Portal de Periódicos da CAPES

Radiofrequency energy treatment of GERD

2003; Elsevier BV; Volume: 125; Issue: 3 Linguagem: Inglês

10.1016/s0016-5085(03)01132-6

1528-0012

Peter J. Kahrilas,

Esophageal and GI Pathology

This issue of Gastroenterology contains a landmark paper by Corley et al.1Corley D.A. Katz P. Wo J.M. Stefan A. Patti M. Rothstein R. Edmundowicz S. Kline M. Mason R. Wolfe M.M. Improvement of gastroesophageal reflux symptoms after radiofrequency energy a randomized, sham-controlled trial.Gastroenterology. 2003; 125: 668-676Abstract Full Text Full Text PDF PubMed Scopus (300) Google Scholar reporting on a sham-controlled trial of radiofrequency energy treatment (the Stretta [Curon Medical, Sunnyvale, CA] procedure) of gastrointestinal reflux disease (GERD). As a quick synopsis, the study concludes that, “… (Stretta treatment) improved GERD symptoms and quality of life compared with a sham procedure, but did not decrease esophageal acid exposure or medication use at 6 months.” Given the seeming banality of this result, why is this such a landmark study? The answer to that lies not in the impact of the findings, but in what they reveal about the approval process for device-based as opposed to pharmaceutical-based therapeutics. Since the entry of the Stretta procedure (and a myriad of other intraluminal therapies for reflux disease) onto the therapeutic landscape almost 4 years ago, patients and physicians have witnessed a barrage of news reports, demonstrations, presentations, and testimonials regarding the novelty and efficacy of these therapies. Now, 4 years later, we have the first modicum of controlled clinical data on any one of them and it reports that, “… (Stretta treatment) improved GERD symptoms and quality of life compared with a sham procedure, but did not decrease esophageal acid exposure or medication use at 6 months.” No doubt important findings, but hardly worth highlighting in the evening news. So how is it that, although a therapy was “approved” more than 3 years ago, we now find ourselves critically evaluating its effectiveness for the first time? The answer to this quandary lies in the absence of a level playing field regarding the evaluation and monitoring process for pharmaceutical as opposed to device-based therapeutics in the United States. Whereas pharmaceuticals undergo great scrutiny and subsequent oversight by the Food and Drug Administration (FDA) regarding their initial approval and subsequent marketing practices, devices (such as Stretta) do not. As a medical device, the Stretta system is regulated under a risk-based system that classifies instruments from I (lowest risk) to III (highest risk). Stretta is a class II (moderate risk) device. Class II devices come to market through the 510(k) process requiring supporting data that the device is “substantially equivalent” to a predicate device (one marketed before the current law or one previously found to be equivalent to such a medical device). Via this process, the Stretta System was cleared by the FDA on April 18, 2000 as follows: “the Stretta System is intended for general use in the electrosurgical coagulation of tissue and intended specifically for use in the treatment of GERD.” The “substantial evidence” reviewed by the FDA (www.fda.gov/cdrh) was an open-label, single-arm, nonrandomized clinical study involving 47 patients in 6 clinical sites who were treated and followed up for 6 months.2Triadafilopoulos G. Dibaise J.K. Nostrant T.T. Stollman N.H. Anderson P.K. Edmundowicz S.A. et al.Radiofrequency energy delivery to the gastroesophageal junction for the treatment of GERD.Gastrointest Endosc. 2001; 53: 407-415Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar From that trial it was concluded that: “Overall, the clinical study results show that patients tolerate the Stretta treatment of GERD well without general anesthesia, with shorter hospital stay, and with less recovery time compared to published reports on fundoplication. Thus, the clinical data show that the Stretta device is safe and effective for treatment of GERD and the risk-benefit profile is substantially equivalent to that of fundoplication surgery.” In the view of this author, that was an extremely premature judgment. With a disease that can have a 60% placebo response rate,3Kahrilas P.J. Gastroesophageal reflux disease.JAMA. 1996; 276: 983-988Crossref PubMed Google Scholar it is simply not good enough to equate a 60% symptomatic response in an uncontrolled trial to existing data on alternative therapies that utilized different trial design, different clinical endpoints, and very different patients. Although this author is sometimes critical of fundoplication surgery, that criticism stops far short of equating it to Stretta treatment. Published data have consistently demonstrated that fundoplication surgery can produce excellent results in the treatment of esophagitis. However, no such claim can be made for Stretta treatment. Radiofrequency treatment in GERD has never