Cerebrospinal fluid cutaneous fistula after neuraxial anesthesia: an effective treatment approach

1 Introduction

Neuraxial anaesthesia and analgesia is a common anaesthetic choice for inferior limb orthopaedic procedures, providing good surgical conditions and postoperative pain control with a low incidence of complications.

Cerebrospinal fluid (CSF) cutaneous fistula is a rare but potentially serious complication of epidural analgesia [1], [2], [3]. It consists of the abnormal output of CSF from the subarachnoid space through the epidural insertion site, usually after epidural catheter removal [4]. It can lead to symptoms of intracranial hypotension, such as postural headache and associated symptoms (nausea, vomiting, photophobia or phonophobia) or severe complications such as meningitis or subdural haematoma [3], [5]. Therefore, early recognition and management are crucial.

There are few cases described in the literature, and there are no specific recommendations as to its diagnosis and treatment. We present a case of a CSF cutaneous fistula following epidural analgesia in a 77-year-old man that was successfully treated with a simple suture at the epidural site.

2 Case report

A 77-year-old male, with history of hypertension, type 2 diabetes and prostate cancer, was proposed for total right knee arthroplasty.

The anaesthetic plan was a combined neuraxial blockade (subarachnoid and lumbar epidural), two separate needles technique. It was performed at the L4-L5 level, midline approach and on the first attempt without complication. The epidural space was found using an 18G needle at 5 cm using loss of resistance with air, and the catheter was inserted 4 cm in cephalic orientation; the epidural catheter was used for post-operative pain management. For the subarachnoid block, a 27G pencil tip needle was used and 9 mg of isobaric bupivacaine 5 mg/mL and sufentanil 0.002 mg were administered. The surgery was completed uneventfully.

Adequate post-operative pain control was provided by an epidural infusion ofropivacaine. Anticoagulation was initiated with enoxaparin 40 mg lid for venous thromboembolism prophylaxis.

On the fourth post-operative day, the epidural catheter was found exteriorised. The Acute Pain Unit evaluated a moderate amount of fluid leakage from the epidural catheter insertion site, which required frequent dressing changes. This fluid was collected and the glucose level determined, 143 mg/dL, was consistent with the diagnosis of CSF cutaneous fistula. A sample was sent for cytological analysis to confirm the diagnosis. The patient had no symptoms consistent with CSF hypotension, and he denied headache, photophobia, nausea and vomiting; there were no neurological signs, and he remained afebrile. The initial therapeutic approach consisted of bed rest, hydration and prophylactic antibiotic therapy with vancomycin, and the patient was closely monitored for symptoms of liquor hypotension and meningeal signs.

On the sixth post-operative day, as fluid drainage persisted, it was decided to close the fistula with a simple suture at the epidural insertion site. After skin closure, there was no more fluid drainage, and the dressing remained clean and dry. The patient initiated progressive intermittent uprisings to 30–90° that were well tolerated. Five days later, the stitches were removed and no drainage was observed. The patient remained asymptomatic and was discharged uneventfully.

3 Discussion

There is a lack of information about CSF cutaneous fistula related to anaesthetic procedures and its aetiology is not clear [4]. Certain risk factors have been described: several attempts to locate the epidural space; accidental puncture during insertion of the catheter; spinal and epidural combined approach; the use of systemic or epidural steroids that have a known suppressive effect and delay the healing of the tract [3], [4], [5], [6], [7]. In our case, a combined spinal and epidural block was performed.

When fluid drainage is observed through the epidural insertion site, the first step should be making a differential diagnosis between CSF and interstitial fluid. There are two main diagnostic methods: measurement of glucose and protein concentrations and beta-2-transferrin electrophoresis [6]. The presence of glucose and low protein levels suggests CSF [1]. This is a quantitative colorimetric test, which is easy and quick to perform but with low specificity. Beta-2-transferrin electrophoresis is used to confirm the diagnosis, which has high sensitivity and specificity for CSF. Beta-2-transferrin is an iron transport protein that can be found in CSF, perilymph or aqueous humour [6]. However, this is a time consuming and expensive method; therefore, it is not always requested.

There is no standardised therapeutic approach for CSF cutaneous fistula, and the selected treatments vary. Initial treatment is conservative: bed rest (which reduces the loss of CSF and intracranial hypotension), fluids, analgesia and antibiotic prophylaxis. Some authors argue against the use of antibiotics in patients with no signs of infection [5]. When these measures are not enough and the fluid leak persists, other options are used, such as cutaneous suture at the epidural site or blood patch [4]. A cutaneous stitch is a simple, easily feasible option that leads to fluid accumulation and creates a tamponade effect, promoting the closure of the tract [6], [7]. Epidural blood patch consists of injecting autologous blood into the epidural space near the CSF leak, which forms a blood clot and closes the leak. However, this is a procedure with possible risks, and it cannot be used in patients on anticoagulation [8]. Both techniques reduce the risk of meningitis [4]. When central nervous system infections are present, the cutaneous suture technique would also minimise the risks of more invasive approaches, such as blood patch or surgery.

In this case, the patient was asymptomatic and no neurological signs were present. An initial conservative approach was chosen. Given the persisting fluid leaking, we performed a cutaneous suture at the epidural puncture site that provided successful closure of the CSF cutaneous fistula.

4 Conclusion

CSF cutaneous fistula is a rare but potentially life-threatening complication of the epidural technique. However, there are a lack of recommendations for its management. An initial conservative approach is consensual, but when it fails, there are no standardised guidelines and different treatments are proposed.

In this case, the combination of conservative measures and cutaneous suture of the epidural insertion site proved to be an effective treatment approach to this entity.