Transatlantic Conference on Clinical Trial Guidelines in Peripheral Arterial Disease
Clinical Trial Methodology
Abstract—Guidelines for the clinical development of drugs in peripheral arterial disease (PAD) have been issued by the Food and Drug Administration for the United States and by the regulatory agency of the European Union for Europe. With increasing globalization, transatlantic cooperation in drug research and development is essential for the future and would be substantially facilitated by the existence of transatlantic guidelines. A conference was held in Basel, Switzerland, in November 1997 to discuss the scientific background of the existing guidelines on the basis of published evidence and the extensive knowledge of clinical investigators and experienced regulators. The meeting was attended by 52 invited experts from the United States and Europe, as well as by representatives from the 2 regulatory authorities. The main conclusions from the meeting are presented and may serve as a reference for the future development of transatlantic guidelines for the evaluation of pharmacotherapy in PAD.
Over the years, a number of guidelines for the clinical development of drugs for the treatment of peripheral arterial disease (PAD) have been developed by specialist vascular societies and national health authorities such as the Committee for Proprietary Medicinal Products (CPMP) in the European Union and the Food and Drug Administration in the United States. However, existing guidelines from the United States and Europe have not harmonized. Guidelines by definition are not static documents but should reflect the current knowledge. To create a platform for future discussion and for the potential development of transatlantic guidelines, a conference was held in Basel, Switzerland, on November 15–17, 1997, involving 52 experts from the United States, Europe, and the regulatory agencies of both continents. The conference was organized by William R. Hiatt from the University of Colorado Health Sciences Center, Section of Vascular Medicine, and the Colorado Prevention Center, Denver, Colo; John A. Dormandy, Department of Vascular Sciences, St George’s Hospital, London, UK; Kurt A. Jaeger and Karl-Heinz Labs, Department of Angiology, University of Basel Medical School, Basel, Switzerland; and Claus-Steffen Stuerzebecher, Berlin, Germany. The following text summarizes the main conclusions from the meeting.
PAD: Fontaine Stage II (Intermittent Claudication)
Phase III clinical trials should generally follow a randomized, double-blind, parallel-group study design. Crossover designs should be the exception and may only be acceptable for short trials with short-acting drugs, eg, pharmacokinetic or pharmacodynamic phase I or II studies, but not for phase III trials.
Placebo Control Versus Active-Drug Control
Phase III clinical trials in patients with intermittent claudication with walking distance as the primary outcome measure should generally be placebo controlled. Active-drug controlled trials without a placebo arm will be the exception and may only be considered if the comparator drug has consistently shown superiority over placebo in several trials and the magnitude of the drug effect has been widely accepted by the medical community.
Duration of Treatment and Follow-Up
The duration of treatment will depend on the aim of the study and the end point(s) chosen. Claudication Trials. For claudication studies, a period of ≈6 months is generally accepted. The length of exposure to the drug should be sufficient to ensure that tolerance (if expected) will not develop. The duration of the posttreatment follow-up period will also depend on the study goals. With respect to efficacy in claudication trials, a short follow-up period (eg, not less than 4 weeks) is regarded as sufficient to demonstrate that no sudden loss in claudication distance (rebound) occurs. With respect to safety, long-term experience is required to demonstrate that the clinical benefit of the compound outweighs the risks. If there is any evidence of drug toxicity, the follow-up should be performed in a double-blind, controlled manner rather than an open-label manner. Prevention Trials. CPMP guidelines1 suggest a minimum trial duration of ≥12 months for long-term end points, such as cardiovascular morbidity and mortality or progression of atherosclerosis. However, to obtain sufficient event rates and clinical relevance of the results, a treatment period of ≥24 months is strongly recommended.
Uneven distribution of potential but unknown confounders, such as cardiovascular risk factors, hemodynamic parameters, demographic variables, and others, can usually be avoided by an appropriate randomization process and an adequate sample size. The occurrence of slight imbalances in the data set can be handled by post hoc (analytical) adjustment, provided that the adjustment for the effects of predefined covariables is prospectively planned and specified in the study protocol.
In claudication trials, there is usually no need for stratification. The clinical suspicion that claudicants with diabetes mellitus respond to drug treatment differently from those without diabetes has not been substantiated. Thus, stratification is not required for diabetic patients as long as patients with clinically severe diabetic neuropathy are excluded; the latter is required, because this complication interferes with the assessment of the claudication distances (absolute claudication distance [ACD] and initial claudication distance [ICD]).
