V.A. cooperative study on antiplatelet agents in diabetic patients after amputation for gangrene: I. Design, methods, and baseline characteristics

V.A. cooperative study on antiplatelet agents in diabetic patients after amputation for gangrene: I. Design, methods, and baseline characteristics

V.A. Cooperative Study on Antiplatelet Agents in Diabetic Patients After Amputation for Gangrene: I. Design, Methods, and Baseline Characteristics Joh...

1MB Sizes 0 Downloads 18 Views

V.A. Cooperative Study on Antiplatelet Agents in Diabetic Patients After Amputation for Gangrene: I. Design, Methods, and Baseline Characteristics John A. Colwell, Stephen F. Bingham, Carlos Abraira, James W. Anderson, Hau C. Kwaan, and The Cooperative Study Group

ABSTRACT: This report describes the experimental design, methods, and baseline characteristics of patients enrolled in a Veterans Administration Cooperative Study on the effect of aspirin (325 mg t.i.d.) and dipyridamole (75 mg t.i.d.) (110 patients), or placebo (121 patients) on major vascular outcome variables in noninsulin-dependent diabetic patients with either a recent amputation for gangrene (n = 207) or active gangrene (n = 24). It also describes the baseline characteristics of the patients. A total of 231 patients of 563 screened (41%) were enrolled at 11 partcipafing V.A. Medical Centers during a 39 month period. The median age at entry was 60 years, the median duration of diabetes was 10 years, and weight was 110% of desirable. All patients were men. Sixty-eight percent were treated with insulin and 32% with diet alone. Only 42% were smokers at entry, 40% had retinopathy, 61% sensory neuropathy, 42% hypertension, and 29% had a history of myocardial infarction, angina, and/or congestive heart failure. Thirteen percent had a history of cerebrovascular disease. Despite randomization, the treatment group had an increased frequency of a history of cerebrovascular disease (p = 0.01), diagnosed as stroke (p = 0.03), a finding suggesting that the treatment group was at a slightly increased risk fer vascular disease upon enrollment in the study. Other baseline variables did not differ significantly between the two groups. This study should provide definitive data on the efficacy of these doses of antiplatelet agents in preventing further vascular disease in diabetic men with gangrene or recent amputation for gangrene, using death due to vascular disease and subsequent amputation of the opposite extremity for gangrene as major outcome variables. It should also give useful information on the effect of antiplatelet therapy on other vascular outcome variables such as same side amputations, myocardial infarction, stroke, transient ischemic attack, retinopathy, and renal failure. Finally, the study should provide useful data on the natural history and significance of risk factors in this patient population. KEY WORDS: Diabetes mellitus, antiplatelet agents, Iower extremity vascular disease, V.A. Cooperative Study, amputation, gangrene.

Received August 29, 1983; revised February 2, 1984. Address reprint requests to: John A. Colwell, 109 Bee Street, Charleston, SC 29403 LontroUedClinicalTrials 5:165-183 (1984) © ElsevierScience Publishing Co., Inc. 1984 52 Vanderbilt Ave., New York, New York 10017

165

0197-2456/84/$03.00

166

J.A. Colwell et al.

