Mind the gut: Navigating the complex landscape of gastroprotection in neurosurgical patients

Abstract

Neurosurgical patients, including those with severe traumatic brain injury, spinal cord injury, stroke, or raised intracranial pressure, are at heightened risk for stress ulcers and aspiration pneumonitis, leading to significant morbidity and mortality. These patients are typically managed through both pharmacological interventions [e.g., proton pump inhibitors (PPIs), histamine 2 (H2) antagonists, sucralfate] and non-pharmacological measures (e.g., nasogastric decompression, patient positioning) to mitigate adverse outcomes. The pathogenesis of stress ulcers in neurosurgical patients is multifactorial, but the routine use of stress ulcer prophylaxis remains controversial. While gastric acid suppression with H2 receptor antagonists and PPIs is commonly employed, concerns have arisen regarding the association between elevated gastric pH, bacterial colonization, and ventilator-associated pneumonia. The lack of comprehensive data on gastroprotection in critically ill neurosurgical patients, who face a greater risk than non-neurosurgical counterparts, further complicates this issue. Recent studies, such as one by Gao et al on the efficacy of vonoprazan-amoxicillin dual therapy in elderly patients, highlight the potential of novel therapies, but the influence of pre-existing conditions like Helicobacter pylori infection remains unclear. Non-pharmacological interventions, including nasogastric decompression and early enteral nutrition, are critical in improving outcomes but require further research to refine strategies. This editorial underscores the need for tailored approaches and encourages further investigation into optimal gastroprotective strategies for neurosurgical patients.

Key Words: Gastroprotection; Neurosurgery; Stress ulcer; Proton pump inhibitor; Histamine 2 receptor antagonist; Vonoprazan; Nasogastric tube; Enteral nutrition

Core Tip: Neurosurgical patients, particularly those with severe brain injury and elevated intracranial pressure, face heightened risk of stress ulcers and gastrointestinal complications. Effective management requires a combination of pharmacological and non-pharmacological strategies. While gastric acid suppression is common, its routine use may raise concerns about bacterial colonization and ventilator-associated pneumonia. Recent studies on therapies such as vonoprazan-amoxicillin for Helicobacter pylori as well as combined naso-gastric and naso-intestinal tube application emphasize the need for tailored gastroprotection highlighting critical gaps that warrant further research and individualized approaches in clinical practice.

INTRODUCTION

Stress ulcers and associated upper gastrointestinal bleeding (UGIB) have long been known to complicate the disease course in neurosurgical patients, with Harvey Cushing being one of the firsts to describe this phenomenon[1]. Patients with a spectrum of neurological illnesses such as traumatic brain injury (TBI), spinal cord injury (SCI), intracerebral hemorrhage (ICH)[2,3], ischemic stroke, central nervous system tumors, infections have been shown to develop stress ulceration at a rate that is significantly higher than those without neurological illness[4-7]. The pathophysiology of stress ulceration in this patient population seems to be multifactorial and can be summarized as an alteration in the balance between protective and destructive factors that regulate the gastric mucosal lining (Table 1)[3,8-13]. Destructive factors include gastric acid and pepsin hypersecretion, splanchnic hypoperfusion which are themselves a consequence of uninhibited vagal activity and reduced sympathetic drive[8,14]. Furthermore, risk factors such as age over 60 years, presence of syndrome of inappropriate anti diuretic hormone secretion, pre-existing Helicobacter pylori (H. pylori) infection, mechanical ventilation, as well as corticosteroid use have been shown to predispose patients to develop stress ulcers[2,8,15-17].

Table 1 Risk factors for stress ulcer and upper gastrointestinal bleeding in the neurocritical cohort.

No.	Risk factors
1	Glasgow coma scale < 9[3,10,11]
2	Age > 60 years[8,9]
3	Syndrome of inappropriate anti diuretic hormone[8]
4	Pyogenic central nervous system infection[8]
5	Elevated plasma cortisol[12]
6	Cerebral vasospasm[13]
7	Elevated intracranial pressure[13]
8	Mechanical ventilation > 48 hours[13]
9	Organ failure (renal, hepatic)[9]
10	Hypotension[9]
11	Coagulopathy[9,13]

While stress ulcer prophylaxis (SUP) is clearly indicated in neurocritical patients, the evidence regarding the best strategy for the same is not clear. Studies on pharmacologic strategies involving proton pump inhibitors (PPIs) and/or histamine 2 (H2) antagonists have raised concerns about the safety of routinely using these medications in the intensive care unit (ICU) setting[18]. Notably, there are concerns regarding the potential development of pneumonia and Clostridium difficile infection (CDI)[19,20]. Non-pharmacological strategies aimed at preventing stress ulcers, such as nasogastric decompression, conflict with established principles of post-operative care, particularly the early initiation of enteral feeding, which is known to maintain gut integrity, reduce infection risk, and improve overall recovery outcomes[21,22].

