Randomized, double-blind, placebo-controlled crossover trial of modafinil in the treatment of excessive daytime sleepiness in narcolepsy

There is a need for new stimulant medication for the excessive daytime sleepiness (EDS) seen in many neurologic sleep disorders. Medication is often relatively ineffective, poorly tolerated with significant side effects, associated with a development of tolerance, or has a strong potential for illicit drug use. Modafinil, or 2-[(diphenylmethyl)-sulphinyl] acetamide, represents a novel drug being developed for EDS treatment. The exact mechanism of action is unclear, but may involve centralα₁-adrenergic tone. Modafinil exerts its psychomotor stimulant effect by a mechanism of action different from amphetamine and methylphenidate, both of which induce increased dopamine levels in the caudate nucleus in conjunction with stereotyped behavior and increased locomotor activity, whereas modafinil-induced excessive locomotor activity is not accompanied by either stereotyped behavior or increases in dopamine.¹ The subjective effects of modafinil are also markedly different from those of amphetamine, suggesting that at clinically useful doses modafinil does not have the same abuse liability as amphetamine.² In humans modafinil increases high-frequency alpha waves and decreases delta and theta wave activity, an effect consistent with increased alertness.³ Modafinil has weak peripheral sympathomimetic activity and minimal effects on hemodynamics.⁴ Significant effects on EDS have occurred in controlled and open studies,^5-9 with the drug being well tolerated, with some exposures longer than 7 years.

Although more than 1,000 patients have been exposed to the drug, only a limited number have been studied in double-blind, controlled trials. A 400-mg daily dose was not previously studied under double-blind, placebo-controlled conditions, although in open trials patients have self-titrated their dose up to or beyond the 400-mg daily dose. The objective of the present study was to confirm the short-term efficacy of 200 mg of modafinil daily, and to study the comparative short-term efficacy and safety of 400 mg of modafinil daily in the treatment of EDS associated with narcolepsy.

Methods. Patient selection. All patients met the International Classification of Sleep Disorders (ICSD)¹⁰ diagnostic criteria for narcolepsy and had moderate or severe daytime sleepiness producing a moderate or marked impairment of social or occupational function. At least one Multiple Sleep Latency (MSLT)¹¹ had been done in all patients, with the exception of a small number who had definite cataplexy. Both new and previously diagnosed patients were included. Female patients of child-bearing potential had a negative serum pregnancy test and were using an effective method of birth control. Patients agreed to refrain from operating a motor vehicle and from being involved in any other potentially hazardous activity during the double-blind phases of the study. All patients signed an informed consent form and the protocol was approved by the local research ethics boards.

Patients first underwent a history, physical examination, and screening polysomnogram. Patients were excluded for any of the following reasons: chronic amphetamine treatment in the last 2 months; EDS related to sleep apnea, periodic limb movement disorder, or alcohol consumption; circadian rhythm disorder, insufficient sleep, or other known syndromes causing sleepiness; shift work; a history of head trauma; known hypersensitivity to modafinil; agitated states or severe anxiety; clinical depression; moderate or severe psychosis or dementia; dyskinesia; active peptic ulcers; hyperthyroidism; symptomatic cardiovascular disease; glaucoma; concurrent psychiatric, neurologic, neoplastic, endocrinologic, or infectious disease that might contribute to EDS; illicit drug use; the use of antipsychotic medication; or the use of any medication that might influence sleep (particularly REM sleep) or contribute to EDS, except for anticataplectic medications that were continued in constant dosage. Narcoleptic patients with coexistent, significant sleep apnea (apnea index>10 per hour or respiratory disturbance index >20 per hour) were also excluded from the trial.

Design. This study was a 6-week, three-period, randomized, crossover, placebo-controlled trial conducted in nine centers in Canada. Patients were evaluated at baseline and at the end of each 2-week period. There were no run-in or washout periods. Prescribed stimulants (if any) were continued until the day following the baseline testing. Patients were assigned randomly to one of the six possible treatment sequences in balanced blocks of six, stratified by center. Each patient received each of the following treatments during one of the three 2-week periods: placebo, modafinil 200 mg, or modafinil 400 mg in divided doses (morning and noon). Patients, study personnel, and the sponsor's monitor were all blind to the treatment being received. In order to minimize the chance of carryover effects, efficacy data were only collected for the second week of each 2-week period.

