Postoperative Early Lumbar Drainage Can Reduce the Duration of Fever or Infection in Patients with Complicated Intracranial Tumors after a Long Operation Time
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.355097
Source of Support: None, Conflict of Interest: None
Keywords: Complicated intracranial tumor surgery, complications, fever, infection, lumbar drainage
Yingxi Wu, Yunze Zhang and Yang Wu has contributed equally to the manuscript.
For conventional craniotomy, the average intracranial infection rate is approximately 5%,,,, and the rate of infection does not seem to be high. However, in some complicated or difficult skull base craniotomies in which the anterior clinoid process, orbital wall, petrous apex or occipital condyle need to be removed, this increases the operation time and the frequency of manipulation, which may increase the postoperative incidence of infection or fever. Additionally, complex skull base surgeries provide favorable conditions for the occurrence of intracranial fever or infection because of the opening of the frontal sinus, mastoid or internal auditory canal.,,,
Intracranial fever or infection caused by neurosurgical operations are always the difficulty and focus of treatment, which may increase the hospitalization expense and postoperative hospital length of stay (LOS), further aggravating the patient's condition and increasing morbidity and mortality.,, Although lumbar drainage (LD) has some potential risks, including intracranial hemorrhage, headache secondary to intracranial hypotension and radiculopathy,, it still plays an important role in the treatment of intracranial fever or infection, and its advantages have been elaborated.,,
In clinical practice, it has been observed that for some patients with delayed fever or infection due to aseptic inflammation or infection caused by long operation time or bone foam, hemostatic materials and repair materials after the third day after surgery, continuous lumbar puncture or LD is always needed.,, For some patients who receive complicated and difficult skull base craniotomies, to reduce brain contusion and encephalocele, we place LD tubes to release the cerebrospinal fluid (CSF) during the operation. For some of these patients with a longer operation time (dural opening time of more than four hours), the LD tube is not removed immediately after surgery, but early CSF drainage is performed by the preplaced LD tube. Whether such treatment decreases the incidence or duration of postoperative fever or infection and reduces the postoperative hospital stay is worth exploring.
To determine the effectiveness and safety of early LD in the treatment of postoperative fever or infection, we performed a retrospective cohort study by categorizing patients into two groups in which the dural opening time was more than four hours. The duration of postoperative fever or infection, hospital LOS, and complications were compared between the two groups.
All research protocols were approved by the institutional review board of Tangdu Hospital, and informed consent was obtained from each patient in this study. Patients with intracranial tumors undergoing conventional craniotomy at the Center of Neurosurgery of Tangdu Hospital (Xi'an, People's Republic of China) from January 2019 to December 2020 were included in this study. The inclusion criteria were as follows: (1) age 18–70 years; and (2) dural operation duration of more than four hours. Serious intracranial hypertension, hydrocephalus, heart disorder or pulmonary functional impairment were contraindications for LD. Patients with severe lumbar vertebrae hyperosteogeny or lumbar disc herniation were also excluded from the study. Ethics committee is obtained and the date of the approval is 02-June-2021.
A total of 142 patients were categorized into two groups: The LD group underwent preoperative lumbar drain placement, intraoperative CSF release and postoperative continuous drainage; the control group received routine craniotomy without preoperative LD. The placement of lumbar spinal drainage was completed by senior surgeons following the standardized protocol.
LD implementation and management
Patients received LD before the surgery in the operating room or general ward. Under local infiltration anesthesia, a lumbar puncture needle was placed into the lumbar three-four or four-five interspace, which varied among patients. When CSF outflowed, the drainage catheter was implanted into the subdural space, and the valve was immediately closed to prevent CSF release. Finally, the drainage catheter was tightly fixed onto the back.
