Published ahead of print on October 24, 2002, doi:10.1164/rccm.200207-677OC
© 2003 American Thoracic Society
Respiratory and Cerebrovascular Responses to Hypoxia and Hypercapnia in Familial DysautonomiaNew York University Medical Center, New York University School of Medicine, New York, New York; Clinica Medica 2, Istituto di Ricovero e Cura a Carattere Scientifico S. Matteo and Dipartimento Medicina Interna, University of Pavia, Pavia, Italy; Department of Neurology, University of Erlangen-Nürnberg, Erlangen, Germany Correspondence and requests for reprints should be addressed to Dr. Luciano Bernardi, M.D., Clinica Medica 2, University of Pavia, P.le Golgi 2, 27100 Pavia, Italy. E-mail: lbern1ps{at}unipv.it
Although cardiorespiratory complications contribute to the high morbidity/mortality of familial dysautonomia (FD), the mechanisms remain unclear. We evaluated respiratory, cardiovascular, and cerebrovascular control by monitoring ventilation, end-tidal carbon dioxide (CO2-et), oxygen saturation, RR interval, blood pressure (BP), and midcerebral artery flow velocity (MCFV) during progressive isocapnic hypoxia, progressive hyperoxic hypercapnia, and during recovery from moderate hyperventilation (to simulate changes leading to respiratory arrest) in 22 subjects with FD and 23 matched control subjects. Subjects with FD had normal ventilation, higher CO2-et, lower oxygen saturation, lower RR interval, and higher BP. MCFV was also higher but depended on the higher baseline CO2-et. In the FD group, whereas hyperoxic hypercapnia induced normal cardiovascular and ventilatory responses, progressive hypoxia resulted in blunted increases in ventilation, paradoxical decreases in RR interval and BP, and lack of MCFV increase. Hyperventilation induced a longer hypocapnia-induced apneic period (51.5 ± 9.9 versus 11.2 ± 5.5 seconds, p < 0.008) with profound desaturation (to 75.8 ± 3.5%), marked BP decrease, and RR interval increase. Subjects with FD develop central depression in response to even moderate hypoxia with lack of expected change in cerebral circulation, leading to hypotension, bradycardia, hypoventilation, and potentially respiratory arrest. Higher resting BP delays occurrence of syncope during hypoxia. Therapeutic measures preventing hypoxia/hypocapnia may correct cardiovascular accidents in patients with FD.
Key Words: familial dysautonomia • hypoxia • hypotension • chemoreceptors • autonomic nervous system
Familial dysautonomia (FD), or Riley–Day Syndrome, is an autosomal recessive disorder with extensive central and peripheral autonomic perturbations affecting cardiovascular and respiratory systems (1, 2). Because the disorder affects the development and survival of unmyelinated sensory and autonomic neurons, with sympathetic development more widely affected than parasympathetic development, affected individuals are unable to mount appropriate cardiovascular or catecholamine responses to physical stress, including change of position or exercise (2–5). Postural hypotension, without compensatory tachycardia, as well as episodic hypertension can be striking, especially in the adult population (2, 5, 6). Many of the respiratory disturbances are attributed to chemo- and baroreceptor dysfunction (1, 7, 8). One dramatic clinical manifestation of these dysfunctions is the breath-holding episode that can result in decerebrate posturing before the brainstem triggers breathing (1). Typically, these episodes follow crying or laughing when poorly coordinated respirations result in increases in depth of ventilation with or without increases in respiratory rate. In addition, affected individuals may manifest periodic breathing and central apnea during sleep, lack of tachypnea with respiratory infection, and an inability to adapt to low oxygen environments, such as high altitudes, airplane travel, and underwater swimming (1). Sustained hyperventilation in normal subjects commonly induces a short apneic period (9), but in subjects with FD hyperventilation may induce prolonged apnea or even respiratory arrest (1, 10). Although survival statistics improved appreciably with the establishment of centralized care and better access to supportive treatments, as of 1982, a newborn with FD still had only a 50% probability of reaching 30 years of age (9). The two most common causes of death were pulmonary and unexplained sudden death, with many deaths occurring during sleep (10). Despite the high mortality in this population, only a few studies have analyzed respiratory control in a very small number of these patients (7, 8). Although central depression of respiration was hypothesized (7), it is still not known whether respiratory abnormalities relate to cardiovascular and cerebrovascular dysfunction. This has practical consequences, because a better knowledge of the complex mechanisms regulating cardiorespiratory control, oxygen transport, and cerebral perfusion may benefit not only subjects with FD but also patients with different types of cardiorespiratory and autonomic involvement. In the present study, using standard rebreathing tests, we investigated whether chemoreflex responses are altered in subjects with FD. The cardiovascular and cerebrovascular responses to progressive hypoxia or hypercapnia were evaluated simultaneously. In addition, we studied the time course of these variables in response to hyperventilation to simulate events preceding the breath-holding episodes in subjects with FD, and to understand the mechanisms inducing syncope and respiratory arrest in this population.
