Whistler Summary: “The Slow Rate of Rapid Progress”

Abstract

Gastroesophageal reflux disease (GERD) is now the most common diagnosis made in gastroenterology outpatient practice. Increasing time and resources are spent on its management each year. Many conditions associated with, or the result of, GERD also appear to be on the rise—the most worrisome, esophageal adenocarcinoma, is steadily increasing in incidence, although it remains an uncommon cancer. Now that proton pump inhibitors (PPIs) are widely available, including over the counter in some countries, patients seen in consultation by gastroenterologists are becoming ever more challenging. In view of these trends, a meeting was convened in Whistler, British Columbia, Canada, in the fall of 2006 to review the current state-of-the-art approach to GERD, and to highlight areas of uncertainty in this field. The aim was to focus attention on patient needs, in an effort to provide better symptomatic care. The discussions were free ranging, on a variety of the most difficult problems, and the chairmen were instructed to lead the discussions into the unmet needs. The following pages highlight a summary of these discussions. THE SCOPE OF THE PROBLEM OF GERD AND SPECULATION ON ITS ETIOLOGY GERD is a well-established condition in western countries, and continues to grow in importance. The increase in incidence is now also emerging as a worldwide phenomenon, with GERD and its complications increasingly diagnosed in parts of the world undergoing rapid development, such as in Asia.1 In the United States, GERD has overtaken abdominal pain as the most common gastroenterologic diagnosis made during outpatient office visits. Spending on PPIs accounts for over $10 billion per year in the United States, with 2 of the top 5 best-selling drugs being PPIs.2 Regrettably, there is a paucity of formal cost-benefit analysis to determine whether the current expenditure on GERD (principally on its therapy) represents an efficient use of increasingly scarce healthcare resources. The limited evidence available suggests that erosive esophagitis can be treated more cost-effectively than can nonerosive reflux disease (NERD). The most likely explanation for the increase in GERD is the increasing weight of the population. Weight gain, overweight, and obesity are risk factors for the occurrence of reflux symptoms, erosive esophagitis, and esophageal cancer.3,4 A dose-response relationship between body mass index and reflux symptoms was clearly demonstrated in a recent analysis of the Nurses Health Study, which also showed that the risk of reflux symptoms receded with even modest weight loss.5 Whether certain specific constituents of the diet are to blame for GERD has been the subject of speculation, but consistent associations have not been established. Nevertheless, most evidence supports the idea that citrus, fatty, and spicy foods increase symptoms in GERD, and that a high fiber diet leads to, among other benefits, a decrease in reflux symptoms. Whether this effect is through the relief of constipation or other mechanisms is not known. For coffee lovers, the good news is that this beverage has mixed effects; although it increases acid secretion it aids in the activation of PPIs. The decision to remove particular items from the diet is best made on an individual basis. In the context of the worldwide epidemic of obesity, a holistic approach would be to use the development of GERD as an opportunity to advise the patient on the appropriate interventions for weight control, stressing that this may reduce overall health risks. It remains controversial whether the worldwide trend of declining Helicobacter pylori prevalence is at all important in the GERD epidemic. It seems probable that the loss of this chronic gastric infection in populations leads to more healthy stomachs as the prevalence of chronic gastritis declines. This results in increased acid secretion, which combined with increasing body weight and declining physical activity may result in increasing esophageal acid exposure and an increasing incidence of GERD. Few would advocate not eradicating this bacterium once diagnosed, especially in areas where gastric cancer is common, in view of its established association with gastric neoplasia. However, a different and much debated viewpoint exists that argues against this position. Thus, in areas where gastric cancer is uncommon and adenocarcinoma of the esophagus is increasingly rapidly, there is growing support for not testing for H. pylori or treating it when present, as it may exacerbate GERD symptoms and increase the PPI dose required.6 Numerous medications may cause mucosal injury and esophagitis, but for the most part they do not cause GERD. However, in patients with preexisting GERD, especially when accompanied by a hiatus hernia or lower esophageal sphincter dysfunction, a large number of medications, notably aspirin, NSAIDs, calcium channel blockers, and antiasthmatic drugs, may exacerbate or seriously complicate the