A Case of Concomitant Obstructive Sleep Apnea and Non-Alcoholic Steatohepatitis Treated With CPAP Therapy.

Obstructive sleep apnea syndrome is a disorder of sleep breathing that is a result of recurrent and intermittent hypoxia during sleep induced by the repeated partial or complete collapse of the upper airway, eventually causing chronic intermittent hypoxia. Non-alcoholic fatty liver disease is divided into non-alcoholic fatty liver and non-alcoholic steatohepatitis. Animal and human studies showed that obesity is associated with chronic liver hypoxia, even in the presence of systemic normoxia causing inflammation and release of cytokines. A "two-hit" model has been proposed. The first hit is characterized by insulin resistance and excess hepatic lipid accumulation secondary to abnormal fatty acid metabolism. Oxidative stress and inflammation are thought to comprise the second hit. Gold standard for the diagnosis of non-alcoholic steatohepatitis is a liver biopsy. Many clinical scores and non-invasive tools are used for the diagnosis of non-alcoholic steatohepatitis. Conservative management with lifestyle modifications including diet, exercise and weight loss remains the therapy of choice today. We present a case report of a 39-year-old man who was diagnosed with concomitant non-alcoholic steatohepatitis and severe obstructive sleep apnea. He was started treatment with continuous positive airway pressure and demonstrated excellent adherence to therapy for 6 years, with concomitant obstructive sleep apnea and non-alcoholic steatohepatitis which reversed with prolonged optimal continuous positive airway pressure therapy. Physical examination remained unremarkable except for morbid obesity. His abdominal girth, as well as body mass index, remained unchanged. After 6 years of optimal continuous positive airway pressure therapy, liver enzymes and relevant lipid panel normalized, suggesting reversal of non-alcoholic steatohepatitis.

Introduction

Obstructive sleep apnea (OSA) syndrome is a disorder of sleep breathing that is a result of recurrent and intermittent hypoxia during sleep induced by the repeated partial or complete collapse of the upper airway, eventually causing chronic intermittent hypoxia (CIH) [1]. The adult OSA task force of the American Academy of Sleep Medicine defines clinical OSA characterized by the occurrence of daytime sleepiness, loud snoring, witnessed breathing interruptions or awakenings due to gasping or choking in the presence of at least five obstructive respiratory events - apneas, hypopneas per hour of sleep index (AHI). The presence of AHI of 15 or greater in the absence of sleep-related symptoms is also sufficient for the diagnosis of OSA due to the greater association of this severity of obstruction with important consequences such as increased cardiovascular disease risk [2].

Arbitrary criteria to define OSA syndrome have led to a wide range of incidence of OSA that has been reported from 4% to more than 60% among the adult population [3]. Prevalence of certain risk factors like obesity (body mass index (BMI) > 35 kg/m2), congestive heart failure, atrial fibrillation, treatment-refractory hypertension, type 2 diabetes, nocturnal dysrhythmias, stroke, pulmonary hypertension, high-risk driving populations and preoperative bariatric patients increase the incidence much more than the general population [2].

Non-alcoholic fatty liver disease (NAFLD) is divided into non-alcoholic fatty liver (NAFL) and non-alcoholic steatohepatitis (NASH). They are differentiated from each other by the histological presence of cellular injury and fibrosis in NASH and its absence in NAFLD. Macrovesicular steatosis with lobular and periportal inflammation may be seen in both [4].

Evidence suggests a strong association of metabolic risk factors like obesity, type 2 diabetes mellitus, elevated triglyceride levels, cardiovascular disease and NAFLD with OSA [1, 5-10]. Prevalence of NAFLD is around 30% in the general population, but it doubles among patients with diabetes or obesity [11]. Studies show that OSA syndrome can cause intrinsic liver injury and lead to NAFLD [12, 13]. This can happen independent of obesity. Although there are conflicting results on the effect of continuous positive airway pressure (CPAP) on liver enzymes (used as surrogates for the intrinsic function of the liver in NAFLD), the majority of the studies were confounded, non-randomized and small [14-17]. This case presents NAFLD associated with OSA syndrome, which reversed with optimal CPAP therapy over a prolonged duration of time.

Case Report

The patient was a 39-year-old Hispanic man with medical co-morbidities of morbid obesity, diabetes mellitus, hypertension, hypercholesterolemia and liver dysfunction. He was initially seen in our sleep clinic for symptoms of loud snoring, witnessed apnea, daytime fatigue and hypersomnolence. His Epworth sleepiness scale was 15. He underwent in-lab polysomnography (PSG) that revealed severe OSA with AHI of 50 and the lowest oxygen saturation detected during the study was 63%. During the same PSG, he was successfully titrated to CPAP of 16 cm H2O. Subsequently, his adherence to CPAP therapy was assessed at regular intervals demonstrating excellent adherence. An example of his compliance data for the last year is shown in Table 1.

