Coronary artery disease and reflux esophagitis are common clinical disorders [3]. In their classic presentations, coronary ischemia with exertional angina and reflux esophagitis with heartburn and acid regurgitation can be easily diagnosed. However, the pain in these two disorders is usually atypical. Because coronary artery disease is so serious, patients who have atypical chest pain are first evaluated for this disorder. Patients with strong indications of coronary artery disease (exertional angina-like pain, cardiac risk factors, and abnormal stress test results) have coronary angiography. However, one third to one half of these patients have no substantial coronary lesions and are categorized as having noncardiac chest pain. In this issue, Frobert and colleagues [1] report the results of their investigation into how frequently reflux esophagitis occurs in selected patients with noncardiac chest pain.

Using long-term pH monitoring, the authors found no difference in the esophageal reflux index between patients and controls. In this series, however, 16% of patients and no controls had a positive esophageal acid perfusion test result, suggesting that these patients may have reflux esophagitis or a so-called irritable esophagus [4]. Other studies [4, 5] have reported higher frequencies of reflux esophagitis in patients with noncardiac chest pain than in controls. These differences may reflect meaningful biological associations but may also indicate some bias in the selection of patients who had coronary angiography and the lack of a uniform definition of noncardiac or unexplained chest pain. Thus, although routine testing for reflux esophagitis in patients with noncardiac chest pain (as defined by Frobert and colleagues [1]) may not produce pain, it may be prudent to rule out reflux esophagitis by a therapeutic trial with acid-suppressive therapy, such as omeprazole, 40 mg/d for 4 weeks.

Chest pain may also be caused by esophageal motility disorders. For more than two centuries, clinical investigators have hypothesized that esophageal muscle spasm may cause chest pain [6]. The advent of esophageal motility studies established that in some patients, chest pain is caused by forceful, nonperistaltic esophageal contractions, known as diffuse esophageal spasm [7]. Classic diffuse esophageal spasm, however, is uncommon, and most patients with noncardiac chest pain have only subtle motility abnormalities [8].

Researchers who believe that abnormal esophageal muscle contraction is a cause of chest pain have sought support for their view through three different types of investigation. First, improved techniques for quantifying esophageal pressures showed that patients with noncardiac chest pain can generate high pressures. The esophagus that generates high peristaltic pressure was given a theatrical name, "nutcracker esophagus" [8]. However, whether vigorous peristaltic contractions can actually cause the chest pain remains unclear [9]. Nutcracker esophagus is a manometric diagnosis and may not be responsible for chest pain.

Second, provocative tests that reproduced patients' pain were used to determine whether the pain correlated with changes in esophageal motility. Esophageal acid perfusion that produced heartburn did not produce substantial esophageal motor abnormalities. The original belief that heartburn was caused by esophageal muscle spasm [10] could not be substantiated. Similarly, experimental stress did not produce major diffuse spasm, as was originally speculated [6]. Of the many pharmacologic agents that can affect esophageal motility or cause chest pain (including gastrin, ergonovine, and bethanechol), only edrophonium was thought to be reliable and safe for provocative testing [8]. In Frobert and colleagues' study [1], 16% of the patients developed pain after receiving edrophonium. This drug is a cholinesterase inhibitor that prolongs the action of acetylcholine. Interpretation of the results of edrophonium testing is complicated by the observation that cholinergic stimulation may dilate coronary arteries by releasing endothelial-relaxing factor but may also constrict atherosclerotic vessels. Thus, a positive response to edrophonium provocation does not establish esophagism as the cause of chest pain. Moreover, edrophonium frequently induces chest pain without affecting esophageal motility. At one time, ingestion of ice-cold fluids was used as a provocative test for diffuse esophageal spasm. Although such drinks did produce chest pain, they also inhibited the contractile activity of the esophageal smooth muscle [11]. This finding casts further doubt on the relation between pain and esophageal muscle spasms.

Third, a method for long-term recording of esophageal motility in an ambulatory setting was developed to correlate the patient's experience of spontaneous chest pain with esophageal motility abnormalities [8, 9]. No convincing motor abnormalities that could explain the chest pain have been found with this method. The study by Frobert and colleagues [1] in particular showed no differences in 24-hour motility studies between controls and patients and in patients before and during episodes of spontaneous chest pain.

What conclusions can be drawn about the role of esophageal motility disorders in chest pain? It is clear that the vigorous contractions induced by esophageal food impaction, classic diffuse esophageal spasm, and achalasia can cause chest pain. However, these conditions are easily suspected because of associated symptoms of dysphagia and regurgitation. Classic esophageal spasm without dysphagia or regurgitation is uncommon. As Frobert and coworkers conclude, searching for esophageal motility disorders in patients who have noncardiac chest pain but do not have dysphagia or regurgitation is not likely to be fruitful. The edrophonium test is also not useful in establishing esophageal motility disorder as the cause of chest pain.

