HLA Haplotypes, Polysomnography, and Pedigrees in a Case Series of Patients With Narcolepsy

Summary

An ongoing study of the genetics of narcolepsy ascertains families through a case series of narcoleptic probands using diagnostic criteria consisting of 1) clinical history of excessive somnolence, 2) a mean sleep latency on the multiple sleep latency test (MSLT) of less than 7.9 minutes, 3) the rapid eye movement (REM) sleep-related symptom of cataplexy, 4) nocturnal polysomnography ruling out sleep apnea syndrome, and 5) two or more transitions to REM sleep on the MSLT. All probands and first-degree relatives received clinical and laboratory evaluations as well as human leukocyte antigen (HLA) typing. Demographic characteristics of the 32 probands are as follows: 17 males and 15 females; mean age was 42.1 years (range 13–70 years). The polysomnographic data confirmed daytime sleepiness and increased tendency for REM sleep for the 32 probands. Nocturnal polysomnographic results are as follows: sleep latency, 3.2 minutes; total sleep time, 442 minutes. MSLT results are as follows: sleep latency, 3.1 minutes; REM latency, 6.9 minutes; number of REM periods, 3.2. HLA typing revealed the presence of the HLA haplotypes, DRB1*15 and DQB1*0602, in 21 narcoleptic probands, with two African-Americans having the DQB1*0602 but not the DRB1*15 allele. Among the 57 relatives of the 32 probands, 1/31 females and 7/26 males were found to be affected with narcolepsy (p < 0.02), which suggests a higher diagnostic rate in male relatives. The 21 probands who were positive for the DRB1*15 and DQB1*0602 haplotypes did not differ from the 10 probands who were negative for these alleles in terms of their nocturnal sleep parameters, MSLT findings, or clinical presentation. Three families with multiple individuals affected with narcolepsy are presented. Two families have more than one affected individual who does not have the high-risk HLA haplotype. In one of these families, the disease is segregating independently of any HLA haplotype. In the third family, there is cosegregation with HLA DRB1*15 and DQB1*0602. One family contains a pair of DNA-confirmed, monozygotic twins with narcolepsy who are discordant for cataplexy and have the HLA DR14(Dw9)/DQB1*0503 and DR4(Dw4)/DQB1*0302 haplotypes.

Narcolepsy is a central nervous system disorder that appears to be inherited in some cases. Its estimated prevalence is about 0.03–0.06% in western Europe and North America (1). Honda and colleagues were the first to report that the disorder is statistically associated with subtypes of human leukocyte antigen (HLA) phenotypes, then known as HLA-DR2 (2). This strong association has been confirmed and found to extend also to the HLA-DQ region by several other groups (3–7). Since these observations, the nomenclature for HLA typing has been periodically extended and refined. The current nomenclature (8,9) is more specific for subtypes in the HLA region and now refers to HLA-DRB1*15 and HLA-DQB1*0602 subtypes as the most likely haplotypes for narcoleptics (10). Efforts to further specify and refine HLA markers for narcolepsy continue. Such studies indicate that human narcolepsy has a genetic basis, although the etiology remains elusive and narcoleptics who do not have these HLA markers have been reported (5,7,11).

We are engaged in a long-term genetics study of narcolepsy using a case series model of ascertainment. Our study involves determining the HLA subtype of confirmed narcolepsy probands and their first-degree relatives. We now report on the sleep and HLA characteristics of the first 32 narcoleptic patients who serve as probands. These probands appear to have unambiguous narcolepsy, but only 21 of 31 probands tested (68%) have both the HLA-DRB1*15 and HLA-DQB1*0602 markers for narcolepsy.

