Scientific investigations Idiopathic Hypersomnia: Clinical Features and Response to Treatment Mohsin Ali, M.B.B.S. 1 ; R. Robert Auger, M.D. 2,3 ; Nancy L. Slocumb 2 ; Timothy I. Morgenthaler, M.D. 2,4 1 Massachusetts General/Brigham & Women s Hospitals, Harvard Medical School, Boston, MA; 2 Center for Sleep Medicine, 3 Department of Psychiatry & Psychology, 4 Division of Pulmonary & Critical Care Medicine, Department of Internal Medicine, Mayo Clinic, Rochester, MN Objective: A recent American Academy of Sleep Medicine publication identified a need for research regarding idiopathic hypersomnia. We describe various clinical and polysomnographic features of patients with idiopathic hypersomnia, with an emphasis on response to pharmacotherapy. Methods: A retrospective review of our database initially identified 997 patients, utilizing idiopathic hypersomnia, hypersomnia NOS, and primary hypersomnia as keywords. The charts of eligible patients were examined in detail, and data were abstracted and analyzed. Response to treatment was graded utilizing an internally developed scale. Results: Eighty-five patients were ultimately identified (65% female). Median (interquartile range) ages of onset and diagnosis were 19.6 (15.5) and 33.7 (15.5), respectively. During a median follow-up duration of 2.4 (4.7) years, 65% of patients demonstrated a complete response to pharmacotherapy as assessed by the authors grading schema. Methylphenidate was most commonly used as a first-line agent prior to December 1998, but subsequently, modafinil became the most common first drug. At the last recorded follow-up visit, 92% of patients were on monotherapy, with greater representation of methylphenidate versus modafinil (51% vs. 32%). Among these patients, methylphenidate produced a higher percentage of complete or partial responses than modafinil, although statistical significance was not reached (38/40 [ 95%] vs 22/25 [88%], respectively, p = 0.291). Conclusions: The majority of patients with idiopathic hypersomnia respond well to treatment. Methylphenidate is chosen more often than modafinil as final monotherapy in the treatment of idiopathic hypersomnia, despite the fact that it is less commonly used initially. Further prospective comparisons of medications should be explored. Keywords: Idiopathic hypersomnia, treatment, modafinil, methylphenidate, stimulants Citation: Ali M; Auger RR; Slocumb NL; Morgenthaler TI. Idiopathic hypersomnia: clinical features and response to treatment. J Clin Sleep Med 2009;5(6):562-568. Idiopathic hypersomnia (IH) is a primary disorder of hypersomnolence that is distinguished from narcolepsy by virtue of the absence of electrophysiologic and other features of rapideye-movement (REM) sleep disturbance. Further subcategorization of IH is described within the International Classification of Sleep Disorders Second Edition (ICSD-II), which separates the condition into 2 forms based on whether or not long sleep time is described. Afflicted individuals with this descriptor classically exhibit prolonged unrefreshing sleep bouts and may also describe profound sleep inertia, often referred to as sleep drunkenness (see Table 1). 1 Due in part to the recent publication of the second edition of this manual (2005), the existing IH knowledge base employs discrepant diagnostic criteria. 2-8 Of the 7 reports in the literature to date, none utilizes a diagnostic schema that is entirely consistent with that listed in the ICSD-II, although only 1 study 3 was published subsequent to its inception. 2-8 Inclusion criteria for the Submitted for publication March, 2009 Submitted in final revised form August, 2009 Accepted for publication August, 2009 Address correspondence to: Timothy I. Morgenthaler, M.D., Mayo Center for Sleep Medicine, Gonda 17W, 200 First Street SW, Rochester, MN 55905; Tel: (507) 266-7456; Fax: (507) 266-7772; E-mail: tmorgenthaler@ mayo.edu Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 562 majority of these studies (with the exception of that authored by Bassetti and colleagues 5 ) either do not mandate threshold Multiple Sleep Latency Test (MSLT) values 2-4,6,8 (consistent with the first edition ICSD description) 9 or require MSLT thresholds that are either inconsistent with current requirements 5 or indeterminate in nature. 