course of GERD. The clinical or epidemiologic associations of GERD with various medications may be confounded by indication—for example, the association of the use of some antiasthmatic drugs with having reflux disease—because reflux is common to the development of both GERD and asthma. SYMPTOMS, DEFINITIONS, AND DIAGNOSTIC UNCERTAINTY—GERD, NERD, AND FUNCTIONAL HEARTBURN It is well recognized that there is often dissociation between a patient's GERD symptoms, the endoscopic findings, and measures of acid reflux. Indeed, the previous heavy reliance upon “heartburn” as the cardinal symptom of reflux disease is increasingly seen as flawed, especially in some cultures and languages. Patients with heartburn often have other upper gastrointestinal symptoms, and recent large studies among primary care practitioners show that few distinguish between heartburn and dyspepsia.7 In addition, there is a considerable overlap between symptoms of GERD and those of other functional abdominal disorders, including irritable bowel syndrome. Other complaints also impact their quality of life, including sleep disturbance from overt reflux, or from multiple amnestic episodes leading to sleep fragmentation.8 Recent data suggest that 1 in 4 patients presenting with sleep disturbance may have GERD without classical reflux symptoms.9 This recognition of the multiple clinical consequences of excessive reflux has led to recent efforts to redefine GERD more holistically, most recently as “a condition which develops when the reflux of stomach contents causes troublesome symptoms and/or complications.”10,11 The change in emphasis away from a precise definition of GERD based on frequency of symptoms, toward a patient-based definition relate to the impact of symptoms and/or the presence of complications reflects the patient's perception of variable and multiple symptoms and has ramifications with respect to impairment of quality of life and consumption of health care resources. Symptoms do not distinguish patients with erosive GERD from those who have NERD. Both groups suffer similar severities of impairment of their health-related quality of life and require similar treatment. The increased appreciation of a broad symptom complex in patients with GER has renewed focus on clinical aids to diagnose and assess treatment outcomes using patient questionnaires. A recent review has highlighted the inadequacy of most reflux questionnaires.12 The recently validated ReQuest questionnaire allows patients to document the broad array of symptoms seen in GERD patients, clustered in symptom groups or dimensions, and allows accurate tracking of changes in symptoms with PPI treatment. ReQuest is simple to use, patient driven, and responsive to prompt changes in symptoms.13 It has been successfully used in clinical trials, and has potential as a tool in clinical practice. Whether some version of it may be adapted as a diagnostic tool is yet to be determined. NERD has been considered a heterogeneous syndrome that may include patients with microscopic inflammation, those with an esophagus sensitive to physiologic amounts of reflux and those with functional heartburn in whom nonacidic or weakly acidic reflux may play a role.14 Some NERD patients respond less well to antisecretory therapy. It has become common experience that it is easier to heal esophageal erosions than to abolish symptoms. Indeed, the symptomatic response to PPI therapy in endoscopy-negative patients is significantly inferior to responses in those with erosive disease whose symptoms more consistently respond to acid suppression.15 The development of appropriate diagnostic pathways and alternative treatments for patients considered to have NERD but failing initial PPI treatment, remain areas of pressing unmet need. In practice, this group may include patients with functional GI disease, visceral hypersensitivity or anxiety or depression, who present with somatic symptoms that mimic reflux but do not respond to PPI therapy. The goals of GERD management in primary care patients are to recognize and treat symptoms, reduce risk, and restore quality of life. Symptom pattern evaluation is the key to initial decision making in most cases. However, a particular difficulty, often inadequately addressed by reflux guidelines, is the significant overlap of reflux symptoms with those of other disorders. Epigastric pain for example may be the sole symptom of GERD, peptic ulcer disease, or functional dyspepsia. Diagnostic algorithms need to consider alternative diagnoses and issues related to risk reduction rather than just short-term reduction of symptoms. In this context, a strategy for H. pylori infection needs to be individualized depending on a history of peptic ulcer and the local prevalences of infection and gastric cancer in the community. Another particular challenge is how to determine if GERD is the cause of atypical or extraesophageal symptoms. A trial of PPI therapy with an assessment of response is often assumed to have significant diagnostic