Traumatic head injury in the anticoagulated elderly patient: a lethal combination.

Warfarin is the most common oral anticoagulant used for chronic anticoagulation therapy. Even without any antecedent trauma overanticoagulation can result in intracranial hemorrhage. The triad of anticoagulation with warfarin, age greater than 65 years, and traumatic head injury frequently produces a lethal brain hemorrhage. A retrospective review of more than 2000 patients admitted to the Trauma Service between September 1998 and May 2000 produced 278 patients with head injury and CT-documented intracranial hemorrhage. Of these patients 21 were admitted with an elevated prothrombin time (PT) due to anticoagulation with warfarin. Eighteen patients (86%) were above the age of 70. The most common indications for anticoagulation were atrial fibrillation (71%), deep venous thrombosis (19%), aortic valve replacement (9%), and ischemic cerebral infarcts (9%). Fourteen injuries were the result of a fall, one resulted from a gunshot wound, and one resulted from an assault. The remaining five patients were excluded as their history, workup, and evaluation by neurosurgery suggested a spontaneous bleed leading to fall rather than a fall causing a traumatic bleed. The average Glasgow Coma Score on admission was 11. The average PT and International Normalized Ratio (INR) on admission were 19.2 and 2.99 respectively. Eight of the 16 patients analyzed died. The risk of intracranial hemorrhage with relatively minor head injury is increased dramatically in the anticoagulated patient. A mortality rate of 50 per cent far exceeds the mortality rate in patients with similar head injuries who are not anticoagulated. In addition the risk/benefit equation of anticoagulation for the elderly is more complex and differs from that for younger patients. Perhaps more frequent and judicious monitoring of prothrombin time levels with lower therapeutic ranges (INR 1.5-2) is necessary.

Outcome following microsurgical vascular decompression or partial sensory rhizotomy in 125 cases of trigeminal neuralgia.

Microsurgical vascular decompression (MVD) was used for 92 patients with trigeminal root entry zone (REZ) compression. In 2 patients, abnormal nerve roots were sectioned; 5 had extra-axial tumors, and 26 with no apparent REZ pathology had partial sensory rhizotomy (PSR). Mean follow-up was 26 months; 105 patients, including those with tumors, are pain-free. Nine have infrequent discomfort, requiring no medication. Eleven patients had recurrence, 9 within 1 year; 4 on medication and 3 after reoperation are pain-free. Complications were few. No patient died. Although PSR and MVD had equal efficacy, MVD is favored to relieve trigeminal neuralgia without sensory loss.

Relations among threshold, spike height, electrode distance, and conduction velocity in electrical stimulation of certain medullospinal neurons.

This report describes how the threshold for extracellular, electrical stimulation of cell bodies in the rat's rostromedial medulla depends on the distance to the stimulating electrode. A monopolar microelectrode both delivered current pulses near medullospinal neurons and, after decay of the stimulus artifact, detected whether an orthodromic spike had occurred by collision of that spike with a suitably timed antidromic spike initiated at the thoracic spinal cord. The liminal current and the height of antidromic spikes were noted at a series of vertical electrode positions. Regression analysis was performed to determine whether threshold and the inverse of peak-to-peak spike height varied more as the radial distance or its square. The square relationship provided a much better fit for threshold and a marginally better fit for the inverse of spike height. The spatial decline in excitability (K2) averaged 859 microA/mm2, falling within the range of values found for fibers and cell bodies in other studies. The constant of spatial decline in spike height (C2) in millivolts per square millimeter was positively correlated with K2. Both C2 and K2 were negatively correlated with conduction velocity. From threshold distance curves fitted by regression analysis, the mean separation of sites of spike maxima and threshold minima along each electrode path was 16 micron; the estimated distance from these sites to, respectively, the loci of spike generation and spike excitation were positively correlated and similar. The variation of C2 and K2 with conduction velocity may be due either to an influence of the size and shape of the dendritic tree on the spatial decrement of excitability and spike height or to a confounding in the studied equations of the space-independent effect of the size of a cell body on spike height and excitability.

An estimate of minimum number of brain stem neurons required for inhibition of a flexion reflex.

