Anterior Skull Base Fractures Treated with Surgical Correction: Factors Associated with Development of Infection

Abstract

Objective: Traumatic anterior skull base fractures can be associated with significant morbidity and are managed based on the method of injury, presence of cerebrospinal leak, clinical stability and other factors. Our objective is to determine factors associated with the development of infection in the management of surgical anterior skull base fractures. Methods: We completed a retrospective review of adult patients with traumatic anterior skull base fractures treated operatively with bi-frontal craniotomy from 2012 to 2022. The statistical analysis was completed with Prism 9.0 software for spearman correlation coefficient analysis. Results: Our study had a total of 51 patients. The average time from arrival to surgery was four days. A total of 20 patients were injured via gunshot wound (GSW) and 31 via other methods of injury. A total of 81% of patients were started on antibiotics prior to surgery, and all patients were kept on antibiotics after surgery. Five patients developed infections. Infection was not correlated with time to surgery, length of stay, type/duration of antibiotics, EVD placement/duration, or lumbar drain placement. However, all but one of the infections were in patients injured by GSWs vs. other mechanisms of injury (p < 0.01), and duration, not merely presence, of lumbar drain was correlated with development of infection (p < 0.01). Conclusions: In the treatment of anterior skull base fractures, time to surgery does not significantly increase risk of infection. However, patients with gunshot wounds have higher risks of infection, and thus more aggressive management should be considered.

1. Introduction

Anterior skull base fractures secondary to trauma are associated with significant morbidity and mortality and are managed variably based on method of injury, presence of cerebrospinal fluid leak, clinical stability and other factors. The urgency in treating these injuries is debated, and the literature is sparse. Simple fractures without cosmetic deformity and without significant intracranial injury are typically managed conservatively with or without cerebrospinal fluid (CSF) diversion, depending on the presence of a CSF leak. Typically, skull base fractures that are comminuted, associated with significant intracranial injury or with a CSF leak are managed with craniotomy, debridement of devitalized tissue, cranialization of sinus and peri-cranial/fat graft [1,2].

Few studies have been conducted that specify the time to operating room (OR) and its effect on outcomes. Two prior studies on the subject found no difference in terms of infection risk or other outcomes with regard to early (defined as within 24 h) interventions vs. intervention within two weeks [3,4]. An evidence-based approach guide to how urgently these fractures need to be repaired (in the absence of neurological decline/significant intracranial findings) would be helpful in clinical practice for operating room utilization, coordination with other sub-specialty services, and overall patient safety. Typically, at our level one trauma center, patients are stabilized and ideally taken in 1–2 days, but they are not emergently taken in the absence of life-threatening hematoma or other emergent circumstance. We present a retrospective case study of patients admitted to Memorial Hermann Texas Medical Center (Houston, TX, USA) from 2012 to 2022 with anterior skull base fractures managed operatively.

2. Materials and Methods

We completed a retrospective review of adult (>18 years old) patients with traumatic anterior skull base fractures treated operatively from 2012 to 2022 at the Department of Neurosurgery, Memorial Hermann Hospital, affiliated with the University of Texas Health Science Center at Houston. Clinical data were obtained through a chart review, using the institutional electronic medical record system, and were entered into a secured database. Exclusion criteria were pediatric patients (<18 years old), patients managed non-operatively, and patients who were discharged and returned with delayed infections. Patient information included demographics, injury mechanism, surgical timing and details, involvement of additional surgical services, use and duration of lumbar or external ventricular drains, antibiotic type and duration, and relevant laboratory and microbiological data. The study was approved by the Institutional Review Board (IRB# HSC-MS-14-0319), with a waiver of patient consent granted due to the retrospective nature of the study and anonymized data handling.

Descriptive statistics were used to summarize patient demographics and clinical characteristics. Continuous variables, such as age, time to surgery, and drain duration, were reported as means and ranges. Categorical variables, such as mechanism of injury, presence of infection, and use of drains, were presented as counts and percentages. Comparative analysis between groups (e.g., gunshot wound vs. other injuries, infection vs. no infection) was performed using the Spearman correlation to assess associations between variables such as drain duration and infection. A p-value of <0.05 was considered statistically significant. All analyses were conducted using GraphPad Prism 9.0 software (GraphPad Software, Boston, MA, USA).