Twenty-two stable subjects with FD (11 females; age 11 to 46 years; 25 ± 2 years) and 23 age- and sex-matched, healthy control subjects (11 female; age 12 to 44 years; 26 ± 2 years) were studied at the Dysautonomia Center at New York University Medical Center. The Institutional Review Board approved the protocol, and written informed consent was obtained from subjects or parents of subjects under 21 years of age. Diagnostic criteria, therapy, and resting data are reported in the online supplement. Subjects were studied supine. Electrocardiogram (ECG, D2 lead; Colin Corp., San Antonio, TX), noninvasive blood pressure (BP) (Colin Corp.), respiratory movements (Respitrace; Nims, Fl), oxygen saturation (SaO2) (Ohmeda, Louisville, CO), expired CO2 (Colin Corp.), and midcerebral artery blood flow velocity (MCFV) by a 2 mHz transcranial Doppler probe at a depth 35–55 mm through the temporal window (DWL, Sipplingen, Germany) were monitored. Subjects breathed through a mouthpiece connected to unidirectional valves either in air or into a rebreathing circuit; in the expiratory part, a heated pneumotachograph (Fleisch, Epalinges, Switzerland) was used to quantify expiratory flow. Three tests were performed: (1) isocapnic progressive hypoxia (from resting values to 80% oxygen saturation); (2) hyperoxic hypercapnia (up to 15 mm Hg above resting values and under low oxygen flow to maintain SaO2 at 1–2% above resting level); and (3) monitoring of the period after mild hyperventilation. Parts (1) and (2) provided measurements of chemoreflex sensitivity to O2 and CO2, assessed cardiovascular and cerebrovascular changes induced by these maneuvers, and verified if changing the levels of oxygen alone could modify the cardiovascular response. Part (3) was performed to mimic the typical breath holding reported in subjects with FD and to monitor the cardiovascular changes and the conditions that induced and reversed the apnea. Parts (1) and (2) were performed with a rebreathing circuit; part (3) was performed in 11 subjects with FD and in 7 control subjects by breathing for 2 minutes at 15 breaths/minute (i.e., close to the normal breathing rate to limit the extent of hyperventilation), which lowered CO2 levels by approximately 10 mm Hg. Many patients with FD cannot voluntarily control respiratory excursions. Therefore, this test was performed only in those patients with FD capable of sufficient coordination to follow instructions for deep breathing and in a subset of matched control subjects. The respiratory flow was integrated by software to calculate breath-by-breath tidal volume, minute ventilation, breathing rate, SaO2, and end-tidal carbon dioxide (CO2-et ). The chemoreflex sensitivity to hypoxia or hypercapnia was obtained from the slope of the linear regression of minute ventilation versus SaO2 or CO2-et , respectively (11, 12). Mean values for heart period (RR interval) and systolic BP were obtained during 1 minute before (baseline) and during the last minute of each rebreathing test. The sensitivity of MCFV to CO2-et was calculated by the slope of the linear regression between these two values during hyperoxic progressive hypercapnia. Data presented here are mean ± SEM. Differences were analyzed by analysis of variance mixed design (repeated measures in two subject groups). If overall significant changes were observed (p < 0.05), then significance was tested by Sheffe's test. Correlation between different variables was evaluated by linear regression analysis.
Anthromorphometric data are summarized in Table E1 (see the online supplement). Although the two groups were well matched regarding age and sex, there were significant differences in body size parameters, i.e., weight, height, BMI and body surface area (BSA) (Table E1 in the online supplement). Therefore, resting and chemoreflex respiratory data were corrected for BSA; similar results were obtained when corrections were made using other anthropomorphometric data. Subjects with FD showed moderate anemia; blood samples were not taken from control subjects.