7 As can be seen in Table 1, under the current criteria, such data are compulsory to assign a diagnosis of IH without long sleep. Since prolonged nocturnal polysomnography is typically not performed in clinical practice, such values are also arguably requisite to ascribe a diagnosis of IH with long sleep. Although the debate regarding the requirement of this criterion is meritorious, 3 the variations and/or lack of diagnostic criteria in the accumulated literature raise the possibility that studies included heterogenous patient populations, a prospect deemed further probable by virtue of the fact that IH lacks the hallmark features seen in other primary hypersomnias, such as narcolepsy with cataplexy or Kleine-Levin syndrome. In the largest series to date, Anderson and colleagues recently reported on the clinical and polysomnographic characteristics of 77 patients with IH compared to patients with narcolepsy. 3 Inclusion criteria were described as consistent with those listed in the ICSD-II, but MSLT and nocturnal polysomnography requirements were omitted. Notwithstanding these limitations, significant contributions to the sparse IH treatment literature were provided. The study reported responses to pharmacother-
Table 1 International Classification of Sleep Disorders-2 Criteria for the Diagnosis of Idiopathic Hypersomnia Idiopathic Hypersomnia: Response to Treatment IH with long sleep A. The patient has a complaint of excessive daytime sleepiness occurring almost daily for at least three months. B. The patient has prolonged nocturnal sleep time (more than 10 h) documented by interview, actigraphy, or sleep logs. Waking up in the morning or at the end of naps is almost always laborious. C. Nocturnal polysomnography has excluded other causes of daytime sleepiness. D. The polysomnogram demonstrates a short sleep latency and a major sleep period that is prolonged to more than 10 h in duration. E. If an MSLT is performed following overnight polysomnography, a mean sleep latency of less than 8 min is found and fewer than two SOREMPs are recorded. Mean sleep latency in IH with long sleep time has been shown to be 6.2 ± 3.0 min. F. The hypersomnia is not better explained by another sleep disorder, medical or neurological disorder, mental disorder, medication use, or substance use disorder. IH without long sleep The patient has a complaint of excessive daytime sleepiness occurring almost daily for at least three months. The patient has normal nocturnal sleep (greater than 6 h but less than 10 h), documented by interviews, actigraphy, or sleep logs. Nocturnal polysomnography has excluded other causes of daytime sleepiness. Polysomnography demonstrates a major sleep period that is normal in duration (greater than 6 h but less than 10 h). An MSLT following overnight polysomnography demonstrates a mean sleep latency of less than 8 min and fewer than two SOREMPs. Mean sleep latency in IH has been shown to be 6.2 ± 3.0 min. The hypersomnia is not better explained by another sleep disorder, medical or neurological disorder, mental disorder, medication use, or substance use disorder. Abbreviations: IH, idiopathic hypersomnia; MSLT, multiple sleep latency test; SOREMPs, sleep-onset rapid eye movement periods. 1 apy in 61 patients over a mean follow-up of 3.8 years. Eightynine percent of participants (54/61) commenced treatment with modafinil. Thirty-nine remained on monotherapy (72%, 39/54), and approximately 60% (24/39) demonstrated a sustained improvement in daytime sleepiness (assessed with Epworth Sleepiness Scale [ESS] scores), although 75% (18/24) required higher than recommended doses (i.e., > 400 mg). The remaining patients switched to dexamphetamine because of modafinil side effects (15%, 8/54) or added either caffeine or dexamphetamine due to lack of efficacy with modafinil alone (11%, 6/54). One patient remained on modafinil despite lack of efficacy. Among those who commenced treatment with a medication other than modafinil, 3% (2/61) received dexamphetamine, and 3% (2/61) used either pemoline or methylphenidate. Although their smaller numbers limit comparative