METHODS

To be included in our case series, all probands had to be unsolicited patients who presented to the Scripps Clinic Sleep Disorders Center seeking evaluation and/or treatment. Our proband series was defined so as not to be biased for parameters such as responsiveness to public media, presence of cataplexy, existence of similarly affected relatives, or presence of a particular HLA haplotype. We began these studies in 1985 and used diagnostic criteria that were based on the current practices of the time and the criteria of the Association of Sleep Disorders Centers [ASDC (12)], which consisted of the following: 1) clinical history of excessive somnolence, 2) a mean sleep latency on our diagnostic four-nap multiple sleep latency test (MSLT) of less than 7.9 minutes, 3) cataplexy, 4) nocturnal polysomnography ruling out sleep-related respiratory abnormalities such as apnea and alveolar hypoventilation, and 5) two or more transitions to rapid eye movement (REM) sleep on the MSLT. Our mean sleep latency criterion of less than 7.9 minutes was based on the average sleep latency on the MSLT (mean = 4.7, SD = 3.2) for 100 consecutive narcoleptics from our database. Cataplexy was defined according to the ASDC criteria of 1979 as follows: “A sudden, dramatic decrement in muscle tone and loss of deep reflexes leading to muscle weakness, paralysis, or postural collapse; usually precipitated by an outburst of emotional expression—notably laughter, startle, or sudden physical exercise”. We recognize, as has been noted (13), that, unlike cataplexy, the ancillary symptoms of hypnagogic hallucinations and sleep paralysis are not specific for narcolepsy and also occur in other disorders of excessive somnolence, and that isolated sleep paralysis occurs in as many as 1 in 20 otherwise normal individuals. These symptoms are, thus, not suitable as inclusion criteria. Our criteria for significant sleep-related respiratory abnormality required that obstructive or hypopneic events had to 1) result in at least a 50% reduction in airflow as measured by nasal thermister, 2) be more than 10 seconds in duration, 3) be associated with drops in oxygen saturation as measured by pulse oximetry of more than 4%, and 4) be more frequent than five per hour of sleep. These criteria were based on the current practices of the time and the criteria of the Association of Sleep Disorders Centers (12,14).

All probands and all available first-degree relatives (parents, siblings, and offspring) who were willing to participate were evaluated as their schedules permitted. Evaluations consisted of a clinical sleep disorders interview with one of us (RH), nocturnal polysomnography (NPSG), an MSLT, physical and mental status examinations, Stanford cataplexy scale, Hamilton depression scale, and HLA typing. The analyses of these recordings were performed on measures of sleep latency, REM latency, total sleep time, sleep efficiency, sleep stages, and number of REM sleep periods for the nocturnal studies, and measures of sleep latency, REM latency, and number of REM sleep periods for the MSLT. Student’s t-tests were used to test the hypothesis that the mean value of each sleep parameter is the same for the HLA positive and HLA negative probands; a Bonferroni correction for multiple tests was applied.

RESULTS

To date, 32 narcoleptic patients have satisfied entrance criteria. Demographic characteristics of the 32 probands are as follows: 17 males and 15 females; mean age was 42.12 years (range 13–70 years).

Most probands manifested the entire narcoleptic tetrad (i.e. excessive daytime sleepiness, cataplexy, hypnagogic hallucinations, and sleep paralysis as well as disturbed nocturnal sleep). A 13-year-old and one adult proband (one DRB1*15/DQB1*0602 positive, one not) did not experience cataplexy (6%). However, these two probands had abnormal MSLT, with sleep latency (SL) of 0.8 and 1.4 minutes and two and three REM sleep onsets, respectively. HLA typing revealed the presence of the high-risk haplotype in 21 narcoleptic patients (66%), with two African-Americans having DQB1*0602 but not DRB1*15. Ten probands (31%) had nontypical haplotypes. The results of HLA typing for one proband are pending. The representative results of the NPSGs and MSLTs along with individual results of the HLA typing for all 32 narcoleptic probands are presented in Table 1. The results of the NPSGs and MSLTs (mean SL on NPSGs of 3.2 minutes and mean SL on MSLT of 3.1 minutes) are not surprising for a series of patients with narcolepsy and do electrographically confirm the presence of pathological sleepiness in all of our narcoleptic probands. Pathological tendency toward REM sleep is confirmed by the mean number of REM sleep periods of 3.2 and mean REM latency on the MSLTs of 6.9 minutes.

