Measurement of motor evoked potentials following repetitive magnetic motor cortex stimulation during isoflurane or propofol anaesthesia

V. Rohde*,1, G. A. Krombach1, J. H. Baumert2, I. Kreitschmann-Andermahr1, M. Weinzierl1 and J. M. Gilsbach1

1 Department of Neurosurgery and 2 Department of Anesthesiology, Technical University (RWTH) Aachen, Germany

Corresponding author. E-mail: vrohde@ukaachen.de

Accepted for publication: May 13, 2003


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
Background. Isoflurane and propofol reduce the recordability of compound muscle action potentials (CMAP) following single transcranial magnetic stimulation of the motor cortex (sTCMS). Repetition of the magnetic stimulus (repetitive transcranial magnetic stimulation, rTCMS) might allow the inhibition caused by anaesthesia with isoflurane or propofol to be overcome.

Methods. We applied rTCMS (four stimuli; inter-stimulus intervals of 3, 4, 5 ms (333, 250, 200 Hz), output 2.5 Tesla) in 27 patients and recorded CMAP from the hypothenar and anterior tibial muscle. Anaesthesia was maintained with fentanyl 0.5–1 µg kg–1 h–1 and either isoflurane 1.2% (10 patients) or propofol 5 mg kg–1 h–1 with nitrous oxide 60% in oxygen (17 patients).

Results. No CMAP were detected during isoflurane anaesthesia. During propofol anaesthesia 333 Hz, four-pulse magnetic stimulation evoked CMAP in the hypothenar muscle in 75%, and in the anterior tibial muscle in 65% of the patients. Less response was obtained with 250 and 200 Hz stimulation.

Conclusions. In most patients, rTCMS can overcome suppression of CMAP during propofol/nitrous oxide anaesthesia, but not during isoflurane anaesthesia. A train of four magnetic stimuli at a frequency of 333 Hz is most effective in evoking potentials from the upper and lower limb muscles. The authors conclude that rTCMS can be used for evaluation of the descending motor pathways during anaesthesia.

Br J Anaesth 2003; 91: 487–92

Keywords: brain, repetitive transcranial magnetic stimulation; equipment, monitors; monitoring, motor evoked potentials


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
A major complication of spinal surgery is motor deficit after surgery. To avoid this, spinal cord function can be monitored with somatosensory evoked potentials or spinal cord evoked potentials.1 2 Neither test exclusively measures pyramidal tract function,3 with reports of unchanged potentials during surgery but postoperative paresis.4 In 1985, Barker and coworkers obtained compound muscle action potentials (CMAP) in extremity muscles (motor evoked potentials, MEP) in awake human subjects by motor cortex stimulation with a single magnetic stimulus (sTCMS).5 Following sTCMS, indirect (I) waves and, occasionally, direct (D) waves descend from the cortex and the corticospinal tract to the {alpha}-motoneuron.6 7 Summation of the waves allows the {alpha}-motoneuron to reach threshold and cause a recordable CMAP. Single TCMS has been used to study the descending motor pathways in diseases of the first and second motor neuron. Use of sTCMS for monitoring the pyramidal tract during surgery and in intensive care was less successful, because most anaesthetics inhibit MEP. Anaesthetics probably reduce the descending waves and prevent summation at the {alpha}-motoneuron. High frequency repetitive motor cortex stimulation (rTCMS) is now possible. Between two and four stimuli with a maximum intensity of 2.5 Tesla, with frequencies from 10 to 1000 Hz can be applied to the motor cortex. We studied if high frequency stimulation could generate sufficient descending waves to reach the firing threshold of the {alpha}-motoneuron despite anaesthesia with either isoflurane or propofol with nitrous oxide. Measuring CMAP during isoflurane or propofol anaesthesia could allow rTCMS to be used for monitoring the descending motor pathways during spinal surgery.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
Patients
We studied 27 patients (16 men, 11 women, mean age of 50 yr, range 24–76 yr) with lumbar disc herniation with rTCMS and CMAP recording from the hypothenar and anterior tibial muscle after induction of anaesthesia and before surgery started. We excluded patients with preoperative motor deficits, cardiac pacemakers, cerebrospinal fluid shunts, or a history of epilepsy, head trauma, or brain surgery. The study was explained and informed consent was obtained. The local ethics committee gave approval for testing during surgery, but not for randomization of the anaesthetic regimen. Either isoflurane or propofol/nitrous oxide anaesthesia was chosen according to the preference of the anaesthesiologist.