In prevention trials, the situation is different. There is evidence from studies using antiplatelet agents in patients with diabetes mellitus and vascular disease that fewer (rather than more) person-years of treatment are required to prevent fatal cardiovascular events than for nondiabetic patients with vascular disease.2 3 Thus, stratification for diabetes mellitus (or separate trials for diabetic patients) is recommended.
Run-In Phase and Testing for Baseline Stability
Current guidelines suggest that patients with pronounced baseline variability in the claudication distance be excluded from clinical trials (the absolute claudication distance should not vary by >25% when tested twice on a treadmill over a 3- to 6-week run-in period).1 The reason for this recommendation is to not unduly inflate the size of the patient sample. However, there is little published evidence that variability before entry translates into higher variability during and at the end of the study. Thus, the decision whether to use stability criteria should be left to the investigators and the sponsors responsible for a particular trial. However, it is recommended that patients who have a recognized potential for instability (differences in claudication distances between days, recent phases of deterioration or improvement, or recent surgical or endovascular intervention) be excluded and that a study run-in phase with 2 to 3 treadmill tests for the purpose of treadmill test familiarization be included.
Selection of Patients
Inclusion and exclusion criteria should warrant that the patient sample is representative of a general PAD population.
Inclusion criteria for PAD stage II trials are broadly accepted and are well described in the current European Union (CPMP) guidelines.1 Claudication should be clinically stable and should have been present for ≥6 months; the diagnosis of PAD should have been established and confirmed by hemodynamic measurements (ankle/brachial pressure index [ABI] <0.90); and the location of the lesion(s) should be documented by duplex sonography or angiography. Patients with a recognized potential for claudication instability should be excluded (see Run-In Phase and Testing for Baseline Stability). In trials targeting the claudication distance, there is usually no reason for prerandomization stratification of the patient sample (see Stratification).
The therapeutic value of a structured, organized, and supervised physical training program is well established.4 5 Thus, patients who already participate in or intend to enroll in a supervised group physical training class should be excluded from clinical trials unless the study protocol defines that all patients from both groups will participate.
Decisions regarding unsupervised training should be made and strategies should be set before initiation of a clinical trial, and these should not be changed during the course of the study. It may also be recommended that to allow for a potentially necessary adjustment for different levels of physical activity, the type, frequency, and intensity of any home-based training should be documented in the case record form.
Primary End Points
Because the key symptom of stage II PAD is intermittent claudication, claudication distance on the treadmill should be the primary end point. Cardiovascular morbidity and mortality represent the main risks associated with stage II PAD.4 6 7 8 Prevention studies represent an entirely different category of clinical trials and thus require different study design. From a clinical and an economical point of view, this type of study may even be more important than trials that focus merely on the patient’s walking ability.
ICD Versus ACD
Both parameters (ICD and ACD) are clinically relevant. However, there is published evidence that the reproducibility of the 2 parameters varies with the treadmill protocol used.9 10 11 12 13 14 With constant-load treadmill protocols, the reproducibility of ICD and ACD is similar, whereas with graded protocols, ACD is superior to ICD. Consequently, ACD may be given preference over ICD as the primary end point; this decision would allow it to be independent of the treadmill protocol but is mandatory if a graded test is used.
Assessment of Claudication Distance: Constant-Workload Versus Graded-Workload Treadmill Testing. The preferred method to assess claudication distances is treadmill testing. There are 2 internationally accepted treadmill protocols, ie, the constant-workload protocol, which uses a constant speed and grade (mostly 2 mph [3.2 km/h] and 12% grade), and the graded test, in which the speed is kept constant but the grade is varied, starting horizontally but then increasing in predefined steps (eg, 2%) at predefined intervals (eg, 2 minutes).
The 2 tests differ in that the relationship between total work and walking time follows a linear function with the constant test but increases progressively with the graded test.9 15 This change in work rate explains the main advantages of the graded test: (1) a low workload in the early test phase allows proper differentiation of patients with differing highly limited walking distances, and (2) the continuously increasing workload in late test stages avoids the occurrence of a walking-through phenomenon.
