31.1 Introduction to Lumbar Drains

The pediatric population presents many challenges to the surgeon applying the endoscopic endonasal approach (EEA) for skull base tumor resection. One of the most difficult of these challenges is the repair of skull base defects, as many of these patients are in various phases of skull base and sinonasal development. Many times, defects created by surgery result in high-flow cerebrospinal fluid (CSF) leaks, and repair of these may be limited by pediatric anatomy. Failure of a successful closure may result in pneumocephalus, postoperative CSF leak, meningitis, and other morbidities. To this end, postoperative CSF diversion in the form of lumbar drains may be used in the postoperative period to provide CSF diversion and prevent postoperative CSF leaks. ^{1 ,​ 2} Lumbar drains have been used in the adult population with success; however, this is rarely addressed in the pediatric literature. In addition, CSF-leak risk appears to be higher in pediatric patients compared to adult patients with similar tumors. In our series (unpublished data), we found that the postoperative CSF leak for chordoma was 40% for pediatrics compared with 22% for adults.

The pediatric population, in addition to the challenges presented from the surgical anatomy, presents challenges associated with the use of lumbar drains. The adult produces approximately 500 mL of CSF daily, whereas children (ages 4–13 years) produce only 65 to 150 mL, ³ making small children particularly at risk for overdrainage complications. A smaller body habitus may increase risk of overpenetration of the stylet and needle, causing bleeding from the ventral venous plexus or even penetration of the retroperitoneal cavity. A useful recommendation for needle depth in the pediatric population may be estimated by the following formula: depth of LP = 0.77 cm + (2.56 × body surface area [BSA; m²]). ⁴ Postimplantation challenges exist as well. The pediatric population may be at greater risk of drain dislodgement and noncompliance resulting in over- or underdrainage. To counter this, a well-educated nursing staff is critical for successful postoperative drain care. These are in addition to the known risks of lumbar drain placement, which include postural headaches, nausea, meningitis, tonsillar herniation, intracranial hypotension, and retained catheters. ^{1 ,​ 5 ,​ 6 ,​ 7}

Despite these challenges, several studies have documented successful use of lumbar drains in the pediatric population for a variety of reasons, including fulminant idiopathic intracranial hypertension and traumatic and postoperative CSF leaks. ⁸ Levy et al describe the successful use of lumbar drains for the diversion of CSF in 16 pediatric severe traumatic brain injury patients without complications associated with lumbar drain placement or care. ⁹ Lumbar drainage after surgical intervention is less well documented. Lumbar drainage has been used after operative repair of a traumatic CSF leak. ¹⁰ Zhan published a series of 11 patients who underwent EEA for sellar region pathology and noted the use of a lumbar drain in one patient who developed a postoperative CSF leak. No complications were listed. ¹¹

Di Rocco et al published a series of 28 pediatric patients with anterior skull defects repaired through an endoscopic approach. In patients with a significant preoperative CSF leak, lumbar drains were placed (n = 5). Also, a lumbar drain was placed in one patient who developed a postoperative CSF leak. The drains were in place for 3 to 8 days. They found that lumbar drainage increased hospital stay, constrained the patients, and increased risk of drain-related complications. They recommend drains be placed in cases of abundant CSF leak or in early recurrence of CSF leakage, but they could not draw any definitive conclusions. ¹² The largest series of EEA for pediatric patients included 133 patients and noted that a perioperative lumbar drain was placed in 25 patients, but no complications of drain management was mentioned. ¹³ This series was later revisited and evaluated for risk factors for CSF leak in this population. Drains were not associated with postoperative CSF leak or lack thereof in the series; however, this is likely a reflection of the fact that lumbar drains were only placed in cases with high-flow intraoperative leaks. ¹⁴ This lends support to the usage of drains in this setting, as it seems to lower the risk of lower-flow leaks without drains.

Our clinical practice is to use a lumbar drain in cases of high-flow CSF leak created during EEA (▶ Fig. 31.1). These drains are then kept in place for 72 hours and removed to avoid increased risk of drain-associated meningitis, which can increase with prolonged drainage. Typically, 5 to 10 mL of CSF is removed per hour depending on patient size (similar in adults). Although no association between CSF leak and lumbar drain was discovered during the aforementioned analysis (suggesting appropriate usage of lumbar drains), a recent randomized trial was concluded, demonstrating clear lumbar drain utility in adults.¹⁷ Lumbar drains in this adult population significantly decreased CSF leaks after endoscopic endonasal surgery for pathology in the anterior or posterior cranial fossa. This is likely a result of defect size and was independent of patient BMI or surgical history.

The use of a lumbar drain when a patient develops a postoperative leak is much more controversial. In general, our philosophy has been to NOT use lumbar drainage as the primary modality in this setting. Rather, if postoperative CSF leak is suspected or proven, early re-exploration is recommended. Several studies have shown a success rate of lumbar drainage of 50% as the primary treatment of postoperative leak. ^{15 ,​ 16} While this may avoid re-exploration in a few patients, leak is never confirmed, and the patients in whom it fails end up with significantly prolonged courses and increased risk of meningitis and subsequent complications that are entirely avoidable.

Contraindications to placement of lumbar drains also exist and are important to recognize. Patients with large residual posterior fossa lesions or small posterior fossa cavities are at particular risk for downward herniation after lumbar drain placement. Patients with pneumocephalus may also suffer from overdrainage/herniation with lumbar drain placement. For these reasons, it is important to get postoperative imaging prior to opening of a lumbar drain. Patients with myelomeningocele/meningoceles may not have a lumbar cistern, and a lumbar drain would be contraindicated.

The decision to place a lumbar drain after endoscopic endonasal surgery is largely up to the individual surgeon and defect created after tumor removal. There is some class 1 evidence for the usage of drains in large anterior or posterior fossa defects in adults, but this can only be extrapolated to children. However, given that CSF leaks appear to be more prevalent in the pediatric population, lumbar drain usage in the setting of large or high-flow defects seems prudent.