3. Results

Our study had a total of 51 patients with an average age of 33, and 82% of our population was male. The average time from arrival to surgery was four days, the median time from arrival to surgery was three days, and the range of included patients in analysis was 0 to 18 days. A total of 39% of patients were injured via gunshot wound, but other mechanisms included motor vehicle collisions (MVC) and motor cycle collisions (31%), assault (4%), workplace accidents (10%) and falls (4%). In six patients, the mechanism was not known or did not fit into these main categories. The average time to surgery for patients injured via gunshot wound was three days. All patients underwent bi-frontal craniotomy for repair of skull base fractures (including some with a cribriform plate, additional sinus fractures, etc.) and frontal sinus cranialization and repair. All patients had inner and outer table frontal sinus fractures. Subgaleal drains were used in all cases and were removed within three days of the operation. A total of 21 patients had a lumbar drain placed with an average duration of six days; 13 patients had an external ventricular drain (EVD) placed with an average duration of nine days. This included both pre- and post-operative placement. Only one patient required ventriculoperitoneal shunt placement for hydrocephalus.

A total of 81% of patients were started on antibiotics prior to surgery, and all patients were kept on antibiotics after surgery. A total of 85% of patients had an additional service to neurosurgery involved in the surgery, including plastic surgery, ophthalmology, oromaxillofacial surgery (OMFS), and/or otorhinolaryngology (ENT). Five patients (10%) developed infections. The most common organisms isolated were staphylococcus species (not aureus) and Enterobacter. Patient data are shown in

Table 1. Demographic and treatment data.

Demographics (n = 51)
Mean Age	33+/−16.30
Male Sex	42 (82.35%)
Mechanism of Injury
MVC	13 (24.53%)
GSW	20 (39.22%)
Mean Length of Stay (Days)	13.85+/−8.55
Most Common Comorbidity
Hypertension	3 (5.6%)
Treatment Characteristics (n = 51)
Median time to surgery (days)	3
Mean number of additional Services involved in care	1.21+/−0.72
LD	21 (41.18%)
Mean LD Duration (Days)	5.86+/−3.17
EVD	13 (25.49%)
Mean EVD Duration (Days)	8.77+/−3.70
Surgical infection Agent	5 (9.8%)
Staphylococcus sp. (not aures)	2 (3.92%)
Enterobacter cloacae	2 (3.92%)
Lactobacillus sp.	1 (1.96%)
Received Pre-op Abx	41 (80.39%)

.

The five patients with infections consisted of two cases of meningitis, one case of osteomyelitis, one case of scalp abscess, and one case of infection of hardware to repair the hard palate. Infection was not correlated with time to surgery, number of services involved in surgery, length of stay, type/duration of antibiotics, EVD duration/placement or lumbar drain placement. When patients injured by GSW were analyzed independently, time to surgery remained insignificant. Lumbar drain duration was significantly correlated with increased risk of infection (p < 0.01). Patients with gunshot wounds as their mechanism were significantly more likely to develop infection (p < 0.01).

4. Discussion

Anterior skull base injuries tend to be associated with significant trauma, including gunshots and high-velocity car or motorcycle accidents or assault [2,5]. Some fractures can be managed conservatively; however, these injuries may lead to cosmetic deformity as well as cerebrospinal fluid leak that requires surgical intervention [1]. The reported incidence of cerebrospinal fluid leak in these injuries varies greatly, from 4 to 30% [4]. If fractures are simple and do not require surgical repair, cerebrospinal fluid leaks can be managed with a lumbar drain or EVD. Most authors suggest surgical repair if leakage is still observed after a trial of conservative management [1]. Compound, depressed, or fractures leading to dural tear/brain laceration and/or contusion may need to be repaired surgically [2]. Penetrating injuries through this region should be repaired surgically. Thin cut facial CT scans with multi-planar reconstruction are appropriate for characterization and surgical planning.

Injuries of the anterior skull base involving the frontal sinus are particularly important to recognize because of the dense adhesion of dura to the bone, leading to potential complications, such as CSF leak, brain laceration, mucocele development, infections, and pneumocephalus. Typically, these skull base traumatic injuries are managed with craniotomy, debridement of devitalized tissue, cranialization of sinus and peri-cranial graft. However, many management questions remain without evidence-based guidance, including what the urgency of repair is. This question is important because surgical repairs often involve coordination of specialty services, such as plastic surgery, ENT, and/or ophthalmology as well as neurosurgery. We studied 53 patients with anterior skull base fractures managed operatively to try to help guide the management of these injuries better, focusing on time to OR, mechanism of injury, additional sub-specialty services involved in the case, CSF leak management, and antibiotic use.