The reproducibility of the 2 tests is comparable except with rather short claudication distances. In these cases, the graded test is superior to the constant-load test.15 Both tests can be equally recommended for use in clinical trials, although if either very short or rather long claudication distances are to be tested, the graded test should be preferred.
Prevention Studies: Morbidity and Mortality.
The main risk for PAD patients is related to ischemic cardiovascular events. An adequate end point in prevention studies is a composite end point comprising nonfatal ischemic stroke, myocardial infarction, cardiovascular death, and potentially coronary and carotid revascularization and major amputation, whichever occurs first. To clarify the terminology, reference is made to the World Health Organization (WHO) criteria for the diagnosis of coronary events and strokes (fatal and nonfatal). In the optimal case, total mortality may be the most relevant clinical end point; however, this poses problems of feasibility because of the need for large numbers of patients. The combination of nonfatal cardiovascular morbidity with all-cause mortality reflects a comparison of unequal entities, and inclusion of noncardiovascular morbidity renders clinical trials rather difficult. A combination end point that uses cardiovascular morbidity and cardiovascular mortality has been criticized because survival, irrespective of the cause of death, is what ultimately matters. Such a concern may be mitigated if it can be proven that results for the primary composite end point are statistically significant, that in the optimal case cardiovascular mortality is significantly reduced, and that the change in cardiovascular mortality also favorably influences all-cause mortality.
Secondary End Points
Secondary end points should focus on clinically relevant data that support the study aim. If the primary end point is a composite end point, the components of this composite end point should be evaluated as individual secondary end points.
In claudication trials with walking distance as the primary end point, data on morbidity and mortality must also be collected for safety reasons.
Quality of Life as an End Point
PAD represents just one (peripheral) manifestation of a generalized atherosclerotic process. Usually, affected patients have multiple morbidities, and the assessment of quality of life (QOL) may give a more representative picture of the patient’s perception of health than the exclusive measurement of walking performance. However, at present, a number of unresolved questions prevent the use of QOL as a primary end point. Problem areas include choosing the most appropriate instrument, proper validation of scales, potential compositing of end points, and the definition of what magnitude of change with a specific QOL scale may be considered clinically relevant. At present, QOL should be assessed as a secondary end point.
PAD: Fontaine Stages III and IV (Critical Limb Ischemia)
For the purpose of PAD guidelines, only the chronic and not the acute form of critical limb ischemia (CLI) will be considered.
Definition of PAD Stages III and IV/CLI and CLI-Related Risk
Most vascular specialists in continental Europe use the Fontaine classification to categorize the most severe forms of limb ischemia. The original Fontaine classification was based on clinical information only. The need for a more objective definition of Fontaine stages III and IV (or its Anglo-American equivalent, critical limb ischemia) became evident when it was shown that these patients have an excessively high risk of cardiovascular events, amputation, and mortality. The lack of uniform criteria in reporting the results of studies of CLI precludes the acquisition of reliable data and prevents the comparison of efficacy of different therapeutic strategies. Thus, to clarify, specify, and homogenize the definition of PAD stages III and IV, a consensus document was devised with the input of 8 European vascular specialist societies.16 17 CLI in both diabetic and nondiabetic patients was defined by either of 2 criteria: persistent recurrent distal extremity pain at rest that required the patient to use analgesics for >2 weeks, with an ankle systolic pressure of ≤50 mm Hg and/or a toe systolic pressure ≤30 mm Hg; or ulceration or gangrene of the foot or toes in combination with the hemodynamic criteria listed above. Furthermore, variables describing the compromised microcirculation (such as a transcutaneous Po2 ≤20 mm Hg) were included, particularly to be applied in patients with unreliable pressure readings. The justification of the criteria used was based on the correlation between impaired peripheral hemodynamics and the increased risk of cardiovascular events and amputation.6 18 19 20
