Reproducibility and feasibility of neurophysiological assessment of the sensory trigeminal system for future application to paroxysmal headaches

As the distribution of pain in primary headaches suggests involvement of the trigeminal sensory pathways, trigeminal somatosensory evoked potentials (TSEP) and blink reflexes (BR) may provide important information about their functional integrity. Functional differences between symptomatic and non-symptomatic sides and between measurements during and outside attacks may be particularly informative. These tests should therefore be reproducible and should require a suitable number of patients for future studies in patients with primary, paroxysmal headaches. We performed TSEP and BR twice in 22 healthy volunteers, in order to calculate sample sizes based on reproducibility data. This is, to our knowledge, the first study investigating the reproducibility of TSEP and BR measurements. Latencies of TSEP and BR are appropriate for future studies, as their reproducibility allows practical sample sizes (less than 25 subjects). Duration, amplitude and area parameters of the BR responses were less appropriate for longitudinal studies.     

Trigeminocervical reflex in fibromyalgia patients

OBJECTIVE: To describe the properties of trigeminocervical reflex in normal subjects and in patients with primary fibromyalgia syndrome (PFS) having neck pain. DESIGN: Prospective testing of trigeminocervical reflex. SETTING: University hospital electromyography laboratory in Turkey. PARTICIPANTS: Patients with PFS (n=16) and healthy volunteers (n=20). INTERVENTIONS: Trigeminocervical reflex is a brainstem reflex that is evoked by stimulating the sensory branches of the trigeminal nerve and can be recorded from the neck muscles. Electric stimulation of the supraorbital nerve evokes a reflex response (C3) and early reflex response (C1). The mean latencies of C1 and C3 of patients with PFS were compared with normal values.Main Outcome Measure: The C1 and C3 latencies of trigeminocervical reflex. RESULTS: In healthy volunteers, C3 latency +/- standard deviation was 54.17+/-6.00ms ipsilaterally and 51.25+/-9.26ms contralaterally. The difference was not significant (P=.26). The C1 latency was 17.46+/-4.89ms. In patients with PFS, C1 latency was 13.83+/-4.48ms and the C3 latency was 62.70+/-18.22ms. The difference was not significant between the patients (P=.08) and healthy volunteers (P=.17). CONCLUSION: In patients with PFS having neck pain, trigeminocervical connections were not influenced and some other mechanisms may be responsible for pain in these patients.     

The blink reflex in chronic cluster headache: a comparison with migraine patients suffering from unilateral pain

Objective. To evaluate the blink reflex (BR) in chronic cluster headache (CH) patients. Design. The elecrophysiological data were collected in during the headache-free phase. Setting. Headache patients were recruited from outpatients seen for the first time at the First Neurologic Clinic of Bari University. Patients and participants. Ten CH patients, 19 migraine without aura patients with strictly unilateral headache (MwoA) and 18 normal controls were selected. Measurements and results. The BR procedure was applied. In CH, a significant R2 duration increase was found on the symptomatic side in comparison with MwoA and controls. In both patient groups an early appearance of the R3 component was bilaterally clear. Conclusions. The BR findings confirm the central genesis of CH. The R3 abnormalities suggest a basic dysfunction of the Central control on the trigeminal nociceptive circuits. The R2 involvement on the symptomatic side indicates a unilateral facilitation of the trigeminal-facial connections persisting after the CH bout. Received: 3 January 2000, Accepted in revised form: 2 November 2000     