Efficacy variables. The primary efficacy variables were the mean sleep latency on the Maintenance of Wakefulness Test(MWT),¹² and the mean number of sleep episodes and periods of severe sleepiness reported in a detailed patient diary. The MWT is a modification of the MSLT in which subjects are tested in a comfortable chair with the instruction to try and stay awake, rather than the MSLT, in which subjects are tested lying down with the instruction to try and fall asleep. The MWT was selected rather than the MSLT because it measures the ability to stay awake rather than the rapidity of falling asleep, which is more relevant to what is desired clinically, and because it has been previously shown to be sensitive to treatment.^12,13 The MWT was conducted as first described by Mitler et al.¹² with the following modifications: The test consisted of four, not five, test sessions separated by 2 hours, and each session was terminated at minute 40 if no sleep had occurred, rather than at minute 20. Sleep latency was then scored as 40 minutes. The dependent measures include sleep latency (the interval from the start of the test session to the first 90-second continuous period of stage 1, or the first 30-second continuous period of any other sleep stage) as well as stage 1 latency (the interval to the first 30-second period of stage 1 sleep). During these test sessions the patients sat on identical La-Z-Boy (Monroe, MI) model 505 recliners provided to the centers. The instructions to stay awake given to the patients before lights out were standardized across centers. An effort was made to maintain room temperature within a narrow range and it was recorded at each nap.

Patients were asked to complete daily diaries throughout the second week of each 2-week period. Patients recorded a subjective evaluation of the previous night's sleep, an estimate of the number of hours of sleep, the time the study medication was taken, voluntary sleep episodes (naps and nighttime sleep), involuntary sleep attacks and episodes of severe sleepiness(sleepiness that was severe enough to prevent the carrying out of an activity), the hour of retiring and hour of rising, a subjective evaluation of the effect of sleepiness on the patient's chosen daily activity, and the number of attacks of cataplexy.

Patients completed the Epworth Sleepiness Scale (ESS)¹⁴ at each clinic visit before undergoing polysomnography (PSG). The ESS involves self-assessment of the probability of falling asleep under a range of specified situations ranging from very inactive to very active. The ESS has been validated and proved to be a reasonable predictor of objective sleepiness measure assessed by the MSLT. Every 2 hours between MWT naps a 10-minute four-choice reaction time test (FCRTT)¹⁵ was administered to assess daytime performance deficits.

At the end of each of the three treatment periods patients and physicians compared that period with the patient's prestudy condition, using the following six-level scale: marked effect (complete or nearly complete remission of EDS), moderate effect (partial remission of EDS), minimal effect(slight decrease in EDS that did not substantially alter the status of the patient), no change, minimally worse (slight increase in EDS), and much worse(substantially increase in EDS).

At the end of the trial (week 6) both physicians and patients were asked to rank the three periods in order of preference. If applicable, they were also asked to compare the preferred treatment period with any prestudy stimulant medication they may have been taking.

Tolerance and safety variables. The Profile of Mood States(POMS)¹⁶ was used to measure the effects of modafinil on mood over the second week of each 2-week period. On arrival, before PSG, patients were asked to evaluate their mood "during the past week." The questionnaire consists of a list of 65 words that describe feelings that the subjects ranked according to how they felt. The five-level scale ranges from"not at all" (0) to "extremely" (4). Six factors are derived: tension-anxiety, depression-dejection, anger-hostility, vigor-activity, fatigue-inertia, and confusion-bewilderment. In addition, a less well-validated factor, friendliness, and the total mood disturbance score were derived. The test has demonstrated reliability and validity, and norms have been published for a sample of 2,360 adults.¹⁶

The POMS was administered again before the morning dose (0730 hours to 0800 hours) and also 3 hours after the morning dose (1030 hours to 1100 hours). The patient was asked to evaluate "How you are feeling right NOW" to detect any acute mood effects of the stimulants. Tests were separated by 3 hours as this represents the time to maximal plasma concentrations after a dose of modafinil.

Overnight PSG preceded each MWT testing. It included recording of central and occipital referential EEG, eye movements, submental EMG, chest and abdominal movement, upper airway airflow, transcutaneous SaO₂, and anterior tibialis EMGs. Sleep was staged in standard fashion according to Rechtschaffen and Kales.¹⁷ The effects of modafinil on sleep initiation and maintenance, sleep architecture, periodic leg movements in sleep (PLMS), and on sleep-disordered breathing were evaluated. The subjective effects of modafinil on nocturnal sleep were also evaluated by the patient in the patient diary. Patients were also questioned at each visit regarding the appearance, worsening, or disappearance of any unusual signs or symptoms. Any complaint or symptom, volunteered or observed, was recorded as an adverse clinical event independently of whether or not the adverse event was thought to be associated with the study medication.