Despite the clear and present danger that stress ulcers pose for this patient population, there is a palpable lack of consensus on the topic and the literature regarding gastroprotection in neurocritical patients is vast and riddled with inconsistencies[23]. We aim to summarize some of the major issues and attempt a general synthesis, highlighting the key areas of disagreement, the evidence supporting various prophylactic and therapeutic approaches, and potential pathways for future research and clinical practice improvements in the management of stress ulcers in neurosurgical and neurocritical patients.

PHARMACOLOGICAL APPROACHES

H2 receptor antagonists (H2RAs) have been known at least since the 1990s to be effective in reducing stress ulcers and related bleeding in critically ill patients[24]. However, their efficacy in the neurosurgical patient cohort was unclear at the time. A prospective randomized trial of critically ill neurosurgical patients by Reusser et al[25], in which the authors evaluated the development of ulcer and bleeding endoscopically, concluded that the use of ranitidine for routine SUP may not be necessary. Over the next several years, multiple randomized trials evaluating H2RAs in high-risk neurosurgical patients came out with the conclusion that the use of H2RAs were beneficial in preventing stress ulcers and gastrointestinal bleeding[9,26-28].

PPIs are widely utilized in the hospital setting for the prevention and management of stress ulcers and UGIB. Multiple systematic reviews and meta-analyses have concluded that PPIs are generally more effective compared to H2RAs in preventing stress ulcers and UGIB in the critical care setting[29-31]. Despite the demonstrably better efficacy, safety of PPIs over H2RAs is less clear. The same meta-analyses either found increased or no difference in the rates of pneumonia in patients treated with PPIs compared to H2RAs. A large pharmacoepidemiologic study of over 35000 patients found that PPIs were associated with significantly higher rates of pneumonia and CDI, raising concern regarding the routine use of PPIs[32]. However, the PEPTIC randomized trial, involving a comparable number of patients concluded that there was no statistically significant difference in all-cause mortality or CDI in critically ill patients treated with PPIs over H2RAs[33].

When considering the critically ill neurosurgical patient, these issues are magnified further. Not only are these patients at an elevated risk but there is also a paucity of trials evaluating the aforementioned factors in this specific cohort. In our review of the literature, we found only four single center randomized trials that explicitly compared the safety and efficacy of PPIs vs H2RAs in neurosurgical patients. Three trials found no significant difference between the two classes of medications with respect to mucosal injury[34-36] whereas the other one found PPIs to be better[37]. Still, none of the studies found any difference in the rate of adverse events between the evaluated groups. A meta-analysis on the topic found that SUP proved more effective than placebo or no prophylaxis in preventing UGIB and lowering all-cause mortality, without raising the risk of pneumonia. However, the reliability of this finding was affected by “lack of trials with a low risk of bias, sparse data, heterogeneity among trials, and a concern regarding small trial bias”[38]. It is also important to consider the potential adverse effects of the concomitant use of HR2As with seizure prophylactic agents like phenytoin as well as PPIs with clopidogrel as these drugs affect the cytochrome P450 enzyme system and could lead to unpredictable interactions[39,40].

Therefore, investigating the risk-benefit profiles of these agents is crucial, given their widespread use in everyday practice. It becomes even more relevant in the context of the availability of newer acid suppressive agents such as the recently approved novel potassium competitive blocker, vonoprazan. Vonoprazan has been shown to be more efficacious than PPIs for acid suppression and having a comparable adverse effect profile for treating H. pylori infection, erosive esophagitis, and gastroesophageal reflux disease[41-43]. The recent study by Gao et al[44] on the real-world efficacy of the vonoprazan based regimen in the elderly, offers insights. The study, despite being retrospective in nature, highlights the need for investigating acid suppressive therapies and H. pylori eradication in specific populations. Extending this to the neurosurgical patient cohort, we find that there is paucity of data regarding acid suppressive regimens in different subgroups as well. For instance, age over 60 years has been identified as an independent risk factor in developing stress ulcers and UGIB in neurocritical patients[9,26]. However, several known risk factors for stress ulcer development remain unexplored, particularly the role of preexisting H. pylori infection in the pathogenesis of stress-related mucosal damage. While one prospective study by Maury et al[15] found that H. pylori infection was more common in ICU patients who experienced bleeding compared to controls, the exact nature of the relationship remains uncertain. Additionally, other established risk factors, such as prolonged non-steroidal anti-inflammatory drug therapy, mechanical ventilation, and corticosteroid use[45,46], need to be revisited within the context of specific subpopulations.