The tail-flick reflex elicited by noxious heat in lightly anesthetized rats is known to be prevented by trains of low-amplitude current pulses passed through a monopolar microelectrode in the rostromedial medulla ( RMM ). The effect of the distance from such an electrode on the threshold of cell bodies was described in the preceding paper (11). This paper estimates the density of cell bodies in the RMM and, subsequently, estimates the number of cell bodies excited by the aforementioned pulses, a figure whose upper bound is between 30 and 75. The mean chronaxy for suppression of tail flick was found to be 162 microS. Correspondingly, for activation of spikes in somata of the RMM , it was found to be 170 microS. The axons belonging to these somata, located in the spinal lateral columns, had mean chronaxies of 360 microS. These comparisons favor the idea that cell bodies in the RMM , not axons, mediate the suppression of tail flick. Other evidence for this conclusion is given in the text. Resting activity in the RMM was found to average 6.33 Hz. Thus if the inhibitory process depends only on the instantaneous sum of activity in the many thousands of RMM neurons, all nocifensive reflexes should be continuously suppressed. But since this is not so, the relative timing of spikes in the population may also be critical. The synchronizing effect of electrical stimulation then explains the low number of cells needed to prevent the reflex.

The activity of neurons in the rostral medulla of the rat during withdrawal from noxious heat.

Neurons of the rostral ventromedial medulla (RVM) have been implicated in the modulation of nociceptive transmission. In order to further analyze their role in pain behavior, we studied their activity while eliciting the tail flick reflex with noxious heat. Recording sites were regions in the RVM from which microstimulation (less than or equal to 10 microA, 400 mu sec, 50 Hz continuous pulse trains) inhibited the tail flick reflex. Extracellular unit activity and tail temperature were recorded, stored, and plotted with reference to either the time of tail flick or the time when the stimulating temperature reached 45 degrees C. Neuronal discharges were found to be either increased (on-cells), decreased (off-cells), or unchanged around the time of the tail flick. The decreases in discharge were more closely correlated with the tail flick behavior than with the temperature of the stimulus. These off-cells were located at sites of lowest threshold for tail flick inhibition and tended to be ventral to on-cells. We propose that off-cells must pause if the tail flick is to occur, and that this pausing allows the transmission of nociceptive input through spinal reflex loops.

Evidence that disinhibition of brain stem neurones contributes to morphine analgesia.

Analgesia results when opiates are microinjected into the rostral ventromedial medulla (RVM). This region, which includes the nucleus raphe magnus and the adjacent reticular formation, is rich in immunoreactive enkephalin-containing neurones and terminals, and contains neurones that project to the spinal cord dorsal horn where they inhibit identified nociceptive spinothalamic tract neurones. Although opiates have previously been reported either to excite or inhibit RVM cells, the possibility of an opiate effect being consistent within a physiologically defined subclass has not been examined. Recently we described a class of neurone in the RVM (the off-cell) that abruptly pauses just before a heat-evoked tail-flick reflex. If off-cells are made to fire continuously by direct electrical stimulation of the RVM, the tail-flick reflex does not occur. We report here that analgesic doses of morphine completely eliminate the pause in firing that precedes the tail-flick reflex. We propose that this disinhibition of off-cells in the RVM is a primary process contributing to opiate inhibition of nociceptor-induced reflexes.

Lumbar intrathecal naloxone blocks analgesia produced by microstimulation of the ventromedial medulla in the rat.

In lightly barbiturate-anesthetized rats. low threshold (less than 10 micro A) electrical stimulation within the rostral ventromedial medulla inhibited the tail-flick response to noxious heat. Naloxone applied intrathecally at the lumbar level reversed this inhibition, but the same dose of naloxone applied to the cervical intrathecal space had no effect. Doses of naloxone 1- to 4-fold greater than the intrathecal dose did not antagonize tail-flick suppression when given systemically. Because neither systemic nor intrathecal naloxone had any effect on base-line tail-flick latencies, we conclude that the inhibition of the tail-flick response resulting from microstimulation in the ventromedial medulla is mediated by a spinal opioid synapse.

Naloxone-reversible analgesia produced by microstimulation in the rat medulla.

Using microstimulation of the rostral medulla in the barbiturate-anesthetized rat, a map was constructed of loci for inhibition of the tail-flick response to noxious heat. Low threshold sites (less than or equal to 10 microA) were found in both the nucleus raphe magnus and the nucleus reticularis paragigantocellularis. Chronaxie determinations indicate that analgesia was not produced by activation of large myelinated axons of passage. Systemic naloxone only antagonized the inhibition generated from stimulation at low threshold sites. Inhibition from higher threshold sites, for example from the nucleus reticularis gigantocellularis, was not naloxone reversible. Depending on the area stimulated, either an opioid-or a non-opioid-mediated inhibition results from microstimulation within the rat medulla.