Before the operation, the LD catheter was placed at a height of 10 cm above the level of the external auditory canal. The CSF was released to 40–60 ml after the removal of the bone flap during the operation. After the completion of the surgery, the LD catheter was not removed, and it was kept closed until it needed to be opened. Computed tomography (CT) skull was used for each patient 24 hours after surgery, and the stopcock could be opened for continuous CSF drainage if no hemorrhage or hematoma was found in the surgical field or other parts of the brain. The LD tube was placed at a safe height, and the CSF drainage was controlled at approximately 150 ml/day. Moreover, a stopcock was not allowed to open when the patient got out of bed. If a patient had a serious headache induced by intracranial hypotension, the LD catheter could temporarily be closed or raised to decrease CSF drainage. The LD catheter was replaced if it was fractured or had frequent CSF leaks at the puncture site to prevent retrograde infection of the lumbar subarachnoid. The lumbar catheter also needed to be replaced at the other lumbar space in patients who had fever, whose lumbar catheterization lasted more than ten days and whose intracranial infection was not controlled completely. If the patient's temperature and white blood cell (WBC) count in the CSF stopped decreasing and suddenly rose, the possibility of retrograde infection was considered, and continuous lumbar puncture was implemented instead of LD.
Diagnosis and treatment of intracranial fever or infection
All patients received plain and contrast-enhanced MRI on the third day after surgery. Liver and kidney function, routine blood count, electrolytes, C-reactive protein and calcitonin were tested by collecting venous blood. In addition, routine examination and biochemical detection of CSF were performed on alternative days; antibiotics (ceftriaxone sodium) were used for patients preoperatively and within 24 hours after surgery to prevent infection. For patients with a mastoid or frontal sinus opening or a skull base dural opening, it was essential to observe whether CSF rhinorrhea or otorrhea occurred postoperatively.
Intracranial infection was not generally diagnosed in patients with fever during the first three days; therefore, antibiotics were not used continuously in this interim, and physical cooling or antipyretic analgesic drugs were usually administered. We divided postoperative patients with intracranial fever into the following three categories:
In addition to routine LD on the second day, the LD group was given antibiotics for the last two cases from the fourth day after surgery, and the antibiotics were adjusted according to drug sensitivity. If the patient's temperature was greater than 37.4°C on the fourth day after the operation and the CSF test was abnormal by lumbar puncture in the control group, continuous lumbar puncture or LD was performed, and the use of antibiotics was the same as that of the LD group.
Empiric de-escalation antibiotic therapy was applied to intracranially suspected or confirmed infection in patients who were infused with vancomycin (1 g, Q12H) and meropenem (2 g, Q8H) intravenously., For patients with fever on the third day after the operation, the indications for removal of LD tubes or for stopping lumbar puncture included WBC count <300 × 109/L, glucose >2.2 mmol/L by CSF testing and a normal temperature for three consecutive days in both groups. However, for patients who had no fever for two consecutive days after the third postoperative day in the LD group, the LD tube could be removed following WBC count <300 × 109/L and glucose >2.2 mmol/L by the CSF test.
Patient characteristics (age, sex), tumor location, type of tumor and preoperative coexistent diseases (hypertension, diabetes, liver/kidney dysfunction, etc.) were evaluated and analyzed retrospectively. The primary endpoints in the study were postoperative duration and rate of fever or infection and postoperative hospital LOS, which was designated from the date of surgery to the date of discharge. The rates of death, reoperation and readmission caused by intracranial hemorrhage, CSF leakage and intracranial infection at postoperative day 30 were also calculated in both groups. In addition, postoperative CSF leaks, including rhinorrhea, otorrhea and incision leaks, were compared between the two groups according to postoperative clinical manifestations and magnetic resonance imaging (MRI).
The other outcomes included complications (e.g., headache, nausea, vomiting, pneumocephalus, nerve root pain and numbness, brain herniation and accidental withdrawal of LD tubes) related to LD and other systemic complications (e.g., respiratory, cardiovascular and urinary system complications, epilepsy and venous thrombosis).
All data, including patient characteristics and intraoperative and postoperative outcomes, were obtained and assessed during the hospital stay and follow-up period (30 days). Student's t test was used for continuous data with a normal distribution, and the Mann–Whitney U test was employed for non-normally distributed data. For categorical data, the Chi-squared test or Fisher's exact test was performed (e.g., extent of tumor resection, tumor location, postoperative complications). Significant difference was identified as P < 0.05. All data were analyzed by means of Statistical Package for the Social Sciences (SPSS) software version 19.0.
In our study, a total of 142 patients who underwent tumor craniotomy were included. Among them, 74 patients (31 males and 43 females with mean age 49.7 ± 11.3 years) received preoperative LD, early CSF drainage performed after surgery (LD group), and the other 68 patients (28 males and 40 females with mean age 53 ± 12 years) did not receive preoperative LD (control group) [Table 1]. The baseline characteristics, including sex, age and concomitant diseases, did not significantly differ between the two groups. The proportion of women in each group was higher than that of men, although there was no statistically significant difference in the gender composition between the two groups.