NON-PHARMACOLOGICAL APPROACHES

While pharmacological interventions are key in preventing stress ulcers, non-pharmacological strategies play an equally important role in optimizing patient outcomes. Interventions, such as nasogastric decompression, early enteral nutrition (EN), and appropriate patient positioning, are vital in the management of neurocritical patients. Among these, nasogastric tube (NGT) decompression is a common practice in the critical care setting, helping to relieve gastrointestinal obstruction, manage postoperative ileus, and reduce the risk of aspiration[47]. Although there is limited evidence, NGT decompression is also thought to potentially protect against stress ulcers by maintaining an optimal gastric pH and thus mucosal integrity. More importantly, draining gastric contents through NGT is crucial for neurocritical patients, as increased intra-abdominal pressure has been shown to elevate intracranial pressure (ICP), making effective decompression essential for preventing further complications[48-50].

Early EN is a well-established principle in postoperative care, essential for preserving gut integrity and supporting recovery[21,22]. Despite proven efficacy, intolerance to early enteral feed initiation occurs in neurocritical patients due to reduced mentation (low Glasgow coma scale, sedation) and thus are maintained on parenteral nutrition (PN) during the critical phases following central nervous system (CNS) insult. PN is not without shortcomings and is known to result in hyperglycemia, higher infection rates, and hepatic steatosis[51]. To balance the risk vs benefit of these nutritional strategies, a combination approach seems to work better than either strategy alone. Several recent studies in the neurocritical population have demonstrated that EN + PN strategies are associated with lower complications from stress ulcers, aspiration as well as fewer days of hospitalization and better nutritional status[52-54]. This raises an important question: If nasogastric decompression is crucial for preventing complications from gastric distension, and early enteral feeding is vital for preserving mucosal integrity and promoting recovery, how can these two seemingly conflicting strategies be balanced? The solution may lie in a combination approach, where a NGT is used for gastric decompression while a naso-intestinal tube is employed for enteral feeding. Recent studies have demonstrated the efficacy of naso-intestinal tubes in the neurocritical patient population, showing their ability to initiate enteral feeding while reducing gastric retention, pulmonary aspiration, and the risk of pneumonia[55,56]. However, further research is needed to better define optimal protocols and assess the long-term outcomes of such combined strategies. Ultimately, continued investigation will be essential to refine these approaches and ensure they provide the best balance between gastric decompression and nutritional support for neurocritical patients.

RECOMMENDATIONS FOR FUTURE RESEARCH

High-quality primary evidence on gastroprotection in the neurocritical care population remains scarce. Most available data on the subject are over a decade old (Table 2)[7,9,25-28,34-37], with only one recent randomized controlled trial published in 2019. Our recommendations for the direction of future research are based on the available systematic reviews and meta-analyses on the topic[38,57]. A well-designed, large, randomized trial is the need of the hour to address the issues discussed herein. Patient selection should encompass a broad spectrum of neurocritical illnesses to ensure comprehensive representation. Existing trials have predominantly focused on patients with TBI and ICH, while other conditions such as SCI, subarachnoid hemorrhage, and CNS tumors remain underrepresented. These underrepresented groups account for a substantial proportion of neuro ICU admissions and exhibit risk factors unique to their conditions. For example, corticosteroid use is typically limited to patients with CNS tumors and SCI[3], potentially resulting in a markedly different ulcer risk profile compared to TBI and ICH patients[10-13,22,58]. When it comes to interventions requiring further study, significant gaps remain in understanding the safety profiles of commonly used gastroprotective agents. For example, the association between PPI use and the incidence of nosocomial pneumonia or CDI has not been definitively established. Furthermore, we found only one trial directly comparing PPIs to no prophylaxis[37], limiting conclusions to comparisons between PPIs and H2RAs. In addition to addressing these issues, future studies could adopt multi-arm designs to explore combination strategies, such as early EN, combined NGT and NIT interventions, and the efficacy and safety of emerging agents like vonoprazan. Meta-analyses on the topic have also highlighted inconsistencies in endpoint definitions. While stress ulceration and UGIB were the most commonly studied outcomes, the methods used to assess these endpoints varied significantly. Some studies employed endoscopic evaluation of stress ulcers and gastrointestinal bleeding[8,25], whereas others relied on indirect measures such as gastric pH, occult blood, or coffee ground/frank bloody NGT aspirates[7,34,36]. To enable meaningful comparisons and conclusions about interventions, standardization of endpoints is essential. A potential standardized endpoint suggested is clinically significant bleeding, defined as endoscopic evidence of lesions leading to substantial hemodynamic instability, the need for transfusion, or surgical intervention[13,21,37].

Table 2 Summary of randomized trials on gastroprotection in neurocritical patients.