Premedication and anaesthesia
Premedication with midazolam 7.5 mg orally was given 60 min before anaesthesia. Anaesthesia was induced with propofol 2 mg kg–1, fentanyl 1–2 µg kg–1 and mivacurium 0.1–0.2 mg–1 kg–1 to facilitate tracheal intubation and positioning for operation. In 17 patients, anaesthesia was maintained with propofol 5 mg kg–1 h–1, nitrous oxide 60% in oxygen and fentanyl 0.5–1 µg kg–1 h–1. In 10 patients, anaesthesia was maintained with isoflurane 1.2% and fentanyl 0.5–1 µg kg–1 h–1, and nitrous oxide was not used. These two methods of anaesthesia were routine anaesthetic regimens during the study period and were not changed for the study. Neuromuscular transmission was monitored by train of four stimulation of the ulnar nerve and electrophysiological testing was not started before the mivacurium had stopped working. The time between giving mivacurium and first magnetic motor cortex stimulation was at least 30 min. Arterial pressure, heart frequency, oxygen saturation, and body temperature were measured during anaesthesia and were kept constant.

Repetitive TCMS of the motor cortex and recording of muscle action potential
Stimulation was applied with the flat coil (outer coil diameter 9 cm). The centre of the coil was placed on the head over the motor cortex of the arm and leg. During the entire investigation, the orientation of the coil (tangential to skull, the handle pointing strictly dorsally) was kept unchanged. Stimulation was always performed with the maximum condenser capacity (Magstim QuadroPulse Model 500, The Magstim Company, Spring Gardens, Whitland, Dyfed, UK). Train of four stimuli were used for motor cortex activation with inter-stimulus intervals (ISI) of 3, 4, and 5 ms. This corresponds to stimulation frequencies of 333, 250, and 200 Hz. Each recording was repeated at least once.

CMAP were recorded from the contralateral abductor pollicis brevis and the anterior tibial muscle, using subdermal needle electrodes in a tendon-belly arrangement. The signal was filtered between 30 and 3000 Hz, and amplified to show either 200 or 500 µV per division. For each patient, recordings were superimposed to identify the shortest latency and the greatest amplitude (baseline-to-peak) of the evoked potentials. The maximum amplitude of the potential complex was measured.

Statistical analysis
The individual CMAP latencies and amplitudes for each stimulation setting were used to calculate the mean (SD). Data were compared with a paired t-test. Significance was taken if P<0.05. The percentage of CMAP, which allowed clear identification of the potential offset and measurement of the latency and the baseline-to-peak amplitude in relation to the number of stimuli and frequency were determined. Positive CMAP identification was based on at least two recordings with amplitudes, which were at least twice the amplitude of the baseline noise. If in two successive trials, one CMAP was significant and the other not, stimulation was repeated. If the third CMAP was clearly seen, CMAP identification was considered positive. The amplitude of the baseline noise was measured from peak-to-peak with the sensitivity set to 50 µV per division.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
Hypothenar muscle
No CMAP could be found if anaesthesia was maintained with isoflurane. CMAP measurement was not possible for technical reasons in one of the 17 patients of the propofol/nitrous oxide group, so the data of 16 patients were analysed. In 12 of the 16 patients (75%), a CMAP with clearly defined latencies and stable amplitudes could be recorded from the hypothenar muscle using rTCMS with 333 Hz. When the stimulation frequency was 250 Hz, a CMAP could be obtained in 11 of the 16 patients (69%), and in eight of the 16 patients (50%) using a frequency of 200 Hz (Fig. 1). The mean latency of the EMG response to a frequency of 333 Hz was 25.6 (2.77) ms. With a reduction of the frequency to 250 and 200 Hz the latency increased to 27.2 (3.09) and 29.8 (4.77) ms. This increase was statistically significant (P<0.01) (Table 1). The mean amplitude of the evoked muscle action potential following rTCMS with 333 Hz was 0.56 (0.62) mV, and less with lower frequencies (Table 1).