All patients in the LD group and control group underwent LD placement and tumor craniotomy by the same skilled and experienced medical team. The relevant results of surgery and operative details are shown in [Table 2]. The majority of patients presented with deep-seated, complicated skull base tumors such as cerebellopontine angle (CPA) meningioma or acoustic neuroma, tuberculum sellae and parasellar meningioma, sphenoid ridge, and petroclival meningioma, and foramen magnum and craniocervical junction meningioma or schwannoma. In both groups, there was no significant difference with respect to the categories and location of lesions. The duration of dural opening was more than four hours in all patients in both groups, and the median duration of dural opening was similar between the LD and control groups at 5 (4, 4.5, 7, 13) hours versus 5 (4, 4, 6.5, 13.5) hours (P = 0.642). Of course, the medial operative time also did not significantly differ between the two groups at 7 (6, 6.33, 7.81, 12.5) hours versus 7 (6, 6.67, 7.94, 13.17) hours (P = 0.88). The reoperation rate was similar between the two groups (3/68 [4.412%] in the LD group vs 2/62 [3.226%] in the controls, P = 1), and there was no significant difference. Moreover, there were no patients with 30-day all-cause readmission and no patients who died of any diseases.
Primary outcome measures
Primary clinical outcomes are shown in [Table 3]. Six patients withdrew from the study in the LD group, and six patients withdrew from the study in the control group because of postoperative pulmonary infection, reoperation due to brain herniation or intracerebral hemorrhage or tracheotomy owing to lower cranial nerve injury. There were 22 patients in the LD group and 23 patients in the control group who presented with delayed fever, which was supposed to be caused by intracranial infection or aseptic inflammation. The key indicator measurements are shown in [Table 3]. The percentage of patients with delayed fever in the LD group was lower than that in the control group (22/68, [32.353%] vs 23/62, [37.097%], P = 0.570), but there was no significant difference. The median duration of delayed fever in the LD group was obviously lower than that in the control group (7.762 ± 3.129 days vs 11.73 ± 5.239 days, P = 0.0046). Moreover, there was a significant reduction in the median postoperative LOS (12 [8, 10, 15, 21] days in the LD group vs 15 [9, 13, 20, 28] days in controls).
Secondary outcome measures
Secondary outcome measures are shown in [Table 4] and [Table 5]. All LD tubes were successfully placed following the aseptic procedure, whether before surgery in the LD group (74) or after surgery in the control group (21), due to postoperative intracranial fever or infection. Drain-related complications, including headache (16/74, [21.622%] in the LD group vs 4/21, [19.048%] in controls, P = 0.798), nausea and vomiting (8/74, [10.812%] in the LD group vs 3/21, [14.28%] in controls, P = 0.631) and nerve root pain and numbness (3/74, [4.05%] in the LD group vs 0/21, [0%] in controls, P = 1) exhibited no significant difference between the groups [Table 5]. Compared with the control group without LD tube prolapse, there was accidental prolapse of the LD tube in the LD group. Furthermore, three patients with brain herniation were observed in the LD group compared with one patient in the control group. There was no occurrence of retrograde infection in either group. Other system-related complications, such as pneumonia (3/74, [4.054%] in the LD group vs 3/68, [4.412%] in controls, P = 1), urinary tract infection (1/74, [1.351%] in the LD group vs 0/68, [0%] in controls, P = 1) and venous thrombosis (1/74, [1.351%] in the LD group vs 2/68, [2.941%] in controls, P = 0,607), did not significantly differ between the groups.
Postoperative cerebrospinal fluid leakage
[Table 6] indicates the rate of postoperative cerebrospinal fluid in both groups. The rate of CSF leakage in the LD group was 5.41% (4/74), which was lower than that in the control group (8/68, [11.76%], P = 0.174), although there was no significant difference. This may have been due to the inadequate sample size. One patient had CSF rhinorrhea, and three had CSF incision leaks in the LD group. In contrast with the LD group, the control group had two patients who experienced CSF rhinorrhea, one patient who experienced CSF otorrhea and five patients who experienced CSF incision leakage. Furthermore, two patients in the control group underwent reoperation to repair CSF leaks and were cured.