Cortical somatosensory evoked potentials: effects of positional changes

Supine and sitting lower extremity cortical somatosensory evoked potential (CSEP) examinations were performed in 30 healthy volunteers to establish normal values and to determine variations in SEP parameters caused by positional changes, side of stimulation, and the influence of height. CSEPs were recorded at Cz'-Fz (10-20 international EEG system) after stimulation of the saphenous, peroneal, and sural nerves at the ankle. Each nerve was tested in the supine and sitting positions. CSEP P1, N1 latencies and P1 - N1 amplitudes were recorded. The mean values of these parameters and side-to-side differences were determined for both positions. Parameter changes between the two positions were also determined. No statistically significant difference was found when comparing supine and sitting, or side-to-side values (p greater than 0.05). Regression analyses of P1 latency vs height revealed a significant positive correlation for both positions (p less than 0.0001-0.0045). Mean P1 and N1 latencies (msec) +/- 1SD, and mean P1 - N1 amplitudes (microV) +/- 1SD are as follows: peroneal nerve (n = 119) P1 = 39.5 +/- 2.98, N1 = 48.2 +/- 3.79, P1 - N1 = 1.41 +/- 0.81; sural nerve (n = 119) P1 = 41.3 +/- 4.03, N1 = 50.9 +/- 4.57, P1 - N1 = 1.31 +/- 0.61; saphenous nerve (n = 119) P1 = 41.5 +/- 4.02, N1 = 50.4 +/- 4.10, P1 - N1 = 0.87 +/- 0.30. The normative data generated by this study will help clinicians to distinguish normal variations in lower extremity CSEP parameters.(ABSTRACT TRUNCATED AT 250 WORDS)     

Painful shoulder in hemiplegic patients: a study of the suprascapular nerve

The etiology of painful shoulder in hemiplegic patients has not been fully explained. Since it has been shown that a suprascapular nerve lesion can be associated with "frozen shoulder," we have investigated the possibility that a similar lesion may exist in the painful contracted shoulder of hemiplegic patients. Thirty hemiplegic men with shoulder pain were examined and latencies determined for the suprascapular nerve from the supraclavicular fossa to the supra- and infraspinatus muscles on both sides. Routine EMG was also done for the same muscles. Patients' ages ranged from 45 to 85 (means 67 years) and the duration of hemiplegia from 1 month to 13 years (means 30.9 months). Mean latencies of the noninvolved side were 3.2 +/- 0.5 (SD) and 4.2 +/- 0.7 ms to the supraspinatus and the infraspinatus muscles, respectively. Three patients had latencies greater by two SD on the hemiplegic side than on the noninvolved side. Suprascapular nerve block did not relieve the shoulder pain completely in these patients. Excluding these three, mean latencies on the hemiplegic side were 3.1 +/- 0.4 ms to the supraspinatus muscle and 4.1 +/- 0.6 ms to the infraspinatus muscle. It is concluded that a lesion of the suprascapular nerve is not responsible for the painful contracted shoulder of the hemiplegic patient although such a lesion may exist incidentally.     

The Blink Reflex in Cluster Headache

To investigate the involvement of the trigeminal system in cluster headache, in twelve subjects the electrically-elicited blink reflex during a symptomatic period was examined. In eleven cases, the amplitude of the contralateral R2 response on the symptomatic side was significantly lower, at the same stimulus intensity, than on the asymptomatic side (p = 0.005). The blink reflex can be useful to evaluate biological and drug-induced phenomena in cluster headache.     

Visual function with acupuncture tested by visual evoked potential

Visual evoked potential (VEP) testing is used frequently and is an important ophthalmologic physiological test to examine visual functions objectively. The VEP is a complicated waveform consisting of negative waveform named N75 and N135, and positive waveform named P100. Delayed P100 latency and greatly attenuated amplitude on VEP are known characteristics for diagnosing optic nerve disease. Acupuncture has been used to treat wide clinical symptoms with minimal side effects. The confirmation of the efficacy of acupuncture generally relies on subjective symptoms. There is not much scientific evidence supporting the acupuncture treatments for eye diseases up to today. However, the VEP test can evaluate objectively and numerically the efficacy of the treatment by the acupuncture. We analyzed 19 healthy subjects (38 eyes). The P100 latencies in the group of less than 101.7 msec (total average) before acupuncture stimulations were not different than those after treatment (98.2 +/- 3.0 msec, 98.2 +/- 4.0 msec, respectively, p = 0.88, n = 17), but the latencies in those subjects with longer or equal to 101.7 msec were statistically different after acupuncture (104.6 +/- 2.8 msec, 101.9 +/- 3.7 msec, respectively, p = 0.006, n = 21). These results show that the acupuncture stimulation contributes to the P100 latencies of pattern reversal (PR)-VEP to some subjects who have delayed latencies, and this electrophysiological method is a valuable technique in monitoring the effectiveness of acupuncture therapy in the improvements of visual functions. The purpose of this study is to evaluate the physiological effects by acupuncture stimulations using PR-VEP in normal subjects.     