value for GERD. However, a meta-analysis revealed the sensitivity and specificity of this “PPI test” to be only moderate,16 and diagnostic uncertainty remains a major driver of referral for endoscopy in practice. Although early referral for endoscopy or once-in-a-lifetime endoscopy have proponents as appropriate strategies for the management of reflux in primary care, neither are standard recommendations of consensus groups based on current evidence, nor have they proven to be cost-effective. In referral centers, the focus is on newer diagnostic technologies in an effort to gain greater accuracy in the diagnosis of reflux disease. The “gold” standard for the presence or absence of acidic reflux is the result of 24-hour pH-metry, not conventional endoscopy, results of which are normal in up to 70% of cases. Conventional catheter-based pH-metry is dependent on a pH probe positioned 5 cm above the gastroesophageal junction, and reflux episodes recognized when the pH is less than 4 at this location. However, because the esophageal nerves respond to a pH of 5.0, this standard may be misleading in terms of acid-related symptoms. Alternative positioning of the pH probe, prolonged pH monitoring, or the addition of motility, impedance studies or bilirubin measurements, may all have roles but are not universally available, and have yet to find a validated and cost-effective place in routine care. Similarly, newer endoscopic technologies such as chromoendoscopy, narrow band imaging, and magnification endoscopy aimed at increasing the diagnostic yield of endoscopy, are still to be properly evaluated. Recently, the concept of complete remission of reflux disease has been studied in the context of improving the rigor of outcome assessment in clinical trials of PPIs. Combining the modified LA classification of esophagitis to measure healing of esophageal erosions, with the ReQuest symptom questionnaire to measure symptom relief, has revealed a lower rate of complete remission (defined as healing of erosions and symptom relief) than reported using either endoscopically confirmed healing or symptom response alone.17 Such a tool may become the gold standard for comparative studies of PPIs, and indeed for the assessment of any newer therapeutic modalities in GERD. PATHOGENESIS OF SYMPTOMS IN REFLUX DISEASE Although conceptually, it has been easy to envisage how mucosal damage and erosion can allow gastric acid access to mucosal sensory nerve endings causing what is then perceived as pain, the mechanisms leading to pain perception in NERD remain uncertain. One view is that basically similar mechanisms are responsible for symptoms in NERD and in GERD (by the believers in “microscopic GERD”), whereas others believe that there is a primary perturbation in pain perception in NERD, such as visceral hypersensitivity. These theories are not necessarily mutually exclusive, although the relative contributions of the 2 mechanisms and the degree to which the 2 mechanisms occur in the same patients have not been evaluated, because oral medications able to consistently elevate esophageal pH to 5.0 or greater especially at night are not as yet available. Dilated intercellular spaces, described many years ago in erosive GERD, have more recently been also described in NERD. Microscopic evidence of inflammation (increased neutrophils and other inflammatory cells and increased proinflammatory cytokine production) may be present in biopsies from NERD patients, even when the esophagus looks normal macroscopically. The harder one looks, the more abnormalities one may find. High-resolution endoscopy with narrow band imaging, chromoendoscopy, and other magnification modalities are all being used to determine the extent to which subtle anatomic abnormalities exist in NERD. The fact that the resolution of symptoms does not correlate well with the resolution of erosions in GERD indicates that there is discordance between the presence of macroscopic damage and symptoms. This may be due to the capacity of the esophageal epithelium to resist acid owing to pH regulatory processes in these cells while the sensory nerves lack such mechanisms, and likely express acid sensitive ion channels that are triggered at pH 5.0. There is an important relationship between the esophagus and the brain, in at least some patients. For example, functional magnetic resonance imaging has demonstrated that esophageal acid perfusion can lead to alterations in brain imaging that are quantitatively different in GERD patients and normal controls.18 Visceral hypersensitivity has long been recognized in NERD. There are likely 2 elements involved, one is local and may be related to esophageal microscopic damage, and the second is central, involving hypervigilance. Visceral sensation can be modulated locally: peripheral sensitization induced by acid in the esophagus (or inflammation anywhere in the GI tract) can alter the intensity of perceived sensory responses to other stimuli—this “secondary hyperalgesia” has been shown by