In our series, two patients were excluded as they did not have surgery on initial admission. One patient was offered surgery and left against medical advice and returned with intracranial abscess and obvious infection 67 days later, and one patient failed conservative management with lumbar drain and returned with meningitis and subdural empyema (43 days later). These two patients were not included in subsequent analysis; however, they warrant mention as they demonstrate the consequences of failed conservative management. Patients with life-threatening hematoma and/or neurologic decline were taken emergently (11 patients). The additional patients were all determined to warrant operative intervention due to the presence of inner and outer frontal sinus fractures, but did not necessitate emergent intervention. No patients in this analysis were trialed on conservative management prior to operating. One patient who went to the OR emergently developed an infection. An increased length of time to surgery did not increase risk of infection, and this held true when analyzing gunshot wounds separately from all other mechanisms of injury. However, given the delayed presentation of several patients with infection, timely intervention is appropriate.

The most common mechanism was gunshot wound, and these patients were significantly more likely to develop infections (all five infections included in analysis were injuries via gun shot, p < 0.01). This is similar to other series. As is standard practice at our institution, gunshot remnants were removed only if easily accessible and were not pursued if located deep in the parenchyma. Bhatoe presented a case series of 23 gunshot injuries to the anterior skull base obtained during war time, with 13% of patients developing meningitis and 2 patients later developing intra-cranial abscesses [6]. This is similar to our overall 10% infection rate. Given the higher risk for infection from penetrating trauma, these operations should be performed urgently.

The majority of cases (87%) were completed with at least one additional surgical subspeciality (ENT, ophthalmology, plastics, OMFS) available with assistance, with either repair of complex facial fractures, closure, or both. This is consistent with other series [1,2]. The number of specialties in addition to neurosurgery in the OR did not increase risk of infection, and simultaneous repair of the cosmetic deformity is appropriate in most cases. In cases when the superior border of the orbit has been affected, definitive repair with the appropriate services is essential prior to placement of the pericranial graft to avoid future disruption of the repair.

In the population, 40% of patients had visible CSF leak and required CSF diversion peri- and/or post-operatively, with 21 patients treated with a lumbar drain (LD) and 13 patients with an external ventricular drain (EVD). Two patients had both an EVD and a LD during their hospital stay. Lumbar drains were kept for an average of six days, and external ventricular drains were kept for an average of nine days. Neither the placement or duration of external ventricular drains correlated with the development of infection. Lumbar drain placement did not correlate with infection, but lumbar drain duration did correlate with an increased infection risk (p < 0.01). Only one patient required placement of a ventriculoperitoneal shunt, which is consistent with the current literature, according to which permanent CSF diversion is rarely needed in this population. Various management strategies have been described for the management of CSF leaks by other authors, including simple continuous lumbar drainage, and first a trial with LD before the placement of EVD if the leak persists and a large defect is observed on CT [7,8]. The detailed management strategy of CSF leaks is outside the scope of this article.

Antibiotics were used prior to definitive surgery in 81% of patients. The majority of patients were started on two agents, most commonly ceftriaxone with nafcillin or vancomycin. This practice was not associated with a decreased risk of infection. All patients were started on antibiotics post-operatively, with an average treatment time of five days. Five patients developed infections (10%), with the most common organism isolated being staphylococcus species (not aureus). Patients who developed infections were treated for 2–3 weeks of antibiotics, depending on the organism isolated. Controversy exists in the treatment of skull base fractures with/without visible cerebrospinal fluid leaks with antibiotics. A total of 5 randomized control trials and 17 non-randomized clinical trials showed no effect on the development of infection with antibiotic prophylaxis, and our study confirmed these results [9,10]. At our institution for simple skull base fractures, we do not prescribe prophylactic antibiotics, but we do use antibiotics for penetrating traumatic brain injury. Larger randomized control trials are needed to answer this question.

Our study is limited insofar as it is retrospective in nature and we are not able to randomize patients to different intervals to OR intervention. Additionally, we only studied patients who required operations and did not evaluate fractures managed conservatively, several of which may have also gone on to develop infections.

5. Conclusions

In the treatment of anterior skull base fractures, the time to surgery does not significantly increase the risk of infection. Lumbar drains and EVDs are needed in the management of these patients, but prolonged-duration lumbar drains can be associated with an increased infection risk. Although all patients should be treated as soon as possible, gunshot wounds have a significant association with infection and should be treated expediently and more aggressively.