Study medication. Each patient took two tablets at breakfast and two at lunch. Tablets were supplied in blister strips containing 32 tablets each (i.e., 8 days of study medication). Patients were given a 2-week(+2 extra days) supply at baseline, week 2, and week 4, and were requested to return all unused medication for pill counts.

Statistical analysis. Because there was no washout period, the risk of carryover effects was minimized by using only data from the second week of each 2-week period. The ESS and POMS asked the patients to base their answers on the previous 7 days. Diary data were only recorded for the last 7 days of each 2-week period. PSG/MWT retest sessions were separated by 14 days. The comparative efficacy analysis was conducted on the results of the visits at the end of each of the three 2-week periods. The study was designed so that each treatment was preceded and followed equally as often by each other treatment: "digram balanced." Data were analyzed by ANOVA using the general linear model for a digram-balanced crossover design. Digram-balanced squares are superior to any other Latin squares in separating treatment from carryover effects.¹⁸ If ANOVA detected a significant difference, pairwise comparisons were made using least squares means.

Results. Seventy-five narcoleptic individuals were enrolled at nine different sleep disorder clinics. The range of numbers of patients enrolled per center was four to 14. All diagnoses were established in a sleep laboratory and by a sleep specialist meeting the recently published Canadian Sleep Society standards for PSG.¹⁹ The first patient was enrolled on January 25, 1995; the seventy-fifth patient was enrolled on December 8, 1995. Seventy-one patients completed the full 6 weeks. One patient dropped out after the baseline testing for personal reasons. The other three dropped out due to adverse events (vide infra). Mean (± SD) compliance based on the number of tablets returned at each visit was 98.5± 7.4% on placebo, 97.5 ± 10.4% on modafinil 200 mg, and 99.0± 7.8% on modafinil 400 mg. Two patients misunderstood instructions and did not start taking their medication until the second week of the first 2-week period.

Demographic data. Patients were mostly middle aged (43 ± 16 years) Caucasian (92%) females (62.7%). They had had EDS for 23 ± 16 years. The diagnosis of narcolepsy had been made a mean of 16 ± 13 years after the onset of EDS. Eighty percent of patients met the ICSD diagnostic criteria by virtue of having a clear history of cataplexy(although many no longer had attacks due to their anticataplectic medication). All patients met the PSG and MSLT criteria. The mean diagnostic MSLT latency was 3.5 ± 2.5 minutes with two or more sleep onset REM periods occurring on at least one testing. Concerning the anticataplexy medication continued throughout the study, 28 patients (37%) were being treated with one (or, in two patients, two) of eight different medications: amitriptyline (1), clomipramine (5), fluoxetine (5), imipramine(2), paroxetine (2), protriptyline (8), sertraline (1), and gamma hydroxybutyrate (5).

Efficacy. Sleep latency on MWT. Modafinil, 200 and 400 mg, significantly increased the mean MWT sleep latency compared with placebo by 40% (p = 0.0002) and 54% (p = 0.0001), respectively, with no significant difference between the two doses(table 1). No significant carryover effect (p= 0.593) was detected. Latency to the first period of stage 1 was also significantly increased by daily doses of modafinil 200 mg (p = 0.0043) and 400 mg (p = 0.0243). A greater number of tests was ended due to no sleep on both doses of modafinil than on placebo (p< 0.0005). This difference was most marked for the two afternoon naps. At each individual nap (0930, 1130, 1330, and 1530 hours) sleep latency on the 400-mg dose was clearly longer than on placebo (respectively, p = 0.030, p = 0.019, p = 0.0001, p = 0.0001;figure 1). The 200-mg dose was only different from placebo at the two naps following the noontime dose (p = 0.0762, p= 0.4934, p = 0.0001, p = 0.0004). At all time points from 0 to 40 minutes across the test sessions, a greater percentage of patients were still awake when on 400 mg of modafinil than when taking 200 mg, and more patients remained awake on 200 mg than on placebo (figure 2).

