I. INTRODUCTION The pulmonary circulation has many properties very different from those of the corporal circulation. Pulmonary arterial pressure is much lower than that of the corporal circulation, although the circulatory system of the former is smaller, whereas blood flow is the same as in the corporal circulation. Among the major regulatory mechanisms, hypoxia has a direct vasodilatory effect on the body arteries, whereas it is an important vasoconstrictor for the small pulmonary arteries. The sympathetic nervous system has a significant corresponding effect on the somatic circulation, whereas it has little effect on the pulmonary circulation. However, some humoral factors, including endothelin-I (a vasoconstrictor) and nitric oxide (NO, an important vasodilator), have the same effect on the physical and pulmonary circulation. Sleep apnea is a common disorder that encompasses a range of sleep-related respiratory disorders, including central and obstructive sleep apnea syndrome (OSAS) and nocturnal hypoxia; obstructive sleep apnea syndrome is the most common type of sleep respiratory disorders, and epidemiologic surveys have estimated that the prevalence of obstructive sleep apnea is 4% in middle-aged men and 4% in middle-aged women. Epidemiologic surveys estimate the prevalence of obstructive sleep apnea syndrome to be 4% in middle-aged men and 2% in middle-aged women. Obstructive apnea is accompanied by significant hypoxemia and oscillations in intrathoracic pressure, and is thus considered to have an important negative effect on the pulmonary circulation. In fact, much more is known about the pulmonary circulation during sleep in OSAS and other lung diseases such as chronic obstructive pulmonary disease (COPD) than in normal subjects, due to the need to measure PAP by invasive means and the inability to influence sleep. Little information is available on PAP during sleep in normal subjects: small sample studies have shown that normal subjects can have a mild increase in PAP of 4-5 mmHg during sleep, with no significant differences between sleep periods. In contrast, early studies confirmed that PAP can be significantly increased during sleep in patients with OSAS and emphasized the importance of pulmonary circulation involvement during apnea and its possible role in progression to cardiac failure; pulmonary heart disease is considered an important feature of the syndrome. These studies were conducted mainly in patients with severe OSAS, with the possibility of coexisting obesity hypoventilation. Since then, the prognostic role of OSAS for pulmonary heart disease has been questioned as more and more OSAS has been demonstrated in mild respiratory disturbances as well, and alterations in PAP in apnea have been found to be milder than in the previous studies. Pathophysiologic studies in humans and animals have shed important light on the pathogenic role of apnea in pulmonary circulatory disorders, and relatively large sample sizes have made it possible to evaluate the prevalence of persistent pulmonary hypertension or right heart changes in apnea. Finally, the role of OSAS therapy on pulmonary circulation during sleep and wakefulness has also been studied, although the available information is limited. II Pulmonary hemodynamics during sleep (i) Acute changes in pulmonary artery pressure during obstructive apnea Many investigators have explored the acute changes in PAP in recurrent obstructive apnea, with early findings emphasizing that there can be a significant increase in PAP at the termination of apnea or immediately following apnea, especially during the REM sleep period or during the second half of the night. Studies of pulmonary wedge pressure (Pwp) similarly confirmed a progressive, but usually not excessive, increase during apnea.Pwp reflects the level of left ventricular filling pressures, thus suggesting that apnea-associated increases in PAP are at least partly of the postcapillary type. It has since been reported that PAP and Pwp may be very mildly but nonpathologically elevated in apnea. All of the studies described above used endovascular measurements of PAP and showed that large oscillations in PAP were synchronized with obstructive respiratory effort, thus suggesting that there is a rapid elevation of PAP immediately after the termination of apnea. Direct intravascular measurement of PAP in OSAS would certainly have a superimposed effect on the respiratory-induced oscillations in intrathoracic pressure. The acute changes in PAP during apnea have since been described in greater detail thanks to the establishment of the transmural PAP measurement method. Transmural PAP is measured by measuring the magnitude of intrathoracic pressure by means of intraesophageal pressure as a base control level, which allows estimation of pulmonary artery distending pressure and is independent of intrathoracic pressure. Measurement of transmural PAP pressure using esophageal pressure revealed that PAP appeared to begin to rise during