Resting Ventilation
Chemoreflex Responses to Hypoxia and to Hypercapnia
Chemoreflex sensitivity to hypoxia. The chemoreflex sensitivity to hypoxia was markedly depressed in subjects with FD, even after correction for body size (-0.12 ± 0.04 versus -0.32 ± 0.06 L/minute/m2/%-SaO2; p < 0.011) (Figure 3) . In control subjects, rebreathing increased systolic and diastolic BP's and decreased RR interval (Figures 1 and 4) . Subjects with FD had opposite responses, as both systolic and diastolic BP's decreased significantly but without presyncopal symptoms. The mean RR interval did not change (Figure 4). In fact, RR interval increased in 14 of 22 (64%) subjects with FD. In response to progressive hypoxia, MCFV significantly increased in control subjects but did not change in subjects with FD (Figures 1 and 5 [middle panel]).
BSA-corrected chemoreflex sensitivity for CO2. BSA-corrected chemoreflex sensitivity for CO2 was slightly lower in subjects with FD (0.49 ± 0.11 versus 0.689 ± 0.07 L/minute/m2/mm Hg; p = NS) (Figures 3 and 4). Because subjects with FD had higher resting CO2-et values and similar slopes, their regression line was shifted to the right. During rebreathing, systolic and diastolic BPs increased and RR interval decreased in control subjects and in subjects with FD (Figures 2 and 4). MCFV also increased in both groups to a similar extent (Figure 6 , left panel). The MCFV sensitivity to increases in CO2 was nearly identical in the two groups (2.6 ± 0.6 cm/second/mm Hg in subjects with FD, and 2.4 ± 0.2 cm/second/mm Hg in control subjects; p = NS).
Respiratory, Cardiovascular, and Cerebrovascular Changes Induced by Hyperventilation In both subjects with FD and control subjects, hyperventilation caused the expected decrease in CO2 levels accompanied by a marked decrease in RR interval, a moderate increase in systolic and diastolic BP's, and an increase in oxygen saturation. Hypocapnia induced a marked reduction in MCFV in all subjects (Figure 4). Although hyperventilation caused similar CO2 decreases in both groups (Figure 5, right panel), profound effects were evident in the subjects with FD after termination of hyperventilation (Figure 6). All subjects with FD experienced complete apnea (mean duration: 51.5 ± 9.9 seconds) accompanied by severe desaturation (reaching an average of 75.8 ± 3.5%, Figure 5). After hyperventilation, BPs reached values below resting levels, similar to the response seen during progressive hypoxia, whereas RR interval increased to levels greater than baseline values (Figure 4). In both groups, apnea terminated when CO2-et values were 1–2 mm Hg below baseline levels in subjects with FD (38.6 ± 11 mm Hg versus 39.7 ± 1.1 mm Hg; p = NS) and in control subjects (31.6 ± 1.4 versus 33.7 ± 0.6; p = NS). In the control subjects, termination of hyperventilation induced transitory slowing of respiration (with a pause of 11.2 ± 5.5 seconds; p < 0.008 versus subjects with FD) without desaturation (Figure 5), and BP and RR interval returned toward baseline levels at the end of this short period (Figure 4). During apnea, the MCFV increased in all subjects and in both groups it reached values similar to the MCFV baseline levels (Figure 5, left and right panels), despite the fact that in subjects with FD (but not control subjects) the return to baseline CO2 levels was accompanied by marked oxygen desaturation.
Individuals with FD can succumb to sudden cardiovascular death, but parameters for increased risk have not been identified. Our study contributes to the understanding of FD pathophysiology and suggests mechanisms for fatal cardiovascular events. We have found that in patients with FD, the ventilatory, cardiovascular and cerebrovascular responses to hypoxia are markedly blunted, whereas the responses to hypercapnia are preserved. Respiratory challenges can elicit normal cardiovascular and respiratory responses in individuals with FD as long as they are performed under normo- or hyperoxia, but with hypoxia central depression develops quickly, leading to hypoventilation, bradycardia, and hypotension. Increased resting supine BP combined with maintained cerebrovascular reactivity to CO2 may serve as compensatory mechanisms, and appear to limit the deleterious consequences of this central depression. If the subject is dehydrated or even erect, then hypotension might occur and compensation would be compromised. Reduced ventilatory sensitivity to hypoxia may have other important consequences, such as impaired coronary vasodilation, that can predispose to arrhythmia during prolonged apneic episodes.