INTRODUCTION Diabetic patients have accelerated arteriosclerotic vascular disease w h e n compared to a control population. Atherosclerosis is a major cause of death in patients with diabetes mellitus in large epidemiologic studies [1], in autopsy series [2-4], and after retrospective analyses of clinical events [5]. In the Framingham Study, it was shown that diabetic men die twice as often from cardiovascular disease compared to nondiabetic men of the same age [1]. Peripheral vascular disease is particularly devastating in patients with diabetes mellitus. Retrospective studies have shown that death rates following amputation for gangrene were 35%-57% at 3 years and 50%-73% at 5 years after surgery [6-18]. It has been estimated that from 30%-61% of patients will have a second amputation within 3 years of the first amputation for diabetic gangrene [9,11,12,16,18,19]. The ischemic lesions result primarily from arteriosclerosis in large leg arteries: the diabetic patient has more extensive stenotic and occlusive major artery disease in vessels below the knee than does the nondiabetic patient [20,21]. Recent theories of the pathogenesis of arteriosclerosis have supported a role for blood platelets in the process [22-24]. Altered platelet behavior is present in many diabetic patients. Thus, increased platelet adhesiveness, hypersensitivity to platelet aggregating agents, increased production of the proaggregatory prostaglandin metabolite thromboxane, platelet-plasma interactions, decreased platelet survival in vivo, and increased plasma levels of a platelet-specific protein, beta-thromboglobulin, have all been reported in the diabetic state [25]. It is not yet clear whether the platelet abnormalities in diabetes mellitus precede the vascular disease, are a result of the vascular disease, or bear no relationship to the vascular disease. However, in view of the possible involvement of the blood platelet in arteriosclerosis, and the suggestive findings in other forms of large vessel disease that antiplatelet agents may be of some benefit [26-28], we designed a multihospital, doubleblind, randomized, placebo-controlled clinical trial to determine if treatment of diabetic patients with peripheral vascular disease using antiplatelet agents would alter either the progression of their peripheral vascular disease or mortality. We chose patients with advanced lower extremity vascular disease and used amputation for gangrene of the opposite leg and vascular death as primary outcome variables. Other vascular events such as same side amputations, myocardial infarction, stroke, transient ischemic attack, angina, claudication, and retinopathy were considered to be secondary outcome variables, but were studied equally carefully. At the beginning of the study, it was known that aspirin would block the platelet release reaction in diabetic patients [29], and that dipyridamole had effects in vivo on platelet survival that could be beneficial to patients with extensive vascular disease [30]. There was also evidence that aspirin plus dipyridamole were synergistic in an animal model of diabetes [31]. Prostacyclin had not yet been discovered. Accordingly, the combination of aspirin and dipyridamole was chosen as antiplatelet agents, and dosages similar to those used in other multicenter trials were used.

Amputation in Diabetes: AntiplateletAgents

167

METHODS Experimental Design Patient selection Inclusion criteria. All study patients were required to have a positive diagnosis of diabetes mellitus demonstrated by meeting one of the following criteria:

a. three fasting serum glucose values greater than 130 mg/dl. b. abnormal oral glucose tolerance test using the Wilkerson criteria [32]. Diabetic patients with lower extremity vascular disease also had to meet the criteria for one of the following two strata: a. active gangrene (Group A) due to diabetic vascular disease. Gangrene was defined as necrosis with darkening of tissue and absence of sensation in the affected area. Infection could be present or absent. Patients with ischemic ulcers or ischemic rest pain with a Doppler ratio less than 0.5 in the involved limb were considered as having gangrene. b. amputation (Group B) of part of a lower extremity within the previous 12 months for diabetic gangrene. This was subsequently changed to 18 months because of difficulties encountered in enrolling a sufficient number of patients within 12 months of amputation. Exclusion criteria. Patients were excluded if there was a history of gastrointestinal hemorrhage, active peptic ulcer within the previous two years, bleeding tendency, malignant disease with life expectancy less than three years, or allergy to aspirin or dipyridamole. Patients who required anticoagulant therapy, treatment with oral antidiabetic drugs, or antiplatelet drugs were also excluded. Women, uncooperative or unreliable patients, and patients with proliferative retinopathy, hypotension (blood pressure < 100/60), advanced renal disease (serum creatinine > 3 mg/dl), or persistently inhibited second phase of platelet aggregation were also excluded. During the course of the study, proliferative retinopathy was dropped as an exclusion criterion because there was no evidence that antiplatelet agents had led to retinal bleeding. Patients with bilateral amputations of both lower extremities were excluded, unless the amputation on the remaining leg was for one toe or less. A detailed discussion of the patient selection procedure and other operational aspects of the study was provided in an operations manual, which was discussed with the participants and revised before beginning the study. Patient recruitment. Patients in this study were recruited primarily from diabetic and peripheral vascular clinics and from inpatients at the 11 participating V.A. Medical Centers. Surgical and pathology reports of amputations were routinely monitored in order to locate eligible patients. Potential participants were screened by the participating investigators using the criteria noted above. In addition, investigators thoroughly explained the study to potential partic-