Recent studies have raised doubts as to whether the ankle or toe pressure cutoff points have been chosen correctly and whether the threshold values were too low.21 22 There is published evidence21 indicating a comparable incidence of a combined end point of amputation and cardiovascular death in placebo-treated patients with peripheral pressures <50 mm Hg and those with peripheral pressures <60 mm Hg and/or an ABI <0.60. Comparable findings regarding ulcer healing and amputation were reported elsewhere.23 24 Conversely, patients with peripheral pressures clearly below CLI threshold who did not experience pain at rest or ischemic skin lesions have been described in the literature.25
There is undoubtedly a general association between cardiovascular risk, amputation, and peripheral pressures. It is of primary value to describe the population at risk with the highest possible sensitivity and specificity. The second CLI consensus conference17 used a pressure cutoff of 50 mm Hg at the ankle for this purpose, which, as shown by recent experience,21 22 may be doubted. The problem is that the optimal cutoff is not known, and at present, it is best to give a pressure range (50 to 70 mm Hg). The higher the pressure cutoff chosen, the higher will be the sensitivity but the lower the specificity for the definition of CLI. For clinical trials, use of the upper end of the range will increase the study’s inclusivity and thus the availability of patients, but it will decrease the cardiovascular event rate, as well as the amputation rate. Use of the lower end of the range will result in a narrower (more exclusive) set of study criteria, with high patient-sample homogeneity and a higher event rate but reduced patient enrollment. Some reasons for the uncertainty concerning the correct pressure cutoff are related to problems with the accuracy of measuring peripheral pressures noninvasively, particularly in diabetic patients, patients with chronic renal disease, patients undergoing chronic steroid treatment, and the elderly (age >80 years). Under these circumstances, it remains an open question whether a single fixed cutoff point can be defined at all. At present, reflecting the above discussion and until additional data are available, regulators suggest that CLI patients be stratified with respect to pressure ranges (eg, ankle pressure ≤50 mm Hg versus 50 mm Hg < ankle pressure ≤70 mm Hg, or alternatively, toe pressure ≤30 mm Hg versus 30 mm Hg < toe pressure ≤50 mm Hg) and that the results of clinical trials be reported accordingly.
Because of the nature and progression of CLI, only double-blind, randomized, parallel-group trials are appropriate; crossover designs are not acceptable. Proper randomization will ensure an equal distribution of risk factors and patient background characteristics between the treatment groups. The treatment of concomitant diseases should be continued throughout the trial; a protocol-defined standard regimen for the treatment of concomitant diseases is not feasible, particularly in multicenter, multinational trials. However, such standardization of a concomitant medication regimen is not required if an appropriate randomization procedure is used.
Duration of Treatment and Follow-Up
The duration of treatment will be determined by the pharmacological and toxicological profile and the mode of action of the drug under investigation. The overall duration of the trial will also depend on the end point(s) selected. Whereas a total of ≤6 months may be appropriate as a treatment and follow-up period for ulcer healing, the assessment of limb salvage rates requires a longer period of time (ie, ≥12 months).
A short run-in phase of 3 to 4 days should provide evidence that the disease is roughly stable (ie, that there is no rapid improvement or deterioration). In the majority of cases, a washout period is not considered necessary; however, if the previous treatment included a drug prohibited by the study protocol, a short washout period (of 2 to 3 days) is recommended.
Special stratification procedures are generally not required except as defined in Table 6⇓. The exception may be diabetes mellitus. There is no clear evidence from the literature that diabetic patients react differently to drug treatment than nondiabetic patients.26 27 28 On the other hand, it may be argued that the pathophysiology of diabetic PAD (in which microangiopathy and neuropathy potentially play a major role) differs from PAD caused solely by atherosclerosis. In view of these differences in pathophysiology, and in agreement with European Union guidelines,1 it is suggested that diabetic and nondiabetic patients should be stratified, particularly if the study end point includes the limb salvage rate, cardiovascular morbidity and mortality, and/or quantification of the progression of atherosclerosis.
Placebo-Controlled Versus Active-Therapy Controlled Trials
Based on the European Union guidelines,1 a double-blind, placebo-controlled trial design is a standard requirement. However, if there is a comparator drug that has consistently been proven to be superior to placebo and has shown convincing efficacy, an active-therapy controlled trial may be considered. The aim of clinical trials using an active-therapy control may be to demonstrate superiority or noninferiority of the experimental drug. An appropriate statistical approach for the study aim chosen should be predetermined.
