Sterile cerebrospinal fluid ascites, hydrothorax and hydrocele as a complication of ventriculoperitoneal shunting in an elderly patient

Abstract

An 89-year-old man with a history of multiple abdominal surgeries and ventriculoperitoneal (VP) shunt placement for normal pressure hydrocephalus presented for intractable abdominal bloating and scrotal swelling, for which imaging revealed massive ascites, bilateral hydrocele and small bilateral pleural effusions. Cardiac, hepatic and renal workup were insignificant. Culture and cytology of ascitic fluid were negative for infection or malignancy. Aetiology of the ascites as secondary to Cerebrospinal fluid (CSF) from the VP shunt was confirmed via ligation of the shunt. Sterile CSF ascites, hydrothorax and hydrocele are rare complications of VP shunt for hydrocephalus and are mostly presented in paediatric patients. We report the first known case of concurrent CSF ascites, hydrothorax and hydrocele in an elderly patient. We examine the difficulty of shunt replacement as a diagnostic and treatment modality in this age group and propose the use of reversible shunt ligation as a diagnostic modality.

Background

A ventriculoperitoneal (VP) shunt is the most frequently used treatment for hydrocephalus. Approximately 20 947 shunts were inserted between 2004 and 2013 in the UK and Ireland.¹ Despite the frequency of VP shunt placements, complications are relatively common, occurring in 11%–25% of patients, with shunt obstruction and infection being the most common causes of shunt failure.² Sterile cerebrospinal fluid (CSF) ascites is a rare complication of VP shunt for which a literature review revealed 65 cases in English. Of these cases, the vast majority are of paediatric cases, with only one known reported case of an elderly patient above the age of 65.³ We present a case of sterile CSF Ascites in an elderly man complicated by bilateral pleural effusion and bilateral hydrocele which was diagnosed by ligation of VP shunt.

Case presentation

An 89-year-old patient presented with abdominal distension for the past year. He has a history of gastritis, duodenal erosion, percutaneous endoscopic gastrostomy (PEG) tube placement for oesophageal dysmotility and normal pressure hydrocephalus for which a VP shunt was placed in 2009. He underwent PEG tube placement in 2017 for oesophageal dysmotility. A CT of his abdomen completed a few months after this operation showed a small amount of ascites proximal to his right VP shunt and a right pleural effusion. He underwent a midline incisional hernia repair in 2019 and was readmitted to an external hospital 3 months after for abdominal and scrotal bloating. At that time, he was found to have large abdominal ascites, small bilateral hydrocele and bilateral pleural effusions. Given the patient’s history of torsemide for chronic lymphedema in his legs, he was discharged on an increased dose of torsemide. The patient presented to our hospital 6 months after with reports of worsening abdominal bloating and bilateral scrotal swelling, despite his increased dose of torsemide.

Investigations

CT of the abdomen and pelvis showed massive ascites (figure 1). Ultrasound of his scrotum showed moderate bilateral hydroceles (figure 2). Chest X-ray showed small bilateral layering pleural effusions (figure 3). Paracentesis was completed, removing 2.7 L of clear yellow ascitic fluid with a total protein of 2.5 g/dL, albumin of 1.5 g/dL, glucose of 155 mg/dL and a serum-ascites albumin gradient of 1.3. His blood albumin level was 2.8 g/dL. The normal reference range for blood albumin is 3.5–5.2 g/dL, but it would be expected to be lower for elderly patients. No leucocytes or organisms were isolated from culture and cytology did not show malignancy. A Doppler liver showed no evidence of portal or hepatic thrombosis. CT and ultrasound of the abdomen did not show any signs of liver cirrhosis or metastasis. An echocardiogram showed a left ventricle ejection fraction of 55% and grade 1 dysfunction. Urinalysis was negative for proteinuria, glucosuria or haematuria. The patient’s abdomen gradually reaccumulated fluid, and a repeat paracentesis was done 10 days later, drawing another 1.6 L of similar clear yellow fluid.