Modulation of trigeminal laser evoked potentials and laser silent periods by homotopical experimental pain

Cutaneous laser stimulation activates predominantly the A-delta and C mechano-heat nociceptors. Applied to the perioral region, low intensity CO(2)-laser pulses evoke reproducible trigeminal cortical evoked potentials (LEPs). High intensity CO(2)-laser stimuli induce a reflex response in the contracted jaw-closing muscle, the so-called laser silent period (LSP). Both LEPs and LSP provide a useful tool to study the physiology of the trigeminal nociceptive system. In ten healthy subjects we recorded the subjective ratings of the perioral laser stimulation and the trigeminal LEPs and LSP before, during and after homotopic experimental tonic muscle (infusion of hypertonic saline into the masseter muscle) and tonic skin pain (topical application of capsaicin to the cheek). LEPs were recorded from the vertex at two stimulus intensities: low (1.1 x pain threshold, PTh) and high (1.5 x PTh). LSP from masseter and temporalis muscles were recorded bilaterally through surface electromyographic (EMG) electrodes. CO(2)-laser pulses were applied to the perioral region (V2/V3) on the painful and non-painful side. The amplitude of LEPs increased with higher stimulus intensities (P<0.0001), but were suppressed by 42.3+/-5.3% during experimental muscle pain (P<0.0001) and by 41.6+/-3.2% during skin pain (P<0.0001). No pain-related effects were observed for the N and P latency of the LEPs (P> 0.20). The LSP in the masseter and temporalis muscles had similar onset-latency (80+/-5 ms), offset-latency (111+/-5 ms) and duration (31+/-4 ms). Experimental pain had no effect on the onset- and offset-latency (P>0.05). Experimental pain, whether from muscle or from skin, reduced the degree of suppression (P<0.01) and the area under the EMG curve (P< 0.005) of the LSP. The LSP was still suppressed during the post-pain recordings when the skin pain had disappeared (P<0.05). In all experiments experimental tonic pain decreased the subjective ratings of the perioral laser stimulation (P< 0.001). Experimental tonic pain, either from muscle or from skin, induced bilateral inhibitory effects on the trigeminal laser evoked potentials and brainstem reflex responses and on the subjective ratings of the laser pulses. These effects could be mediated through the activation of segmental and suprasegmental inhibitory systems that may function interdependently.     

Laboratory-free measurement of gait rhythmicity in the assessment of the degree of impairment and the effectiveness of rehabilitation in patients with vertigo resulting from vestibular hypofunction

A portable system for measurement of stride time rhythmicity was developed using the technique of Hausdorff et al (2001a Arch. Phys. Med. Rehabil. 82 1050-6). Measurement was performed for an extended period of walking of 256 steps for each foot outside of the laboratory on 18 normal individuals and 20 patients referred with symptomatic vestibular impairment. Ten of the patients were reassessed following vestibular therapy. Gait rhythmicity measured by standard deviation (SD) stride time was found to be significantly higher in patients with vestibular impairment than in normal volunteers (mean +/- SD 60.3 +/- 39.8 ms versus 21.9 +/- 4.9 ms respectively, P < 0.001, t-test). The ten patients who returned following a course of vestibular rehabilitation displayed significant improvement in SD stride time following therapy (mean +/- SD 57.3 +/- 44.6 ms prior to and 40.9 +/- 23 ms following therapy, P = 0.01, Wilcoxon signed rank test). Gait rhythmicity measurement, specifically measurement of stride time variability, appears to be a powerful assessment tool for objective measurement of extent of impairment and response to therapy in patients with vestibular hypofunction.     