apnea and had a transient rise at the termination of apnea. Unlike intravascular PAP, which changes rapidly and is large, the change in transmural PAP during obstructive apnea is smooth, and the peak pressure of transmural PAP does not occur abruptly at the time of respiratory resumption but begins to rise gradually and progressively in the early to middle part of the apnea. This is an important difference from the change in pressure in the circulation at the onset of apnea, where there is a rapid rise after the termination of apnea. (ii) Temporal characteristics of pulmonary artery pressure changes during sleep Studies have shown that changes in PAP during sleep in patients with OSAS may vary according to the time of night, sleep state and apnea. Early studies have demonstrated that PAP rises progressively during prolonged apneas; the rise is interrupted when the intervals between apneas are prolonged, and returns to its original level during wakefulness. There was a tendency for PAP to increase progressively during shorter intervals of apnea, because the brief ventilation during these intervals was not sufficient to restore blood flow and alveolar oxygen partial pressure to normal levels. The above studies suggest that decreased alveolar oxygen partial pressure may be a major cause of the progressive elevation of PAP during sleep, but none of them continuously monitored changes in oxygen saturation (SaO2), which might have better elucidated the role of hypoxia. Other studies have reported higher PAP in REM than in NREM, possibly because apnea and hypoxia are more severe in REM, but not excluding the role of neural mechanisms associated with REM sleep itself. Finally, the role of the nocturnal time course in determining PAP levels in patients with OSAS has recently been investigated. One study confirmed the tendency of PAP to show a significant but small progressive increase over time during the night and that it was not associated with the characterization of apnea exacerbation during the night; however, another study confirmed that PAP levels are highest during the middle part of the night, when apnea is at its most severe. In conclusion, most studies support that changes in PAP over time during sleep are mainly closely related to the severity of apnea, and that PAP levels are highest during apneas accompanied by lower SaO2. (iii) Changes in cardiac output during obstructive sleep apnea In normal people, their cardiac output decreases when they go from wakefulness to sleep mainly due to the decrease in heart rate. Since cardiac output affects arterial pressure, understanding the changes in cardiac output during apnea will help to understand the mechanism of changes in PAP during apnea. Early studies have demonstrated that the heart enlarges during apnea, and it was hypothesized that cardiac filling and output might be different during apnea. Studies in dogs also found an increase in right heart volume during the end portion of the apnea. Subsequent ultrasound and nuclide studies have confirmed that the diameter of the left ventricle increases at end-systole rather than end-diastole during apnea in humans, suggesting that increased inspiratory effort during apnea can lead to an increase in left ventricular afterload, which in turn affects its emptying. Changes in filling and output of the right and left ventricles during the apnea cycle (i.e., the time between an apnea and its subsequent resumption of ventilation) are opposite. Ultrasound has demonstrated that in OSAS patients with apnea, right ventricular volume increases and left ventricular end-diastolic filling decreases during the inspiratory portion of the obstructive respiratory effort due to septal displacement, whereas the opposite is true during the expiratory portion of the respiratory effort. Animal studies have also confirmed that obstructive inspiratory effort is associated with decreased left ventricular output and increased right ventricular output; however, the changes in mean cardiac output are the same for the physical and pulmonary circulations during each part of the apnea cycle. Thus, during the apnea cycle, whether obstructive or nonobstructive, the changes in left and right ventricular output during inspiration are at least partially averaged by the reverse change during expiration, and it is just not clear whether full compensation of output from the pulmonary and somatic circuits can be achieved after each obstructive respiration or after a longer period of time. Several studies have measured cardiac output during obstructive apnea. Early studies of OSAS hemodynamics using thermodilution reported a decrease in cardiac output during obstructive apnea and an increase during subsequent resumption of ventilation, but in fact this method was not suitable for some events as rapidly altered as obstructive apnea. Impedance cardiography was later used to show that left ventricular output decreases during apnea, but that cardiac output during sleep decreases only during