View larger version (33K):
[in this window]
[in a new window]
 
Fig 1 An example of CMAP obtained from the right hypothenar and the anterior tibial muscle after four-pulse transcranial magnetic stimulation of the left motor cortex during propofol/nitrous oxide anaesthesia. The stimulation frequencies were (A) 333, (B) 250, and (C) 200 Hz.

 

View this table:
[in this window]
[in a new window]
 
Table 1 Latencies and amplitudes of the CMAP of the hypothenar and the anterior tibial muscle following rTCMS of the motor cortex (four stimuli, stimulation frequencies of 333, 250, and 200 Hz) during propofol/nitrous oxide anaesthesia The values are mean (SD). *P<0.05, compared with the mean latency following rTCMS with 333 Hz
 
Anterior tibial muscle
Isoflurane made CMAP recording from the anterior tibial muscle impossible. Measurable responses were recorded in 11 of the 17 patients (65%) given propofol. At 200 Hz, the proportion with a detectable CMAP was 53% (Fig. 1). With a frequency of 333 Hz, the mean latency of the registered muscle action potential was 36.9 (3.61) ms. Mean latencies increased when frequency of stimulation was reduced (Table 1). The mean amplitude of the evoked motor potential was 0.42 (0.34) mV with 333 Hz stimulation and less with lower frequencies, but this decrease was not significant.


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
sTCMS and drug effect
Magnetic MEP has not been widely studied during anaesthesia with volatile anaesthetics or propofol. Volatile anaesthetics suppress MEP following single magnetic stimulation of the motor cortex. In dogs, Young and coworkers found that low doses of isoflurane and halothane abolished CMAP,8 and Glassman and coworkers9 found that halothane 0.5% eliminated the MEP. These results were also found by Yamada et al. in cats.10 In a clinical study of 17 patients, the amplitude of CMAP was reduced with isoflurane 0.5%, and the potentials were lost at more than 1.0%.11 Animal studies suggest, that propofol is less depressant than volatile anaesthetics. A very high dosage (40 mg propofol kg–1 h–1) was needed to suppress CMAP in 15.4% of rats.12 In man, reports vary considerably. Two studies found that CMAP were not detected if propofol was given.11 13 In contrast, Kalkman and colleagues found recordable CMAP after a single dose of 2 mg kg–1.14

To improve measurements after sTCMS, different anaesthetic regimens have been used. Nitrous oxide/opioid anaesthesia with or without ketamine, barbiturates, and etomidate can allow sTCMS and CMAP recording.1519

Rationale for rTCMS
We attempted to obtain MEP by changing the stimulation. In non-anaesthesized humans, sTCMS can cause up to five I waves, which pass down the pyramidal tract and produces a series of excitatory postsynaptic potentials (EPSP) at the {alpha}-motoneuron. The temporal summation of EPSPs brings the {alpha}-motoneuron to a firing threshold and causes a recordable muscle action potential in the target muscle.7 The critical number of descending waves needed to cause a muscle response seems to be at least 3.20 In the cortex, volatile anaesthestics and propofol impair synaptic transmission by increasing the concentration of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) and reducing the release of excitatory neurotransmitters.21 22 This reduces the number of trans-synaptically generated I waves and impairs {alpha}-motoneuron activation during anaesthesia.23 24 We thought that if the number of the descending I waves could be increased again, then magnetic stimulation would allow MEP recording without altering the anaesthetic regimen. We stimulated the motor cortex four times with frequencies from 200 to 333 Hz. We could not show successful activation of the {alpha}-motoneuron if isoflurane 1.2% was used for anaesthesia. This suggests that isoflurane 1.2% inhibits the synapses, so that less than one I wave per stimulus is generated. A reduction of the isoflurane concentration might increase I wave generation and CMAP recordability. We found that the inhibition by propofol was less. Repetitive TCMS with four stimuli generated EMG responses in the muscles of the arm and leg. We conclude that propofol at a dose of 5 mg kg–1 h–1 reduces the number of I waves to one or two per stimulus.