Differences of blink-reflex abnormalities in patients suffering from idiopathic and symptomatic trigeminal neuralgia

We investigated the brainstem blink reflex in patients suffering from idiopathic and symptomatic trigeminal neuralgia to establish possible dysfunction in the reflex cycle and determine eventual differences between the two disease types. The study included 35 patients with idiopathic disease and seven patients with symptomatic disease, their results compared with those of 50 normal controls. We stimulated the forehead afferents of the supraorbital nerve and recorded the response from both orbicularis oculi muscles. We tested latencies of bilateral late components (R2, R2'), irritative R3 component and the incidence of R3 component. The patients with idiopathic trigeminal neuralgia showed normal parameters of blink reflex, except for the greater incidence of R3 component. Patients with symptomatic trigeminal neuralgia showed prolonged latencies of R2, R2' and R3 components when stimulating the afflicted side, but no significant difference in incidence of R3 component compared with the control group. The results indicate that abnormalities of blink reflex are significantly different in the two groups of patients. The high incidence of R3 component seems to be typical of idiopathic disease, whereas the prolonged latencies of late reflex components after stimulation of the afflicted side seem to be typical for symptomatic disease. These results suggest that testing the blink reflex may prove a significant aid in distinguishing the idiopathic and symptomatic disease types.     

Circulating nociceptin levels during the cluster headache period

The trigeminal innervation of the dura and its vessels has a prominent role in the mechanism of cluster headache. Nociceptin, an opioid neuropeptide, is the endogenous ligand of the OP-4 receptor, with both algesic and analgesic properties depending on the site of action. Nociceptin and its receptor are expressed by trigeminal ganglion cells where they co-localize with calcitonin gene-related peptide, a marker peptide of the trigeminovascular neurones. Nociceptin inhibits neurogenic dural vasodilatation, a phenomenon related to trigeminovascular activation. To explore its possible involvement in cluster headache, we studied circulating levels of nociceptin when attack-free during the cluster period, and also after the termination of the cluster period, using radioimmunoassay. In 14 cluster headache patients nociceptin levels during the cluster period were significantly lower than in age-, and sex-matched controls (4.91 +/- 1.96 vs. 9.58 +/- 2.57 pg/ml, P < 0.01). After the termination of the cluster period nociceptin levels (8.60 +/- 1.47 pg/ml) were not statistically different from controls. Nociceptin levels did not correlate with age, length of disease or episode length. Lower nociceptin levels during the cluster period may result in a defective regulation of trigeminal activity that might not protect sufficiently against the attacks.     

Optimal interelectrode distance in sensory and mixed compound nerve action potentials: 3- versus 4-centimeter bar electrodes

OBJECTIVE: To evaluate the amplitude and latency for 3-cm versus 4-cm distance between the active and reference electrodes (electrode separation) used to obtain normative sensory and mixed compound nerve action potential data. DESIGN: Prospective, unblinded clinical test evaluating 3 nerves: mixed median and ulnar across wrist (8 cm), and radial antidromic sensory (10 cm). SETTING: University and private practice electrodiagnostic laboratories. PARTICIPANTS: One hundred six adult volunteers without known neuropathy. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Peak-to-peak amplitude and onset and peak latencies. RESULTS: Mean onset latencies +/- standard deviation (SD) were equal for 3-cm and 4-cm separations (median, 1.6+/-0.2 ms; radial, 1.7+/-0.2 ms; ulnar, 1.5+/-0.2 ms). Mean peak latencies were also equal for 3-cm and 4-cm separation for radial (2.2+/-0.2 ms) and ulnar (1.9+/-0.2 ms) studies but differed for the median study (3 cm, 2.0+/-0.3 ms; 4 cm, 2.1+/-0.3 ms; P<.0001). Mean amplitudes +/- SD with 3-cm and 4-cm separations were, respectively, 101+/-39 microV and 103+/-39 microV (P=.0434) for the median, 47+/-17 microV and 48+/-16 microV (P=.0209) for the radial, and 52+/-28 microV and 55+/-29 microV (P=.0001) for the ulnar study. These differences were statistically significant but clinically insignificant. CONCLUSIONS: The results support a hypothesized difference in amplitude but not latency between 3- and 4-cm separation. Clinically, however, the magnitude was insignificant.     

