Helicobacter pylori Eradication Therapy Research: Ethical Issues and Description of Results

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2010;8: PERSPECTIVE Helicobacter pylori Eradication Therapy Research: Ethical Issues and Description of Results DAVID Y. GRAHAM Department of Medicine, Molecular Virology and Microbiology, Michael E. DeBakey Veterans Affairs Medical Center and Baylor College of Medicine, Houston, Texas See related article, Sonnenberg A et al, on page 1894 in Gastroenterology. As an infectious disease, the approach to anti Helicobacter pylori therapy differs from other common gastrointestinal conditions because treatment success of more than 90% to 95% should be expected and the reasons for treatment failure can always be understood. Neither comparisons with another regimen nor randomization are required to identify a highly successful therapy. Treatment success should be judged first in relation to outcome (ie, >95% or grade A). Inclusion of a known inferior regimen in a clinical trial is generally unethical. If the use of a known inferior drug is required by a regulatory agency, subjects must be given full and accurate information regarding expectations with each regimen; there can be no deceptions. Comparative trials should be restricted to highly successful treatments (ie, comparisons of different doses, durations, compliance, cost, and so forth). Success should be judged as ordered categories such as <85%, 85% 89%, 90% 94%, or >95% and statistically equivalent regimens with the same grade success (ie, 90% 94% [Grade B]) are inferior to those higher category (ie, >95% [Grade A]) regimens. Only grade A or B regimens should be prescribed. Here we discuss anti H pylori eradication studies from the prospective of an infectious disease with the goal of providing recommendations regarding changes in approach and in reporting that should help resolve the ethical issues and make the results of clinical trials more useful to clinicians. Keywords: Ethics; Clinical Equipoise; Scoring. Helicobacter pylori are susceptible in vitro to a large number of antimicrobial agents; however, in practice, eradication has proven to be difficult. Effective therapy generally requires the simultaneous administration of several antimicrobial agents often accompanied by an antisecretory drug. By 2000, it was recognized that treatment success of at least 90% or greater could be reliably achieved. However, this degree of success was fleeting because the initially successful regimens were undermined by increasing antimicrobial resistance. 1,2 Although this general decline in the effectiveness of the most widely recommended therapies was not accompanied by clear changes in recommendations from various professional or expert groups, 3 it did prompt clinical trials with many different antibiotic combinations. Few were designed based on the traditional approaches to treatment of an infectious disease, and designs more typical for other common gastrointestinal diseases (eg, reflux esophagitis, constipation, dyspepsia, irritable bowel syndrome, and so forth) were generally used. In retrospect, many may have been unethical. Here we discuss anti H pylori eradication studies from the prospective of an infectious disease with the goal of providing recommendations regarding changes in approach and in reporting that should help resolve the ethical issues and make the information more useful to clinicians. Clinical Research Regarding Treatment of Common Bacterial Infections Medical research fundamentally differs from medical practice. The obligation of the physician in clinical practice is to provide optimal care to the individual patient. In contrast, clinical trials are scientific experiments designed to increase generalizable scientific knowledge so as to promote the medical good of future patients. Although patients may derive benefit from the treatment offered as part of clinical research, such benefit is not the primary goal of the research. 4 Fundamentally, patient care deals with patients and clinical research deals with subjects. 5 The procedures involved in research also differ from those involved in practice. First, research follows a defined protocol that includes restrictions on all the elements of the study including who is included or excluded, the details of treatment including the drugs to be used, their doses, frequency of administration, and so forth. Research typically also involves clinically unnecessary procedures and tests specifically designed to assess the safety and outcomes. Research often includes randomization, masking, and the use of placebo, such that neither the physician nor the subject know exactly what was given. 6 There are a number of specific ethical issues surrounding human clinical research. Emanuel et al 7 proposed 7 ethical requirements for clinical research (Table 1). The key element for an ethical clinical trial is the presence of an honest null hypothesis, which requires clinical equipoise. 8,9 Clinical equipoise fundamentally consists of a genuine uncertainty regarding the merits of one treatment compared with another in relation to a particular population. 8 Clinical research should be designed to solve that uncertainty such that at its conclusion the results should be convincing enough to resolve the current disturbance in clinical equipoise. Abbreviation used in this paper: PPI, proton pump inhibitor by the AGA Institute /$36.00 doi: /j.cgh