Comparison of intracranial infected or noninfected fever in both groups
The outcomes of intracranially infected or uninfected fever are shown in [Table 7]. In the LD group, three patients were diagnosed with intracranial infection, six patients were diagnosed with suspected intracranial infection, and 13 patients were diagnosed with aseptic inflammation. In contrast with the LD group, the control group had four patients with intracranial infection, five patients with suspected intracranial infection, and 14 patients with aseptic inflammation. Pathogens detected by CSF culture in four patients in the control group included Staphylococcus aureus, Staphylococcus epidermidis, Acinetobacter baumannii and Sphingomonas paucimobilis. Similarly, pathogens were found in three patients in the LD group; these organisms included Staphylococcus epidermidis, Acinetobacter baumannii and Pseudomonas aeruginosa. These Gram-positive cocci and Gram-negative bacilli were common bacteria in central nervous system (CNS) infections.
Intracranial fever or infection is one of the common complications after craniotomy, especially that performed for deep intracranial tumors. Due to the complicated operation and long operative time, the rate of fever or infection is higher, and LD or continuous lumbar puncture is effective for the treatment of intracranial fever or infection. Coincidentally, deep tumor resection often requires a preplaced LD tube to release CSF during surgery to avoid brain contusion, enlarge the corridor to the tumor, and provide an adequate operation space for tumor resection. Our study confirmed that early CSF drainage with a preplaced LD tube after surgery can prevent postoperative delayed fever and infection safely and effectively.
The incidence of postoperative neurosurgical infection is 2.04%–8.6%, and this difference may be due to the different types or locations of operations included in various studies.,,,,, The longer the operation time, the higher is the infection rate. The duration of surgery was confirmed to be related to intracranial infection by not only retrospective cohort studies but also prospective randomized multicenter trials., To date, the largest randomized study of 132,000 patients performed by Aditya V et al. indicated that the surgical duration was closely linked to postoperative intracranial infection; furthermore, with increased operation time, the rate of infection also increased greatly. In our previous clinical work, we found that the dural opening time had a higher correlation with infection or fever than the whole operation time. In the current study, we used a dural opening time of four hours as the inclusion criterion but not an operation time of four hours because dural opening was crucial to the occurrence of intracranial infection or infection. In our series of cases, the median time of dural opening was five hours, and the median time of surgery was seven hours from skin incision to bandaging of the incision. In our study, we found that the infection rate was approximately 12% in the LD group, which was slightly higher than the overall infection rate reported in previous literature. In addition, we also found that the rate of delayed fever in the LD group was 30%, suggesting that 18% of postoperative patients had aseptic inflammation, which accounted for a large proportion of postoperative patients with intracranial tumors. Both infection and aseptic inflammation will further increase the difficulty and complexity of treatment for postoperative patients with complex tumors, which poses a great challenge to our existing treatment methods.
The efficacy of continuous LD on the treatment of postoperative intracranial fever or infection has been widely established by many recent studies.,, Ren et al. retrospectively reviewed a largescale retrospective analysis of 1062 patients with postsurgical meningitis and found that intravenous antibiotic treatment combined with LD was effective and safe for patients. Multiple studies also indicated that early continuous LD was useful for decreasing complications (e.g., fever, vasospasm, hydrocephalus, cerebral infarction) related to subarachnoid hemorrhage by discharge of the bloody CSF and breakdown products of erythrocytes in the subarachnoid space, which may lead to intracranial aseptic inflammation.,,
Similar to the results reported for early LD in the treatment of arachnoid hemorrhage, our findings were that early LD could significantly reduce the duration and rate of postoperative delayed fever and postoperative LOS compared with the corresponding outcomes in the control group. Although the rate of postoperative delayed fever was not statistically significant, there may be a significant difference after expanding the sample size. The reason early LD lowers the duration of postoperative delayed fever is that LD not only promotes a reduction in the bacterial concentration in CSF or inflammatory factors leading to aseptic inflammation but also provides a time window for early treatment. Moreover, for patients with fever caused by aseptic inflammation, initial moderate LD was beneficial for reduced intracranial pressure and mitigated cerebral vasospasm caused by erythrocytes or breakdown products of erythrocytes.