Effect of Hypoxia on Cardiovascular Function: Development of Central Ventilatory Depression Subjects with FD frequently have elevated supine BP and impaired cerebral autoregulation (16, 17). Elevated supine BP may play a compensatory role during hypoxia by preventing an excessive reduction in cerebral perfusion. This hypothesis was supported by our observation that MCFV failed to increase in subjects with FD during hypoxia but remained adequate despite decreases in both systolic and diastolic BP's. This suggests that despite directionally unfavorable responses of BP and HR to progressive hypoxia, the subject with FD can maintain relatively compensated cerebral blood flow that may limit the extent of cerebral hypoxia during the posthyperventilatory apnea. A substantial proportion of our subjects with FD (12/22) had a gastrostomy to compensate for oral incoordination so that fluid requirements could be met. Thus, the improved fluid balance of our subjects with FD may have contributed to limiting the deleterious cardiovascular effects previously seen during hypoxia (8).
Is the Response to Hypercapnia Normal in Subjects with FD?
Effect of Apnea on Cardiorespiratory Regulation The progressive increase in CO2 in subjects with FD caused an increase of MCFV, suggesting cerebral vasodilation (Figures 5 and 6). This cerebral vasodilatation, together with higher baseline BP levels, limited or delayed the extent of cerebral hypoxemia and central depression and allowed the subjects to reach a threshold sufficient to stimulate breathing. Because the first breath after apnea generated a CO2 level that was just below the resting CO2 level, it is likely that that termination of apnea was due to stimulation of the central (CO2) chemoreflex (18). The relationship between MCFV and CO2 was not changed by apnea in subjects with FD and remained similar to that of control subjects and to that obtained during hyperoxic hypercapnia, despite the fact that subjects with FD developed a profound hypoxemia. This confirmed that cerebral blood flow in subjects with FD was not sensitive to hypoxia. Thus, blunted hypoxic MCFV may contribute to central depression during hypoxia. The rapid development of hypoxemia during apnea or during breathing in hypoxia likely results from a combination of factors. The subject with FD has a smaller thorax and vital capacity due to a physically smaller body habitus and limited chest wall expansion caused by kyphoscoliosis. In addition, oxygen carrying capacity is decreased by the common presence of anemia (Table E1 in the online supplement). These factors may contribute to prolonged apnea after hyperventilation, as well as poor tolerance of environments with low partial pressure of oxygen, such as pressurized airplane cabins and high altitudes.
Prolonged Apnea May Predispose to Arrhythmias in Patients with FD The clearly abnormal responses to hypoxia in patients with FD speak to the importance of the autonomic nervous system in facilitating an appropriate physiologic response. The attenuation of this sympathetic vasoconstriction in the setting of FD is an important aspect of understanding what is happening in these patients, particularly during apnea, when the sympathetic vasoconstriction should be most marked. The relevance of these data are further emphasized by the fact that during both hypoxia and apnea the direct vasodilatory effects of hypoxia (13) are less opposed by sympathetic vasoconstriction, hence predisposing patients to hypotension during hypoxia and especially during apnea.
Autonomic Dysfunction Is Totally Organic—or Is It Also Functional?
Study Limitations
Conclusion Hyperventilation can be dangerous in subjects with FD because the resulting apnea can provoke rapid hypoxemia and cardiorespiratory depression. If there is an insufficient compensation by hypoxia-induced cerebral vasodilation, or if hypotension develops, then irreversible cardiovascular changes and even death might occur. Coronary flow abnormalities induced by hypoxia and hypoventilation may also contribute to or overlap with these mechanisms. Administration of oxygen at low flow, and assurance of adequate hydration, appear to be beneficial; in addition, techniques increasing the sensitivity to hypoxia (26) appear worth testing. The results of this study have relevance not only for the case of FD but can also extend to other more common pathologies. The occurrence of central depression during hypoxia is frequent in patients with cardiovascular disorders and those with diabetes, as well as in patients with other types of autonomic neuropathies. This condition is probably due to an impairment of cerebrovascular responsiveness to various stimuli, of which hypoxemia may be a common condition, precipitating cardiovascular accidents.
Supported by a grant from the Deutsche Akademie der Naturforscher Leopoldina (B.S.) and by grants from the Dysautonomia Foundation, Inc. (F.B.A.). This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form July 9, 2002; accepted in final form October 18, 2002
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