apnea during the REM sleep phase, when the heart rate can reach lower levels than during the NREM phase. However, this technique cannot reliably obtain data after apnea because hyperventilation after apnea affects the impedance signal detected. Finally, nuclide imaging to determine left ventricular volume changes and intravascular flowmetry to determine pulmonary artery blood flow velocities have shown that both left and right ventricular outputs are reduced to levels that are not adequately compensated for by an increase in heart rate when ventilation is resumed, and thus cardiac output is temporarily reduced. Other studies in humans and animal models have also shown that both left and right ventricular output tend to decrease on resumption of respiration: however, due to the compensatory effect of heart rate, there may not be a significant change in cardiac output after apnea. The difference between artificially induced apnea in animals and spontaneous apnea in humans, the differences in techniques used, and the limited accuracy of some methods all contribute to the discrepancy in results. In conclusion, the cardiac output of obstructive apnea is characterized by a decrease in ventilation recovery, but this is accompanied by an increase in heart rate, thus making the net effect on cardiac output still unclear. There is no evidence that changes in cardiac output during the apnea cycle have any significant effect on changes in PAP. Furthermore, it is not clear what the mean absolute cardiac output is during the apnea cycle. Given this consideration, it has been demonstrated that during the NREM period, mean cardiac output in dogs increases by approximately 17.9% during cyclic obstructive apnea compared with normal breathing, but it is not clear if this is also true in humans. (iv) Mechanisms of altered pulmonary hemodynamics during sleep Several factors have been proposed to contribute to changes in PAP during sleep in patients with OSAS: hypoxemia, hypercapnia and acidemia, neurologic reflexes, and changes in intrathoracic pressure. 1, Hypoxemia Many severe obstructive sleep apneas have periodic PAP elevations immediately following hypoxia, and this reversible PAP fluctuation begins with a decrease in the partial pressure of blood oxygen, thus representing a direct vasoconstrictive effect of hypoxia on small pulmonary arteries. The effects of hypoxia have been most intensively studied because it has become clear that the pulmonary vasculature responds to alveolar low partial pressure of oxygen. The first hint of the effects of hypoxia came from the observation that inhaled high-flow oxygen reduced the highest PAP levels at night, and it was later reported that changes in transmural PAP during apnea were significantly correlated with SaO2 levels and that intermittent inhalation of oxygen affected transmural PAP levels during apnea, probably because of the difference in SaO2 levels between inhaled air and oxygen during apnea. Artificially induced apnea in dogs and complete elimination of hypoxia by oxygen inhalation resulted in a more delayed increase in PAP. Although it is clear that hypoxia is an important stimulus for contraction of small pulmonary arteries, physiological studies have revealed that the characteristics of the response of small pulmonary arteries to hypoxia do not conform to the typical apnea model in which elevations in PAP are immediately followed by changes in partial pressure of oxygen of up to several seconds. The rapid or slow profile of the time course of hypoxic pulmonary vasoconstriction has been examined and confirms that the pulmonary vascular response, although initiated after seconds, is not complete until at least several minutes, and that it takes the same amount of time after hypoxia is lifted for PAP to return to basal levels. Thus, the role of hypoxia in determining changes in PAP in apnea is to be interpreted in the light of the different characteristics of the pulmonary vascular response to hypoxic stimuli. One study compared the changes in transmural PAP in periodic apneas in different groups: whether there were apneas with a smooth but smaller decrease in SaO2 (NREM sleep period) or apneas with a significant decrease in SaO2 (which may occur during REM sleep period). As expected, the highest PAP occurred at the end of apnea in the second group. Interestingly, in the first group, PAP could be gradually reduced to a stable basal value immediately after each apnea and before the start of the next apnea, whereas in the second group, PAP remained at a high level at the start of the next apnea. On the basis of these findings and the temporal nature of the hypoxic vasoconstrictor response, the response of PAP to hypoxia in obstructive apnea may be as follows: during an apnea, if a sufficiently low level of hypoxia is reached, PAP may begin to rise but does not necessarily rise to the same level if the hypoxia is prolonged longer than the apnea itself; whereas after the apnea is terminated, because oxygen usually returns to the normal level within a few seconds PAP does not return to basal levels unless the preceding vasoconstrictor response is weak. This explanation also applies to the temporal characteristics of nocturnal PAP changes: PAP rises progressively in severe hypoxic apneas with short intervals between apneas, whereas PAP is smoother in mild hypoxic apneas or when the intervals between apneas are longer. Individual differences in response to hypoxia may also make the intensity of PAP changes different, and it has been demonstrated that OSAS patients respond differently to hypoxia when awake. 