Peritoneal Sarcoidosis

Case Report

A 61-year-old African-American man with a medical history of well-controlled diabetes, pulmonary sarcoidosis, and a 2-year history of intractable hiccups was referred to our hepatology clinic for evaluation of new-onset ascites. A review of his medical records revealed multiple complications and extensive work-up of his intractable hiccups. The hiccups began acutely 2 years prior without an apparent trigger. They had been occurring approximately 3 times per minute, with occasional pauses of up to several hours, and were interfering with eating and sleeping. Vomiting triggered by the hiccups resulted in a 50-lb weight loss, as well as the development of new erosive esophagitis that resolved with proton pump inhibitor therapy. Other complications included recurrent aspiration with resultant lung abscesses, for which the patient was hospitalized repeatedly.

Initial work-up of the patient’s hiccups included a high-resolution chest computed tomography scan, which showed mild mediastinal lymphadenopathy as well as scattered pulmonary, hepatic, and splenic lesions consistent with his known sarcoidosis. Brain magnetic resonance imaging was normal.

Exhaustive efforts at medical therapy of the patient’s hiccups included chlorpromazine, prochlorperazine, gabapentin, phenytoin, baclofen, promethazine, metoclopramide, codeine, fentanyl, duloxetine (Cymbalta, Eli Lilly), mirtazapine, citalopram, alprazolam, omeprazole, rabeprazole, and prednisone, in addition to hypnosis and homemade remedies, all of which produced only transient relief or were completely ineffective. Given the patient’s continued severe symptoms, a unilateral phrenic nerve resection was performed, again without lasting relief of symptoms.

In the hepatology clinic, attention was turned to the patient’s new ascites. The ascites had developed rapidly over the previous 4 weeks without obvious precipitants. There was no abdominal pain and no other associated symptoms. The patient had never smoked but admitted to a long history of alcohol abuse before stopping 4 years prior. There was no family history of liver disease or of recurrent hiccups. Physical examination was notable for evidence of a large volume of ascites without appreciable hepatomegaly. Cardiac examination was normal, and there was no edema.

Laboratory work-up revealed an aspartate aminotransferase level of 32 U/L, alanine aminotransferase level of 32 U/L, alkaline phosphatase level of 124 U/L, total bilirubin level of 0.5 mg/dL, albumin level of 3.2 g/dL, platelet count of 333,000 per microliter, and international normalized ratio of 1.2. In addition, alpha-fetoprotein was undetectable. Iron panel and brain natriuretic peptide levels were normal, and HIV and viral hepatitis serologies were negative.

A paracentesis was performed and returned a clear, slightly bloody fluid. There were 27,000 red blood cells per mm³ and 279 white blood cells per mm³, 48% of which were polymorphonuclear. Serum-to-ascites albumin gradient was 0.9 g/dL, with an ascitic fluid protein of 4.4 g/dL. Repeated ascitic fluid cytologic examinations were unrevealing. A liver biopsy was performed and revealed confluent, noncaseating portal granulomas with focal nonbridging fibrosis, which is consistent with hepatic sarcoidosis (Figure 1). At the time of the biopsy, free hepatic venous pressure was 13 mmHg and wedged hepatic venous pressure was 16 mmHg, making transhepatic venous pressure normal at 3 mmHg. An echocardiogram was also normal.

Figure 1.

Liver biopsy showing confluent noncaseating portal granulomas.

Out of concern for possible malignancy, an exploratory laparoscopy was performed. This examination revealed multiple small light-yellow glistening nodules and a thin fibrinous plaque on the peritoneum (Figure 2). Pathologic examination revealed noncaseating histiocytic granulomas with fibrosis, which is consistent with sarcoidosis (Figure 3). Fungal and acid-fast bacilli cultures were negative.

Figure 2.

Laparoscopic image of glistening nodules on the peritoneal surface.

Figure 3.

Biopsy of peritoneal nodules showing noncaseating histiocytic granulomas.

Following the exploratory laparoscopy, the patient was placed on prednisone therapy for 2 months. His ascites improved, and at 4 months postprocedure, he had no appreciable ascites. However, his hiccups continue unabated.

discussion

Sarcoidosis is a chronic granulomatous disease of unknown cause and is most common in young adults, particularly African Americans.¹ Most patients have pulmonary involvement producing cough, dyspnea, and, eventually, restrictive lung disease in severe cases. The heart, skin, eyes, and nerves may also be affected.

Hepatic involvement in sarcoidosis is common, with approximately two thirds of patients having liver involvement on autopsy.^2,3 When hepatic disease is present, granulomas tend to be portal or periportal, sometimes producing presinusoidal portal hypertension. Inflammation surrounding the granulomas may produce periportal inflammation, which can rapidly progress to cirrhosis.⁴ Although the patient in this case had portal granulomas and early fibrosis, the high ascitic protein and low serumascites albumin gradient suggested that portal hypertension was not the cause of his ascites. This was confirmed by a normal transhepatic venous pressure gradient.

In contrast to the common finding of hepatic involvement, peritoneal sarcoidosis is quite rare, with an estimated 19 cases reported in the English literature.⁵ The most common presentation of peritoneal sarcoidosis is new-onset ascites. The serosal surfaces may be diffusely studded with granulomas, or granulomas may form around surgical scars.⁶ The combination of high-protein ascites and peritoneal nodules may be mistaken for a fungal, tubercular, or malignant cause.⁵ Most cases of ascites due to peritoneal sarcoidosis appear to resolve within several months; it is not clear whether steroid use may accelerate this resolution.⁷

Intractable hiccups (singultus) can be a vexing clinical problem. A broad array of causes has been reported and nearly as many treatments have been proposed, yet our pathologic understanding of hiccups remains minimal. One theory posits that there exists a hiccup reflex arc with a hiccup center in the brain stem near the respiratory centers.⁸ In addition to other pathways, the phrenic and vagus nerves serve as both afferent and efferent components of this hiccup reflex arc.

Dozens of causes of intractable hiccups have been reported. Causes such as central nervous system lesions, infections, and metabolic disturbances can be seen as relating to the disturbance of the proposed hiccup center. Stimulation of the phrenic and vagus nerves is a more common situation in the gastroenterology patient. Esophagitis, peptic ulcer disease, pancreatitis, inflammatory bowel disease, hiatal hernia, and gastric distention are some of the reported triggers of hiccups.⁹ We think that this patient’s peritoneal lesions were likely the cause of his hiccups.

A similarly wide variety of treatments for intractable hiccups has been proposed. Homemade remedies are diverse and well known, and pharmacologic treatment may involve antipsychotics, anticonvulsants, muscle relaxants, and antidepressants, among others.⁹ Gastroesophageal reflux should be treated if present and is also a common complication of intractable hiccups. Ablation of the phrenic nerve in hopes of interrupting the hiccup reflex arc has been reported, but was unsuccessful in our patient.¹⁰

The case described illustrates two unusual manifestations of sarcoidosis. Peritoneal sarcoidosis should be added to the long list of possible causes of intractable hiccups. Although rarely reported, peritoneal sarcoidosis also must be included in the differential diagnosis of high-protein ascites.