Subjective sleepiness. Modafinil, 200 and 400 mg, significantly reduced the self-assessed mean number of involuntary sleep episodes and severe somnolence (i.e., that sufficient to interfere with normal daytime activities) by 24% (p = 0.013) and by 26% (p = 0.007), respectively, with no difference between the two doses (see table 1). No significant carryover effect (p = 0.815) was detected. Modafinil did not appear to deter the patients' ability to nap when desired, as there was no difference in the number of voluntary naps among the three treatment regimens. Patients were asked to choose a single daily activity that was the most affected by their EDS. Both doses of modafinil lessened the effect of daytime somnolence on these daily activities (p = 0.019 and p = 0.051 for the 200-mg and 400-mg doses). The mean daily number of cataplectic attacks was low with no significant difference among the three treatments: placebo, 0.45 ± 1.70; modafinil 200 mg, 0.23 ± 0.66; and modafinil 400 mg, 0.16 ± 0.52.

Epworth Sleepiness Scale. Compared with placebo, modafinil 200 mg (p = 0.018) and modafinil 400 mg (p = 0.0009) both significantly decreased the likelihood of falling asleep as measured by the total ESS (table 2). In none of the score's eight situations did the mean score indicate that patients were less likely to fall asleep on placebo than on the 200-mg dose. In none of the eight situations did the mean score indicate that patients were less likely to fall asleep on the 200-mg dose than on the 400-mg dose. Patients reported being less likely to fall asleep "sitting reading" while taking 200 mg (p = 0.033) and 400 mg (p = 0.005) of modafinil compared with placebo. They also reported being less likely to fall asleep "watching TV" after 200 mg(p = 0.002) and 400 mg (p = 0.001) of modafinil. These first two items on the ESS were the most sensitive to change. They were also the two daily activities selected most often by patients in their diary as the activity most affected by their sleepiness. Compared to placebo, modafinil 200 mg (p = 0.027) made patients less likely to fall asleep "sitting inactive in a public place." The effect of modafinil 400 mg was marginal (p = 0.068) compared with placebo. As a "passenger in a car," patients were less likely to sleep while on modafinil 400 mg(p = 0.013) than placebo. The result on 200 mg was in between the placebo and modafinil 400 mg, but was not statistically different from either. No other individual item scores were significantly different among the groups.

Performance testing. On the FCRTT there were no differences in the mean daily reaction time or the number of gaps. Overall ANOVA detected a significant difference in the mean daily errors (p = 0.021); however, no individual paired comparison was significant. Compared with placebo there was a negligible 11% decrease in errors on modafinil 200 mg(p = 0.500), and a marginal 20% decrease on modafinil 400 mg(p = 0.074). Although scores on all three variables were in the expected rank order, the differences were small and the variations large. Reaction times, gaps, and errors also did not differ among the groups at 0930, 1130, and 1330 hours. However on the last session, at 1530 hours(approximately 3 hours after the second dose of modafinil, corresponding to its maximal plasma concentration), performance improved on both doses of modafinil, but not on placebo. At 1530 hours patients had 41% fewer gaps on modafinil 200 mg compared with placebo (p = 0.026), and 44% fewer gaps on modafinil 400 mg (p = 0.027).

Treatment preference. At the end of each treatment period patients compared that period with their prestudy condition (37 had been untreated, 29 had been taking methylphenidate, eight were on pemoline, and one was on mazindol). Among the patients taking no stimulant medication prior to the study, a decrease in EDS was reported by 80% of patients while on modafinil 400 mg, by 66% while on modafinil 200 mg, and by 34% on placebo. Among the patients who had been taking stimulants before the study, an improvement in EDS was reported by 53% of patients while on modafinil 400 mg, by 50% while on modafinil 200 mg, and by 25% on placebo. The percentage of patients reporting a deterioration on study medication was placebo, 58%; modafinil 200 mg, 25%; and modafinil 400 mg, 33%. The investigator evaluations were consistent with those of the patients.

At the conclusion of the crossover, patients were asked to state their first and second preference for one of the three treatment periods. In 84% of patients one of the modafinil doses was chosen as the best period. As a first choice, 50% of patients preferred the 400-mg period, 34% chose the 200-mg period, and 16% chose placebo. As a first or second choice, the 400-mg dose was selected by 80% of patients, the 200-mg dose by 84%, and placebo by 36% of patients.