TABLE 1.

Representative results of the NPSG and MSLT for 32 narcoleptic probands

		NPSG						MSLT			HLA haplotypes
Kindred no.	ID	SL (minutes)	TST (minutes)	% SE	REM L (minutes)	% REM	REMs	SL (minutes)	REM L (minutes)	REMs	DRB1*15 (+/−)	DQB1*0602 (+/−)
NRC001	2.2	3.0	463.0	93.0	18.5	31.1	6	3.3	14.3	2	−	−
NRC002	2.6	3.5	284.2	61.1	1.5	22.7	5	1.8	2.9	4	+	+
NRC003	2.1	1.5	336.8	80.4	103	19.6	3	6.6	8.6	3	+	+
NRC004	2.1	2.5	533.7	94.0	64.1	18.7	7	1.4	12.4	2	−	−
NRC005	2.3	10	408.0	79.7	80.9	26.2	4	6.9	10.4	3	+	+
NRC006	2.1	0.5	426.2	86.3	52.1	19.3	5	1.8	1.1	4	+	+
NRC007	2.1	2.0	654.8	94.9	142.1	20.6	8	0.8	11.0	2	+	+
NRC008	3.3	1.5	400.9	85.1	2.0	16.0	5	3.4	2.3	4	−	−
NRC009	2.1	0.5	261.8	68.4	66.8	12.8	2	2.3	12.8	2	+	+
NRC010	2.3	2.5	385.0	86.9	1.0	18.9	5	1.7	1.5	4	+	+
NRC011	2.2	2.0	376.4	69.7	1.0	24.4	7	5.3	6.3	3	+	+
NRC012	2.1	3.5	505.1	88.2	69.9	14.5	5	2.5	6.8	3	+	+
NRC013	2.1	12.4	464.8	92.8	8.9	28.2	4	1.3	3.4	4	+	+
NRC014	2.2	2.0	412.7	90.9	166.8	21.2	4	6.9	4.8	4	NA	NA
NRC015	2.3	2.0	372.6	80.6	4.0	33.0	5	7.0	4.5	3	+	+
NRC016	2.2	2.5	342.3	82.3	55.3	18.5	4	5.3	4.8	4	−	−
NRC017	2.1	11.0	598.3	92.9	72.3	16.8	5	6.6	11.6	3	+	+
NRC018	2.1	1.5	471.3	86.2	68.6	18.0	5	5.1	8.0	3	−	+
NRC019	2.1	4.0	442.5	88.3	83.7	14.7	5	1.4	9.5	3	−	−
NRC020	2.2	1.5	443.6	74.8	0.0	36.9	5	0.6	1.9	4	+	+
NRC021	2.1	0.5	497.3	97.3	84	24.7	5	5.1	11.4	3	−	−
NRC022	2.1	0.0	478.9	83.6	2.5	25.1	6	0.3	6.8	3	+	+
NRC023	2.1	3.0	526.8	90.5	70.9	18.4	5	1.4	8.0	3	+	+
NRC024	2.1	1.0	351.0	96.0	1.0	19.0	4	0.9	8.3	3	−	+
NRC025	2.3	1.0	451.0	96.6	54.4	28.1	5	0.8	3.3	4	−	−
NRC026	2.2	0.5	471.0	92.4	3.0	19.3	4	2.9	3.8	4	+	+
NRC027	2.1	1.0	420.6	69.4	0.0	32.8	5	0.9	6.3	3	+	+
NRC028	2.2	9.6	389.4	67.1	295.1	21.4	2	5.8	10.6	3	+	+
NRC029	2.2	0.5	483.0	97.0	0.5	30.0	4	2.4	0.5	4	+	+
NRC030	2.1	8.3	484.5	94.6	74.8	21.9	6	3.6	11.8	2	−	−
NRC031	2.2	2.5	571.0	97.9	66.8	26.3	6	2.9	4.1	4	−	−
NRC032	2.1	3.0	444.0	89.5	58.4	14.2	5	2.3	7.4	3	−	−
MEAN		3.2	442.3	85.9	55.4	22.3	4.9	3.1	6.9	3.2

