Is acetazolamide really useful in the management of traumatic cerebrospinal fluid rhinorrhea?
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.156280
Source of Support: None, Conflict of Interest: None
Background: Traumatic cerebrospinal fluid (CSF) rhinorrhea is a serious and potentially fatal condition as it may lead to meningitis. As acetazolamide decreases CSF production and hence CSF pressure, it has been proposed that the medication may help in curing CSF rhinorrhea. There is no definitive evidence, however, that acetazolamide is actually beneficial in treating traumatic CSF rhinorrhea. The aim of this study was to determine if the administration of acetazolamide in patients of head trauma with CSF rhinorrhea was beneficial in decreasing the duration of CSF rhinorrhea. The acid-base and electrolyte changes caused by the drug were also studied.
Keywords: Head injury; CSF rhinorrhoea; acetazolamide
Cerebrospinal fluid (CSF) rhinorrhea is commonly encountered in patients sustaining a head injury, especially those with basilar skull fractures.  Most of these CSF fistulae resolve spontaneously. Some require a longer time to stop and a few may need surgical intervention. Addition of acetazolamide has been suggested in the management of traumatic CSF rhinorrhea.  Its usage is justified by the fact that it reduces CSF production by 48%. ,,,, The actual role of acetazolamide in the amelioration of traumatic CSF rhinorrhea is not clear as, till date, very few formal studies have ben conducted investigating its impact in the management of CSF fistulae following head trauma. The purpose of this study was to investigate whether or not the administration of acetazolamide in head trauma patients with CSF rhinorrhea was beneficial in decreasing the duration of CSF rhinorrhea. The acid-base and electrolyte changes caused by acetazolamide admininstration were also studied.
In this prospective randomized, clinical trial, we evaluated 44 patients with traumatic CSF rhinorrhea who were referred to the trauma center of a tertiary care hospital over a period of one and a half years. Of these, 21 patients were randomly assigned to the experimental group and 23 to the control group. The experimental group received acetazolamide while the control group was not administered the medication. Two patients were excluded from the study because they died before a reasonable assessment could be carried out. However, their acid-base and electrolyte derangements following acetazolaminde admininstration were significant. At follow-up, the duration of CSF leakage, period of hospitalization, incidence of meningitis, electrolyte and acid base disturbances, number of surgeries and lumbar drainage instituted in both the groups were recorded. The lottery method of simple randomization was used to allocate the patients to both the groups.
The patients included in the study were those greater than 5 years of age presenting to the trauma centre with a history of head trauma and diagnosed with acute CSF rhinorrhea. The following types of patients were excluded from the study: Patients less than 5 years of age, patients with a history of remote trauma and delayed (more than 7 days after head trauma) CSF leak, and patients with a large defect at the base of skull that clearly required surgical intervention.
The patients in the control group were managed conservatively with a strict bed rest, head elevation of atleast 30 degrees, and avoidance of sneezing, coughing and straining. None of them were administered acetazolamide. The patients in the experimental group were also managed conservatively on the same principles. In addition, acetazolamide (250 mg qid in adults and 8-30 mg/kg/day in children in divided doses) was also admininstered to them. The approximate duration of conservative management in both the groups was arbitrarily defined as 10 days after which they were taken up for either lumbar drainage or surgery.
A complete physical examination of the patients with head trauma was performed to detect any sign of CSF leakage, and a brain CT scan was undertaken immediately after stabilizing the patient. Those who met the inclusion criteria entered the study. With a checklist designed for each patient, we registered and compared the following data between the two groups: The demographic data, the duration of CSF leakage, the incidence of meningitis during admission (based on the presence of headache, fever and neck stiffness), the need for surgical intervention, the need for lumbar drainage and the location of the precise site of CSF leakage. The follow up period was fixed at 90 days from the onset of CSF rhinorrhea and the patients were followed up for upto 3 months in both the groups.
The data was entered in SPSS version 16. To determine the correlation between our quantitative data, we used the independent t-test, Mann-Whitney test and Fisher's exact test. For our qualitative data, we used the Chi- square test. A P value <0.05 was considered significant. The mean number of electrolyte fluctuations occurring in each treatment group and the days required for CSF leak to stop was recorded in both the treatment groups.
The age range was from 9 to 75 years in the experimental group and 16 to 65 years in the control group. The t-test showed that the mean age in our two groups (34.48 years in the experimental, and 32.04 years in the control group) did not differ significantly (P = 0.597). Of the 44 patients, 38 were men and 6 women. The gender distribution of the patients in the two groups was nearly the same, that is, there were 20 male patients and 1 female patient in our experimental group, and 18 male and 5 female patients in our control group. This was not significantly different based on the Chi-square test (P = 0.101).