Twenty-seven of the 29 patients who had been on methylphenidate before the study completed the crossover and compared their preferred treatment period with their previous therapy. Four of these patients selected placebo as the preferred treatment period. The preferences of the 23 remaining patients were marked for modafinil, five patients; moderate for modafinil, five patients; mild for modafinil, two patients; no preference, one patient; mild for methylphenidate, four patients; moderate for methylphenidate, three patients; and marked for methylphenidate, three patients. Of the eight patients who had been on pemoline prior to the study, seven chose one of the modafinil doses as their preferred treatment period and one chose the placebo period. None of seven patients preferred their previous treatment over modafinil. Their preferences were marked (three patients), moderate (one patient), mild (two patients), and none (one patient). The sole patient who had been on mazindol before the study had a marked preference for modafinil 400 mg.

Tolerance and safety. On the POMS modafinil did not increase anger-hostility or confusion-bewilderment compared with either placebo or normative scores. The 400-mg dose increased tension-anxiety compared with placebo (p = 0.028), but the scores were still well below both the baseline scores and the normative score for this factor. The vigor-activity and fatigue-inertia scores were not different among the groups, but the rankings were in the expected order of placebo, 200 mg, 400 mg, and normative data.

The predose morning POMS "NOW" factor scores after 2 weeks of modafinil were not different from placebo. In contrast to the "PAST WEEK" test given the previous night, patients on the 400-mg dose of modafinil did not report higher tension-anxiety scores. The acute effects of modafinil on mood were not found to be different from those of placebo. No acute effects of modafinil were detected on tension-anxiety, anger-hostility, or confusion-bewilderment. Vigor-activity increased after modafinil 200 mg(p = 0.018) and 400 mg (p = 0.012).

Consistent with the subjective diary reports, 2 weeks of treatment with either daily dose of modafinil had no effect on PSG measures of sleep initiation, sleep maintenance, sleep architecture, or on respiratory or movement measures in sleep (table 3). In addition, neither dose of modafinil was shown to have an adverse effect on subjective sleep quality assessed in the sleep log.

No significant drug effects were detected on blood pressure or heart rate in either normotensive or hypertensive patients. No serious unexpected adverse events occurred during the study. Three patients dropped out due to adverse effects, two during placebo (one rash, one palpitations), and one after 2 days on 400 mg modafinil (nausea, headache, anxiety).

Fifty-two subjective adverse events were reported a total of 226 times by 64 different patients. The number of adverse events decreased from the first to the last 2-week treatment period: 87 (39%) were reported during the first period, 77 (34%) during the second, and 62 (27%) during the third. This decrease was due to the 400-mg treatment for which 47% of adverse events were reported in the first period, 37% in the second, and 15% in the third, suggesting that some tolerance develops to the adverse effects of modafinil. The frequency by period was relatively constant for the two other groups. The distribution of adverse effects across treatments was in the expected rank order of placebo, 51 complaints (23%); modafinil 200 mg, 75 complaints(33%); and modafinil 400 mg, 100 complaints (44%). The nervous system was the most frequently affected (36.3%), followed by the body as a whole(32.7%), and the digestive system (17.7%). No other system accounted for more than 5% of the complaints. Many such complaints were transient in nature. For instance, headache lasted for less than 7 days with either dose of modafinil, as did nausea. Nervousness lasted less than 7 days in all but one patient, in whom it lasted 13 days (200-mg dose).

No significant differences were found between placebo and 200 mg of modafinil for any of the reported adverse events. The 400-mg dose caused more nausea (p = 0.039) and more nervousness (p = 0.007) than either of the other treatments. Headaches, dizziness, and dry mouth also appeared to be more frequent on 400 mg of modafinil, but the differences were not statistically significant (p = 0.187; p = 0.07; p = 0.163, respectively). Dyspepsia and vasodilatation(flushing) occurred more frequently on modafinil 200 mg than on placebo; but again, differences were not statistically significant and there was no dose relationship.

Discussion. EDS in patients with narcolepsy was significantly decreased by modafinil at both doses (200 and 400 mg per day) compared with placebo. The results showed a dose-effect relationship and were significantly different from placebo on four different measures of EDS: the objective MWT, the recording of self-assessed periods of involuntary sleep attacks plus marked somnolence or sleep in the patient diary, the subjective evaluation of the effects of somnolence on a selected daily activity, and the self-assessed ESS. This overall clinical and laboratory efficacy combined with low side effects confirms results of other studies on the effects of modafinil in narcolepsy,^3,5-9 almost all of which were not double-blind crossover trials.