2. Hypercapnia and Acidemia As for the role of hypercapnia and acidemia on PAP, little is known. It has been suggested that hypercapnia may lead to elevated PAP when ventilation between apneas is ineffective, and that hypercapnia caused by oxygenation during sleep may possibly explain why PAP is not reduced in some patients with OSAS even though SaO2 has improved. Although hypercapnia usually enhances the response to hypoxia, its role in PAP changes in apnea has not been demonstrated. The study confirmed that the pulmonary vascular response to hypoxia during wakefulness was enhanced by hypercapnia in only 4 of 20 patients with OSAS, a finding that does not support an important role for hypercapnia in OSAS-related changes in PAP. 3, Neurologic reflexes Another factor that may affect PAP in OSAS is the sleep phase, as the level of PAP in the REM phase of the obstructive apnea cycle was found to be higher in than in the NREM phase. Studies have confirmed a negative correlation between SaO2 and PAP in both REM and NREM phases, with the slope of the regression line being the same in both sleep phases, but with an upward shift of the regression line in the REM phase, at least in some patients. Thus, it is possible that neural mechanisms contribute to the increase in pulmonary vascular tone, especially during REM sleep periods. However, these data can also be explained by the aforementioned temporal characterization of the hypoxic pulmonary vascular response: in many of the typical REM sleep-phase apneas, their significant decrease in blood oxygen causes the small pulmonary arteries to constrict and then, because of the short intervals between the apneas, they do not have enough time to fully relax, and so may lead to the observed upward shift in the SaO2-PAP correlation. In addition, it was confirmed that the REM sleep period in dogs does not have much effect on the change in PAP during apnea. Similarly, PAP was not affected by the onset of microarousals, a finding obtained in dogs that may also apply to humans. In contrast, the change in transmural PAP during apnea in humans is very slow and does not show an abrupt change, which is in contrast to the abrupt increase in body circulation pressure shown at the termination of apnea, and could possibly be due to the effect of microarousal at the termination of apnea. Although adrenergic receptors are abundant in the walls of small pulmonary arteries, and they can lead to altered tone and even remodeling of the pulmonary vasculature when stimulated, the effects of the sympathetic nervous system on pulmonary hemodynamics are poorly defined and little understood. No neural mechanism has been demonstrated to influence changes in PAP in apnea. The changes in PAP in apnea in OSAS patients with Shy-Drager syndrome or treated with atropine are the same as in other OSAS patients. The above results were also verified in dogs, suggesting that the sympathetic effect, if any, on OSAS-induced changes in PAP is minimal. 4. Intrathoracic pressure The mechanical effect of changes in negative intrathoracic pressure on pulmonary hemodynamics during apnea is obvious. It is well established that there is a concomitant increase in right cardiac output and decrease in left ventricular output during the inspiratory portion of an obstructive respiratory effort. The effect of altered intrathoracic pressure on PAP was first explored in a study that monitored both right atrial pressure and PAP. Subsequent studies have found that there can be a transient proportional increase in PAP after a decrease in esophageal pressure and that changes in esophageal pressure can lead to changes in transmural PAP. The hemodynamic effects of changes in intrathoracic pressure can be explained by changes in ventricular anterior and posterior load. Decreases in intrathoracic pressure begin with increases in venous return to the right atrium and in right ventricular filling and output, which can subsequently lead to some increase in PAP. However, further decreases in intrathoracic pressure can lead to collapse of the extrathoracic veins, thereby limiting venous return. In addition, a decrease in intrathoracic pressure leads to an increase in left ventricular afterload, which can lead to an increase in pulmonary blood volume and pulmonary capillary pressure. No increase in pulmonary blood volume during obstructive apnea has been demonstrated