Table 1

Patient’s Compliance Data for the Last 1 Year

Days used in the 365 days	365 days
PAP device	Automatic positive airway pressure
PAP pressures	10 - 18 cm
Average number of hours used per day	7 h 32 min
Average leak	3.7 L/M
Residual AHI	0.7

His family history was unremarkable for any liver or sleep disorders. He denied toxic habits including consumption of tobacco, alcohol or any other recreational drugs. He did not receive any blood transfusions. The abdominal girth, as well as BMI, remained unchanged in the range of 39.7 - 40.6 kg/m2 for more than 10 years. He felt that his overall health and exercise tolerance have improved after starting CPAP therapy.

Physical examination remained unremarkable except for morbid obesity with BMI 40 kg/m2. Initial laboratory tests performed between September 2011 and June 2012 are illustrated in Table 2. Figure 1 shows the trends of liver enzymes and lipid panel.

Table 2

Laboratory Values

Hematology	Value	Biochemistry	Value
White cell count (k/µL)	9.2	Creatinine (mg/dL)	0.8
Neutrophils (%)	58.7	Blood urea nitrogen (mg/dL)	11
Lymphocyte (%)	32	Serology
Red blood cell (mil/µL)	5.51	ANA	Negative
Hemoglobin (g/dL)	17.2	AMA	Negative
Hematocrit (%)	51.6	Hepatitis B surface antigen	Negative
Platelets (k/µL)	205	Hepatitis B core antibody	Negative
		Hepatitis C virus antibody	Negative
Activated partial thromboplastin time (s)	38.7	Thyroid function
Prothrombin time (s)	10.9	TSH (mIU/L)	1.02
INR	1.0	FT3 (ng/dL)	166
Ferritin (ng/dL)	587	FT4 (ng/dL)	1.25

Figure 1

Trends of liver enzymes and lipid panel. AST, aspartate transaminase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; TGs, triglycerides; LDL, low-density lipoprotein.

Ultrasonography of the liver showed deep-echo attenuation with hepatomegaly, highly suggestive of fatty infiltration of the liver. Computed tomography (CT) of the abdomen concurred with these findings. Following this, he underwent CT-guided liver biopsy that demonstrated moderate macrovesicular steatosis with inflammation, hepatocyte ballooning degeneration and portal fibrosis. On the basis of this, he was diagnosed with NASH stage 2.

After approximately 1 year of CPAP therapy, liver enzyme levels had decreased. Finally, at 6 years liver enzymes and relevant lipid panel normalized as graphically represented in Figure 1. Serum gamma-glutamyl transferase levels decreased from 68 IU/L to normal levels over 6-year period. Other pertinent labs including that of HbA1c and HDL also improved and normalized. His lifestyle remained unchanged since he was diagnosed with OSA. There were no additional medications used for the treatment of NASH in this patient. Our case stands out for its prolonged CPAP therapy follow-up with the evidence of optimal compliance. As such, it can potentially reverse NASH, which is contradictory to what was found in few available randomized control trials (RCTs), which were characterized by small size, poor design and short duration of therapy [14-18].

Discussion

The liver receives its blood supply from the highly oxygenated hepatic artery and a poorly oxygenated portal vein, thus causing an oxygen gradient in the hepatic lobule. The partial pressure of oxygen in the periportal blood is 60 - 65 mm Hg and that in the peri-central venular blood is 30 - 35 mm Hg. This makes liver sensitive to hypoxia and different parts of the hepatic lobule respond differently to hypoxia. Periportal hepatocytes play a predominant role in oxidative energy metabolism, glucose production, and urea and bile biosynthesis. Peri-central venular hepatocytes take up roles in glucose uptake and storage as glycogen, glutamine formation and xenobiotic metabolism [19, 20]. Animal studies have demonstrated that hypoxia alone aggravates and accelerates the progression of NASH by up-regulating the expression of lipogenic genes, by down-regulating genes involved in lipid metabolism and by decreasing insulin sensitivity [21]. Experiments on mice revealed liver to be most sensitive for acute intermittent hypoxia causing reciprocal oxygen oscillations when compared to other tissues. Acute intermittent hypoxia increases the oxidative stress and adipose inflammation irrespective of body fat content. Animal and human studies showed that obesity is associated with chronic liver hypoxia, even in the presence of systemic normoxia causing inflammation and release of cytokines [22]. On this basis, it is speculated that OSA syndrome might synergize with obesity to accelerate the process of hepatic dysfunction and death, aggravating liver and metabolic disease. Nocturnal hypoxia has been linked to decreasing adiponectin level that is implicated in cardiovascular diseases and OSA syndrome [23, 24]. Hypoxia-induced release of cytokines and reactive oxygen species increases the release of nuclear factor kappa B (NF-κB), which is implicated in OSA syndrome. CIH has been associated with endoplasmic reticulum stress that leads to accumulation of unfolded proteins at the molecular level [25].