NPSG, nocturnal polysomnography; MSLT, multiple sleep latency test; HLA, human leukocyte antigen; SL, sleep latency; TST, total sleep time; SE, sleep efficiency; REM, rapid eye movement; REM L, latency to first REM period; REMs, number of REM periods; NA, not available.

With respect to current differential diagnostic practices, we examined the degree to which this case series of narcoleptic probands met the current diagnostic criteria as specified by the International Classification of Sleep Disorders [ICSD (15)]. We found that all probands met one or the other of the following two sets of ICSD criteria: 1) recurrent daytime naps or lapses into sleep that occur almost daily for at least 3 months, and sudden bilateral loss of postural muscle tone in association with intense emotion (cataplexy); and 2) a complaint of excessive sleepiness or sudden muscle weakness; associated features that include sleep paralysis, hypnagogic hallucinations, automatic behaviors, and disrupted major sleep episode; polysomnography that demonstrates one or both of the following: sleep latency less than 10 minutes or REM sleep latency less than 20 minutes; an MSLT that demonstrates a mean sleep latency of less than 5 minutes; two or more sleep-onset REM periods; and absence of any medical or psychiatric disorder that could account for the symptoms. None of the probands were diagnosed with any additional sleep disorders such as sleep apnea. In particular, none met the current diagnostic criteria as specified by the ICSD (15) for sleep apnea.

Table 2 compares the 21 probands who had the typical HLA haplotype with the 10 probands who did not have the haplotype in terms of various sleep parameters. There were no significant differences detected between the HLA-positive and the HLA-negative groups on any sleep parameter listed. The HLA-negative group appeared to have a higher mean sleep efficiency, but this difference only approached statistical significance.

TABLE 2.

Mean values for NPSG and MSLT results for the 21 HLA-”positive” (+) and 10 HLA-”negative” (−) narcoleptic probands

	NPSG						MSLT
HLA	SL (minutes)	TST (minutes)	% SE	REM L (minutes)	% REM	REMs	SL (minutes)	REM L (minutes)	REMs
+	3.3	433.8	82.8	49.8	22.8	4.7	3.1	6.4	3.2
−	2.9	463.0	91.9	56.2	21.4	5.4	3.0	8.1	3.1

NPSG, nocturnal polysomnography; MSLT, multiple sleep latency test; HLA, human leukocyte antigen; SL, sleep latency; TST, total sleep time; SE, sleep efficiency; REM, rapid eye movement; REM L, latency to first REM period; REMs, number of REM periods.