Furthermore, we compared the incidence of postoperative CSF leakage between the two groups, and the rate of CSF leakage (5.405%) after surgery in the LD group was lower than the rate (11.765%) in the control group, although there was no significant difference between the two groups. The results that preoperative LD may provide an effective approach to reduce the risk of CSF leakage in skull base procedures were supported by previous studies.,, Bien et al. reported on 150 patients undergoing posterior fossa surgery and showed that preoperative LD reduced CSF fistula formation by 23%. A retrospective review by Allen et al. demonstrated that 76.2% of 63 patients had resolved CSF fistulas with LD. In our series, skull base bone erosion by tumors or removal by a microgrinding drill was commonly encountered; if a dural opening was suspected or proven intraoperatively, LD catheter placement could reduce the pressure gradient and promote the recovery of multilayer repairing tissue affiliated with the skull base dura mater. The fact that the occurrence of postoperative CSF leakage increased the potential risk for intracranial infection also highlighted the importance and necessity of using LDs. However, in patients with high-flow CSF fistulas or larger defects of the skull base dura, even if an LD catheter is inserted into the lumbar cistern, the CSF fistula may not be controlled completely, and reoperation may need to be implemented for leak cessation.,
Even though satisfactory outcomes were achieved in the LD group, certain drain-related complications observed in the current study were in accordance with some contemporary reports.,,, The common clinical symptoms in patients were intracranial hypotension, headache and nausea caused by overdrainage, which could be alleviated by intravenous or oral saline solution and temporarily closing the LD tube. Another uncommon presentation was pneumocephalus caused by intracranial atmosphere gas accumulation, for which most patients had no obvious symptoms or had only mild neurological symptoms and signs, but there were also rare patients in whom air accumulation progressed to tension pneumocephalus, leading to the occurrence of transtentorial herniation. Nerve root pain and numbness are also common complications secondary to LD that could be mitigated or eliminated by regulating the depth of the LD tube or withdrawing LD. The most severe LD-related neopathy was thought to be epidural hematoma followed by brain herniation caused by the pressure difference between the supratentorial and infratentorial parenchyma, which promoted brain sagging and resulted in wedging into the tentorial hiatus., There were no patients with epidural hematoma in the two groups; however, three patients in the LD group and one patient in the control group were identified to have brain herniation caused by LD, and there were no deaths among them. In the LD group, one patient exhibiting only brain herniation without hematoma was cured with external decompression, and other two patients recovered well through conservative treatment. In the control group, the patient received conservative treatment and was cured. All four patients had contemporary mild-to-moderate neurologic disorders, including headache, mental confusion and slight hemiplegia.
Although perioperative LD brought about some complications, it was acceptable and manageable. We did our best to minimize the rates of drain-related morbidity by performing the following strategies: (1) keeping the Siphon segment of the LD tube at least 10 cm in height above the external auditory canal in order to maintain normal intracranial pressure and prevent overdrainage (100–150 ml/day); (2) adherence to aseptic techniques in LD placement and CSF collection from the valve; (3) if there is leakage into the incision, timely replacement of the dressing or even of the LD tube; (4) attention given to abnormal symptoms and signs in patients and immediate use of CT scanning; (5) LD duration of no more than ten days, otherwise, replacement of the tube; (6) high-quality nursing care for patients with LD.
However, there are a few limitations in our study. First, our sample size was relatively limited because the rate of deep complicated skull base lesions with a longer operation time in all routine craniotomies was low. Second, our findings may not be applicable to patients undergoing relatively simple craniotomies with a short operation time, and routine LD placement may not achieve an obvious improvement in decreasing the rate of postoperative intracranial fever or infection. Last, the retrospective study may not provide well-powered evidence for the use of early LD. Therefore, a large-sampled randomized controlled trial involving different medical centers needs to be designed to further evaluate the utility of early LD in complicated intracranial tumors.
Postoperative early LD is conducive to the treatment of deep-seated complicated lesions following a longer operation time. In clinical practice, this procedure has led to significant reductions in the duration of postoperative infection or fever and in the postoperative LOS, while no additional complications have occurred. Careful and standardized LD and management processes ensure the safe and effective implementation of the process. Furthermore, we will conduct a prospective randomized controlled trial to provide a higher level of evidence to validate our results in future studies.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]