Ventriculoperitoneal shunt tube infection and changing pattern of antibiotic sensitivity in neurosurgery practice: Alarming trends
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.185408
Source of Support: None, Conflict of Interest: None
Introduction: Infection associated with a ventriculoperitoneal shunt is a severe complication with a high morbidity and substantial mortality. There are no guidelines to choose antibiotics in case of shunt infection. Most surgeons use antibiotics of their choice whereas limited centres follow their own antibiotic policy. An alarming increase in antibiotic resistance has led to rising morbidity and mortality.
Keywords: Antibiotic resistance; hydrocephalus; infection; ventriculoperitoneal shunt
Hydrocephalus is a very commonly encountered entity in neurosurgical practice in both pediatric and adult population. Despite the growing use of endoscopic third ventriculostomy, ventriculoperitoneal (VP) shunt has been the treatment of choice for hydrocephalus since the invention of the shunt valve by John Holter in 1959. Cerebrospinal fluid (CSF) shunting procedures provide a rapid means of normalizing intracranial pressure and can prevent neuronal damage as well as other detrimental sequelae., Recently, adjustable pressure valves and antibiotic-impregnated shunt catheters have been used to continue the trend of catheter development to ensure better outcomes.,,,,,,
Despite significant developments in the technology and design of VP shunt systems, shunt failure remains a significant problem in neurological surgery. Many studies have been performed to investigate the complication rates and causes of failure in both children and adults.,,,,,,,, Some of these studies have suggested that there may be specific risk factors for shunt failure such as a younger age, male sex, socioeconomic status, and ventricular catheter location. In addition, several reports have calculated shunt survival rates at different time points after the initial shunt placement.,,,,
Infection associated with a VP shunt is a severe complication with a high morbidity and substantial mortality, with infection rates ranging from 2 to 27%, often with a poor outcome.,,,,, Shunt-associated infections are most frequently (65%) caused by coagulase-negative Staphylococcus.,, Gram-negative bacteria are the next most frequent pathogens, accounting for 19–22% of cases., Several independent risk factors for shunt infection have been identified, including previous shunt-associated infection, shunt revision for dysfunction, postoperative CSF leakage, an advanced age, duration of the shunt placement operation, experience of the neurosurgeon, and use of a neuroendoscope.,,
In addition, the use of antibiotics in the treatment of shunt infection varies substantially. The indiscriminate use of antibiotics and the lack of proper guidelines for treatment have contributed to an increase in antibiotic resistance.
We present a retrospective review of patients who underwent VP shunt surgery for hydrocephalus with the aim to analyze the incidence of shunt revision and the patterns of shunt infection. We also analyzed the change in antibiotic sensitivity pattern in the past 6 years.
A 6-year (January 2010 to December 2015) retrospective analysis of patients who underwent VP shunt placement at our institute was performed. The demographic data of all cases were tabulated for etiology, age, and clinical presentation.
The case records of those patients who underwent revisions of VP shunt for clinical suspicion of shunt infection, and in whom shunt tube and CSF were sent for culture and sensitivity, were further analyzed for the pattern of shunt infection. In our study, the revision of shunt tube with new shunt assembly was only practiced when the same assembly could not be reinserted. Thus, a new shunt assembly was placed in cases where the shunt tube was exposed or there was evidence of redness or swelling over the skin along the shunt tract. In the cases in whom revision of the shunt tube was performed for shunt obstruction because of an omental plug, blood clot or debris, or due to shunt breakage or malposition, the same assembly was reinserted. In the latter situation, therefore, the samples were not sent for any microbiological analysis.
The processing of the samples for culture was done by the standard techniques, and the identification of the organism along with its sensitivity pattern was performed using the Vitek 2 system (Biomerieux). The primary etiology of hydrocephalus, bacterial growth, and changes in the antibiotic sensitivity patterns to various antibiotics were further analyzed reviewing the data obtained over the last 6 years.
A total of 1186 VP shunts were performed in 6 years from January 2010 to December 2015. There were 756 males (63.8%) and 429 females (36.2%). More than half of these cases (54.4%) were under the age of 10 years [Table 1].