Sympathetic skin responses from frontal region in migraine headache: a pilot study

Frontal sympathetic skin responses (F-SSRs) were recorded to investigate sympathetic nervous system activity in migraine headache (MH). Thirty-five patients with unilateral MH and 10 healthy volunteers were studied by evoking bilateral F-SSRs with electrical stimulation of the median nerve in attack, post-attack and interictal periods. The mean latencies were longer and the maximum amplitudes were smaller on the symptomatic side compared with the asymptomatic side ( P  < 0.05 for both amplitude and latency) in attack and in interictal periods. In five patients, F-SSRs were absent bilaterally, in four patients the responses were absent only on the symptomatic side during the attack period. In the post-attack period, F-SSRs on the symptomatic side had higher amplitudes and shorter latencies compared with the asymptomatic side ( P  < 0.01 for both amplitude and latency). There is an asymmetric sympathetic hypofunction on the symptomatic side in attack and interictal periods, whereas there is a hyperfunction in the post-attack period.     

The blink reflex in chronic tension-type headache, migraine, and cervicogenic headache

R1 and R2 blink reflex latencies were investigated blind in 10 patients with cervicogenic headache, 11 patients with chronic tension-type headache, 11 patients with migraine, and 9 headache-free controls. There were no R1 or R2 latency differences between the four groups. The latency of R1 increased significantly with headache duration in tension-type headache patients. Shorter R1 latencies were found on the symptomatic side than on the non-symptomatic side in cervicogenic headache patients. The results suggest that a state of hyperactivity may be present in the ipsilateral trigeminal nucleus in cervicogenic headache. Hypoactivity, however, may develop over time in tension-type headache.     

The blink reflex in migraine

The blink reflex is an objective and useful method to study the trigeminal system. It was recorded in 43 migraine patients and the findings compared with those of 31 healthy controls. The latencies of the R1 component were in the normal range in both groups. The R2 latencies ranged between 30 and 32 ms in the control group. In contrast, more than half of the patients with migraine had R2 latencies between 32 and 35 ms in the migraine group. Some migraine patients had latencies above 35 ms. The R2 latency was statistically significantly different between controls and migraineurs (p < 0.0001). Our findings indicate that trigeminal afferents and/or polysynaptic pathway in brainstem may be altered in migraine.     

Occipital nerve blockade in chronic cluster headache patients and functional connectivity between trigeminal and occipital nerves

Headache syndromes often involve occipital and neck symptoms, suggesting a functional connectivity between nociceptive trigeminal and cervical afferents. Although reports regarding effective occipital nerve blockades in cluster headache exist, the reason for the improvement of the clinical symptoms is not known. Using occipital nerve blockade and nociceptive blink reflexes, we were able to demonstrate functional connectivity between trigeminal and occipital nerves in healthy volunteers. The R2 components of the nociceptive blink reflex and the clinical outcome in 15 chronic cluster headache patients were examined before and after unilateral nerve blockade of the greater occipital nerve with 5 ml prilocain (1%) on the headache side. In contrast to recent placebo-controlled studies, only nine of the 15 cluster patients reported some minor improvement in their headache. Six patients did not report any clinical change. Exclusively on the injection side, the R2 response areas decreased and R2 latencies increased significantly after the nerve blockade. These neurophysiological and clinical data provide further evidence for functional connectivity between cervical and trigeminal nerves in humans. The trigeminocervical complex does not seem to be primarily facilitated in cluster headache, suggesting a more centrally located pathology of the disease. However, the significant changes of trigeminal function as a consequence of inhibition of the greater occipital nerve were not mirrored by a significant clinical effect, suggesting that the clinical improvement of occipital nerve blockades is not due to a direct inhibitory effect on trigeminal transmission.     