Cataplexy in the HLA-negative group appeared to be clinically similar to cataplexy in the HLA-positive group. For example, the cataplexy and other findings for patient MVC, a 52-year-old female, are not dissimilar from those of the HLA-positive narcolepsy population. MVC was diagnosed with narcolepsy at age 16. She presented the symptoms of daytime sleepiness with sleep attacks, cataplexy, hypnagogic hallucinations, and sleep paralysis at age 14. Over time, the attacks became more severe. She also developed disturbed nocturnal sleep. Attacks were usually provoked by positive emotions: laughter, hearing or telling a joke, excitement, elation, making a quick verbal response in a playful or funny context. Negative emotions such as embarrassment, anger, stress, or tension could also provoke attacks, as well as emotions such as surprise, being moved by something emotional, or being startled. The attacks could be complete, with a fall to the ground, or partial, involving muscles of the legs, arms, neck and shoulders, or jaw. Her attacks are always bilateral in the knees but sometimes felt more strongly on one side of her upper body. The duration of cataplexy is usually seconds but may extend for 1–2 minutes. Attack frequency is once per week. MVC has been treated with anticataplectic medications such as imipramine and protriptyline and has experienced marked reduction in the frequency and severity of cataplexy. The clinical manifestation of cataplexy was confirmed by the Stanford cataplexy questionnaire, reviewed independently by two experienced clinicians. The severity of MVC’s symptoms of narcolepsy with cataplexy was also confirmed by a very high score of 389 on the narcolepsy status questionnaire (16). This scale ranges from 60 (almost never experiencing symptoms) to 420 (almost always experiencing symptoms while engaging in any of 12 activities). The polysomnography results confirmed the diagnosis of narcolepsy (NPSG SL of 3.0 minutes and MSLT SL of 2.3 minutes, REM latency of 7.4 minutes, and three sleep-onset REM periods). These results are presented in Table 1 (Kindred no. NRC032, ID 2.1). The HLA typing results showed DRB1*0103, DRB1*11, DQB1*0301, DQB1*X. The probable haplotype is DRB1*0103/DQB1*0301 and DRB1*11/DQB1*0301.

PRELIMINARY OBSERVATIONS ON THE KINDRED

Although the focus of this report is the polysomnographic and HLA characteristics of the probands, some observations can now be made on the relatives of three probands whose families have been extensively studied. Inspection of the pedigrees of the 32 probands disclosed that there was an average of five first-degree relatives per proband for a total of about 160 individuals. At this writing, not every first-degree relative has been available for evaluation or willing to participate due to factors such as death, illness, time constraints, and geographic separation.

A total of 57 first-degree and more distant relatives have undergone complete evaluation so far. These individuals include 18 mothers, 22 fathers, seven siblings, and 10 offspring. Demographic characteristics of these first-degree relatives are as follows: 26 males and 31 females; mean age was 39.1 years (range 10–83 years). Of the 57 relatives for whom studies are complete, eight meet criteria for narcolepsy. This indicates a significant risk factor for individuals closely related to a narcoleptic proband in our sample.

The evaluation of first-degree relatives is relatively complete for families NRC005, NRC008, and NRC031; pedigrees are shown in Figs. 1–3. The variability of symptomatology and of symptom severity is presented in the tables below each figure. Among the 57 relatives, 1/31 females and 7/26 males are affected. This difference is statistically significant (Fisher’s exact test; two-tailed; p < 0.02) and suggests a higher diagnostic rate in male relatives. The greater percentage of males affected is reflected in both the siblings and the parents of the probands. This higher rate cannot be due to the usually cited referral bias of males from the workplace being referred more often for daytime sleep problems, because all family members were systematically studied in the sleep laboratory, irrespective of gender.

FIG. 1.

The HLA haplotypes for family NRC005: a) DRB1*15, DQB1*0602; b) DR17, DQB1*0201; c) DR4, DQB1*0301; d) DRB1*15, DQB1*0602/0603. The nomenclature shown above is that used by the University of California, Los Angeles Tissue Typing Laboratory. Abbreviations in the accompanying table are as follows; DX, diagnosis of narcolepsy; EDS, excessive daytime sleepiness; HH, hypnagogic hallucinations; SP, sleep paralysis, NS, nocturnal sleep disturbance.

FIG. 3.

The HLA haplotypes for family NRC031: a) DRB1*15, DQB1*0602; b) DR16, DQB1*0502; c) DR4(Dw14), DQB1*0302; d) DR14(Dw9), DQB1*0503. The nomenclature shown above is that used by the University of California, Los Angeles Tissue Typing Laboratory. Abbreviations in the accompanying table are as follows: DX, diagnosis of narcolepsy; EDS, excessive daytime sleepiness HH hypnagogic hallucinations; SP, sleep paralysis; NS, nocturnal sleep disturbance.