A total of 259 (21.8%) patients underwent shunt revision with tuberculous meningitis [83 (24.2%)] and pyogenic meningitis [13 (25.4%)] as the leading etiology [Table 2].
According to the criteria defined in the 'Material and Methods' section, the shunt assembly was changed completely or partially (distal or proximal end) in 156 patients, and was sent for microbiological analysis along with the CSF sample.
In 79 (50.6%) patients, there was growth of bacteria in the shunt tubing, or CSF, or both [Table 3]. The maximal growth of bacteria was seen in samples of those patients who had earlier undergone shunt insertion for post-tuberculous hydrocephalus [41 (51.9%)] followed by pyogenic meningitis [Table 4]. In those patients who underwent shunt revision, the cytology and microscopic findings were not suggestive of tuberculosis, hence only an aerobic culture was done (culture for Mycobacterium tuberculosis was not performed). Samples sent for microbiological analysis revealed growth in the CSF only in 36 (23.1%) samples, both in the CSF and shunt tube in 27 (17.3%) samples, and in the shunt tube alone in 16 (10.2%) samples. A single organism was grown in 56 (70.8%) samples and there was polymicrobial growth in 23 (29.1%) samples [Table 5].
Among the isolates, it was found that Staphylococcus aureus [65 (82.3%)] followed by coagulase-negative Staphylococcus [22 (25.3%)] were the most common isolates, followed by Escherichia More Details coli, that was isolated from17 (21.5%) samples. In 1 (1.3%) patient, Aspergillus fumigatus was isolated from the scrapings present inside the abdominal end of the shunt tube [Table 6].
The clinical features of patients with shunt revisions whose shunt tubing, CSF, or both, showed growth of an infective organism were retrospectively analyzed. The most common symptom was headache present in 62 (78.5%) patients, followed by vomiting in 60 (75.9%). Other common presenting complaints were fever in 58 (73.4%), decreased consciousness in 50 (63.3%), seizures in 21 (26.6%), redness of skin over the shunt tube in 20 (25.3%), abdominal pain in 13 (16.5%), and blurring of vision in 11 (13.9%) patients [Table 7].
The sensitivity pattern of bacterial organisms revealed that the gram-negative organisms were more sensitive to broad-spectrum antibiotics such as meropenem or imipenem and tigecycline and least sensitive to ticarcillin. The gram-positive bacteria isolated were more sensitive to the broad-spectrum antibiotic, linezolid and least sensitive to roxithromycin.
The trend of antibiotic sensitivity to gram-negative bacteria from 2010 to 2015 showed a decrease in sensitivity to all the quinolones. The average sensitivity to the quinolones in 2015 was only 20%. Similarly, there has been an increasing resistance to all the cephalosporins with an average sensitivity rate of 30–40%. Imipenem, meropenem, and tigecycline have a sensitivity of more than 60%, which has resulted in them becoming the most commonly used antibiotics. Similar trends were seen for quinolones and cephalosporins for gram-positive bacteria, with linezolid and teicoplanin showing sensitivity of more than 60% [Figure 1] and [Figure 2].
Shunt infection poses a major threat in the treatment of hydrocephalus. Early detection and diagnosis of shunt infection is usually difficult because of its nonspecific clinical features. Most of the time, the clinical presentations are the same as that of a shunt malfunction. Headache, vomiting, fever, and progressive deterioration of consciousness were the most common findings among our patients. If the above mentioned signs are present in the postoperative period, VP shunt malfunction should be suspected. In such cases, immediate neuroimaging and CSF studies should be performed to determine whether meningitis is present with shunt malfunction.
The incidence of complications following a VP shunt placement is reported to be in the range of approximately 20–40%.,, In our study, the complication rate was 21.8%. The complications occurred because of shunt infections, malposition of the ventricular or the abdominal end, blockage of the abdominal end by an omental plug, blockage of the shunt by debris or blood clot, breakage of the shunt assembly, migration of ventricular end of the tube, and exposure or extrusion of the shunt tube.