analyses, the response rate for those using drugs other than modafinil and/or combinations of drugs was similar to those using modafinil alone, at 57%. Three patients (5%, 3/61) from this study abstained from pharmacotherapy entirely. In an effort to further expand the literature regarding IH, we studied the clinical and polysomnographic characteristics of a larger number of patients, adhering more stringently to ICSD-II and other criteria. Because the paucity of treatment studies in particular has recently been highlighted, 10,11 subjective response to pharmacotherapy was examined in detail. Our higher number of IH patients treated with agents other than modafinil (specifically methylphenidate) afforded an opportunity for comparison that has not been previously realized. METHODS A computerized search for patients with IH seen at the Mayo Center for Sleep Medicine (encompassing 1996-2007) was performed utilizing idiopathic hypersomnia, hypersomnia NOS, and primary hypersomnia as keywords. The initial search identified 997 candidates for whom clinical and polysomnographic data were reviewed. Those who did not meet IC- SD-II diagnostic criteria (Table 1) were subsequently excluded. All subjects met specified MSLT criteria, with tests performed on the day after polysomnography and scored according to standard clinical guidelines. 12 Specific polysomnographic disqualifiers included an apnea hypopnea index of 5 or greater, a respiratory-effort-related arousal index of 10 or greater, or a periodic limb movement index of 15 or greater. Additional exclusion criteria included failure to attend a follow-up visit after initiation of pharmacotherapy, endorsement of cataplexy (past or present), and/or diagnoses of dementia, traumatic brain injury, stroke, extrapyramidal syndromes, multiple sclerosis, chronic pain, hypothyroidism, substance abuse, and hepatic or renal failure. Those on sedatives or narcotic analgesics were also excluded from further analyses, as were patients who were unable to discontinue antidepressant medications within the specified washout period prior to MSLT. Patients with depression were included only if their illness was judged to be in remission. This was ascertained solely on the basis of the treating psychiatrist s and/or board-certified sleep specialist s (if he/she were also a psychiatrist) consent to proceed with temporary antidepressant discontinuation. Eighty-five patients within the database fulfilled the inclusion criteria. Detailed clinical notes were available for all patients and were reviewed. Subjective treatment response was assessed predominantly by reviewing the language used by physicians or patients to report progress during the follow-up visits. Responses were divided into 3 groups: (1) complete response, which correlated with adjectives such as excellent, great, or entirely satisfactory, provided there was no change in pharmacotherapy; (2) partial response, which correlated with descriptions such as doing better or improved, but in the setting of increased dose adjustment of stimulant medications; and (3) poor response, with such associated descriptors as still sleepy and with a subsequent switch to another medication. The validity of our outcome measurement scale was assessed by evaluating interrater agreement of scoring by 2 blinded independent raters (RA, TM). We found initial agreement to be high (Cronbach s Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 563
M Ali, RR Auger, NL Slocumb et al Table 2 Baseline Characteristics of Patients with Idiopathic Hypersomnia Characteristics No. or median (IQR) Subjects, no. 85 Sex ratio (female/male) 55/30 Age of symptom onset, y 19.6 (15.5) Age at diagnosis, y 33.7 (15.5) Hours of sleep reported a 8.0 (1.4) Actigraphic sleep time b 7.5 (1.0) BMI, kg/m b,c 25.4 (6.2) Initial ESS d 17.0 (7.0) IQR, interquartile range. a Only 84 patients had reported sleep time recorded. b Only 61/85 had actigraphy. c Only75/85 had initial body mass index (BMI) recorded. d Only 51/85 had initial Epworth Sleepiness Scale (ESS) data available. α = 0.8021). After independent scoring, identified differences in actual scores were discussed among the 3 physicians (MA, RA, TM) until agreement was