of neostigmine increased the nausea score in the recovery room (Figure 2). Similarly, the incidence of treatment for nausea was greater in patients who received neostigmine (61%) than in those who did not (29%). However, the incidence of nausea treatment was unrelated to neostigmine dose (the incidence of nausea treatment was 54% after 25 [micro sign]g, 64% after 50 [micro sign]g, and 64% after 75 [micro sign]g).
The analgesic effect of intrathecal neostigmine was dose independent. All doses of neostigmine reduced VAS pain scores in the recovery room to a similar degree (Figure 3). All doses of neostigmine prolonged the time-until-first-morphine-dose by a similar amount. The time-to-first-morphine-dose (median, 25th to 75th percentiles) was 3 h (3 to 4 h) after saline administration, 4 h (4 to 6.5 h) after 25 [micro sign]g neostigmine administration, 4 h (3 to 5.5 h) after 50 [micro sign]g, and 6 h (4 to 8.5 h) after 75 [micro sign]g (P < 0.05 for all neostigmine groups vs. saline by Kaplan-Meier analysis). Similarly, there was a dose-independent reduction in patient-controlled analgesia morphine use during the first 8 h after intrathecal injection of neostigmine (Figure 4).
This is the first, large, dose-response study of low doses of intrathecal neostigmine for postoperative analgesia, and it provides new information regarding dose-related analgesia, nausea, and hemodynamic actions of this agent in different patient populations at the three sites.
Intrathecal neostigmine, at a dose of 100 [micro sign]g, produces mild analgesia (30-40% reduction in pain scores to a noxious cold stimulus) lasting 3 or 4 h in volunteers, whereas 50 [micro sign]g produces no detectable analgesia. However, a painful stimulus, such as occurs after surgery, might increase the apparent potency of neostigmine. This is confirmed by studies with sheep that showed an increase in acetylcholine concentrations in lumbar cerebrospinal fluid during painful electrical stimulation. In addition, the antinociceptive effect of intrathecal neostigmine is enhanced in the first 2 days after lumbar laminectomy surgery in sheep. The current study showed significant analgesia, as defined by a reduction in morphine use, with administration of 25-50 [micro sign]g intrathecal neostigmine in women after surgery, suggesting that a similar phenomenon may occur in humans.
An alternative explanation to the effect of small doses of neostigmine in the current study relates to the mechanisms of action of systemic opioids in producing analgesia. Intravenous morphine injection is thought to activate descending spinal inhibitory pathways, some of which are noradrenergic, and, as such, morphine injection increases concentrations of norepinephrine in lumbar cerebrospinal fluid in animals and humans. Spinally released norepinephrine produces analgesia in part by activating spinal cholinergic neurons to release acetylcholine, and, as such, intravenous morphine increases cerebrospinal concentrations of acetylcholine. All patients in the current study received intravenous morphine by patient-controlled analgesia, and it is possible that smaller doses of intrathecal neostigmine were more effective in these patients than in others because the patients were receiving morphine, which stimulated acetylcholine release on which neostigmine could act.
Regardless of the mechanism, the current study suggests that doses of intrathecal neostigmine of less than 50 [micro sign]g reduce intravenous morphine requirements for 8 h after vaginal hysterectomy. Previous literature does not consistently report a dose-dependent analgesic effect of neostigmine. As reviewed previously, large doses (200 [micro sign]g) uniformly produce analgesia, but also nausea. With one exception, 100 [micro sign]g neostigmine also produces analgesia, but in most cases it is associated with an increased incidence of nausea. The current study corresponds with previous data from open-label studies in women after cesarean section, and in patients after orthopedic procedures, that doses of more than 100 [micro sign]g also may be effective. The clinical usefulness of this reduction in morphine use for 8 h may be limited. Preliminary data suggest that a much longer duration of analgesia can be obtained by coadministration of intrathecal neostigmine with morphine, reflecting the synergistic interaction between these agents by intrathecal administration. 
Intrathecal neostigmine produces nausea in volunteers and patients at doses >or= to 100 [micro sign]g. The current study suggests that doses as small as 25 [micro sign]g may increase the requirements for antiemetic treatment. It is conceivable that the use of propofol during surgery for sedation reduced nausea after intrathecal neostigmine from what would be seen with concomitant propofol administration.
Intrathecal neostigmine administration diminishes hypotension from spinal bupivacaine [12,13]and from clonidine in animals and from epidural clonidine in humans. The mechanism of this interaction is thought to be augmentation by neostigmine of stimulation from spinally released acetylcholine on preganglionic sympathetic neurons. This stimulation may counteract the sympatholytic actions of the local anesthetic, bupivacaine, or of the [Greek small letter alpha]2-adrenergicagonist clonidine. We did not observe a reduced incidence of hypotension in women receiving neostigmine, suggesting that clinically useful doses of intrathecal neostigmine would not affect intrathecal bupivacaine-induced hypotension.
In conclusion, 25-75 [micro sign]g intrathecal neostigmine produces a dose-independent reduction in morphine use for 8 h after vaginal hysterectomy and a dose-independent increase in the need for antiemetic treatment. These data suggest that intrathecal neostigmine alone is unlikely to produce complete analgesia after surgery and that even doses less than 50 [micro sign]g may increase the incidence of postoperative nausea.