esophageal acid perfusion in the distal esophagus sensitizing the esophagus to pain from balloon distention proximally.19 Psychologic stress does not increase GER, but it does promote hypervigilance and translation to consciousness. This has been shown by the effect of provoking anxiety, through viewing images of a fearful face, on perceived gastric sensation.20 Important questions relate to whether or how esophageal sensitization can become permanent, and whether it can lead to a global derangement of visceral hypersensitivity—issues that would seem to be relevant to the high incidence of overlap between GERD and other functional GI disorders. What are the peripheral neurophysiologic abnormalities responsible for visceral hypersensitivity and altered motor abnormalities in GERD? Increasing evidence exists for alterations in gastroesophageal neuromuscular function in reflux disease. For example, GERD patients exhibit earlier and wider opening of the gastroesophageal junction in response to balloon dilation, especially in those with a hiatus hernia.21 Furthermore, the intragastric distribution of meals is altered in GERD patients and when postprandial gastric emptying is delayed, the occurrence of transient lower esophageal sphincter relaxations is prolonged, and there are more relaxations with acid reflux events. Although ineffective esophageal peristalsis has been recognized for some years, it has recently become clear that only severe impairments contribute to excessive esophageal acid exposure. The pathophysiology of symptom development in reflux disease has generally been investigated by individuals or teams focused on a particular mechanism. What is needed are large multiparameter investigations of patients from across the whole spectrum of GERD. There should be simultaneous measurement of abnormal esophageal acid exposure, changes in impedance, bile reflux, and careful and detailed manometric measurement, together with high-resolution endoscopy to evaluate histology, and measurement of inflammatory mediators; studies of psychologic and of both central and peripheral nerve function should also be carried out in these same patients, a demanding task even for very specialized centers. Only then can one attempt to categorize symptom groups and correlate them with measurable abnormalities, and start to evaluate whether a “lumping” or a “splitting” approach to the spectrum of GERD is appropriate. Such an undertaking may be of more than academic importance, as it is likely that with the future development of new classes of drugs to treat GERD, the “one size fits all” concept of acid suppression alone that is currently employed clinically may be superceded by the use of agents that target peripheral or central neurotransmission, address underlying motility abnormalities, suppress inflammation, or deal with specific components of the gastric refluxate. BARRETT ESOPHAGUS AND ESOPHAGEAL CANCER Recent United Kingdom census data indicate that esophageal cancer has now overtaken gastric cancer as the fifth commonest cause of cancer death. Furthermore, the majority (70%) of esophageal tumors are adenocarcinomas.22 The dominant model invoked to explain the increasing incidence of esophageal adenocarcinoma in westernized countries in the last few decades is the sequence from reflux esophagitis through Barrett esophagus (BE) with intestinal metaplasia, and thence to dysplasia and cancer. The progression paradigm has been instrumental in driving recommendations to perform surveillance in patients with BE to detect dysplasia and to prevent cancer. However, certain observations continue to sit uncomfortably with the reflux-Barrett's-cancer model. For example, patients without GERD at initial endoscopy do not appear to develop it subsequently, even with prolonged follow-up for up to 20 years. This casts doubt on GERD being the cause of BE. In addition, most esophageal adenocarcinomas are diagnosed in patients without a prior diagnosis of BE, and many occur in the absence of any visible intestinal metaplasia in resected specimens. Could BE represent a congenital lesion that, in the presence of reflux disease predisposes to cancer, rather than being primarily due to GERD?23 Are there genetic or infectious influences that predispose to cancer only in a minority of cases of BE? Some uncontrolled circumstantial evidence points to a role for acid inhibition in preventing progression of BE to dysplasia and adenocarcinoma24 and has been justified by some as a rationale for employing high-dose PPI therapy in patients with BE as a “chemopreventive” strategy. More compelling as a chemopreventive strategy would be the use of aspirin, because epidemiologic studies have consistently shown an inverse relationship between the use of this drug and the occurrence of esophageal adenocarcinoma. A large multicenter and multifactorial study to assess the relative merits of aspirin and acid inhibition in preventing cancer in patients with BE has recently begun and will also have arms with and