in dogs, whereas elevated pulmonary wedge pressure has been found in human obstructive apnea. Acute pulmonary edema caused by acute left heart failure, possibly due to a sudden increase in afterload during apnea, has also been reported, which is rare in OSAS, but experiments in dogs suggest that subclinical pulmonary edema may be more common. However, although it is thought that LV afterload can be estimated from body circulation pressures, no studies have confirmed a correlation between LV afterload and transmural PAP. Similarly, the decrease in output after apnea could be interpreted as secondary to some mechanical effect. As for the right ventricle, the decrease in output at the end of inspiration was significant, suggesting that expansion of the lungs may have led to compression of the heart or increased pulmonary vascular resistance and output limitation. In contrast, SaO2 reduction and microarousal did not have any effect on the reduction of cardiac output after apnea. In animal experiments, inhibition of the increase in pressure in the body circulation after apnea by pharmacologic blockade of the autonomic nervous system prevented the decrease in left ventricular output, suggesting that the increase in left ventricular afterload after apnea may be the determining factor for the decrease in its output. 5, Humoral factors Humoral factors may also be involved in regulating the effects of obstructive apnea on PAP, but the available information is still very scarce. Studies of atrial natriuretic peptide and transmural PAP during sleep in patients with OSAS did not find a correlation between hormones and blood pressure levels. In conclusion, the available findings suggest that changes in pulmonary hemodynamics in obstructive apnea are mainly the result of mechanical effects of changes in intrathoracic pressure and fluctuations in SaO2. Mechanical effects are present in all obstructive apnea events, affecting ventricular filling and output and leading to rapid but usually mild fluctuations in transmural PAP, whereas fluctuations in SaO2 may have little effect on PAP during brief and mildly hypoxic apneas, whereas a series of prolonged and severely hypoxic apneas can lead to a significant and progressive elevation of PAP with a time course slower than that of SaO2 change itself tends to be slower. The most likely scenario at this time would seem to be that PAP changes in OSAS are primarily due to increased venous return during any apnea, due to microarousals after an apnea, and that the more prolonged elevations overnight are due to hypoxia. (v) Immediate effect of treatment of obstructive sleep apnea on pulmonary circulation during sleep Since obstructive apnea affects the pulmonary circulation, treatment of apnea must also affect pulmonary hemodynamic indices, as demonstrated in several studies in the 1970s, where it was demonstrated that after tracheotomy in patients with OSAS the elevation of the peak PAP pressure could be prevented, although there is still some mild increase in PAP during sleep. The role of continuous positive airway pressure (CPAP) therapy is even more complex because, in addition to eliminating apnea, positive pressure ventilation may also affect venous return, ventricular afterload, and cardiac output. The effect of CPAP therapy on the pulmonary circulation during sleep in patients with OSAS has only been evaluated in terms of the level of PAP: like tracheotomy, CPAP therapy also prevents higher peak PAP pressures. Other studies have evaluated the effect of CPAP therapy on hemodynamics during wakefulness in patients with OSAS and have found that CPAP increases intravascular PAP levels if 10 cm H2O is applied, but this may be due to an increase in intrathoracic pressure; in fact, transmural PAP does not change at 10 cm H2O and decreases at 15 cm H2O.Positive pressure at 10 cm H2O cardiac output did not change or was reduced during ventilation, but there was interindividual variation in this reduction, which rarely resulted in a lower than normal cardiac output. Because obstructive apnea itself leads to a reduction in cardiac output and especially arterial oxygen, this reduction in cardiac output, sometimes caused by CPAP, must be evaluated; in summary, it can be concluded that most patients with OSAS who are treated with CPAP during sleep receive a beneficial effect on tissue oxygen supply due to a significant improvement in SaO2 compared with untreated patients. III. Pulmonary hemodynamics during wakefulness (I) Pulmonary artery pressure during wakefulness in patients with OSAS It was once thought that PAP should be persistently elevated in OSAS as well as in the physical circulation, i.e., patients with OSAS during wakefulness may also have pulmonary hypertension, which was considered a common complication of OSAS until the 1970s. However, subsequent studies have confirmed that pulmonary hypertension is present in less than half of OSAS patients at rest, and most of these are only mild and may not be of