reached, and the resultant scores were used for analyses (Table 4). Forty-nine patients (58%, 49/85) had ESS scores recorded at both initial and subsequent visits. Changes in these scores were correlated with the remainder of the subjective treatment responses when possible. Statistical Analysis Continuous variables are summarized as mean ± SD, or as median (interquartile range). Comparative analyses were performed using Kruskall-Wallis rank sum or t tests for continuous variables, and Fisher exact or χ 2 test for categorical variables as appropriate. We considered p < 0.05 statistically significant. RESULTS Clinical Characteristics of Patients with IH Eighty-five patients (65% female) were ultimately identified, with a median (interquartile range) age at onset of 19.6 (15.5) years and median age at diagnosis of 33.7 (15.5) years, without significant sex differences. Median body mass index was 25.4 kg/m 2 (6.2). Median initial ESS score was 17.0 (7.0). See Table 2. Difference: Actigraphic Sleep-Subjective Sleep 6 4 2 0-2 -4 Mean Difference = -0.99 (0.18) p<0.001 HrsSlp/Night -6 6 7 8 9 10 11 12 Mean: [(Actigraphic Sleep + Subjective Sleep)/2] Figure 1 Difference between patients reported sleep time and actigraphically recorded sleep time. At the time of initial consult, average sleep time was obtained by review of the sleep schedule with the patient. Prior to polysomnography, actigraphy was obtained in 61 (72%) of our patients. Above, we compare the subjective sleep duration with the subsequent actigraphically estimated sleep time, using the Bland-Altman method. The vertical red line represents 7.94 hours, the average of the mean between all paired subjective and actigraphic hours of sleep. The horizontal red line represents the mean difference between actigraphically measured and reported sleep, and the dotted red lines represent the SEM (-0.99 ± 0.18 hours [p < 0.001]). The data show that overreporting of sleep time increased in conjunction with increased estimated sleep time. Although 19% (16/85) of patients subjectively reported more than 10 hours of daily sleep on average, paired sample analysis (using the method of Bland-Altman) demonstrated subjective overestimate of this value during initial consultation (compared with actigraphic measurements) by 0.99 ± 0.17 hours, with a tendency to show greater differences as total sleep time increased (Figure 1). Both this observation and the fact that none of the patients had extended polysomnography (see Table 1, criterion D of IH with long sleep) led to the decision to analyze subjects as a complete group (i.e., without subcategorization). Table 4 Response to Treatment Grade Response Required Criteria Patients, Comments from visit notes Treatment No (%) alteration 1 Complete Response Great, excellent, entirely satisfactory, very well indeed 2 Partial Response Reasonably well controlled, doing better, better overall 3 Poor Response Still sleepy, residual sleepiness, has not done well Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 564 Median (IQR) change in ESS (p = 0.003 a ) None 55 (65) -9 (4) Dose increase 22 (26) -6 (9) Medication changed 8 (9) +2 (9) Abbreviations: IQR, interquartile range; ESS, Epworth Sleepiness Scale. a Kruskal-Wallace test indicates differences across 3 groups. Pairwise comparison using Wilcoxon rank sum test between Complete and Partial Response, p = 0.009; between Partial and Poor Responses, p = 0.219
Idiopathic Hypersomnia: Response to Treatment Limiting Side Effects 100% 90% 80% Proportion of Complaints 70% 60% 50% 40% 30% 20% 10% 0% Amphetamine Dextro Dextroamphetamine Methamphetamine Methylphenidate Modafanil Pemoline Drugs Anxiety Diarrhea Dizzy GI other Headache High cost Hypertension Incr. Tension Insomnia Nervous Palpitations Pregnancy Restlessness Weight Loss Nausea Figure 2 Limiting side effects. Side effects or complaints that limited the ability to increase medication dose or that required discontinuation of the medication are shown above. Notably, the high cost of medication was a common limiting complaint for modafinil and methamphetamine. Hypertension developed in some patients taking methamphetamine, modafinil, and pemoline; causality is not established, however, as