Walking strategy in diabetic patients with peripheral neuropathy

OBJECTIVE: Diabetic neuropathic patients show a peculiar loading pattern of the foot, which led us to hypothesize that a substantial modification exists in their deambulatory strategy. The aim of the present study was to support this hypothesis by quantifying the changes of the loading patterns and by monitoring the excursion of center of pressure (COP) during gait. RESEARCH DESIGN AND METHODS: -A total of 21 healthy volunteers (C) and 61 diabetic patients were evaluated: 27 diabetic subjects without neuropathy (D), 19 with neuropathy (DN), and 15 with previous neuropathic ulcer (DPU). A piezo-dynamometric platform was used to record the foot-to-floor interaction by measuring loading time and the instantaneous COP position during the stance phase of gait. RESULTS: Loading time was significantly longer in neuropathic patients than in control subjects (DPU: 816.8 +/- 150 ms; DN: 828.6 +/- 152 ms; D: 766.5 +/- 89.9 ms; C: 723.7 +/- 65.7 ms; P < 0.05). COP excursion along the medio-lateral axis of the foot clearly decreased from C to DPU groups (C: 6.41 +/- 0.1 cm; D: 4.88 +/- 0.2 cm; DN: 4.57 +/- 0.1 cm; DPU: 3.36 +/- 0.1 cm; P < 0.05) as well as COP excursion along the longitudinal axis for the DPU group only (C: 26.6 +/- 1 cm; D: 26.9 +/- 1 cm; DN: 27.2 +/- 1 cm; DPU: 24.2 +/- 1 cm; P < 0.05). COP integrals were significantly reduced for all pathological classes (DPU: 14.2 +/- 8 cm(2); DN: 25.8 +/- 6 cm(2); D: 27.7 +/- 3 cm(2); C: 38.6 +/- 6 cm(2); P < 0.05). CONCLUSIONS: The accurate quantification of loading patterns and of COP excursions and integrals highlights changes of foot-to-floor interaction in diabetic neuropathic patients. The decreased medio-lateral and longitudinal COP excursions and corresponding changes of loading times and patterns support our hypothesis that a change in the walking strategy of diabetic patients with peripheral neuropathy does occur.     

Myocardial function and hemoglobin oxygen affinity during hyperglycemia in the fetal lamb.

To determine the effects of maternal hyperglycemia on fetal hemodynamic and cardiac function, a study was carried out on nine chronically catheterized fetal sheep. In six fetuses, glucose was infused intravenously with an initial dose of 5 mg/kg per min. Data were compared with controls. This dose was gradually increased to 16 mg/kg per min by the fifth day. The initial blood glucose was 14.7 +/- 3.0 mg/dl and increased to 54.6 +/- 16.4 mg/dl by the last day of the infusion period (P less than 0.001). The PO2 decreased from a baseline of 20.25 +/- 3.40 to 15.88 +/- 5.24 mmHg (P less than 0.01). Similarly significant decreases were also observed for the blood O2 content and O2 hemoglobin saturation: 8.5 +/- 1.7 to 6.4 +/- 2.2 ml/dl and 62.3 +/- 13.6 to 46.1 +/- 17.6%, respectively, during hyperglycemia (P less than 0.01). The duration of the preejection period (PEP) before the start of the experiment was 45 +/- 4 ms; a final value of 57 +/- 10 ms was obtained (P less than 0.01). However, the electromechanical delay and ejection time (ET) showed no significant variation. The ratio of the PEP/ET increased from 0.31 +/- 0.04 to 0.38 +/- 0.07 (P less than 0.01) during hyperglycemia. The reticulocytes increased from 1.4 +/- 1.8 to 3.1 +/- 2.9% (P less than 0.05) and the 2,3-diphosphoglycerate decreased from 4.4 +/- 1.1 to 2.8 +/- 1.2 mumol/g hemoglobin (P less than 0.005). This study demonstrated that fetal hyperglycemia depresses myocardial function in the fetal lamb. The changes in cardiac function could not be explained by the small drop in O2 saturation.     