In principle, all patients with proven CLI (see Definition of PAD Stages III and IV/CLI and CLI-Related Risk) in whom there is no sudden improvement or deterioration [see section on Duration of Treatment and Follow-Up under PAD: Fontaine Stages III and IV (Critical Limb Ischemia)] are eligible for CLI trials. European Union guidelines1 stipulate that only those patients who are ineligible for vascular reconstruction should be included. Although understandable for ethical reasons, this requirement will result in a selection of patients with end-stage disease in whom it is highly unlikely that clinical efficacy of any therapeutic measure would be demonstrated. A compromise could be to accept patients eligible for vascular reconstruction in clinical trials, as long as the trial design warrants that surgical or endovascular procedures are not withheld or unduly delayed; the special design of these types of studies must be accounted for in the statistical analysis plan.
Primary End Points
CLI patients may suffer from pain at rest, may have ischemic lesions, may require amputation, and will have an increased risk for cardiovascular morbidity and mortality. Thus, the CLI patient should be evaluated in a comprehensive fashion. The primary end point may focus on pain at rest, ulcer healing, or amputation, but all other variables (including cardiovascular and total mortality) should also be considered, at least as secondary end points. The primary end point may be single, composite, or based on response criteria.
Single End Points
Pain at Rest.
If pain intensity is chosen as a primary end point, it should be assessed objectively, preferably by use of visual analog scales. Pain at rest remains a soft end point influenced by variables such as mood, motivation, and environmental and other factors. Pain assessment should always be done by the same assessor at the same time of the day, preferably at the time of trough plasma levels of the drug under investigation. Because analgesia is difficult to quantify and the type and dose of analgesics are likely to change during the course of the trial, pain relief must be defined as “complete relief of pain with no use of analgesics.”
To ensure clinical relevance, ulcer healing must be defined as healing of all ulcers of both legs (all ulcers epithelialized, as assessed by an independent physician; photographic documentation, even if standardized, is considered insufficient). Only patients with “flat surface” or “transdermal” ulcers should be admitted. Ischemic cracks between the toes or on the heel cannot be used as measurable end points. The ulcer status at baseline may be documented by measurement of the cumulative total ulcer area, eg, by use of a dual-acetate technique.29 Partial ulcer healing or healing of a reference ulcer only is of doubtful clinical relevance and should not be used as a clinical end point.
The rate of major amputations (through or above the ankle) can be considered a primary end point or part of a primary end point (see also Composite End Points). The amputation rate is usually considered to be one of the “hardest” end points in CLI trials. However, because the criteria for performing amputation may vary, particularly in multicenter and multinational studies, the interpretation of study results may be limited. Existing guidelines1 recommend that amputation criteria be predefined in the study protocol. Even if one adheres to this recommendation, rules may sometimes be broken, because the decision to amputate is highly individual and depends on the patient’s individual risk pattern, general condition, and other factors. Thus, guidance included in a protocol will be of limited value and may introduce a level of pseudoaccuracy not in line with a real-life situation.
Because the underlying cause of CLI is generalized atherosclerosis, both legs (and not only the index leg) must be considered for the assessment of amputation rates.
Composite End Points
Mortality alone, whether cardiovascular or total, is rarely used as a single primary end point in CLI studies. In addition to the prevention of death, the status of the leg is of primary concern. Thus, a composite end point, eg, amputation and death or amputation, systemic morbidity (such as ischemic stroke and myocardial infarction), and death, should be preferred over mortality alone.
Response-Based End Points
The trial end point may be based on response criteria, and a responder definition may be applied to both the treatment and the placebo/comparator groups, with optimal response defined as the patient being alive, having both legs, having no wound or pain, and not taking analgesics. This end-point concept would allow consideration of both the time to response and the duration of response before the inevitable late failure process will occur. It is suggested that the number of “good days” be counted as a measure of response, ie, the time period for which the response criterion applies in a given follow-up period. Such an approach may be a conceptual step forward.
Secondary End Points
Secondary end points should focus on clinically relevant data that support the study objective. Whatever the primary end point, information on cardiovascular morbidity and mortality, as well as on all-cause mortality, must be collected over a sufficiently long period of time (in view of the generally pessimistic prognosis of CLI patients, this will rarely exceed 12 months). If the primary end point is a composite end point, the components of this composite end point should be evaluated individually as secondary end points.
QOL as an End Point
As with intermittent claudication, the assessment of QOL would provide a good tool to quantify the patient’s well-being if sufficiently validated scales were available. However, the problems described with QOL in intermittent claudication also apply to CLI. At present, QOL may only be considered as a secondary end point.