2 December 2010 ISSUES REGARDING H PYLORI THERAPY 1033 Table 1. Ethical Requirements for Clinical Research 7 Scientific or social value Scientific validity Fair subject selection Favorable risk-benefit ratio Independent review Informed consent Respect for enrolled research participants Research subjects should never be exposed to the risks of valueless research (ie, if the answer already is known or the question is trivial, then there can be no honest null hypothesis, and the clinical trial should not be conducted). 4,8,9 An honest null hypothesis also simply cannot rely on a so-called standard of care, guidelines published by various groups, or even Food and Drug Administration marketing approval (eg, Ellenberg and Temple 10 noted [Food and Drug Administration] marketing approval does not establish treatment as standard of care even in the United States and certainly not in other countries ). An honest null hypothesis requires that the investigator critically analyze the information currently available as well as attempt to discover information that would be relevant to the treatment decisions. 8,9 As we will show later, in our opinion, an honest null hypothesis may have been lacking in many recent H pylori treatment trials. Randomization is considered a cornerstone of clinical research. The goal of randomization is to prevent selection bias; however, despite its importance, randomization is not an absolute requirement for a good trial. The need for randomization depends on the study question and the expected results. Although randomization prevents selection bias, it does not eliminate patient-to-patient variations. Randomization is especially useful when treatments are compared and the margin of difference is expected to be narrow (eg, 5% 10%). 11 As we will show later, the sense of requirement for randomization and the need to compare may be in part responsible for treatment trials that included a proven inferior regimen. 9 A final advantage of randomization is that the estimate of the treatment effect is likely to be close to the true effect. 12 The role of randomization in the design of antibiotic therapy for common infectious diseases is discussed later. Infectious Disease Research Versus Other Clinical Research in Gastroenterology Many human conditions have variable and often unpredictable responses to therapy, and the effects of treatment are often small. In many conditions there is also a high rate of spontaneous improvement such that a placebo may be required to determine whether the therapy has any benefit. For treatments in which the effect is reliably superior to placebo, one generally prefers an active control rather than a placebo. The active control is typically the best currently available therapy with the goal of the experiment being to prove noninferiority. If the new therapy is thought to possibly be superior to the best available, one typically performs a randomized head-to-head comparison. Importantly, if the comparison involved an inferior therapy there can be no valid null hypothesis and randomization would be unethical (see later). Studies of antimicrobial therapy for infectious diseases expect treatment success to be high (eg, 95%) and typically are either active control equivalence trials or nonrandomized studies of the new agent using historical controls. The proportion who undergo spontaneous H pylori eradication or with placebo is essentially 0%, allowing one to use historical controls. 3,13 Success is therefore defined in terms of prespecified end points such as greater than 90% or 95%. Although a comparator is not required in infectious disease research, randomization is required when 2 or more highly effective therapies are compared to reduce the risk of selection bias and to ensure that any systematic differences are owing to the treatment itself. However, one would not attempt to prove that a new therapy that overcame resistance was superior to an older one that had become ineffective because of increasing resistance (eg, streptomycin vs multidrug therapy for tuberculosis). Proof that a new therapy was effective (eg, 95% success) would not be enhanced further by showing that another regimen was inferior, especially because the reason for the difference already was known. Factors Responsible for H pylori Treatment Failures Most factors that influence outcome in H pylori treatment studies are known and can be controlled for (Table 2). The most important cause for poor outcome in optimized regimens is the presence of antimicrobial resistance. None of the factors in Table 1 is random, unknown, or not measurable, such that the reason for a poor outcome almost always can be identified. Thus, the approach to treatment studies with H pylori as an infectious disease is markedly different than with other common gastrointestinal and nongastrointestinal conditions such as nonulcer dyspepsia, constipation, or depression. The fact that one can prospectively and retrospectively control for these factors implies that randomization primarily would be of value only when one or more of the factors was the object of the research (ie, comparison of duration for a known highly successful therapy). Ethical Issues With Anti H pylori Therapy We will focus on 2 issues: the requirement for an honest null hypothesis and the need for fully informed consent. As noted earlier, by definition, a trial is unethical if it compares a Table 2. Some Factors Known to Influence Outcome in H pylori Treatment Studies Treatment-specific factors Manufacturer of the drugs Doses, duration, frequency of drug administration, formulation Relation of drug administration and meals Type and dose of adjuvant therapies Study-specific factors Methods to detect the infection and to confirm eradication Prevalence of antimicrobial resistance Patient-specific factors Compliance Side effects Genetic polymorphisms that alter study drug metabolism