Evidence for a frontal cortex role in both auditory and somatosensory habituation: a MEG study

Auditory and somatosensory responses to paired stimuli were investigated for commonality of frontal activation that may be associated with gating using magnetoencephalography (MEG). A paired stimulus paradigm for each sensory evoked study tested right and left hemispheres independently in ten normal controls. MR-FOCUSS, a current density technique, imaged simultaneously active cortical sources. Each subject showed source localization, in the primary auditory or somatosensory cortex, for the respective stimuli following both the first (S1) and second (S2) impulses. Gating ratios for the auditory M50 response, equivalent to the P50 in EEG, were 0.54+/-0.24 and 0.63+/-0.52 for the right and left hemispheres. Somatosensory gating ratios were evaluated for early and late latencies as the pulse duration elicits extended response. Early gating ratios for right and left hemispheres were 0.69+/-0.21 and 0.69+/-0.41 while late ratios were 0.81+/-0.41 and 0.80+/-0.48. Regions of activation in the frontal cortex, beyond the primary auditory or somatosensory cortex, were mapped within 25 ms of peak S1 latencies in 9/10 subjects during auditory stimulus and in 10/10 subjects for somatosensory stimulus. Similar frontal activations were mapped within 25 ms of peak S2 latencies for 75% of auditory responses and for 100% of somatosensory responses. Comparison between modalities showed similar frontal region activations for 17/20 S1 responses and for 13/20 S2 responses. MEG offers a technique for evaluating cross modality gating. The results suggest similar frontal sources are simultaneously active during auditory and somatosensory habituation.     

Symptomatic benefit of supplemental oxygen in hypoxemic patients with terminal cancer: the use of the N of 1 randomized controlled trial.

The symptomatic effects of oxygen on the dyspnea of terminally ill cancer patients have not been clearly established. Therefore, the decision to administer oxygen is frequently made on an individual basis. We report on the use of N of 1 randomized clinical trial to compare the symptomatic benefit of respiratory failure due to terminal ovarian cancer. A 53-yr-old female patient underwent 6 randomized double-blind crossover trials between oxygen 5 L/min delivered by mask and air 5 L/min delivered by mask. Each treatment period lasted 5 min. The mean saturation of oxygen during the baseline period was 84% +/- 3%, versus 84% +/- 4% on air (P,NS) and 94% +/- 4% on oxygen (P less than 0.001). The mean visual analogue scale (VAS) for dyspnea was 77 +/- 4 during the baseline period as compared to 51 +/- 7 after air (P less than 0.001), and 40 +/- 5 after oxygen (P less than 0.001). The patient and the investigator chose oxygen blindly in 5 of 5 cases, air in 0 and 1 case, and no choice was made in 1 and 0 case, respectively. In 4 cases the patient considered the difference between oxygen and air to be of "much importance," and in one case to be of "moderate importance." No significant treatment, period, or interaction between period and treatment was detected for the difference in the VAS. We conclude that oxygen was significantly better than air for symptomatic improvement in this patient. The N of 1 technique is a simple and reliable method for the assessment of individual patients' response to oxygen.