the development of hypertension may have been coincidental. Nonetheless, the treating physicians did not feel comfortable continuing the medication under these circumstances. Family history was available in 49/85 patients (58%), 18 (37%) of whom reported at least 1 family member with hypersomnolence. Information regarding naps was also obtained in 58% of patients, with 36 (73%) responding affirmatively, and experienced as nonrestorative in 72%. The mean nap duration was 1.46 ± 0.86 hours. Among patients for whom there was specific commentary regarding sleep drunkenness, 12 of 18 (66%) responded affirmatively. Information about sleep paralysis was available in 69 of 85 (81%) patients, 7 (10%) of whom responded affirmatively. The presence of hypnagogic or hypnopompic phenomena was ascertained in 78% of patients, only 4% of whom responded affirmatively. Nocturnal Polysomnography and MSLT Polysomnographic findings are summarized in Table 3. For nocturnal assessments, the median of the mean initial sleep latency was similar for males and females ( 6.8 [7.8] vs 7.0 [11.4] minutes, [p = 0.555]), but on MSLT, the mean sleep latency for males was lower than that for females (2.7 [ 3.5] and 4.0 [3.7] minutes, respectively [p = 0.0215]). Thirteen patients (15%) had 1 sleep-onset REM period (SOREMP) on either Table 3 Polysomnographic Features of Patients with Idiopathic Hypersomnia Parameters Measures Initial sleep latency, min 7.0 (8.5) Sleep efficiency, % 89.9 (8.0) Initial REM latency, min 82.0 (55.3) Sleep stage, % of TST 1 6.0 (4.1) 2 55.8 (10.4) 3 15.3 (7.8) REM 22.0 (7.9) TST, min 454.0 (62.2) AHI, no/h 1.0 (2.0) RAI, no/h 1.4 (2.6) PLMI, no/h 0.8 (3.7) SL, min b 4.8 (2.5) Data are reported as median (interquartile range) Abbreviations: REM, rapid eye movement sleep; TST, total sleep time; AHI, apnea-hypopnea index; RAI, respiratory arousals index; PLMI, periodic limb movement index. a Based on polysomnography. b Mean initial sleep latency (SL) on the multiple sleep latency test (MSLT). Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 565
M Ali, RR Auger, NL Slocumb et al Table 5 Mean Daily Doses of Medications Used Medication Mean total daily dose a Modafinil 367.4 ± 140.9 Methylphenidate 50.9 ± 27.3 Dextroamphetamine 35.7 ± 44.4 Methamphetamine 36 ± 17.1 Amphetamine/dextroamphetamine 79.3 ± 30.6 Pemoline 66.9 ± 36.6 Sodium oxybate a 4.5 Caffeine 200 a Total daily doses are shown in milligrams, except sodium oxybate, which is shown in grams. We report various clinical, polysomnographic, and treatment data on the largest and most homogenous IH cohort to date. We applied ICSD-II criteria more stringently than did a recent study utilizing this classification schema, 3 specifically requiring adherent MSLT values. We also mandated a polysomnography respiratory-effort-related-arousal index of less than 10 as an inclusion criterion and rigorously addressed the confounding influence of insufficient sleep by use of actigraphy and/or sleep logs, obtained for a minimum of 1 week prior to polysomnography, and available in 72% and 99% of participants, respectively. Our entry criteria were otherwise similar to those employed by Anderson and colleagues. 3 When comparing our group s demographics with those from other studies with relatively rigorous inclusion and exclusion criteria, 3,5 we found a strikingly similar age of onset and diagnosis, occurring during the late teens and mid-30s, respectively. Similar to the findings of Bassetti and colleagues, our cohort included a disproportionate number of females, which was not found in the more recent study by Anderson and associates. 3,5 Among those for whom a family history was documented, 37% reported a family member with excessive daytime sleepiness, which is similar to what was reported by Anderson and Bassetti (34% and 38%, respectively). 3,5 The occurrence of daytime naps (73%) with a nonrestorative quality (72%) is in the range of what has been reported previously (60%-99% and 77%-78%, respectively). 