Propofol was used together with nitrous oxide, which has no suppressive effect at the cortical level, but modulates the function of spinal inhibitory interneurons and {alpha}-motoneurons.25 In addition to cortical inhibition, isoflurane at 1.2% reduces the excitability of {alpha}-motoneurons.26 The present study did not allow the effect of spinal inhibition on CMAP generation to be measured, using four-pulse magnetic stimulation in anaesthesized patients. However, the spinal inhibitory effects of nitrous oxide and isoflurane at these dosages are probably small in comparison with the effects of propofol and isoflurane on the cortex.

Stimulation characteristics
We cannot infer how rTCMS generates sufficient I waves to activate the {alpha}-motoneuron during propofol anaesthesia. The initial stimulus could serve as a conditioning stimulus and lower the excitation threshold of the corticomotoneuron (intracortical excitation), so that subsequent stimuli generate a sufficient number of descending waves.20 27 It is also possible that the excitation threshold of the corticomotoneuron remains constant, and that each stimulus generates a small number of descending waves which are ineffective.28 By repetitive stimuli with summation of the waves, a descending volley is formed which depolarizes the {alpha}-motoneuron. Summation of activity is supported by the effect of the inter-stimulus interval on the evoked muscle action potentials. With an ISI of 3 ms CMAP were found in 75%, but they were only found in 50% with an ISI of 5 ms. After single motor cortex stimulation, the latency of the first I wave is 2.9 ms. The subsequent I waves follow in intervals of 1.4 ms.29 If we assume that one to two waves are generated per stimulus during propofol anaesthesia, the physiological I wave volley would be almost exactly imitated, if four stimuli with an ISI of 3 ms are applied.

Other studies of rTCMS
Gugino and coworkers30 used rTCMS in three patients anaesthesized with propofol and fentanyl. EMG responses were found when stimulating the motor cortex with three and four pulses (ISI of 2.5 ms). Others used double-pulse and quadruple-pulse stimulation at 500 Hz (ISI of 2 ms) in 40 patients giving increasing doses of propofol and remifentanil,31 which was stopped when intubation became possible. Increasing doses of propfol and, to a lesser extent, remifentanil incompletely suppressed the EMG-response after single and multiple stimuli. Our study supports these observations that anaesthesia with propofol allows recording MEP after repetitive TCMS. However, the previous studies did not indicate the best settings and pulse numbers to allow MEP to be recorded during surgery.

Clinical aspects
We cannot assess EMG-responses during anaesthesia in healthy volunteers. We studied neurologically intact patients with prolapsed lumbar discs requiring surgery. In spite of the encouraging results of the present study, we would not use rTCMS for routine monitoring during lumbar disc surgery because the risk of postoperative motor deficits is small. In addition, the hypothenar and anterior tibial muscles are supplied by different nerve roots, so that MEP could continue despite root damage during disc surgery. Postoperative deficits are more likely during cervical and thoracic spinal surgery, so we will study the role of rTCMS for motor tract monitoring during these operations. The device for rTCMS is bulky, but we found that it can be placed beside the patient’s head without interfering with the neurosurgeon’s and anaesthesiologist’s working space, using a specially designed clamp. To assess the practicability of this approach, we are comparing rTCMS with repetitive transcranial electric stimulation, the routine method for intraoperative monitoring of the descending motor pathways.32 33


    Conclusion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
Isoflurane anaesthesia in usual doses does not allow transcranial magnetic motor cortex stimulation and recording of MEP, even if four magnetic stimuli are rapidly repeated. The inhibition by propofol at potential generation is substantial, but less pronounced than that of isoflurane: four-pulse magnetic stimulation generates measurable potentials in most patients. The efficacy of four-pulse magnetic stimulation depends on frequency and is greatest at a frequency of 333 Hz. Further improvement of stimulation may be possible. A rapid sequence of more than four magnetic stimuli may overcome inhibition by isoflurane. A device that can deliver more than four stimuli at frequencies of 200 Hz or more is currently not available. Reducing isoflurane or propofol/nitrous oxide dose might increase motor evoked potential detection, but this was not studied. We conclude that rTCMS with four magnetic stimuli allows assessment of the descending motor pathways in most patients without changing a standard propofol/nitrous oxide anaesthesic regimen.