3 1034 DAVID Y. GRAHAM CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 12 known (at the time) inferior therapy with a known superior therapy because there can be no honest null hypothesis. By 2001, 2 meta-analyses encompassing more than 53,000 subjects showed that the success with triple therapies (a proton pump inhibitor [PPI], amoxicillin, and either clarithromycin or metronidazole) had decreased to an unacceptably low level (ie, 80%). 14,15 This finding has been confirmed repeatedly by many subsequent studies (eg, Graham and Fischbach 3 and Fischbach et al 16 ). In 2000, sequential therapy consisting of a 10-day regimen starting with a dual therapy (a PPI and amoxicillin) followed by a triple therapy (a PPI, clarithromycin, and a nitroimidazole) was identified as an effective alternative 17 and by 2003 there were 3 studies with more than 700 patients that had been published confirming that sequential therapy reliably provided treatment success of greater than 90% Together, these studies provided unequivocal proof that, at least in Italy, sequential therapy was a successful treatment regimen. Sequential therapy also was compared with legacy triple therapy. The pivotal study was a multicenter study that randomized more than 1000 subjects to sequential or legacy triple therapy. The study was designed to detect a difference of 9% in the eradication rate between the standard 7-day (assumed to have an eradication rate of 80%) and the new 10-day regimen (estimated to have an eradication rate of 89%) with a power of 0.90 and a significance level of 0.01 (.010, 2-sided). The investigators were aware that triple therapy was likely to no longer be a successful regimen stating, Recent studies have shown disappointing results with these regimens. A meta-analysis performed by Janssen et al 15 showed a pooled intention-totreat eradication rate of 79% for PPI triple therapies using either amoxicillin (95% confidence interval, 24 95) or nitroimidazole (95% confidence interval, ). 15 Even lower eradication rates, ranging from 60% to 75%, have been reported in several countries, including Italy and France. 20,21 The comparison conclusively showed that sequential therapy was not only effective but also was superior to legacy triple therapy (92% vs 74%, respectively). 19 That study was followed up in 2004 by additional randomized controlled trials with approximately 450 patients confirming the reliability of the excellent results with sequential therapy and the poor results with legacy triple therapy even when the duration of triple therapy was prolonged to 10 days. 22,23 Therefore, by 2003 to 2004 the data were unequivocal that, at least in Italy, legacy triple therapy provided unacceptably low success rates and that it was inferior to sequential therapy. Clearly, clinical equipoise had been upset such that additional studies comparing sequential therapy with legacy triple therapy, now proven to be an inferior therapy, would be unethical. This requirement also would extend to any ongoing comparison trials in which the legacy triple therapy arm would need to be discontinued because a valid null hypothesis was no longer present. This new information also must be provided to the institutional review boards and to all patients receiving therapy and the consent form would need to be modified accordingly. However, this may not have happened as we will show later. Informed Consent Informed consent is one of the basic principles of clinical research. The Helsinki accord states that in medical research involving human subjects, the well-being of the individual research subject must take precedence over all other interests. By definition, the subject must receive any information that the investigator has that might affect his or her decision to enter a trial or to continue in an ongoing trial. The (at least) annual review of ongoing trials should include a question concerning whether new information has appeared since the previous review that would affect the decision to enter or continue in the study. Whether it would be ethical to offer a subject the opportunity to enter a study in which one regimen was known to be inferior is unclear and even then there can be no deceptions. Freedman 8 believes that such an offer itself would not be ethical, Even if a patient would consent to an inferior treatment, it seems to me in violation of competent medical practice, and hence of ethics, to make the offer. H pylori Treatment Trials Since Because the investigators in subsequent Italian sequential-triple therapy comparisons also typically were investigators in the pivotal study, it is clear that those results were well known to them 17 and the fact that legacy triple therapy was inferior therapy was used to reduce the sample sizes to accommodate the expected larger differences How to handle the informed consent was more difficult. In the Vaira et al 24 study, the subjects and institutional boards were told that the comparison was with the current standard (italics added) or a new therapy that might have higher eradication rates. It was not disclosed whether the institutional review boards and consent forms were updated as the results of additional studies became known. 