The following observations can be made from the present data: 1) two families have more than one affected individual, i.e. familial narcolepsy, who does not have the high-risk HLA haplotype (refer to pedigrees NRC008 and NRC031); 2) in one of these families (NRC008) the disease is segregating independently of HLA, i.e. is not linked to the HLA region nor associated with it; 3) one family with multiple affected members shows cosegregation with the HLA high-risk haplotype (refer to pedigree NRC005); and 4) an interesting pattern of excessive daytime sleepiness within all three families.

The pedigree of family NRC008 shows the presence of monozygotic twins. Although the twins themselves report that they are “identical”, objective confirmation of monozygosity was obtained by DNA testing (17,18). Molecular testing using six highly polymorphic microsatellite (CA)_n repeats indicated that both twins have identical profiles at all markers tested and that the probability of monozygosity is greater than 98%. However, the clinical manifestations of the narcoleptic symptoms in these twins are not identical. They are concordant for pathological daytime sleepiness, hypnagogic hallucinations, and disturbed nocturnal sleep but discordant for cataplexy. A report on these twins was recently presented and published (19).

DISCUSSION

The results of the NPSGs and MSLTs, not surprisingly, confirm pathological sleepiness and abnormal tendency toward REM sleep in all of our probands. Clinical manifestation of narcolepsy is also clear, with 91% of the probands exhibiting cataplexy. With respect to familial trends, our limited number of relatives of probands (n = 57) suggests a higher diagnostic rate in male relatives. We have no obvious explanation for this finding, and further analyses are planned when a larger number of families are available for study.

A critical finding is that 10 of 32 probands do not have the high-risk HLA haplotype. The HLA typing results indicate that narcolepsy may exist without the presence of the typical HLA DRB1*15/DQB1*0602 haplotype. This implies that narcolepsy, or some forms of narcolepsy, do not require the presence of a particular HLA haplotype and strongly suggests that there are subtypes of the disease not distinguishable by existing clinical criteria (20).

The percentage of present cases, 31%, who do not have the HLA DRB1*15/DQB1*0602 markers may appear higher than that reported by others. One explanation for this discrepancy may be the fact that our narcoleptic probands were ascertained in chronological order of presentation at a sleep disorders center and were not solicited on the basis of specific symptomatology. The degree of dissociation between narcolepsy and HLA markers is comparable with that found in a multicenter clinical trial involving over 500 patients (21) and by other investigators (Moldofsky, personal communication). However, the report by Mignot et al. (21) does suggest that the likelihood of a patient having the HLA markers increases with the severity of the cataplexy symptom as measured by a questionnaire. Clearly, more research on the relationship between cataplexy and the diagnosis of narcolepsy is necessary.

Current data support the conclusion that some proportion of narcolepsy cases are inherited, but apparently not all cases. This raises the problem of defining with greater precision the proportion of inherited cases and the clinical differentiation between inherited and sporadic cases, if such exists. Also, our data support the conclusion that there are at least two familial forms of narcolepsy, one possibly linked to the HLA region (NRC005) and one definitely assorting independently of the HLA region on chromosome 6p (NRC008).

FIG. 2.

The HLA haplotypes for family NRC008: a) DR14(Dw9), DQB1*0503; b) DR13(Dw18), DQB1*0604; c) DR4(Dw4), DQB1*0302; d) DR7, DQB1*0303; e) DR7, DQB1*0201; f) DR4(Dw4), DQB1*0301. The nomenclature shown above is that used by the University of California, Los Angeles Tissue Typing Laboratory. Abbreviations in the accompanying table are as follows: DX, diagnosis of narcolepsy; EDS, excessive daytime sleepiness; HH, hypnagogic hallucinations; SP, sleep paralysis; NS, nocturnal sleep disturbance.