The MWT is one of the most robust measures of daytime sleepiness, and it directly measures the ability of subjects to remain awake on request during the daytime. The lengthening of MWT sleep latency on modafinil was dose dependent. Others have shown a similar lengthening of sleep latency at the single dosage of 300 mg per day.^7,8 These changes appear comparable with those described for the MWT in patients on methylphenidate.¹³ Not only was sleep latency lengthened, but there was a marked increase in the number of test sessions during which patients did not fall asleep, a measure not often reported for this test. The effect of increased alertness was confirmed by the patient diaries, daily activity sheets, and ESS. However, daytime performance on the FCRTT showed only a trend in the expected direction, but was not significantly improved. This is in contrast to another modafinil study that measured this variable.⁶ In both studies, however, the variability of narcoleptics in this performance test was high, making it relatively insensitive.

At either dose modafinil did not show any effects on mood as assessed by the POMS. The only apparent possible change was for the subcategory of vigor-activity, which, in assessing the acute effects using the NOW version, was increased on modafinil at both doses. Reports of a similar significant effect exist for amphetamines^2,20 and methylphenidate,²⁰ with a trend in this direction reported elsewhere for modafinil.²

While decreasing diurnal somnolence, modafinil in our study did not have any adverse effects on nocturnal sleep either as assessed objectively by PSG or as reported by patients in their diary. Two weeks of treatment with modafinil had no effects on blood pressure and heart rate. Although the 400-mg dose was not significantly more effective than the 200-mg dose for its clinical response, it did cause more nausea and nervousness. A 400-mg daily dose may also increase the frequency of headache, dizziness, and dry mouth. Finally, patients generally showed a preference for modafinil over previous stimulant medication taken.

The study raises some methodological issues. The appropriateness of the chosen statistics in such a study is of relevance. We tested the homogeneity of the individual MWT test session data in respect to the statistical approach chosen. It was found that the first, second, and fourth MWT nap had homogeneous sleep latency figures. We reran the analyses excluding the"no-sleep" naps. The same significant results emerged and the same main results were found. It may also be questioned why the significant improvement of MWT sleep latency was not associated with a parallel result in performance measures and whether the latter might reflect the particular performance test selected. It should be pointed out that EDS is a complex state or spectrum of states that can involve changes in any or all of subjective assessments of sleepiness and mood, enhanced physiologic sleep pressure (in narcolepsy for both REM and non-REM sleep), impaired arousal, cognitive capacities, performance abilities, autonomic status (affecting pupillary control), and no doubt a number of other dimensions. Similar dissociations have been described between most of these dimensions of sleepiness including between sleep latency and performance measures. The lack of therapeutic response for performance relative to sleep latency may be because of a number of factors. These include the incomplete full response of sleepiness measured by MWT into the normal range, the incomplete sensitivity of performance tasks to physiologic sleepiness and, perhaps above all, the relatively high variability in performance-dependent measures in a population exhibiting increased moment-to-moment variability as well as increased overall levels of sleepiness/alertness.

In summary, as little as a 200-mg daily dose of modafinil is an effective and well-tolerated treatment of EDS in patients with narcolepsy. The long-term effectiveness is under evaluation. Our results and the current literature on modafinil would suggest a number of recommendations for its usage if it becomes available as a prescription drug. Patients who are previously untreated might reasonably be started on this medication for excessive sleepiness, as it is already known to have fewer side effects than other available stimulant medications (methylphenidate, pemoline, and amphetamines) combined with a lower frequency of development of tolerance in open trials. In our series of 75 patients a daily dose of 200 mg was as effective as 400 mg. However, individual response varies and the higher dose might be necessary in some patients. Those patients with narcolepsy who are already well controlled by other stimulants should remain on their medication unless bothersome side effects or tolerance develops, at which point a trial on modafinil would be reasonable.

Acknowledgments

The authors thank the technologists most involved from their centers. These include Irene Bowen (Toronto), Leanne Demeules (Calgary), Sandra Halko(London), Anik Gosselin (Montreal), Peter Hamilton (Vancouver), Kathy Lutley-Borland (Ottawa), Zoe Pouliot (Winnipeg), Terry Walker (Halifax), as well as research assistant Dominique Petit (Montreal).