without endoscopic surveillance. The investigators responsible for this trial are to be congratulated for initiating a trial that may provide a base of evidence for future management of BE.25 Longitudinal studies to establish the natural history of reflux disease are also important in defining whether the commonly quoted 0.5% risk of cancer development in BE is accurate. Demographic and clinical features associated with cancer in BE and in the predisposition to esophageal adenocarcinoma remain poorly defined, and further studies are needed to target the population at risk. Progress in understanding the molecular pathogenesis of esophageal adenocarcinoma remains disappointingly slow. This field is hampered, in part, by a lack of suitable animal models. The predilection for cancer to arise at the gastroesophageal junction may be partly explained by the highly acidic local environment converting the nitrite, derived from dietary sources or saliva after bacterial conversion in the mouth, to potentially mutagenic nitrosating compounds.26 THERAPY The success of PPIs in healing erosions and improving symptoms in GERD has been impressive. For physicians, erosive esophagitis has been a satisfactory entity to treat because there is a relatively predictable endoscopic response, usually accompanied by improvements in symptoms. However, it is clear that even among patients whose erosions heal, a proportion continue to have significant symptoms. The shift of focus from healing endoscopic lesions to symptom relief and improving quality of life (as assessed by global questionnaires) paves the way for improved GERD therapies in the future. The mainstay of GERD therapy now, and for the foreseeable future, has been to reduce the acidity of the gastric refluxate. However, for reasons that remain unclear, a proportion of patients, particularly in studies from the United States, have little or no response to PPIs, even allowing for poor compliance with medication. In the oft-quoted American Gastroenterology Association's survey, 53% of patients reported that they were amenable to improved symptom control, and the majority of those polled reported that heartburn at night was particularly troublesome.8 It is unknown how many among these “PPI nonresponders” would benefit from improved acid inhibition. Although we have a generally agreed pH target for healing erosive esophagitis, there are no data defining the pH target for optimal symptom control. Currently available oral PPI therapy cannot maintain intragastric pH above 4 throughout the 24-hour period. When PPIs are given at the approved time of 30 to 45 minutes before breakfast, or the evening meal if on twice a day therapy, gastric pH at night usually becomes more acidic (termed by some “nocturnal acid breakthrough” or “nocturnal acid recovery”). This is because the half-life of the drug is only 60 to 90 minutes and thus it is not present when pumps are active at night. Nevertheless, in many individuals the long-lasting effect of the covalent inhibition of the pump is sufficient to elevate nighttime acidity and prevent nocturnal symptoms. The importance of this acid breakthrough in the stomach with respect to control of GERD symptoms is uncertain, given that what is important is intraesophageal pH. Most studies to date show little if any relationship between nocturnal intragastric pH, nocturnal intraesophageal pH, and symptoms of reflux. Perhaps only when effective control of acid secretion can be achieved around the clock, it will be apparent whether the patient's nocturnal symptoms are indeed owing to nocturnal esophageal acid exposure. The potential adverse effects of 24-hour control of acid secretion such as sustained hypergastrinemia, enterochromaffinlike cell hyperplasia, and increased susceptibility to enteric infections also remain to be determined. PPIs have a relatively slow onset of action owing to the requirement for activity of their target, the proton pump, and the impossibility of activating all pumps at any one time. Thus there is interest in whether agents that elevate pH more rapidly could replace them. Such agents would be particularly attractive for on-demand therapy. Another issue that still remains unclear, despite vast experience with PPI use over the past 2 decades, is how to get patients to decrease their dose once they have been treated with a PPI. Finally, are PPIs to be considered lifelong treatment? While a “step-down” or tapering approach is usually attempted, only a minority can remain off all treatments. In practical terms, patients often take their PPIs on-demand but how this compares with continuous long-term low-dose therapy remains to be determined. New acid inhibitory drugs are in development and may be of value in the clinical setting. Extension of the plasma dwell time of PPIs employing a new generation of longer half-life PPIs or using a new class of chemically metered absorption derivatives of PPIs can provide more prolonged and profound acid inhibition around the clock.27 Competitive