Use of cyanoacrylate glue for persistent post-dural puncture cerebrospinal fluid leak.

Sir,

We report a case of post-dural puncture cerebrospinal fluid (CSF) leak, managed by liquid tissue adhesive.

A nine year-old girl (12 kg) with a history of multiple laparotomies presented for surgical repair of vesico-cutaneous fistula and ileostomy closure. The anaesthetic plan was general anaesthesia (GA) with epidural analgesia for post-operative pain relief. Epidural catheter insertion was attempted under GA at T8-T9 inter-space with 18-G Tuohy needle. There was an accidental dural puncture at a depth of 2 cm and consequently epidural catheterization was abandoned. The puncture site was covered with sterile dressing and the surgery completed under balanced GA with intravenous opioids for peri-operative analgesia. Post-operatively, on day 3 clear fluid leaking from epidural puncture site was noticed during dressing change. Laboratory analysis of the fluid showed composition consistent with CSF with no microbial growth. There were no clinical signs or symptoms such as lower back pain, lower extremity weakness or pain, decreased lower extremity reflexes, inability to stand, headache, dizziness, vision problems or seizures. The rate of leak of CSF was approximately one drop every 3 or 4 seconds and the leak increased on crying. The leak did not respond to pressure dressing and bed rest. On the 5th post-operative day, the patient was posted for re exploration for anastomotic bowel leak under GA. We applied a liquid tissue adhesive (Dermabond; Ethicon, Inc., Somerville, NJ, USA) (2-octylcyanoacrylate) aseptically on the epidural skin puncture wound. The glue was allowed to polymerize and a dressing applied. The CSF leak ceased after its application. Unfortunately, 8 days later the child had a second bout of sepsis with bowel leak and underwent another re-exploration under GA. She succumbed to refractory septic shock 3 days after this surgery.

Persistent CSF leak from epidural insertion site is uncommon in children even though epidural collections have been reported after diagnostic lumbar puncture.[1] In our case probably a large dural defect, infection and nutritional deficits contributed to poor healing of epidural puncture site leading to CSF fistula. The epidural fat has less fibrous stroma in children compared to adults, facilitating the tracking of CSF along epidural space.[2] Patients frequently recover spontaneously with bed rest, hydration, analgesics and psychological support. Epidural blood patch is the next option but in our case it was negated due to bacteraemia and sepsis. The rationale for skin closure is to reduce the risk of meningitis by stopping the leak.[3]

Cyanoacrylate monomers are sterile liquid tissue adhesives for topical application to hold close easily approximated skin edges from surgical incisions and simple, clean lacerations. They have been used safely and successfully for closure of intractable entero-cutaneous fistulae[4] and post-operative CSF leaks in children.[5] The post-dural puncture CSF leak should be slow enough to allow the skin to dry for bonding and small enough not to require invasive surgical measures. The application of the adhesives at the CSF leak site can be rapidly and painlessly done at the bedside, using sterile precautions. The technique obviates the need for anaesthesia, sedation and an operating room. Allergy, infected wounds and elevated CSF pressure would be relative contraindications for the technique. If the CSF fistula remains unresolved, neurosurgical referral is required to consider surgical closure.

In conclusion, we propose that tissue adhesives can be used for persistent CSF leaks caused by iatrogenic needle punctures, unresponsive to conservative management.