Postoperative infection of shunts occurs in 2–27% of cases in most neurosurgical units throughout the world.,,,,, Isolation of bacteria in shunt infection along with determination of their sensitivity pattern is paramount in managing these cases. Shunt revision was done in 21.3% patients, and among the samples sent for microbiological analysis, 79 (50.6%) patients had evidence of bacterial growth. This high incidence was due to the fact that these cases were referred to us from various places, and in many of these patients, the initial surgeries was done at other hospitals. A poor nutritional status appeared to be a common denominator.
The remaining 77 (49.4%) cultures were negative. In these patients, there was either no infection, or infection was present but the organism was resistant to detection and hence did not grow on culture, or there could have been an anaerobic infection (an anaerobic culture was not performed in the study). Another reason for a negative culture could have been because the patient was on antibiotics. While 27 (17.3%) patients had growth of a bacterial organism in the shunt tubing as well as in the CSF, in 36 (23.1%) patients there was growth of bacterial organism only in the CSF, and in 16 (10.2%) patients, the growth occurred in the shunt tubing only. It highlights the fact that CSF may be a better sample to document infection than the shunt tube alone.
In patients undergoing a shunt procedure, the infectious pathogen will most likely be a microorganism from the resident bacterial flora of the skin, nasopharynx, or the external auditory canal. However, the possibility of nosocomial infection should not be ruled out. Many studies have suggested that coagulase-negative staphylococci are the major pathogens in shunt infections, followed by S. aureus.,,, However, Gram-negative bacteria are also responsible for 7–24% of all VP shunt infections., This study demonstrates that S. aureus (82.3%) and coagulase-negative Staphylococcus (25.3%) were the most commonly isolated bacteria. E. coli (21.5%), Klebsiella species (12.7%), and Citrobacter species (7.6%) were the isolated Gram-negative bacteria. Aspiration of gastric contents is common in unconscious patients and is frequently complicated by bacterial superinfection. The routine use of proton pump inhibitors leads to bacterial colonization of the stomach with aerobes, especially the Gram-negative aerobes, which may lead to infections caused by them. A relatively high proportion of mixed infections (14.7%) was detected in our study. Such infections are more complicated and require appropriate antibiotics to eradicate them.
In our study, infections (41 out of 79) and revisions (83 out of 259) were the highest in the post-tuberculous meningitis hydrocephalus group than in other conditions. The reasons for this increase may have been the poor general condition of these patients as well as the presence of higher protein and cellular content of the CSF, leading to more frequent shunt malfunctions. The reported rate of shunt malfunctions in these patients has been 22–43%.,, An infection rate of 15.6% was reported by Sil and Chatterjee  in the cases having tuberculous meningitis.
There has been an alarming rise in the resistance pattern of the Gram-negative bacteria to cephlasporins and quinolones. Aminoglycosides also have shown a decrease in sensitivity. However, imipenem, meropenem, and tigecycline were still sensitive in this group. Similarly, for gram-positive bacteria, the increasing resistance trends to cephalosporins and quinolones is a major concern. However, linezolid and teicoplanin are still sensitive in this group. This may be because of the poor antimicrobial stewardship practices (the antibiotics to which the bacteria are developing resistance are more commonly available in the government hospital supply and are prescribed to patients more often after surgical intervention) prevalent in the hospital. In addition, the patients admitted to our institute are at a high risk of antibiotic resistance due to numerous contacts with hospital fomites and other patients, immunosuppression, and prior exposure to antibiotics.
It is already known that reinserted shunts due to previous infections may have a greater chance of becoming infected., In our study, there was a tendency for a higher infection rate of reinserted shunts due to the existence of prior infections, compared to that of noninfectious causes.
Shunt infection is a common neurosurgical problem. It can occur in all types of hydrocephalus; however, it is seen more often in the postinfective etiology, in particular tuberculosis. These cases can be difficult to treat because the response to commonly used antibiotics is dismal. Based on our findings of a high prevalence of developing resistance to commonly used antibiotics, we suggest the use of meropenem and teicoplanin with ceftriaxone as an empirical therapy in suspected cases of infections in patients undergoing shunt revision. In addition, because of the changing trends, the specific treatment guidelines should be based on the culture reports.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]