Basel PAD Clinical Trial Methodology Group
M. Aschwanden, University of Basel Medical School, Basel, Switzerland; J.J.F. Belch, Ninewells Hospitals & Medical School, Dundee, UK; D. Bergqvist, University of Uppsala Medical School, Sweden; M. Blumenthal, Bristol Myers Squibb, Princeton, NJ; H. Bounameaux, University of Geneva Medical School, Geneva, Switzerland; E.P. Brass, Harbor-UCLA Medical Center, Torrance, Calif; G. Brevetti, University of Napoli Frederico II, Naples, Italy; M. Cachovan, Herz-Kreislaufklinik Bad Beversen, Bad Beversen, Germany; S.A. Carter, St Boniface General Hospital, Winnipeg, Manitoba, Canada; M. Catalano, University of Milan, Milan, Italy; D.L. Clement, University of Gent Medical School, Gent, Belgium; M.H. Criqui, University of California, San Diego, Calif; J. Covinsky, Hoechst Marion Roussel, Bridgewater, NJ; E. Diamantopulos, Evangelismos State General Hospital, Athens, Greece; C. Diehm, Karlsbad Clinic, Academic Hospital of Heidelberg University, Heidelberg, Germany; J.A. Dormandy, St George’s Hospital, London, UK; D.E. FitzGerald, James Connolly Memorial Hospital, Dublin, Ireland; B. Forbes, Otsuka America Pharmaceuticals, Rockville, Md; S. Fredd, Food and Drug Administration, Rockville, Md; E. Groechenig, Landeskrankenhaus Feldkirch, Feldkirch, Austria; R. Hertrampf, Federal Institute for Drugs and Medicinal Devices, Berlin, Germany; W.R. Hiatt, University of Colorado Health Sciences Center, Denver, Colo; A.T. Hirsch, Cardiovascular Division, UMHC, Minneapolis, Minn; U. Hoppe, Federal Institute for Drugs and Medicinal Devices, Berlin, Germany; G. Ingenito, Otsuka America Pharmaceuticals, Rockville, Md; K.A. Jaeger, University of Basel Medical School, Basel, Switzerland; C. Jeanneret, University of Basel Medical School, Basel, Switzerland; G.G. Koch, University of North Carolina, Chapel Hill, NC; K.H. Labs, University of Basel Medical School, Basel, Switzerland; R.J. Lipicky, Food and Drug Administration, Rockville, Md; F. Mahler, University of Bern Medical School, Bern, Switzerland; M.D. McDaniel, Department of Veterans Affairs, White River Junction, Vt; P. Marrott, Berlex Laboratories, Montville, NJ; G. Moneta, Oregon Health Sciences University, Portland, Ore; L. Norgren, University of Lund Medical School, Lund, Sweden; J.W. Olin, Cleveland Clinic Foundation, Cleveland, Ohio; E. Pilger, University of Graz Medical School, Graz, Austria; A. Pilgrim, Sanofi Pharmaceuticals, Paris, France; J.M. Porter, Oregon Health Sciences University, Portland, Ore; J. Regensteiner, University of Colorado Health Sciences Center, Denver, Colo; J.B. Ricco, Society of Vascular Surgery, Poitiers, France; H. Rieger, Aggertal-Klinik, Hospital for Vascular Diseases, Engelskirchen, Germany; R. Roberts, Hamilton Civic Hospital Research Center, Hamilton, Ontario, Canada; J. Roehmel, Federal Institute for Drugs and Medicinal Devices, Berlin, Germany; D. Sacks, Reading Hospital and Medical Center, West Reading, Pa; L. Sagnard, Sanofi Pharma, Gentilly, France; K.L. Schulte, Vascular Medicine, Evangelisches Krankenhaus, Berlin, Germany; S. Schroeder, Schering AG, Berlin, Germany; F. Spengel, Klinik Feldafing, Feldafing, Germany; D.E. Strandness, Jr, University of Washington, Seattle, Wash; C.S. Stuerzebecher, Schering AG, Berlin, Germany; M. Tschoepl, University of Basel Medical School, Basel, Switzerland.
The Trans-Atlantic Conference on Clinical Trial Guidelines was sponsored by unrestricted educational grants from Hoechst Marion Roussel, Sanofi Pharmaceuticals, and Schering AG (main sponsors), as well as Otsuka America Pharmaceuticals and Schwarz AG.