Recently, the search for the cause of unexplained chest pain has gone beyond the specific viscera, including the heart and esophagus. As we near the close of the twentieth century, focus has shifted from this "planar" consideration of the sources of chest pain to a more three-dimensional concept: Chest pain is considered to arise not only in a specific organ but also in the nociceptive afferent nerves and the modulatory, perceptual, and localizing pain pathways in the central nervous system [3, 12]. Thus, chest pain could be caused by a true noxious occurrence in the esophagus itself, a decrease in the threshold of esophageal sensory nerve endings (nociceptors), or an intrinsic disorder of nociceptive pathways in the central nervous system. This scheme can be likened to the sounding of a fire alarm in the case of a real fire, a sensitive alarm activated by smoke without fire, or a defective alarm, respectively.

Changes in the perceptual threshold of esophageal nociceptors can be investigated by stretching the esophageal wall with a balloon. Esophageal balloon distention has been known to cause chest pain since the late 1800s [13], but the use of this technique to distinguish esophageal from cardiac chest pain has not been very successful [14]. Recently, graded distention of an esophageal balloon was used to determine whether patients with unexplained chest pain are more sensitive to distention than controls. Patients in these studies are reported to have exaggerated perception of pain compared with controls [15]. However, the results of such studies have not been consistent. This inconsistency may be due to the fact that balloon volume is not a reliable stimulus for evoking a nociceptive response through esophageal tension because, among other factors, the variable diameter of the esophagus in different persons results in a nonuniform tension stimulus.

In this issue, Rao and colleagues [2] report the findings of studies in patients with unexplained chest pain in whom ischemic heart disease, reflux esophagitis, classic esophageal motility disorder, and other obvious causes of chest pain had been excluded. They found that these patients have discomfort at lower pressures than normal persons and concluded that patients with unexplained chest pain have reduced thresholds for pain in the esophagus. In addition, however, even pressure measurements do not provide a good stimulus measure of the tension. However, the "barostat" used by Rao and coworkers also included an impedance planimeter, which estimates the diameter of the esophagus and thereby allows calculation of wall tension. Examination of Rao and colleagues' data on circumferential wall tension (Rao SS. Personal communication) showed that although patients with unexplained chest pain as a group have chest symptoms at esophageal wall tension values lower than those in normal controls, the individual values of the esophageal tension that elicited chest sensations in the two groups almost completely overlapped. Rao and colleagues also report that in patients with unexplained chest pain, the esophagus has a larger diameter, is less compliant, and is more reactive than the esophagus in controls. These data provide substantial evidence for esophageal hyperalgesia in patients with unexplained chest pain, although the esophageal distention test may not be useful in an individual patient because of the overlap of results with those of normal controls. Studies such as that by Rao and coworkers are important because they provide a technique for applying a quantifiable stimulus to mechanosensitive nociceptors.

From a clinical perspective, patients with noncardiac chest pain (as defined by a normal coronary angiogram) in whom clinically diagnosable reflux esophagitis and esophageal motility disorders such as achalasia and classic and diffuse esophageal spasm have been excluded should be categorized as having unexplained chest pain. In patients with unexplained chest pain, neither edrophonium nor esophageal acid perfusion graded balloon distention testing are currently helpful in establishing the esophagus as the cause of chest pain.

Where do we go from here? One of the important questions to be answered is whether the hyperalgesia seen in patients with unexplained chest pain is confined to the esophagus or whether other organs are also involved. Some evidence suggests that patients with unexplained chest pain also have hyperalgesia of the heart [16]. Further studies are needed to identify whether the hyperalgesia is caused by changes in the biomechanical properties of the esophagus and its reactivity or by all forms of esophageal stimuli [4], including excessive release of chemical mediators of nociception such as bradykinin, or whether it is caused by true sensitization of esophageal nociceptors.

More information on esophageal nociceptors themselves is also needed to understand the mechanisms of esophageal hyperalgesia. Recent quantitative stimulus-response studies in experimental animals have delineated three types of tension-sensitive sensory receptors in the esophagus [17]. Low-threshold bradykinin-insensitive mechanoreceptors are likely to mediate only non-noxious reflex activities; high-threshold bradykinin-sensitive mechanoreceptors are likely to be specific nociceptors; and intermediate-threshold receptors with a wide dynamic range are active in both non-noxious and noxious ranges of tension stimuli [18]. These and other studies [19, 20] have provided new insights into visceral nociception. Investigation of the central nociceptive pathways for visceral pain has also recently become possible [21] and may provide useful information on the cause of unexplained chest pain. Some of this work has suggested that unexplained chest pain may be a general disorder of visceral nociception that is also associated with behavioral problems [22]. Thus, although details of the mechanism cannot yet be described, it appears safe to predict that in the next century, specific disorders of the nociceptive sensory pathways themselves or their modulators will be identified as the cause of otherwise unexplained chest pain.