168

J.A. Colwell et al. ipants, gave them copies of the consent form to review at their leisure, and answered any questions about the study. Patient enrollment. Eligible patients were asked to sign a consent form before being randomized. Consenting patients were randomly assigned to one of two treatment groups. The active treatment group was prescribed aspirin (325 mg) and dipyridamole (75 mg) to be taken three times daily. The control group was prescribed placebos of identical appearance to be taken three times daily also. All patients were given a list of aspirin-containing drugs to avoid and were supplied with acetaminophen (325 mg tablets) to take for minor pains as needed. Study drugs were donated and prepared in individual patients kits by Boehringer Ingelheim Ltd. (dipyridamole) and Winthrop Laboratories (aspirin), and supplied regularly to the VAMC pharmacies by the Central Research Pharmacy Coordinating Center. Treatment assignments were randomly generated for both Group A and Group B within each center. The random assignments were constrained to provide for balance between the two treatment groups over time within each participating center. Participating investigators and patients were not informed about specific treatment assignments, which were known only to the Biostatistical Support Center. Specific arrangements were made for breaking the code, should an emergency arise. Patient visits. Patients at each participating facility were scheduled to be seen at quarterly intervals for a minimum of three years after enrollment. All surviving patients were to be followed to a common completion date. At each visit, a history, physical examination, and laboratory tests were to be done. Compliance was checked by pill counting. Standard medical care for diabetes was provided, including diet, exercise, and/or insulin therapy, but no attempt was made to rigorously control blood glucose values. Therapy for hypertension, congestive heart failure, and other medical diseases was given during the study according to standards prevailing in each hospital or clinic. Outcome variables. Major vascular events were recorded in detail. Surgical reports, pathology interpretations, and hospital summaries were required on all amputations and deaths. Death certificates and final hospital progress notes were also obtained routinely for all patients who died. Yearly electrocardiograms were obtained and read separately by two cardiologist consultants without knowledge of patients' treatment, in order to document recognized myocardial infarctions and to identify any that were unsuspected. In addition, all events compatible with acute myocardial infarction were reported. This information included history, hospital summary, serial ECGs and serial cardiac enzyme tests. Standard criteria for the diagnosis of ischemic heart disease were used [33]. Other vascular events, including lower extremity bypass procedures, strokes, and transient ischemic attacks were also documented. History of angina and/ or claudication was recorded. Clinical descriptions of these outcome variables were available in the operations manual. Retinopathy was assessed at each quarterly visit by the physician, and the patient was referred for ophthalmologic evaluation whenever retinopathy was suspected. Lower extremity

169

Amputation in Diabetes: Antiplatelet Agents

Doppler readings were performed at each visit, and photographs of lower extremity lesions were made when present.

Sample size and data analysis. Original sample size calculations used a onetailed test with alpha error of 0.05 and beta error of 0.20. Within each strata, a 30% reduction in frequency of events (gangrene or amputation for gangrene) was postulated. These assumptions led to the following original sample size requirements:

Group A (Gangrene) B (Amputation)

Estimated frequency of events % 50 60

Desired frequency % 35 42

# Patients 270 18__66 456

Recruitment in Group A fell well below the predicted numbers, and recruitment was also slow in Group B. By one year after the first patient was enrolled, it was apparent that the required 270 patients in Group A could not be enrolled in the time allotted for recruitment. Since these patients were similar to those in Group B in regard to the predicted risk for subsequent amputation of the opposite leg for gangrene, Groups A and B were combined. Also based upon recruitment projections, a second decision at this time was to combine the major outcome variables of vascular deaths and amputation of the opposite extremity for gangrene. We estimated from retrospective studies [6-19] that 30% of the patients would die of vascular causes and that an additional 30% would have an amputation of the opposite extremity in 3 years. Sample size requirements based on these changes were recomputed in October 1978, with the following result:

Group A&B

Estimated frequency of events % 60

Desired frequency % 42

# Patients 186

a = 0.05, 13 = 0.20, one-tailed test. With an anticipated dropout rate of 20%, a total sample size of 232 would be required with enrollment to be completed by May 1980. This protocol change was reviewed and approved by the Operations Committee and by the V.A. Cooperative Studies Evaluation Committee. While this was a logical decision t o make, it resulted in a design that had limited p o w e r to detect an effect of antiplatelet therapy on opposite side amputations or vascular deaths independently. Further, we could no longer expect to detect any effect of antiplatelet therapy on preventing amputations of the same limb in patients with gangrene (Group A). However, we reasoned that these Group A patients could be combined with those patients w h o would be enrolled after a recent amputation of one to three toes, and that the effect of antiplatelet therapy on subsequent amputations on the same side could be

170

I.A. Colwell et al. analyzed. Finally, we recognized that we had little knowledge of the expected numbers of other major vascular events such as strokes, transient ischemic attacks, and myocardial infarctions to expect in this population. By pooling all major vascular events, however, we could expect to increase the power of the study with this sample size. The principal data analysis technique that is used for this study is the Cox proportional hazards regression model [34]. The major outcome variable is time to vascular death and opposite limb amputation. Included as covariates within the Cox model are baseline characteristics that are either significant vascular risk factors or are substantially unbalanced between treatment groups. All patients will be analyzed in the treatment group to which they are randomized.

Laboratory determinations. Platelet aggregation was performed at each participating hospital at each patient visit. Both the technician and investigator were masked by coding the tubes of blood, which were drawn by a second technician. In each run, a set of coded d u m m y specimens were run. Some of these were drawn from volunteers who had taken aspirin the evening before the test, and some were from volunteers who had taken no aspirin or antiplatelet agents for two weeks before the test. The success of blinding procedures are checked by questioning patients, technicians, and investigators after the study is completed. Platelet aggregation was performed at each center by the method of Born [35]. Instead of adenosine diphosphate, the aggregating agent used was epinephrine in a final concentration in platelet-rich plasma of 50 ~M. The maximum change in optical density was measured in duplicate runs for each plasma sample. The tracing of platelet aggregation was sent to the Hematology Reference Laboratory for evaluation. Serum samples for glucose, cholesterol, and triglycerides were drawn after an overnight fast at each patient visit and sent frozen to the Centralized Biochemical Laboratory. Glucose was determined by the glucose oxidase technique and cholesterol and triglyceride levels were measured using the semiautomated method standardized by the Lipid Research Clinics Program [36]. Quality control was established with the cooperation of the Lipid Standardization Laboratory, Center for Disease Control, Atlanta, Georgia. At each visit, another blood sample was drawn 2 hours after the ingestion of drug (or placebo). Plasma samples were sent frozen to the Hematology Reference Laboratory for determination of plasma salicylate [37] and dipyridamole [38] levels. Serum samples were also sent to the hospital clinical laboratory for multichannel analysis to screen for drug toxicity at each visit; qualitative tests for urine protein were done upon enrollment and at the last patient visit by the hospital clinical laboratory. Study organization. Ten Veterans Administration Medical Centers were the original participants in this study. Funding was provided by the Veterans Administration Cooperative Studies Program. Because of patient recruitment problems, one facility was dropped and two facilities were added after the study had begun. It was estimated that these two facilities would be needed