Figure 1.

Contrast-enhanced axial CT images of the abdomen. (A) Prior to ventriculoperitoneal shunt ligation demonstrates a large amount of ascites. (B) After reopening of the ventriculoperitoneal shunt demonstrates reaccumulation of the intra-abdominal free fluid. Red arrow: intra-abdominal free fluid. White arrow: the ventriculoperitoneal shunt catheter.

Figure 2.

Sonogram of the scrotum demonstrates small bilateral hydroceles.

Figure 3.

Chest radiographs. (A) Prior to ligation of ventriculoperitoneal (VP) shunt demonstrates bilateral layering pleural effusions (white arrows). (B) Postligation of the VP shunt demonstrates resolution of the bilateral pleural effusions. (C) Postreopening of the VP shunt demonstrates recurrence of small bilateral pleural effusions, larger on the left than the right (red arrows).

Differential diagnosis

The most common cause of ascites is liver cirrhosis in the setting of portal hypertension (accounting for 85% of ascites), which was ruled out in this patient given his lack of alcohol use, normal liver function enzymes and non-significant abdominal CT and ultrasound findings. Malignant ascites (accounting for 10% of cases) was ruled out given his negative cytology, negative signs of metastases on imaging and lack of suggestive symptoms such as fatigue or recent weight loss. Other causes including heart failure, nephrotic syndrome and infection were also ruled out with normal findings on echocardiogram, urinalysis and ascitic fluid culture.

Given his history of repeated abdominal surgeries and timing of the ascites accumulation around the time of those surgeries, CSF ascites due to peritoneal irritation was considered.

Treatment

Neurosurgery was consulted and determined there was no stable site for replacement of shunt given the patient’s age and comorbidities. A decision was made to ligate the shunt with the option of reversing the ligation should the patient develop neurological symptoms. The procedure was completed under local anaesthesia at the bedside. The shunt catheter was accessed inferior to the clavicle where it was easily palpated below the skin and was ligated via knots by two silk sutures. Shunt ligation would verify whether the patient’s ascites source was from the VP shunt and also act as a potential long-term minimally invasive treatment.

Outcome and follow-up

The patient’s follow-up abdominal ultrasound and chest X-ray 1 month after discharge showed resolution of the ascites and pleural effusion (figures 3 and 4). The patient also stated that his scrotal swelling had improved significantly. However, he noted he had worsening of his chronic headaches after shunt ligation. CT of his head did not reveal any changes (figure 5) but given the patient’s headache and his desire to reopen his shunt, his VP shunt was reopened. The patient was readmitted 2 months following reopening of his shunt for generalised weakness. A CT scan of his abdomen at the time revealed reaccumulation of a moderate amount of ascites and bilateral pleural effusions (figure 6). The patient stated at the time that his stomach did feel bloated, but it had not been uncomfortable enough that he would want a paracentesis completed. On follow-up 7 months after shunt reopening, the patient reinforced that while he had experienced chronic abdominal bloating, it had not bothered him enough to desire a repeat therapeutic paracentesis.

Figure 4.

Sonographic evaluation of the left lower quadrant. (A) Prior to ligation of the ventriculoperitoneal (VP) shunt demonstrates large amount of ascites adjacent to bowel loops. (B) Postligation of the VP shunt demonstrates resolution of the ascites. (C) Postreopening of the VP shunt demonstrates recurrence of amount of ascites. White arrows: free fluid. Red arrow: bowel loops.

Figure 5.

Brain CT without contrast. (A) Prior to ventriculoperitoneal (VP) shunt ligation. (B) Post-VP shunt ligation and (C) postreopening of the VP shunt demonstrates stable right sided VP shunt, ventriculomegaly and bilateral subdural hygromas.

Figure 6.

Contrast-enhanced axial CT images of the chest. (A) Prior to VP shunt ligation demonstrates small bilateral pleural effusions. (B) Postreopening of the VP shunt demonstrates development of moderate bilateral pleural effusions. White arrows: pleural effusions.