any clinical significance. In contrast, pulmonary hypertension during exercise is much more common, but unlike the resting state, it is predominantly of the postcapillary type, suggesting that the left ventricular response to exercise has been blunted. There is evidence that this only occurs in the presence of daytime hypoxia, and that OSAS patients with pulmonary hypertension have worse daytime blood gases at rest than those without pulmonary hypertension. Longer periods of hypoxia appear to be necessary for chronic elevation of PAP; a transient decrease in blood oxygen that is quickly and completely restored does not in itself appear to be sufficient to cause chronic elevation of PAP. However, explanations for the pathogenesis of this persistent type of pulmonary hypertension have varied from study to study; Strasbourg et al. suggest that pulmonary hypertension at rest in OSAS is due to concomitant pulmonary dysfunction (e.g., small airway obstruction, neuromuscular disease, or extreme obesity). Sustained pulmonary hypertension is uncommon in most patients with simple OSAS, so OSAS patients prone to daytime pulmonary hypertension are usually those with underlying lung disease, hypoxia and hypercapnia, congestive heart failure, or obesity. However, the investigators found that daytime hypoxemia was, on average, less severe in patients with OSAS than in patients with chronic obstructive pulmonary disease (COPD) for the same degree of pulmonary hypertension, and therefore concluded that persistent pulmonary hypertension was not likely to be caused by a sleep-breathing disorder alone, but rather by mild daytime hypoxemia associated with moderate-to-severe nocturnal hypoxia. Indeed, studies confirming the high prevalence of pulmonary hypertension in OSAS patients did include a high proportion of COPD patients. Other authors have emphasized that certain patients with OSAS can have elevated pulmonary artery pressures even if their daytime partial pressure of blood oxygen is not guaranteed with chronic daytime hypertension. There are also studies confirming that in OSAS patients without lung disease, noninvasive PAP measurements show elevated PAP in some patients as well, but the pulmonary hypertension in this case is so mild that it may not be of any clinical significance. These studies support some of the previously believed views that OSAS itself may be an important contributor to pulmonary hypertension, mainly due to the remodeling of the pulmonary vasculature resulting from repeated nocturnal hypoxia. One of these studies also confirmed that other confounding factors such as appetite-suppressing drugs or left ventricular insufficiency in the presence of postcapillary pulmonary hypertension did not influence the outcome. Another study found that daytime and nighttime hypoxia in patients with pulmonary hypertension were the same as in non-pulmonary hypertensive patients, but that the pulmonary vascular response to hypoxia was more pronounced in the former, probably due to pulmonary vascular remodeling and more pronounced alveolar ventilation/perfusion dysregulation. Findings on the correlation between indicators of the severity of OSAS and pulmonary hypertension have been inconsistent. The development of persistent daytime pulmonary hypertension does not usually correlate well with the severity of the sleep apnea disorder itself per se (apnea/hypopnea index, AHI); even the level of nocturnal SaO2, which is most likely to be associated with pulmonary hypertension, can be different or the same in the presence or absence of pulmonary hypertension. An unanswered question is whether there is a correlation between humoral factors in OSAS and pulmonary hypertension during wakefulness. Many studies have demonstrated changes in the levels of certain factors that can influence pulmonary hemodynamics. Among vasodilating substances, it has been demonstrated that urinary natriuretic peptide and nitric oxide levels are elevated and decreased, respectively, in patients with OSAS; both of these levels improve after OSAS treatment. However, the results of studies on changes in the strong vasoconstrictor substance, endothelin I, have been inconsistent. It remains to be explored whether these or possibly other factors may play a role in the development of pulmonary hypertension in OSAS. Another possible reason that is not at all clear is that individual predisposition from a genetic basis may contribute to the different remodeling of the pulmonary circulation in response to hypoxia. It is still very unclear what are the prognostic indications for the development of pulmonary hypertension in OSAS. In a group of unselected patients with OSAS who were followed for 5 years after diagnosis and treatment with CPAP, univariate analyses showed exercise PAP and resting PAP to be significant risk factors for death, as well as a history of prolonged heavy smoking, spirometric indices, and PaO2. However, a similar study has not yet been done in patients with untreated