No	Ref.	Year	Study population	N	Intervention arm 1 (n)	Intervention arm 2 (n)	Intervention arm 3 (n)	Endpoints	Findings
1	Halloran et al[28]	1980	Severe TBI patients	50	Cimetidine (26)	No prophylaxis (24)	-	GIB (NGT aspirate positive for blood)	Overt GIB significantly lower in arm 1 compared to arm 2
2	Reusser et al[25]	1990	TBI and surgically managed ICH patients	40	Ranitidine (21)	No prophylaxis (19)	-	Gastric pH > 4, endoscopic findings of ulcer, blood in NGT aspirates	No difference in rate of stress ulcers or GIB between both arms
3	Metz et al[9]	1994	TBI patients	167	Ranitidine (86)	No prophylaxis (81)	-	GIB (NGT aspirate positive for blood or malena/hematochezia). Nosocomial pneumonia	Overt GIB significantly lower in arm 1 compared to arm 2. No difference in rates of pneumonia
4	Burgess et al[27]	1995	TBI patients	34	Ranitidine (16)	No prophylaxis (18)	-	GIB (NGT aspirate positive for blood or malena/hematochezia)	Overt GIB significantly lower in arm 1 compared to arm 2
5	Chan et al[26]	1995	Non traumatic neurosurgical patients	101	Ranitidine (48)	No prophylaxis (52)	-	Serial endoscopic assessment to determine bleeding requiring transfusion or surgery	Overt GIB significantly lower in arm 1 compared to arm 2
6	Misra et al[7]	2005	ICH	141	Ranitidine (45)	Sucralfate (49)	No prophylaxis (47)	GIB (hematemesis, NGT aspirate positive for blood or malena/hematochezia), 1 month mortality	No difference in GIB or 1 month mortality between arm 1, arm 2 compared to arm 3
7	Brophy et al[34]	2010	All patients admitted to the neurosurgery ICU	51	Lansoprazole (28)	Famotidine (23)	-	The pH ≥ 4 and % of time gastric residuals was < 28 mL	Arm 1 achieved pH ≥ 4 and fewer heme-positive aspirates more often than arm 2 only on day 1
8	Liu et al[37]	2013	Surgically treated ICH patients	165	Omeprazole (58)	Cimetidine (54)	No prophylaxis (53)	UGIB hematemesis, aspiration of coffee-ground material from NGT or positive gastric occult blood or fecal occult blood testing, with or without hemodynamic instability from gross bleeding needing transfusion). Nosocomial pneumonia incidence	Lower rate of UGIB in arm 1 compared to arm 2 and 3. No difference in rate of nosocomial pneumonia
9	Lee et al[35]	2014	Neuro ICU patients admitted post-op or Canonical Variates Analysis management	60	Esomeprazole (30)	Famotidine (30)	-	Overt GIB and VAP	1 patient in arm 2 developed GIB and 1 from each arm developed VAP
10	Senapathi et al[36]	2019	TBI patients with Glasgow coma scale < 10	56	Omeprazole (28)	Ranitidine (28)	-	Time to gastric pH > 3.5-5.0 and occult from NGT aspirates	No significant difference in endpoints between both arms

GIB: Gastrointestinal bleeding; ICH: Intracerebral hemorrhage; ICU: Intensive care unit; NGT: Nasogastric tube; TBI: Traumatic brain injury; UGIB: Upper gastrointestinal bleeding; VAP: Velocity average path.

RECOMMENDATIONS FOR CLINICAL PRACTICE

Given the complex nature of gastroprotection in neurocritical patients, clinical practice must involve a balanced approach that judiciously integrates both pharmacological and non-pharmacological strategies. While PPIs have demonstrated greater efficacy than H2RAs in preventing stress ulcers and associated bleeding, the potential risks, such as pneumonia and CDIs, are not yet definitively established and should not cause significant concern until further data become available. An NGT can effectively manage gastric decompression, which has the dual advantage of lowering ICP and preventing aspiration while a naso-intestinal tube may facilitate early enteral feeding, promoting recovery. Early initiation of enteral feeding particularly, is supported by substantial evidence and should be prioritized, as it seems to be the most effective strategy for preserving gut integrity and minimizing gastrointestinal complications, and prolonged hospital stays.

CONCLUSION

Neurocritical patients represent a unique subgroup of critically ill individuals with distinct risks for stress ulcers and GIB. Consequently, findings from other critically ill cohorts may not directly apply to this vulnerable population due to their differing risk profiles. Current clinical practice guidelines are based on studies that are outdated and lack comprehensive evidence in many areas. There is an urgent need for high-quality studies to evaluate gastroprotective strategies specific to neurocritical patients. As we advance toward more personalized care, tailored approaches are essential to address their unique clinical challenges, optimize outcomes, and minimize complications.