In recent years, many observational human studies evaluating the relationship of OSA syndrome and NAFLD have been published. Heterogeneity in terms of definitions of NAFLD, as well as cutoff points for acute intermittent hypoxia to characterize OSA syndrome among these observational studies, limits its clinical applicability. Obesity and lack of liver biopsies among all subjects were the biggest confounding factors among most of these studies preventing us from drawing independent association of OSA syndrome and NAFLD [12, 26-32]. Tatsumi et al evaluated 83 non-obese patients and concluded that oxygen desaturation during sleep is a risk factor for progressive liver disease from the accumulation of fat. They, however, employed serum type III procollagen as a marker of liver fibrosis. Another study used liver ultrasound for evaluation of NAFLD among 106 patients with severe OSA syndrome and found NAFLD in two-thirds of patients, and among them, 70% had severe OSA syndrome [27]. Polotsky et al further showed that nocturnal oxygen desaturation contributed to insulin resistance and to different histological features with varying degrees of liver injury, including hepatocyte ballooning, inflammation and fibrosis, but failed to show hepatic steatosis per se. On multivariate analysis, AHI, oxygen desaturation index (ODI), lowest desaturation values and percentage of sleep duration with SpO2 < 90% independently predicted NAFLD after adjustment for BMI, weight and insulin resistance. Furthermore, the best predictor of the severity of NAFLD was the duration of hypoxia during sleep [32]. Another observational study by Aron-Wisnewsky et al found that when ODI was used to stratify severity of OSA syndrome among 101 morbidly obese patients, the prevalence of NAFLD was 77%. Among patient with NAFLD, close to 90% had evidence of fibrosis on histology. In this study, ODI was strongly associated with fibrosis and NAFLD activity score (NAS) [13, 33].

Pathogenesis of NAFLD is poorly understood. A “two-hit” model has been proposed. The first hit is characterized by insulin resistance and excess hepatic lipid accumulation secondary to abnormal fatty acid metabolism. Oxidative stress and inflammation are thought to comprise the second hit [1]. RCTs and meta-analysis confirm that CPAP treatment for more than 3 months has shown to improve insulin resistance [14, 18, 34, 35]. Decrease in adiponectin level and increase in leptin level are associated with OSA syndrome. Visceral fat even in non-obese patients with OSA syndrome is associated with insulin resistance. There is some evidence to suggest that their levels are somewhat corrected with prolonged CPAP therapy [36-40]. CIH, which is the distinctive feature of OSA syndrome, is proposed to stimulate peripheral lipolysis and inhibit lipoprotein clearance [1]. Lin et al reported a meta-regression analysis with almost 2,000 patients and showed that CPAP therapy improved cholesterol and triglyceride levels [41]. Another meta-analysis with only RCTs also showed similar results [42].

CIH triggers oxidative stress that leads to increased release of radical oxygen species like superoxide. There are other pro-inflammatory markers like C-reactive protein, interleukin-8 and tumor necrosis factor alpha. CPAP therapy has shown to decrease these levels [43-46]. It has been suggested that, because all these studies did not exclusively include NAFLD patients, CPAP therapy is potentially a preventive measure for liver disease among patients with normal liver function at baseline [1].

NAFLD is characterized by a spectrum of histological findings from simple steatosis to steatosis with necroinflammation (also called NASH). NAFLD can be diagnosed by imaging studies such as ultrasound, CT or magnetic resonance imaging [47]. Prevalence of NASH ranges from 5% to 33% in the general population to 90% among the morbidly obese [48, 49]. Historically, the gold standard for the diagnosis of NASH is a liver biopsy. Many clinical scores like NAS and SAF (steatosis, activity and fibrosis) are used for the diagnosis of NAFLD [50, 51]. Non-invasive tools like fatty liver index, serum algorithm and elastography for diagnosis of NASH have been recently validated [51-56]. The clinical significance of NASH is that it is associated with varying degrees of fibrosis, extremes of which can progress to cirrhosis and hepatocellular carcinoma [57, 58]. In our patient, NASH was diagnosed by liver biopsy.

OSA syndrome must be suspected in patients with excessive daytime sleepiness, fatigue, non-refreshing sleep, nocturia, morning headache, irritability and memory loss [59-61]. Only about 50% of patient with OSA syndrome are obese. Up to 25% of patients with sleep-disordered breathing present neither objective or subjective symptoms of OSA syndrome [62, 63]. Similarly, patients with NAFLD are largely asymptomatic with the common presentation of incidental abnormal liver function tests. They are, however, considered to be part of the metabolic syndrome with obesity and insulin resistance. They are also associated with diabetes mellitus and hypertension [64].