OSAS. Autopsies of 10 extremely obese patients found that they may have had alveolar hypoventilation syndrome and sleep apnea during their lifetime, and that the cause of death was respiratory circulatory failure combined with pulmonary heart disease; these 10 cases were also compared with 10 obese patients who may not have had obesity hypoventilation syndrome, and, unlike the latter, the former frequently showed heavy pulmonary bruising and intra-alveolar hemorrhage, which was consistent with hypoxemia that may be secondary to chronic left heart failure and capillary proliferation. However, it is difficult to apply these findings to the general population of patients with OSAS because these patients are extremely obese and their underlying disease is severe enough to cause death. (ii) Right heart hypertrophy and insufficiency In addition to pulmonary vascular changes, OSAS may also affect the morphology and function of the right heart, which is unfortunately technically difficult to evaluate and accuracy is often limited. Despite this, a number of studies have applied echocardiography to right heart morphology in patients with OSAS and have shown a wide variation in the prevalence of right heart hypertrophy, ranging from 0% to 71%. The studies that did not find right heart hypertrophy only included OSAS patients with normal daytime blood gas analysis. Whereas other studies considered the severity of sleep disordered breathing as a possible major determinant of right heart hypertrophy, their possible effect on daytime lung function was not evaluated. Experimental data suggest that right heart hypertrophy in patients with OSAS may be due to the mechanical effects of prolonged upper airway obstruction. Evaluation of right heart function was first performed in patients with OSAS with associated lung disease. The classification of right heart failure (RVF) is mainly based on clinical criteria. From these studies it can be concluded that abnormal daytime blood gas analysis is the most important determinant of right heart failure in patients with OSAS. Other studies have evaluated right heart function in patients with OSAS excluding those with significant pulmonary disease, one of which found that the risk factors for RVF may be waking hypercapnia or an AHI >40; whereas another study concluded that even in the absence of obesity, pulmonary failure, or LV insufficiency, OSAS itself may lead to RVF.It is important to note that in both of these studies the right heart function failure was defined solely on the basis of reduced right heart ejection fraction (RVEF) and not on the basis of clinical criteria, and therefore the degree of ventricular changes that determined a patient’s right heart insufficiency was much less severe than in the studies that defined it on the basis of overt clinical symptoms and/or signs of right heart failure. Unfortunately, investigators have not always evaluated the morphology or function of the PAP and RV at the same time.Noda et al. found that right heart hypertrophy was present only in patients with pulmonary hypertension. In contrast, Sanner et al. reported no relationship between pulmonary hypertension and right heart insufficiency, and Fletcher et al. demonstrated that pulmonary hypertension and right heart insufficiency tended to be seen in the same patients, but that the RVEF appeared to respond better to treatment than the PAP. In conclusion, although many of these studies have some shortcomings, there is data suggesting that there are some differences in the pathogenesis of the daytime elevation of the PAP and the RV abnormalities in patients with OSAS. abnormalities in OSAS patients with some differences in the pathogenesis of the disease. Mild RV abnormalities may be at least not exclusively associated with daytime pulmonary hypertension, but seem more likely to be the result of sleep apnea disorders and their associated nocturnal hemodynamic disturbances. This hypothesis is supported by the fact that RV function responds much better to OSAS treatment than to daytime PAP. It is also possible that the sympathetic hypersensitivity typical of OSAS patients does not affect PAP but only results in RV impairment. In contrast, permanent impairment of measurable respiratory function in the awake state seems to be essential for the development of most severe RV abnormalities, which are associated with elevated pulmonary circulatory pressures. (iii) Role of long-term OSAS treatment on pulmonary hemodynamics Mild reductions in PAP have been observed in individual cases of OSAS patients receiving long-term treatment. However, in studies of large samples of OSAS, neither tracheotomy nor CPAP therapy revealed a significant reduction in PAP, either at rest or during exercise; on the contrary, both children and adults showed a significant increase in RVEF after long-term therapy, and some studies have confirmed that this improvement occurs only in patients who had very poor pre-treatment indices. Long-term CPAP