3,5 In the Anderson study, with the exception of cataplexy, the single most useful factor in the clinical history that distinguished IH from narcolepsy was nap duration of longer than 60 minutes, which had 87% sensitivity and specificity in their cohort. In our study, mean nap duration was nearly 1.5 hours. The occurrence of vivid dreams (14%), sleep paralysis (10%), sleep drunkenness (66%), and hypnagogic or hypnopompic hallucinations (4%) in our cohort add to the wide range of values reported by the 2 previous studies at 25% to 40%, 4% to 40%, 21% to 52%, and 5% to 43%, respectively. 3, 5 Whether the frequency of these phenomena varies between IH subtypes remains to be elucidated. The degree of objective sleepiness seen in our cohort was comparable with that described by Bassetti and colleagues (MSLT mean sleep latency 4.3 ± 2.1 minutes) 5 but was more severe than the values described by Anderson and colleagues (MSLT mean sleep latency 8.3 ± 3.1 minutes), 3 presumably due in large part to the fact that the latter study did not require MSLT values within its inclusion criteria. The mean sleep latency for male patients was lower than that of females within our cohort, which warrants further investigation. Thirteen of 85 patients (15%) had 1 SOREMP during either polysomnography or MSLT, which is a higher frequency than that reported by both the Anderson and Bassetti studies (3% and 4%, respectively), although it is unclear whether Anderson and colleagues asovernight polysomnography or MSLT, but none had 2 or more SOREMPs. Treatment and Outcome The median duration of follow-up was 2.4 (4.7) years, and the median number of patient visits was 6 (3). Fifty-five (65%) patients reported a complete response to pharmacotherapy, 22 (26%) reported a partial response, and 8 (9%) reported a poor response (Table 4). The change in ESS score was statistically different across response groups, with a change of -9 (4) in complete responders, -6 (9) in partial responders, and +2 (9) in nonresponders (p = 0.003; see Table 4). Five patients were dissatisfied with their treatment response, despite the fact that their ESS scores decreased by at least 6 points. Five patients described subjective improvements in daytime alertness despite an increase (2/5) or a lack of change (3/5) in ESS scores and therefore did not proceed with a change in treatment. The mean total daily doses of prescribed medications are shown in Table 5, and limiting side effects are depicted in Figure 2. Forty-nine patients (58%) tried only 1 medication, and 36 (42%) tried at least 2 medications. The total number of patients tried on any given medication, the number of patients receiving a particular treatment at the last recorded visit, and the associated treatment responses are shown in Table 6. Methylphenidate was most commonly used as a first-line agent prior to December 1998 but was replaced by modafinil subsequent to that date. Modafinil was prescribed to 50 patients (59%) overall, but of those receiving monotherapy during the last recorded follow-up visit (92%, 78/85), only 25 (32%) remained on modafinil. Eighteen of these individuals (72%) reported complete symptomatic relief, 4 (16%) reported partial symptomatic relief, and 3 (12%) reported no benefit. Methylphenidate was prescribed in 61 of the 85 patients (72%) overall and comprised 51% (40/78) of monotherapy regimens during the last visit. Twenty-five of 40 patients (63%) reported a complete response, 13 of 40 (33%) reported a partial response, and 2 of 40 (5%) reported a poor response. Of those on modafinil or methylphenidate monotherapy at last visit, methylphenidate produced a higher percentage of complete or partial response than did modafinil, although statistical significance was not reached ( 38/40 [95%] vs 22/25 [88%], respectively [p = 0.291]). Modafinil monotherapy was most often provided as a single morning dosage, whereas methylphenidate monotherapy was typically provided in divided dosages (3-4 times daily) of regular-release or as a combination of sustained-release and regular-release formulations. The remainder of monotherapy regimens (17%, 13/76) consisted of various amphetamine preparations, pemoline, and caffeine, as listed in Table 6. Eight percent (7/85) were receiving combination therapies at the last recorded visit. Responses to these less commonly prescribed regimens can be viewed in the accompanying table (Table 6). DISCUSSION Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 566