27 Because the early sequential therapy studies generally did not always examine the mechanism for improved success, one could argue that additional studies were needed. However, even if that were the case, additional comparisons with legacy triple therapy were neither needed nor indicated and would be unethical. Because the success of triple therapy declined in proportion to the increase in clarithromycin resistance and sequential therapy was effective in the same population that triple therapy was ineffective, the most likely explanation is that sequential therapy provided good eradication in the face of clarithromycin resistance, possibly because of the addition of the additional antimicrobial, metronidazole/tinidazole. The effect of resistance on the success of triple therapy already had been established and there was no reason to reconfirm those results. The most efficient method to evaluate the question with sequential therapy would have been to directly test sequential therapy in patients with known clarithromycin or metronidazole resistance. It would have been unethical to prospectively test patients with dual resistance because the results with dual PPI and amoxicillin therapy also were well known (ie, dual resistance converted sequential therapy into PPI plus amoxicillin dual therapy just as clarithromycin resistance did with triple therapy). Alternatively, sequential therapy could be evaluated in a nonrandomized trial in which specimens for pretreatment susceptibility testing were obtained and later evaluated. This would further confirm the effectiveness of sequential therapy and also test why it was more effective than legacy triple therapy. Besides the ethical issues surrounding randomizing patients to the inferior triple therapy, 9 this had the additional disadvantage of not being able to use those patients to evaluate

4 December 2010 ISSUES REGARDING H PYLORI THERAPY 1035 variations of sequential therapy that might improve it from an average B result (ie, 90% 94% success) to an A result. 28 A Paradigm Shift Is Required for Design of Studies for the Evaluation of H pylori Eradication Therapy Although, up to this point, we have focused here on studies comparing legacy triple therapy and sequential therapy, there are many other examples of clinical trials in which known inferior therapies were given to subjects and many meta-analyses of the results of these trials. 3,29 H pylori infection is possibly the only common disease in gastroenterology in which the treatment success of experimental therapy is expected to be 95% or greater and in which the causes of treatment failure are easily identified. The focus on many other common gastrointestinal diseases (eg, dyspepsia, irritable bowel syndrome, constipation, and so forth) is to identify a therapy better than placebo (ie, the main focus is whether one therapy is statistically better than another). H pylori therapies are best judged in relation to absolute outcome where 90% or more is good, 95% is excellent, and less than 90% is unacceptable (bad). 3,13,30 Investigators should be required to explain why good or excellent results were not obtained. This focus on better than has resulted in bad results being recommended as equivalent when a better conclusion would have been that they both were unacceptably low Therapy should be chosen only based on what is effective locally. If that is unknown, pretreatment susceptibility testing (tailored therapy) and posttreatment confirmation of cure must be performed to obtain that information. The importance of continued posttreatment testing is emphasized by the fact that even in Italy, the success of sequential therapy seems to be declining probably as resistance, possibly as dual resistance, increases. 24,25 References 1. Fischbach L, Evans EL. Meta-analysis: the effect of antibiotic resistance status on the efficacy of triple and quadruple first-line therapies for Helicobacter pylori. Aliment Pharmacol Ther 2007; 26: Graham DY, de Boer WA, Tytgat GN. Choosing the best anti- Helicobacter pylori therapy: effect of antimicrobial resistance. Am J Gastroenterol 1996;91: Graham DY, Fischbach L. Helicobacter pylori treatment in the era of increasing antibiotic resistance. Gut 2010;59: Miller FG, Brody H. A critique of clinical equipoise. Therapeutic misconception in the ethics of clinical trials. Hastings Cent Rep 2003;33: Miller FG, Rosenstein DL. The therapeutic orientation to clinical trials. N Engl J Med 2003;348: Miller FG, Rosenstein DL. Reporting of ethical issues in publications of medical research. Lancet 2002;360: Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA 2000;283: Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317: Tremaine WJ. Equipoise, H. pylori, and Musketeers choosing an appropriate standard clinical care regimen. Helicobacter 2008; 13: Ellenberg SS, Temple R. Placebo-controlled trials and activecontrol trials in the evaluation of new treatments. Part 2: practical issues and specific cases. Ann Intern Med 2000;133: Retsas S. Treatment at random: the ultimate science or the betrayal of Hippocrates? J Clin Oncol 2004;22: Wieand S, Murphy K. A commentary on treatment at random: the ultimate science or the betrayal of Hippocrates? J Clin Oncol 2004;22: Graham DY. Efficient identification and evaluation of effective Helicobacter pylori therapies. Clin Gastroenterol Hepatol 2009; 7: Laheij RJ, Rossum LG, Jansen JB, et al. Evaluation of treatment regimens to cure Helicobacter pylori infection a meta-analysis. Aliment Pharmacol Ther 1999;13: Janssen MJ, Van Oijen AH, Verbeek AL, et al. A systematic comparison of triple therapies for treatment of Helicobacter pylori infection with proton pump inhibitor/ranitidine bismuth citrate plus clarithromycin and either amoxicillin or a nitroimidazole. Aliment Pharmacol Ther 2001;15: Fischbach LA, Goodman KJ, Feldman M, et al. Sources of variation of Helicobacter pylori treatment success in adults worldwide: a meta-analysis. Int J Epidemiol 2002;31: Zullo A, Rinaldi V, Winn S, et al. A new highly effective short-term therapy schedule for Helicobacter pylori eradication. Aliment Pharmacol Ther 2000;14: Hassan C, De Francesco V, Zullo A, et al. Sequential treatment for Helicobacter pylori eradication in duodenal ulcer patients: improving the cost of pharmacotherapy. Aliment Pharmacol Ther 2003;18: Zullo A, Vaira D, Vakil N, et al. High eradication rates of Helicobacter pylori with a new sequential treatment. Aliment Pharmacol Ther 2003;17: Perri F, Villani MR, Festa V, et al. Predictors of failure of Helicobacter pylori eradication with the standard Maastricht triple therapy. Aliment Pharmacol Ther 2001;15: Bigard MA, Delchier JC, Riachi G, et al. One-week triple therapy using omeprazole, amoxicillin and clarithromycin for the eradication of Helicobacter pylori in patients with non-ulcer dyspepsia: influence of dosage of omeprazole and clarithromycin. Aliment Pharmacol Ther 1998;12: De Francesco V, Faleo D, Panella C, et al. Sequential eradicating therapy: a treatment that does not discriminate Helicobacter pylori strains in patients with nonulcer dyspepsia? Am J Gastroenterol 2002;97: De Francesco V, Zullo A, Hassan C, et al. The prolongation of triple therapy for Helicobacter pylori does not allow reaching therapeutic outcome of sequential scheme: a prospective, randomised study. Dig Liver Dis 2004;36: Vaira D, Zullo A, Vakil N, et al. Sequential therapy versus standard triple-drug therapy for Helicobacter pylori eradication: a randomized trial. Ann Intern Med 2007;146: Paoluzi OA, Visconti E, Andrei F, et al. Ten and eight-day sequential therapy in comparison to standard triple therapy for eradicating Helicobacter pylori infection: a randomized controlled study on efficacy and tolerability. J Clin Gastroenterol 2010;44: Zullo A, Gatta L, De Francesco V, et al. High rate of Helicobacter pylori eradication with sequential therapy in elderly patients with peptic ulcer: a prospective controlled study. Aliment Pharmacol Ther 2005;21: Graham DY, Yamaoka Y. Ethical considerations of comparing sequential and traditional anti Helicobacter pylori therapy. Ann Intern Med 2007;147: Graham DY, Lu H, Yamaoka Y. Therapy for Helicobacter pylori infection can be improved: sequential therapy and beyond. Drugs 2008;68: Graham DY, Go MF. Unethical research relating to Helicobacter pylori? Gut 1995;36: Graham DY, Lu H, Yamaoka Y. A report card to grade Helicobacter pylori therapy. Helicobacter 2007;12:

5 1036 DAVID Y. GRAHAM CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No Dong J, Yu XF, Zou J. Azithromycin-containing versus standard triple therapy for Helicobacter pylori eradication: a meta-analysis. World J Gastroenterol 2009;15: Vakil N, Lanza F, Schwartz H, et al. Seven-day therapy for Helicobacter pylori in the United States. Aliment Pharmacol Ther 2004;20: Fuccio L, Minardi ME, Zagari RM, et al. Meta-analysis: duration of first-line proton-pump inhibitor based triple therapy for Helicobacter pylori eradication. Ann Intern Med 2007;147: Zagari RM, Bianchi-Porro G, Fiocca R, et al. Comparison of 1 and 2 weeks of omeprazole, amoxicillin and clarithromycin treatment for Helicobacter pylori eradication: the HYPER study. Gut 2007; 56: Reprint requests Address requests for reprints to: David Y. Graham, MD, Michael E. DeBakey Veterans Affairs Medical Center, RM 3A-320 (111D), 2002 Holcombe Boulevard, Houston, Texas dgraham@bcm.tmc.edu; fax: (713) Conflicts of interest The authors disclose the following: Dr Graham is an unpaid consultant for Novartis in relation to vaccine development for treatment or prevention of Helicobacter pylori infection. Dr Graham is also a paid consultant for Otsuka Pharmaceuticals regarding diagnostic testing and until July 2007 was a member of the Board of Directors of Meretek Diagnostics, the manufacturer of the 13C-urea breath test. Until November 2009, Dr Graham received royalties on the Baylor College of Medicine patent covering materials related to the 13C-urea breath test. Funding This material is based on work supported in part by the Office of Research and Development Medical Research Service Department of Veterans Affairs. Dr Graham is supported in part by Public Health Service grant DK56338, which funds the Texas Medical Center Digestive Diseases Center and R21DK The contents are solely the responsibility of the author and do not necessarily represent the official views of the VA or the National Institutes of Health.