therapy has the potential to improve daytime blood gas analysis in addition to eliminating apnea, but whether this has a role in hemodynamic improvement in patients with OSAS is unclear. IV.CONCLUSION The established articles on pulmonary hemodynamics in OSAS represent a very important starting point in the field. Research on the periodic changes in PAP and Pwp in obstructive apnea, the effect of hypoxia on PAP changes during sleep, the possible progressive elevation of PAP during sleep, and the possible role of all of these factors in the formation of pneumogenic heart disease that is only partially reversible by OSAS treatment needs to be continued to this day. More modern studies have partially downplayed the previous overemphasis on acute alterations in pulmonary hemodynamics in obstructive apnea and have recognized that changes in transmural PAP are very slow in these events. It is also recognized that mild hypoxic events associated with very transient and sometimes negligible PAP changes are different from severe hypoxic events that result in significant and sometimes progressive PAP elevations. However, this does not mean that the level of PAP during sleep is determined by the level of SaO2: the concentration of circulating substances that affect pulmonary vascular tone, such as urinary natriuretic peptide, nitric oxide or epinephrine, as well as certain possible genetic factors, may have a different effect on the hypoxic response of an individual, and therefore even if the arterial hypoxemia due to apnea is the same, its PAP may arrive at different levels. As with PAP, cardiac output in patients with OSAS varies continuously during the cycle of obstructive apnea, but its changes during each cycle are transient and variable. In contrast to cardiac output during regular breathing, there is less clarity about possible changes in mean cardiac output during repeated obstructive apneas and the relationship between repeated mechanical stress on the right and left ventricles during obstructive apnea and impaired long-term cardiac function. The prevalence of persistent pulmonary hypertension in patients with OSAS who come to the sleep center is estimated to be around 20%. However, there is controversy as to whether pulmonary hypertension is simply the result of pulmonary dysfunction, left ventricular insufficiency, and/or morbid obesity that may be associated with OSAS or whether recurrent apnea itself contributes to the development of pulmonary hypertension. Studies based on the correlation between the severity of sleep disordered breathing and the level of PAP during wakefulness may not be helpful in resolving this issue. As discussed above, different patients may show different responses to hypoxia during apnea; therefore, the correlation between daytime PAP levels and sleep apnea disorders cannot be represented by AHI or nocturnal hypoxia, but rather by changes in PAP during nocturnal apnea. In addition to the relationship with AHI and SaO2, the relationship between PAP in sleep in response to apnea may more clearly reflect the effect of sleep disordered breathing on PAP during wakefulness, but this is difficult to validate in large samples. Future advances in technology will make these studies more clinically practical. RV alterations in OSAS have been studied much later than in PAP. RV hypertrophy has been studied in only a few studies applying cardiac ultrasound, with little histologic evaluation. The small number of studies on this issue, the widely varying results obtained, the selection of patients, and the different designs of some of the studies have prevented us from drawing conclusions about the incidence of RV hypertrophy and its most important risk factor in OSAS. Studies of RV function are also scarce, but the results seem to be more consistent, with the available data showing that reduced RVEF can occur even in the absence of pulmonary hypertension and can be reversed by OSAS therapy, whereas definite right heart failure (hepatomegaly, peripheral edema, and high jugular venous pressures) requires a combination of both pulmonary dysfunction and pulmonary hypertension. Points to note: 1. In obstructive apnea, PAP and Pwp may be elevated. Elevations can range from very mild to severe, and from very brief to several minutes in duration. Significant elevation of PAP over a prolonged period of time is often seen in a series of consecutive apneas with significant hypoxia. 3, Cardiac output may change during apnea, but because of its transient nature, whether it is increased or decreased does not seem to be important. 4, Severe pulmonary hemodynamic impairment occurs only in patients with comorbid lung disease or severe obesity. 5, Moderate pulmonary hemodynamic impairment is seen in patients with OSAS whose blood gas analysis is near normal or mildly abnormal. 6, There is no evidence that CPAP therapy restores daytime pulmonary hypertension to normal levels. 7, Long-term CPAP therapy may increase right heart ejection fraction.