The mechanism of injury was divided into three groups: Road traffic accident (RTA), fall and assault. Thirty five of the 44 patients in both the groups (17 in the experimental group and 18 in the control group) had a RTA, five (1 in the experimental group and 4 in the control group) had a fall, and 4 (3 in the experimental and 1 in the control group) were assaulted. There was, however, no significant difference in these figures between the groups (P value = 0.25, not significant).
Most of the patients in this study (68.18%) had a mild head injury, while 15.91% of them each had a moderate and severe head injury, respectively. Thirty of the 44 patients had a mild head injury (13 in the experimental and 17 in the control group), 7 had a moderate head injury (6 in the experimental and 1 in the control group), while rest of the seven had a severe head injury (2 in the experimental and 5 in the control group). Both the groups were, therefore comparable with respect to the severity of injury having a P value of 0.07 which was not significant.
In 37 (84.1%) of the 44 patients, an associated skull fracture was observed on the non-contrast head CT scans. The most common fracture associated with CSF leakage was that of the frontal bone involving the frontal sinus. This was followed in incidence by the petrous bone fracture (9.09%). A total of 22 patients out of 44 (50%) had pneumocephalus. The distribution of frontal bone fracture, temporal bone fracture and pneumocephalus in both the experimental and the control groups were homogenous with a P value of 0.393, 0.252 and 0.365 respectively, that were not significant.
Antibiotics were given to 10 patients out of the 44 patients for other associated injuries. Of these, 7 patients belonged to the experimental group and 3 patients to the control group. There was no significant difference between the two groups with respect to whether or not antibiotics were administered (P = 0.07). We recorded the incidence of meningitis in both the groups. Only 1 patient (in the experimental group) out of the total of 44 (2.27%) patients developed meningitis. This patient who developed meningitis had not received any prophylactic antibiotics.
After 10 days following the trauma, CSF leakage resolved in all of the patients in the control group, while 3 out of the 20 patients (15.0%) in our experimental group continued to exhibit CSF leakage. Only three patients out of a total of 42 (7.14%) patients had a persistent CSF leak not responding to the conservative treatment. All of them belonged to the experimental group wherein acetazolamide was administered. The rest of the 39 (93.86%) patients had cessation in their CSF rhinorrhea with conservative management. Out of three patients with a persistent CSF rhinorrhea, after 10 days of conservative management, one of the patients was placed on a lumbar drain following which his rhinorrhea stopped. His lumbar drain was removed after 3 days. The second patient underwent an endoscopic, endonasal repair of his cranial base defect following which his CSF rhinorrhea stopped. At follow up, there were no further episodes of CSF rhinorrhea. The third patient still suffers from intermittent CSF rhinorrhea, has had an episode of meningitis and is now under otorhinolaryngology follow up with a plan for an endoscopic repair of his CSF leak.
The median duration of CSF rhinorrhoea in the experimental group (where acetazolamide was administered) was 5.0 days, and in the control group (where acetazolamide was not given), 4.0 days (statistically significant [P = 0.029] by the Mann-Whitney test). In fact, all the patients in both the groups had resolution of their CSF rhinorrhea by the 15 th day following the episode of trauma except for 1 patient in the experimental group who still had active CSF rhinorrhea until the last follow up visit. At discharge, there was no significant difference in both the groups with respect to the Glasgow Coma Score (GCS) and CSF rhinorrhoea. At a follow up of one month, none of the patients in either group had an active CSF rhinorrhea; and, after 90 days of follow up, only 1 patient belonging to the experimental group had a recurrence of CSF rhinorrhea [Figure 1].
The hospital stay in days (HSID) between the two groups were compared by dividing HSID into 3 categories: Less than 7 days, 7-10 days and more than 10 days. There was a significant difference in the duration of hospital stay between the experimental and the control groups. Five patients in the experimental group and 13 in the control group had a HSID of less than 7 days, 6 patients in the experimental group and 5 in the control group had a HSID between 7-10 days, and 10 patients in the experimental group and 3 in the control group had a HSID greater than 10 days (P = 0.025). However, this finding of a longer duration of hospital stay in the experimental group was probably due to the fact that some of the patients in the study group had external injuries other than head injury, such as a facio-maxillary injury, for which they underwent surgery, prolonging their discharge from the hospital.