    Acknowledgement
 
This work was supported by a grant of the Tumorstiftung Kopf-Hals to the first author.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
1 Kai Y, Owen JH, Lenke LG, Bridwell KH, Oakley DM, Sugioka Y. Use of sciatic neurogenic motor evoked potentials versus spinal potentials to predict early-onset neurological deficits when intervention is still possible during overdistraction. Spine 1993; 18: 1134–9[ISI][Medline]

2 Nuwer MR, Dawson EG, Carlson LG, Kanim LEA, Sherman JE. Somatosensory evoked potential spinal cord monitoring reduces neurologic deficits after scoliosis surgery: results of a large multicenter survey. Electroencephal Clin Neurophysiol 1995; 96: 6–11[CrossRef][Medline]

3 Su CF, Haghighi S, Oro JJ, Gaines RW. ‘Backfiring’ in spinal cord monitoring. High thoracic spinal cord stimulation evokes sciatic response by antidromic sensory pathway conduction, not motor tract conduction. Spine 1992; 17: 504–8[ISI][Medline]

4 Lesser RP, Raudzens P, Lüders H, et al. Postoperative neurological deficits may occur despite unchanged intra operative somatosensory evoked potentials. Ann Neurol 1986; 19: 22–5[ISI][Medline]

5 Barker AT, Freeston IL, Jalinous R, Jarratt JA. Magnetic stimulation of the human brain and peripheral nervous system: an introduction and the results of an initial clinical evaluation. Neurosurgery 1987; 20: 100–9[ISI][Medline]

6 Amassian VE, Quirk G, Stewart M. A comparison of corticospinal activation by magnetic coil and electrical stimulation of monkey motor cortex. Electroencephal Clin Neurophysiol 1990; 77: 390–401[CrossRef][ISI][Medline]

7 Day BL, Thompson PD, Dick JP, Nakashima K, Marsden CD. Different sites of action of electrical and magnetic stimulation of the human brain. Sci Lett 1987; 75: 101–6

8 Young SS, Sylvestre AM. The effect of anesthesia with isoflurane, thiopental, methohexital, propofol or nitrous oxide on magnetic motor evoked potentials in the dog. Prog Vet Neurol 1992; 3: 91–4

9 Glassman SD, Shields CB, Linden RD, Zhang YP, Nixon AR, Johnson JR. Anesthetic effects on motor evoked potentials in dogs. Spine 1993; 18: 1083–1089[ISI][Medline]

10 Yamada H, Transfeldt EE, Tamaki T, Torres F, Iaizzo PA. The effects of volatile anesthetics on the relative amplitudes and latencies of spinal and muscle potentials evoked by transcranial magnetic stimulation. Spine 1994; 19: 1512–7[ISI][Medline]

11 Schmid DU, Boll J, Liechti S, Schmid J, Hess CW. Influence of some anesthetic agents on muscle response to transcranial magnetic cortex stimulation: a pilot study in humans. Neurosurgery 1992; 30: 85–92[ISI][Medline]

12 Fishback AS, Shields CB, Linden RD, Zhang YP, Burke D. The effects of propofol on rat transcranial magnetic motor evoked potentials. Neurosurgery 1995; 37: 969–74[ISI][Medline]

13 Taniguchi M, Nadstawek J, Langenbach U, Bremer F, Schramm J. Effects of four intravenous anesthetic agents on motor evoked potentials elicited by magnetic transcranial stimulation. Neurosurgery 1993; 33: 407–15[ISI][Medline]

14 Kalkman CJ, Drummond JC, Ribberink AA, Patel PM, Sano T, Bickford RG. Effects of propofol, etomidate, midazolam, and fentanyl on motor evoked responses to transcranial electrical or magnetic stimulation in humans. Anesthesiology 1992; 76: 502–9[ISI][Medline]

15 Ghaly RF, Stone JL, Levy WJ, Roccaforte P, Brunner EB. The effect of etomidate on motor evoked potentials induced by transcranial magnetic stimulation in the monkey. Neurosurgery 1990; 27: 936–42[ISI][Medline]