inhibitors of the proton pump (acid pump antagonists or potassium competitive acid antagonists) are also being evaluated; they require continuously high plasma levels because they block parietal cell secretion reversibly. However, because they arrest the pump cycle as soon as they bind and do not require acid secretion, they provide a higher intragastric pH more rapidly than PPIs. Their structure is such that they can readily be formulated as time-release tablets so that their plasma half-life is greatly prolonged. Improved understanding of the biology and pharmacology of the proton pump may also aid in the development of acid inhibitory compounds. For example, understanding how the pumps traffic through the cell (the beta subunit appears important), and how precisely secretion is appropriately terminated in the parietal cell, are still largely unknown. Naturally occurring compounds (such as ouabain for the Na, K ATPase) that inhibit various other members of the P2-type ATPases have been identified, but a natural or endogenous ligand for the proton pump (another member of this class of ion pumps) has not yet been found. Of course, the principal abnormality in GERD is not excessive acid production, but rather the failure of normal physiologic processes regulating the operation of the lower esophageal sphincter and the clearance of refluxate from the esophageal lumen. There has been an increasing interest in understanding abnormalities of motility in the stomach, gastroesophageal junction, and esophageal body that ultimately cause reflux and increased attention is being given to the peripheral and central sensory pathways involved in the mediation of esophageal pain. Understanding the complex neurophysiologic pathways involved should lead to the identification of receptors and other pathways that could be potentially targeted by novel therapies, including blockade of esophageal pain receptors. Promotility agents that increase esophageal acid clearance and reduce the volume of reflux, while restoring lower esophageal sphincter pressure, remain in the early stages of clinical trials. It is noteworthy that the γ-amino butyric acid-B agonist, baclofen, showed a beneficial effect on the lower esophageal sphincter even though its effects on reflux have been minor and its clinical use limited by side effects. Analogs of this compound are in development. Endoscopic and surgical therapies for the lower esophageal sphincter are possible and are intuitively attractive, because they appear to address the root cause of the disease. However, most patients find that surgery fails to fully resolve their symptoms and they frequently require ongoing medical therapy for reflux, in addition to acquiring new symptoms from the surgery itself.28 Several endoscopic approaches to improve LES function have been developed in recent years but have not been evaluated extensively in formal randomized long-term clinical trials. Indeed, some have been withdrawn from the market owing to the unexpected complications and this has even led to a plea for a moratorium on further endoscopic therapies until they have been extensively evaluated in appropriately designed clinical trials.29 Certainly, surgical or endoscopic therapy should only be performed after careful evaluation and patient selection, rigorous training of operators, and after appropriately informing patients of possible outcomes. One group to benefit might be those with troublesome confirmed nocturnal reflux unresponsive to PPI therapy. CONCLUSIONS Many challenges remain for the management of GERD and its complications. Among the most pressing, are whether clinical or pathophysiologic subcategorization of patients within the GERD spectrum is useful. This might have implications for therapeutic approaches. If not, a more global approach to patients with the many symptoms of reflux should be considered. Improving our understanding of the neuronal circuitry between the brain and the esophagus that are important in the etiology of symptoms appears worthwhile. There is a pressing need to develop therapies that provide better symptom relief than do currently available drugs. A more holistic attitude to the GERD patient may help us to determine whether comprehensive control of symptoms is feasible. The growing incidence of esophageal adenocarcinoma is serious and important. In the past, there has perhaps been overreliance on the esophagitis-Barrett-dysplasia-cancer sequence, but it is clear that, although this dictates current practice, it has not been a particularly useful model in esophageal cancer prevention. Lastly, large multicenter studies have been established to address the natural history of Barrett's, to determine those at highest risk of cancer, and to establish the potential impacts of chemoprevention or endoscopic surveillance on the outcome of the condition. For the present, as we do not know how to prevent esophageal cancer, our goal should be to strive for a patient who does not voluntarily complain about reflux symptoms.