been adequately compared to medical therapy, surgical therapy, or a true control. Perhaps the last clause is surprising given the object of this editorial, but remember, the current trial compared Stretta treatment and as occasion requires (prn) medical therapy to Sham treatment and prn medical therapy. No study to date has evaluated Stretta as a stand-alone therapy. Thus, with the evidence currently in hand, it is still not possible to assess the relative efficacy of Stretta treatment to other GERD therapies, medical or surgical. Nonetheless, just as we endeavor to be evidence-based in our assessment of medications, so should we be in the evaluation of devices. With the report published in this issue of Gastroenterology, such an assessment is now possible for the Stretta procedure making it unique amongst the entire family of endoluminal GERD therapies. For a more comprehensive analysis of the entire literature on Stretta, the reader is referred to a recent review.4Kahrilas P.J. Technology review radiofrequency therapy of the lower esophageal sphincter for treatment of GERD.Gastrointest Endosc. 2003; 57: 723-731Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar Two mechanisms of action have been proposed for Stretta treatment; scarring of the esophagogastric junction (EGJ) and neurolysis in the region of the EGJ. Scarring (collagen deposition) could “tighten” the EGJ, limiting the occurrence and/or volume of reflux. Neurolysis could reduce sensitivity to noxious stimuli, inhibit transient lower esophageal sphincter relaxation (tLESR), or increase LES pressure. With respect to LES pressure, the bulk of evidence has shown no effect. Thus, mechanistic studies of Stretta treatment have focused on the hypotheses regarding scarring at the EGJ, sensitivity, or tLESR elicitation. No human histopathological data exist demonstrating the tissue effects of Stretta treatment but controlled histological animal data do exist.5Kim M.S. Holloway R. Dent J. Utley D.S. Radiofrequency energy (RFe) delivery to the gastric cardia inhibits triggering of transient lower esophageal sphincter relaxations and gastroesophageal reflux in dogs.Gastrointest Endosc. 2003; 57: 17-22Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar The stomach and esophagus from 2 cm proximal to the z-line to 3 cm distal was examined in 11 experimental dogs 7 months after Stretta treatment and compared with the equivalent specimen from 2 control animals. Treated animals showed marked hypertrophy, as well as fibrosis within the muscle and the gastric cardia was significantly (63%) thickened. However, it should be noted that the Stretta protocol used was more intensive to the cardia than that currently recommended for application in man.6Curon Medical Inc.The Stretta procedure.Endoluminal delivery of temperature-controlled radiofrequency energy for the treatment of gastroesophageal reflux disease. Procedure summary, patient selection, clinical data. 2002Google Scholar Neurolysis in the region of the EGJ could destroy sensory or motor nerve endings. Given the absence of histopathologic data demonstrating neurolysis after Stretta treatment, supporting evidence is based upon physiological data. With respect to visceral sensation, one recent trial (published only as an abstract thus far) suggests reduced sensitivity after Stretta treatment.7Arts J. Van Olmen A. D’Haens G. Sifrim D. Lerut A. Rutgeerts P. Janssens J. Tack J. Radiofrequency delivery at the gastroesophageal junction in GERD improves acid exposure and symptoms and decreases esophageal sensitivity to acid infusion.Gastroenterology. 2003; 124: A-19Google Scholar Utilizing the outcome measure of time to perception of heartburn during the installation of 0.1N HCl (Bernstein test), pretreatment values of 9 ± 6 minutes increased to 17 ± 11 minutes 6 months posttreatment. Of particular interest, all subjects had a positive Bernstein response before treatment while 4 of 13 were insensitive despite 30 minutes of acid perfusion afterward. The other potential effect of neurolysis pertains to tLESRs. Transient LES relaxation is a vago-vagal reflex triggered by gastric distention with an afferent receptive field centered in the gastric cardia and to which a majority of reflux events are attributable. Electrophysiological data suggest that the relevant vagal afferent nerves terminate with specialized endings named intraganglionic laminar endings, deformity sensitive neural transducers.8Zagorodnyuk V.P. Chen B.N. Brookes S.J. Intraganglionic laminar endings are mechano-trandsduction sites of vagal tension receptors in the guinea-pig stomach.J Physiol. 2001; 534: 255-268Crossref PubMed Scopus (220) Google Scholar Thus, a potential mechanism of action for the Stretta procedure is disruption of this vagal mechano-sensitive triggering mechanism for tLESRs. The effect of Stretta on tLESR elicitation has been studied in both canine5Kim M.S. Holloway R. Dent J. Utley D.S. Radiofrequency energy (RFe) delivery to the gastric cardia inhibits triggering of transient lower esophageal sphincter relaxations and gastroesophageal reflux in dogs.Gastrointest Endosc. 