Table 6 Treatment Response According to Drug Used Idiopathic Hypersomnia: Response to Treatment Medication Exposed to drug at any time, No. (% total) Remaining on drug at last visit, No. (% total exposed) Patients Complete response, No. (% of total at last visit) Partial response, No. (% of total at last visit) Poor response, No. (% of total at last visit) Modafinil 50/85 (59) 25/50 (50) 18/50 (36) 4/50 (8) 3/50 (6) Methylphenidate 61 (72) 40 (66) 25 (41) 13 (21) 2 (3) Dextroamphetamine 7 (8) 2 (29) 0 (0) 0 (0) 2 (29) Methamphetamine 5 (3) 3 (60) 3 (60) 0 (0) 0 (0) sessed SOREMPs during both nocturnal polysomnography and MSLT, as the Bassetti group affirmed. 3,5 It remains unknown whether these values differ between IH subtypes. Our treatment data further contribute to the scarce knowledge base regarding patients with IH. In contrast with the study by Anderson and colleagues, we did not rely primarily on the ESS to gauge treatment response. As described by their group, interpretive limitations are incurred with this approach, as many patients endorsed subjective improvement without concomitant drops in ESS scores. 3 Taking these discrepancies into account, we assessed the subjective response to pharmacotherapy utilizing detailed chart reviews (including ΔESS if scores were available) as described above, which also took into account whether an adjustment in the medication regimen transpired. This report therefore represents the most detailed treatment analysis of patients with IH to date, with duration of follow-up (median 2.4 years) similar to that of the Anderson study and with a median of 6 patient visits. When including only those who demonstrated a complete response (utilizing any medication regimen), our success rate (65%) appears comparable with that described by Anderson and colleagues. In contrast with their study, the majority of our patients were treated with approved and recommended doses, although a select few required higher doses. These cumulative results appear to detract from the historical notion that IH is relatively refractory to treatment, 1 although it is not definitively known how these response rates compare with those reported for other primary disorders of hypersomnolence or whether treatment response varies according to IH subtype. Interest- Amphetaminedextroamphetamine 8 (9) 4 (50) 2 (25) 2 (25) 0 (0) Pemoline 7 (8) 3 (43) 3 (43) 0 (0) 0 (0) Caffeine 1 (1) 1 (100) 1 (100) 0 (0) 0 (0) Modafinil/ methylphenidate Modafinil/ dextroamphetamine Methylphenidate/ dextroamphetamine Methylphenidate/ amphetamine- Dextroamphetamine Methylphenidate/ sodium oxybate 3 (4) 3 (100) 2 (67) 1 (33) 0 (0) 1 (1) 1 (100) 0 (0) 1 (100) 0 (0) 1 (1) 1 (100) 0 (0) 1 (100) 0 (0) 2 (2) 2 (100) 1 (50) 1 (50) 0 (0) 1 (1) 1 (100) 0 (0) 0 (0) 1 (100) ingly, our study also demonstrated inconsistent responses with respect to subjective treatment benefit and reported ESS. Five patients (6%) remained dissatisfied with their treatment response despite the fact that their ESS scores decreased by at least 6 points. An additional 5 patients described improvements in daytime alertness despite an increase (n = 2) or a lack of change (n = 3) in ESS scores. Our large number of patients treated with agents other than modafinil (particularly methylphenidate) afforded an opportunity for comparative analyses, which has not been accomplished previously. According to our data, methylphenidate monotherapy may be more frequently effective than modafinil as monotherapy, although statistical significance was not established. Further supporting this notion, however, modafinil was prescribed to 59% of patients overall (and became the first-line agent in most patients after its arrival in the marketplace) but represented only 32% of those receiving monotherapy. In contrast, methylphenidate was prescribed to 72% overall and comprised 51% of monotherapy regimens during the last recorded visit. Because the high cost of modafinil was a commonly cited adverse effect (Figure 2), it is not possible to extricate the relative contributions of medication impotency and economic burden from this observation. Although 19% of patients were initially subcategorized into a diagnosis of IH with long sleep by clinicians based upon subjective reports of daily sleep time, closer scrutiny of actigraphy and sleep-log data revealed that these classifications were inconsistent with the current ICSD-II diagnostic criteria. In fact, our data often showed a discrepancy between the IH patients Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 567