- Copyright © 1999 by American Heart Association
The European Agency for the Evaluation of Medicinal Products. Note for guidance on the clinical investigation in the treatment of chronic peripheral arterial occlusive disease. Publication No. CPMP/EWP/233/95 final. For more information write the European Agency for the Evaluation of Medicinal Products, 7 Westferry Circus, Canary Wharf, London E14 4HB, United Kingdom. E-mail firstname.lastname@example.org
Colwell JA. Aspirin therapy in diabetes. Diabetes Care. 1997;20:1767–1771.
Yudkin JS. How can we best prolong life? Benefit of coronary risk factor reduction in non-diabetics and diabetic subjects. BMJ. 1993;306:1313–1318.
Hiatt WR, Regensteiner JD, Hargarten ME, Wolfel EE, Brass EP. Benefit of exercise conditioning for patients with peripheral arterial disease. Circulation. 1990;81:602–609.
Weidinger P, Bachl N. Fünf Jahre Wiener Erfahrung mit ambulanten Claudicatiogruppen. In: Diehm C, Gerlach HE, eds. Bewegungstherapie bei peripheren arteriellen Durchblutungsstörungen. München, Germany: Zuckschwerdt Verlag; 1988:16–27.
Vecht RJ, Nicolaides AN, Brandao E, Daskalakis G, Groshar G, James J, Kenyon JR, Eastcott HHG. Resting and electrocardiographic findings in patients with intermittent claudication. Int Angiol. 1982;1:119–121.
Hiatt WR, Hirsch AT, Regensteiner JG, Brass EP. Clinical trials for claudication: assessment of exercise performance, functional status, and clinical end points. Circulation. 1995;91:1–8.
Labs KH, Roessner M, Aschwanden M, Jeanneret C, Gehrig A, Jaeger KA. Reproducibility of fixed load treadmill testing. J Vasc Investig. 1998;4:55–60.
Labs KH, Nehler M, Roessner M, Jaeger KA, Hiatt WR. Reproducibility of treadmill testing in peripheral arterial disease: a comparison of a constant load with a graded treadmill protocol. Vasc Med. In press.
Second European Conference on Critical Leg Ischemia. Circulation. 1991;84(suppl 4):1–26.
Second European consensus document on chronic critical leg ischaemia. Eur J Vasc Surg. 1992;6(suppl A):1–28.
Smith GD, Shipley MJ, Rose G. Intermittent claudication, heart disease risk factors, and mortality. Circulation. 1990;82:1925–1931.
Newman AB, Siscovick DS, Manolio TA, Polak J, Fried LP, Borhani NO, Wolfson SK. Ankle-arm index as a marker of atherosclerosis in the cardiovascular health study: Cardiovascular Heart Study (CHS) collaborative research group. Circulation. 1993;88:837–845.
Belch JJF, Diehm C, Henry M, Rieger H, Rudofsky G, Frings M, Hornig F, Söhngen W. Evaluating treatment of critical limb ischaemia: for the European PARTNER concept study groups. Crit Ischaem. 1993;2:4–12.
Carter SA. Role of pressure measurement. In: Bernstein EF, ed. Vascular Diagnosis. 4th ed. St. Louis, MO: CV Mosby; 1993:486–512.
Guilmot JL, Diot E. Treatment of lower limb ischaemia due to atherosclerosis in diabetic and non-diabetic patients with Iloprost, a stable analogue of prostacyclin: results of the French multicentre trial. Drug Invest. 1991;3:351–359.
Diehm C, Hübsch-Müller C, Stammler F. Intravenöse Prostaglandin E1-Therapie bei Patienten mit peripherer arterieller Verschlusskrankheit (AVK) im Stadium III: eine doppelblinde, plazebokontrollierte Studie. In: Heidrich H, Böhme H, Rogatti W, eds. Prostaglandin E1 Wirkung und Therapeutische Wirksamkeit. Heidelberg, Germany: Springer; 1988:133–143.
Belch JJF, Söhngen W. Peripheral arterial disease. In: O’Grady J, Joubert PH, eds. Handbook of Phase I/II Clinical Drug Trials. Boca Raton, Fla: CRC Press; 1997:249–278.