2003; 57: 17-22Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar and human experiments.10 Both of these demonstrated a significant reduction in distention induced tLESR frequency at 3 months and 6 months posttreatment respectively. In the canine experiment the tLESR rate decreased from 4 to 3 per hour in a 1-hour postprandial recording during air infusion into the stomach. In the human trial, the postprandial tLESR rate diminished from 6.8 to 5.2 per hour in 20 GERD patients. Once again, the Stretta protocol used was more intense in the cardia than the current manufacturer recommendation. Esophagitis is greatly reduced or eliminated by treatment with antisecretory medications, particularly proton pump inhibitors (PPIs). Thus, it is impossible to interpret a treatment effect on esophagitis unless antisecretory medication has either been previously withdrawn for a sufficiently long period for disease equilibrium to be achieved or, alternatively, held constant throughout a trial. None of the Stretta clinical trials has used either strategy. In fact, a secondary outcome measure of several trials, including the one reported in this issue of Gastroenterology, was of the degree to which antisecretory medication usage could be reduced as mandated by symptom control. Consequently, no meaningful data exist on the efficacy of Stretta in healing esophagitis. Illustrative of this, 11% of the active treatment group of the sham-controlled trial had erosions at entry compared with 20% at the 6-month time point; comparable numbers for the sham arm were 21% and 22% respectively. It is also noteworthy that a >2 cm hiatal hernia was an exclusion criterion for this trial, a stipulation that would tend to exclude most individuals prone to high-grade esophagitis. Abnormal esophageal pH-metry was an entry criterion for this Stretta trial, which also served as a secondary outcome measure. Reading from Tables 1 and 2 in the paper, median esophageal acid exposure for the active treatment group was 9.5% at entry compared with 9.9% at the 6-month time point; comparable data for the sham arm were 9.9% and 10.7% respectively. Attempting to explain this seeming lack of effect, a post hoc analysis was performed and showed that responders (>50% reduction in one of the quality of life measures) improved their acid exposure values by a median of 3.5% compared with a 1.7% worsening among nonresponders. Given that the cutoff for abnormal pH utilized was >4%, even with this post hoc subgroup analysis well over half of the responders still had abnormal esophageal acid exposure. The primary outcome measure of this trial (and all but one of the uncontrolled trials of the Stretta procedure) was of heartburn relief along with the improvement in quality of life measures. Evident in Table 2 of the paper, a significant improvement was achieved and, along with the reduction in heartburn, quality of life improved. PPI usage was reduced to a similar extent in both arms of the study and, despite that, the sham arm exhibited a numerical reduction in all heartburn severity measures. The controlled and/or objective clinical and physiological evaluations of Stretta treatment circa 2003 can be summarized as follows. Stretta treatment has been evaluated as a treatment for heartburn only in GERD patients with abnormal esophageal acid pH-metry and mild or absent esophagitis. Study designs have uniformly evaluated Stretta treatment as an adjunct to acid suppressive therapy, not as a stand-alone treatment. Heartburn severity and associated quality of life measures were significantly improved amongst treated individuals followed for a period of up to 12 months. Another treatment effect is of a slight reduction in the frequency of tLESRs. Stretta treatment has not been shown to reduce the severity of esophagitis or increase LES pressure. Data on esophageal acid exposure as measured by esophageal pH-metry showed no overall improvement in the sham-controlled trial but a 3.5% median reduction in a post hoc subgroup analysis of responders. PPI usage was reduced following Stretta treatment, but it was reduced to a similar degree amongst sham treated individuals. The common element of the treatment effects observed with Stretta (symptom reduction and reduced tLESR elicitation) is that both are neurally mediated. Given the absence esophagitis healing data and the (at best) modest reduction in esophageal pH-metry, it seems reasonable to conclude that symptom improvement after Stretta treatment is probably not attributable to reduced gastroesophageal reflux. Rather, data suggest that Stretta treatment effects probably result from neurolysis in the region of the EGJ. The dominant treatment effect is likely attributable to diminished esophageal sensitivity while a secondary, less consistent, effect is probably related to diminished elicitation of tLESRs in the postprandial period.