M Ali, RR Auger, NL Slocumb et al perceived total sleep time and the actigraphically determined sleep time, with overestimation of the subjective compared with the objective report (Figure 1). Similar findings have been reported previously. 13 Even if subjects had met this criterion for consistently extended total sleep time, we likely would have infrequently identified patients who met full diagnostic criteria for IH with long sleep using current criteria, as our laboratory does not typically prolong pre-mslt polysomnography in an effort to identify those so afflicted. Standardized adherence to both actigraphic and polysomnographic criteria may therefore result in more accurate classification of IH subtype. It remains unknown whether this would create different profiles in terms of clinical and polysomnographic characteristics and, on a related note, whether a more careful distinction is required by clinicians. There are various limitations to our study. First, due to its retrospective nature, we had limited information on clinical symptoms with the exception of excessive daytime sleepiness. Although open to criticism, our chart-based grading schema was essential to document treatment response, given previously described inadequacies of the ESS and lack of availability of this scale in many of our patients. Clearly, a prospective trial comparing treatment regimens using validated measures for sleepiness, alertness, and quality of life is still needed. Differentiation of clinical characteristics and treatment response according to IH subcategorization (and possibly gender) would also be desirable. ACKNOWLEDGMENTS The authors would like to acknowledge Sarah M. Bonnema and Lori L. Solmonson for their assistance in the preparation of this manuscript. REFERENCES 1. International Classification of Sleep Disorders. Diagnostic and Coding Manual, 2nd ed Westchester, IL: American Academy of Sleep Medicine; 2005. 2. Aldrich M. The clinical spectrum of narcolepsy and idiopathic hypersomnia. Neurology 1996;46:393-401. 3. Anderson KN, Pilsworth S, Sharples LD, Smith IE, Shneerson JM. Idiopathic hypersomnia: a study of 77 cases. Sleep 2007;30:1274-81. 4. Baker TL, Guilleminault C, Nino-Murcia G, Dement WC. Comparative polysomnographic study of narcolepsy and idiopathic central nervous system hypersomnia. Sleep 1986;9:232-42. 5. Bassetti C, Aldrich M. Idiopathic hypersomnia: A series of 42 patients. Brain 1997;120:1423-35. 6. Billiard M, Merle C, Carlander B, Ondze B, Alvarez D, Besset A. Idiopathic hypersomnia. Psychiatry Clin Neurosci 1998;52:125-9. 7. Komada Y, Inoue Y, Mukai J, Shirakawa S, Takahashi K, Honda Y. Difference in the characteristics of subjective and objective sleepiness between narcolepsy and essential hypersomnia. Psychiatry Clin Neurosci 2005;59:194-9. 8. van den Hoed J, Kraemer H, Guilleminault C, et al. Disorders of excessive daytime somnolence: polygraphic and clinical data for 100 patients. Sleep 1981;4:23-37. 9. International Classification of Sleep Disorders, Revised: Diagnostic and Coding Manual. Westchester, IL: American Academy of Sleep Medicine; 2000. 10. Wise MS, Arand DL, Auger RR, Brooks SN, Watson NF. Treatment of narcolepsy and other hypersomnias of central origin. Sleep 2007;30:1712-27. 11. Morgenthaler TI, Kapur VK, Brown T, et al. Practice parameters for the treatment of narcolepsy and other hypersomnias of central origin. Sleep 2007;30:1705-11. 12. Carskadon M, Dement WC, Mitler M, Roth T. Guidelines for the Multiple Sleep Latency Test (MSLT): a standard measure of sleepiness. Sleep 1986;9:19-24. 13. Bradshaw DA, Yanagi MA, Pak ES, Peery TS, Ruff GA. Nightly sleep duration in the 2-week period preceding multiple sleep latency testing. J Clin Sleep Med 2007;3:613-9. Journal of Clinical Sleep Medicine, Vol.5, No. 6, 2009 568