Discussion

CSF Ascites is a rare complication of VP shunt resulting from either overproduction of CSF or inability of the peritoneum to absorb CSF. Extensive production of CSF can result from choroid plexus hyperplasia.^4–7 Inability of peritoneal absorption could be secondary to peritoneal irritation secondary to history of surgery,^{8 9} allergy to shunt material^{10 11} or high CSF protein due to craniopharyngioma or optic pathway gliomas.^{12 13} Of the 65 reported cases of sterile CSF Ascites, only 10 were above the age of 18 and one above the age of 65. Our patient had a history of repeated abdominal surgeries coinciding around the time at which his ascites developed. This suggests an aetiology of peritoneal inflammation leading to inability of CSF absorption.

The accumulation of sterile CSF fluid in the pleural space or scrotum without migration of the VP shunt is even rarer in the literature with 14 reported cases of CSF hydrothorax and 16 cases of CSF hydrocele. Two cases of CSF hydrothorax were over the age of 65 and no cases of CSF hydrocele were identified in the elderly population. Causes of CSF hydrothorax without shunt migration were postulated by Hadzikaric et al to involve two factors: malabsorption of CSF by the peritoneum and open communication—including micro-communications—between the peritoneal and pleural cavities.¹⁴ CSF hydrocele without shunt displacement into the scrotum require a connection between the peritoneum and scrotum. These cases are mostly found in infants below the age of 2 due to patency of the processus vaginalis. Celik et al noted that 23.8% of children with VP shunt procedures develop clinical inguinal manifestations including hydrocele and hernia.¹⁵ Our patient had a history of bilateral inguinal hernia repair, thereby creating potential routes for CSF passage into the scrotum.

Given the rarity of CSF ascites, hydrothorax or hydrocele, a thorough workup to rule out cardiac, hepatic, renal and all other pertinent causes is necessary. The utility of beta-2 transferrin is unclear for determining the source of ascites. Due to the presence of the VP shunt tip in the peritoneum, analysis of intra-abdominal fluid is likely to be positive for beta-2 transferrin irrespective of the source of fluid build-up. In the case of hydrothorax, there are conflicting evidence on the role of beta-2 transferrin testing. Patel et al described the presence of false-negative beta-2 transferrin results in a case of CSF hydrothorax, while Lofrese et al demonstrated its role in determining CSF leakage into the pleural space.^{16 17} Due to our patient’s age and lack of pulmonary symptoms, a thoracentesis was not performed. Definitive diagnosis and treatment most commonly involve ventriculoatrial shunt, ventriculocholecystic shunt or ventriculopleural shunt. However, given the patient’s age and comorbidities, shunt replacement was not possible. In light of his stable hydrocephalus for more than 10 years, the shunt was ligated, which verified the VP shunt as the source of his fluid build-up. Ligation of nonadjustable VP shunts may act as a minimally invasive diagnostic and treatment tool for suspected sterile CSF ascites in the elderly population.

In this report, we present the first known reported case of concurrent CSF ascites, hydrothorax and hydrocele in a patient above the age of 65. We also described the first time use of reversible VP shunt ligation as a minimally invasive method of verifying CSF ascites.

Learning points.

• While most cerebrospinal fluid (CSF) ascites occur in paediatric patients, suspicion for CSF ascites should be raised in elderly patients with ventriculoperitoneal (VP) shunts and a history of peritoneal irritation.

• Given the elevated risk of surgical complications from VP shunt replacement in elderly patients and indetermined reliability of B2-transferrin, VP shunt ligation for nonadjustable VP shunts could be used as minimally invasive tool to confirm CSF Ascites.

• Concurrent irretractable pleural effusions and hydroceles are likely due to flow of CSF from the peritoneum into the pleural space and scrotum.

• There is an elevated risk of CSF hydrocele in elderly patients with histories of inguinal hernias.

Ethics statements

Patient consent for publication

Obtained.