Clinical features like neck circumference, Mallampati score, use of questionnaires like Epworth sleepiness scale, Berlin questionnaire, STOPBANG (snoring, tiredness, observed apnea, hypertension, BMI, age, neck circumference and gender), etc., have shown 80-90% sensitivity but only 35% specificity. The gold standard for diagnosis of OSA syndrome is in-lab PSG; however, in the recent past, more evidence suggests that less expensive and less elaborate home sleep testing may be just as good in certain situations [65-67]. CPAP is the first choice of treatment across all grades of severity of OSA syndrome and is recommended to be offered to all patients [68].

There are no FDA approved drugs for NASH. Conservative management with lifestyle modifications including diet, exercise and weight loss remains the therapy of choice today. Many therapeutic trials have shown little success. Some of the drugs that have been tested include liver-directed therapy (vitamin E, pentoxyphylline, pioglitazone, etc.) for NAFLD [69-71]. Hence, in this current scenario, we are desperate to explore and find more definitive treatment. Treatment of OSA syndrome with CPAP has demonstrated a decrease in most of the metabolic risk factors [72-75]. CPAP treatment directed for OSA syndrome is thought to negate some of the downfalls of NAFLD. RCTs with sham controls to evaluate the effect of CPAP on liver function failed to normalize enzymes or improve liver fat quantified by radiology imaging and other non-invasive measures. However, the limitations of these studies included not being adequately powered, with variable cutoffs for AHI to define OSA syndrome, variable diagnostic modalities for NAFLD and duration of CPAP therapy were rather too short. The longest duration of CPAP therapy among these studies was 3 months [14-17]. Another randomized study that evaluated the effect of CPAP treatment on liver fibrosis, using the FibroMax score (non-invasive testing called biopredictive tests that work by using an algorithm based on the patients’ sex, age, weight, height and specific blood biomarkers) did not demonstrate any impact on reducing fibrosis; even though, it had more than 100 patients [16]. Contrary to the conclusions drawn from these randomized controlled trials, many observational and prospective cohort studies have associated adherence to CPAP therapy with improvement in liver enzymes and radiological steatosis [76-78]. Kim et al also concluded that there was a potential for reduction in liver fibrosis [79]. An important observation in all these studies was the duration of CPAP therapy which ranged from 1 to 3 years. A recent meta-analysis of 192 OSA syndrome patients showed that CPAP therapy for a minimum of 3 months was associated with statistically significant decrease in aspartate transaminase (AST) and alanine aminotransferase (ALT) [80]. One study by Chin et al with a sample size of 40 patients found that AST significantly decreased after single night CPAP therapy [81]. CPAP treatment significantly decreased the cumulative incidence as well as the overall risk for liver disease based on the large Taiwanese retrospective study with a follow-up duration of 10 years. This suggests that prolonged CPAP treatment plays a role in delaying progression and decreasing incidence of liver disease [82].

CPAP compliance was not adequately assessed in many RCTs looking into the event of CPAP therapy on NAFLD. This is one of the major limitations that inhibit any direct conclusion with respect to the effectiveness of CPAP therapy on resolution or reversal of NAFLD. CPAP compliance has varied definitions, for example, the Medicare criteria. In our practice, we find that the most beneficial outcomes are among patients with near maximum possible compliance. This, however, requires diligent and close long-term follow-up. Compliance among OSA syndrome patients in the real world has been shown to be around 20-50% [83-85]. Even by our practice standards, the definition of compliance used in these studies is suboptimal.

Unfortunately, the evidence to suggest positive outcomes of CPAP therapy among patients with liver disease has not been promising. Inadequate CPAP compliance and lack of long-term follow-up may account for the scarcity of evidence supporting positive outcomes of CPAP therapy among patients with NAFLD.

Our case has multiple important clinical implications. First, it is prudent to assume that OSA syndrome is associated with NAFLD. Although past studies have had contradicting results, they were flawed in design, non-randomized and small sized. More recent, better-designed studies have established a stronger link between the two. Second, a trial of treatment with CPAP in patients with OSA syndrome suspected to have NAFLD can help with diagnosis, treatment, follow-up and prognosis since definitive diagnosis of NASH requires a liver biopsy which is invasive and associated with complications. Third, treatment with CPAP must continue for a prolonged period of time as there is evidence to suggest that the beneficial effects of CPAP in liver disease need treatment adherence for a minimum of 3 months. Fourth, we recommend that every patient diagnosed with NAFLD should be screened for the presence of OSA. Fifth, we suggest the need for more RCTs assessing progression and potential reversal of NASH based on optimal compliance with CPAP therapy over significantly longer periods, as suggested by our case. Finally, CPAP treatments among patients who have OSA syndrome may be a preventive measure for NAFLD.