The electrolyte and acid base disturbances between the two groups were also recorded. In our study, there was a significant difference between the experimental and the control groups in terms of pH (P = 0.014), HCO3 (P = 0.001), base deficit (P = 0.001) and serum potassium (P = 0.002). The mean serum pH levels in the experimental group were 7.367 and 7.339 at days 3 and 5 following the episode of trauma, respectively; in comparison, in the control group, the mean serum pH levels were 7.404 and 7.401 respectively, at days 3 and 5. The mean serum bicarbonate levels on the third day in the experimental group were 17.37mEq/L compared to 21.62mEq/L in the control group. Similarly, the mean serum potassium levels in the experimental group were 3.44 and 3.19mEq/L respectively, at days 3 and 5 compared with a mean serum potassium levels of 3.85 mEq/L and 3.95 mEq/L when recorded at the same days respectively, in the control group. These results showed that in the experimental group, serum pH, serum bicarbonate and serum potassium levels progressively decreased with continued use of acetazolamide as compared to the control group in which they were stable. The patients in the experimental group, however, did not manifest symptoms of metabolic acidosis and hypokalemia probably because the mean serum pH level in these patients at day 3 was 7.367 which was in normal range, and at day 5, was 7.339 that was slightly below the normal level (this was despite the fact that there was a significant difference between the mean serum pH levels of the two groups). As per the study protocol, potassium supplementation was instituted when the serum potassium level was lower than 3 mEq/L.
CSF rhinorrhoea is the leakage of CSF from the subarachnoid space into the nasal cavity due to a defect in the dura mater, bone and mucosa. It is a serious and potentially fatal condition because of an increased risk of meningitis and brain abscess. It is estimated that meningitis develops in approximately 10%-25% of patients with this disorder. Among the latter group where meningitis develops, there is nearly a10% mortality. Approximately 80% of all cases of CSF rhinorrhea are caused by head injuries that are associated with cranial fractures.  Around 85% of post traumatic CSF fistulas close spontaneously (within 48 hours in 68% of cases and within 1 week in 85%).  The mechanism of spontaneous cessation of the CSF leak is thought to be the accumulation of blood products and/or initiation of inflammatory adhesions at the site of the dural breach and the associated skull fracture. 
The management of traumatic CSF rhinorrhoea may be broadly divided into two types: Conservative management and surgical treatment. Surgery is reserved for the treatment of CSF leak that neither ceases spontaneously nor responds to a CSF diversion.  Conservative management includes bed rest, head elevation, avoidance of straining and in some cases, a lumbar drain to decrease CSF pressure. The goal of these measures is to reduce active flow through the leak, reduce CSF pressure, and allow healing of the defect to seal the leak, thus avoiding a surgical intervention.  Acetazolamide also decreases CSF production. Therefore, many investigators have advocated the routine use of acetazolamide in all patients with a spontaneous CSF rhinorrhoea and an evidence of increased CSF pressure.  Most of the literature, however, does not clearly define the precise role of acetazolamide in the conservative management of post-traumatic CSF rhinorrhoea. Recently Abrishamkar et al.  conducted a randomized trial that showed that an early acetazolamide administration as a part of the conservative measures instituted in patients who sustain a skull-base fracture can be useful in preventing CSF leakage; and, in shortening the duration of leakage in patients who already have either an otorrhea or a rhinorrhea. In our study, both the experimental and control groups were well matched in terms of age, sex, mechanism of injury, GCS, and the type of skull fracture. In contrast to the above mentioned study, the median duration of CSF leakage in our experimental group was of 5.0 days compared to 4.0 days in the control group, a difference that was statistically significant when evaluated by the Mann Whitney test (P = 0.029). This indicated that acetazolamide did not significantly reduce the duration taken for the resolution of CSF rhinorrhoea thus proving that there is no advantage in the prophylactic administration of acetazolamide in an attempt to reduce the duration of traumatic CSF rhinorrhoea. Furthermore, at discharge, there was no significant difference in the time taken for the resolution of CSF rhinorrhea in both the groups when the patients with corresponding GCS were compared. One month following the trauma, none of the patients in either group (except for one patient in the experimental group who had a recurrence of CSF leak) had an active CSF rhinorrhoea. Furthermore, the use of acetazolamide caused unwanted side effects like metabolic acidosis and hypokalemia.
In conclusion, there was no advantage in adding acetazolamide to the conservative management of traumatic CSF rhinorrhoea. The CSF leak stopped spontaneously or with a lumbar drain within a short span of time even when acetazolamide was not administered. On the other hand, giving the medication adversely caused acidosis and electrolyte imbalance. Therefore, in traumatic CSF rhinorrhoea, the practice of routinely giving acetozolamide should be reconsidered. Furthermore, the addition of acetazolamide to the conservative treatment of traumatic CSF rhinorrhoea did not lead to a decreased hospital stay.
Limitations of the study
The limitation of this study was that due to the nature of the study design, it was not a double blinded one. The small number of cases in the subgroups makes this a pilot study indicating the need for a larger patient recruitment to make the statistical conclusions more meaningful. There was no attempt to quantity the leak in terms of the volume of CSF leak or the size of the bony defect which could have had a bearing on the success of treatment.