16 Kalkman CJ, Drummond JC, Patel PM, Sano T, Chasnut RM. Effects of droperidol, pentobarbital, and ketamine on myogenic transcranial magnetic motor-evoked responses in humans. Neurosurgery 1994; 35: 1066–71[ISI][Medline]

17 Kawaguchi M, Sakamoto T, Shimizu K, Ohnishi H, Karasawa J. Effect of thiopentone on motor evoked potentials induced by transcranial magnetic stimulation in humans. Br J Anaesth 1993; 71: 849–53[Abstract]

18 Lee WY, Hou WY, Yang LH, Lin SM. Intraoperative monitoring of motor function by magnetic motor evoked potentials. Neurosurgery 1995; 36: 493–500[ISI][Medline]

19 Ubags LH, Kalkman CJ, Been HD, Koelman JH, Ongerboer de Visser B. A comparison of myogenic motor evoked responses to electrical and magnetic transcranial stimulation during nitrous oxide/opioid anesthesia. Anesth Analg 1999; 88: 568–72[Abstract/Free Full Text]

20 Rodi Z, Deletis V, Morota N, Vodusek DB. Motor evoked potentials during brain surgery. Pflügers Arch Eur J Physiol 1996; 431 [Suppl.]: R291–2[Medline]

21 El-Beheiry H, Puil E. Anaesthetic depression of excitatory synaptic transmission in neocortex. Exp Brain Res 1989; 77: 87–93[ISI][Medline]

22 Keane PE, Biziere K. The effects of general anesthetics on gabaergic synaptic transmission. Life Sci 1987; 41: 1437–48[CrossRef][ISI][Medline]

23 Burke D, Hicks R, Gandevia SC, Stephen J, Woddforth I, Crawford M. Direct comparison of corticospinal volleys in human subjects to transcranial magnetic and electrical stimulation. J Physiol 1993; 470: 383–93[Abstract]

24 Hicks RG, Woodforth IJ, Crawford MR, Stephen JPH, Burke DJ. Some effects of isoflurane on I waves of the motor evoked potential. Br J Anaesth 1992; 69: 130–6[Abstract]

25 Zentner J, Thees C, Pechstein U, et al. Influence of nitrous oxide on motor-evoked potentials. Spine 1997; 22: 1002–6[CrossRef][ISI][Medline]

26 King BS, Rampil IJ. Anesthetic depression of spinal motor neurons may contribute to lack of movement in response to noxious stimuli. Anesthesiology 1994; 81: 1484–92[ISI][Medline]

27 Kalkman CJ, Ubags LH, Been HD, Swaan A, Drummond JC. Improved amplitude of myogenic motor evoked responses after paired transcranial electrical stimulation during sufentanil/nitrous oxide anesthesia. Anesthesiology 1995; 83: 270–6[ISI][Medline]

28 Inghilleri M, Berardelli A, Cruccu G, Priori A, Manfredi M. Motor potentials evoked by paired cortical stimuli. Electroencephal Clin Neurophysiol 1990; 77: 382–9[CrossRef][ISI][Medline]

29 Kernell D, Wu CP. Responses of the pyramidal tract to stimulation of the baboon’s motor cortex. J Physiol 1967; 191: 653–72[ISI][Medline]

30 Gugino LD, Aglio LS, Potts G, et al. Perioperative use of transcranial magnetic stimulation. Techniques Neurosurg 2001; 7: 33–51[ISI]

31 Scheufler KM, Zentner J. Total intravenous anesthesia for intraoperative monitoring of the motor pathways: an integral view combining clinical and experimental data. J Neurosurg 2002; 96: 571–9[ISI][Medline]

32 Pechstein U, Cedzich C, Nadstawek J, Schramm J. Transcranial high-frequency repetitive electrical stimulation for recording myogenic motor evoked potentials with the patient under general anesthesia. Neurosurgery 1996; 39: 335–44[ISI][Medline]

33 Calancie B, Harris W, Brindle GF, Green BA, Landy HJ. Threshold-level repetitive transcranial elctrical stimulation for intraoperative monitoring of central motor conduction. J Neurosurg 2001; 95: 161–8[ISI][Medline]