Amputation in Diabetes: Antiplatelet Agents

171

to reach recruitment goals in the time allotted. Overall management of the study has been the responsibility of the Study Chairman and the Study Administrator at the Charleston, South Carolina V.A. Medical Center. Data processing and statistical analysis has been the responsibility of the Coordinating Center at the Perry Point, Maryland V.A. Medical Center. The Central Research Pharmacy at the Albuquerque, New Mexico V.A. Medical Center has been responsible for drug distribution. Two reference laboratories were required in this study. The Hematology Reference Laboratory at the Chicago (Lakeside) V.A. Medical Center was to determine plasma salicylate and dipyridamole levels and to evaluate platelet aggregation curves. The Biochemical Reference Laboratory at the Charleston V.A. Medical Center was to determine serum glucose, triglyceride, and cholesterol levels. Five committees and the Study Group met periodically to review various aspects of this study as it progressed. The Study Group was composed of all participating investigators and consultants. At each meeting, this group discussed the study's progress and any problems that were being encountered. Suggestions for improvement in the study were made and proposed changes were presented and discussed. The Executive Committee was the management group and major decision-making body for the operational aspects of this study and acted as its Publications Committee. The Operations Committee was an independent monitoring committee composed of experts from a variety of relevant scientific disciplines, including biostatistics. It provided a continuing critical and unbiased evaluation of the study's progress at meetings every 6-9 months during the study. For these meetings, it was provided with unblinded data summaries and numerous interim analyses. It could recommend discontinuation of the study for such reasons as poor performance by participants, small likelihood of clinically significant results, definitive therapeutic effects on major end points, or unacceptable side effects of drug therapy. The Human Rights Committee was responsible for ensuring that patients' rights and safety were not being violated. The Major End Points Committee reviewed all deaths, amputations, myocardial infarctions, strokes, and transient ischemic attacks in a blinded fashion and made judgments as to whether they were due to vascular or nonvascular causes, based on criteria established before analyses were performed. The Enrollment Committee was also masked, and reviewed all patients entered into the study to determine if enrollment criteria were met. Site visits were made at least once to each participating station by the Study Chairman and Study Biostatistician during the study. Site visits by representatives of the Human Studies Committee were also made at less frequent intervals.

RESULTS Patient Enrollment The first study patient was enrolled in this study on March 1, 1977, and the last patient was enrolled on May 6, 1980. A total of 231 patients were enrolled in this 39-month period, 110 in the drug treatment group, and 121 in the placebo group. This equalled 41% of all patients screened.

172

]. A. Colwell et al.

Table 1

Screening a n d Enrollment S u m m a r y A

Number of patients screened Number enrolled Number excluded Reason for exclusion ° Uncooperative or unreliable b Refused to enter study Proliferative retinopathy Treatment with oral antidiabetic drugs antiplatelet drug Allergic to study drugs Active ulcer within last 2 years Serum creatinine >3 mg/dl Anticoagulant therapy 2nd phase of platelet aggregation inhibited Other

B

C

D

Hospital E F G

K

Total

88 29 47 66 24 56 70 68 44 30 41 M 11 25 26 16 23 37 16 17 20 6 54 18 22 40 8 33 33 52 27 10 35

H

I

J

563 231 332

12

4

1

6

3

4

8

14

8

2

13

75

12 6

3 1

0 3

11 2

0 0

5 8

6 0

0 9

3 2

1 2

2 0

43 33

0 1 2

1 3 0

1 2 1

1 0 0

0 0 1

4 1 1

6 0 1

4 5 0

0 0 0

0 0 0

0 0 4

17 12 10

1 0 1

0 1 1

0 0 0

0 1 0

0 1 0

0 0 0

0 1 1

8 2 2

0 1 0

0 0 0

0 0 0

9 7 5

1 4

1 0

1 0

1 1

0 0

0 1

0 0

0 1

0 1

0 1

0 1

4 10

°Patients with more than one reason for exclusion are not counted. blncludes mentally incompetent, alcoholic, behavioral disorder.

A total of 332 patients were screened a n d excluded for various reasons. A l t h o u g h 107 patients were excluded for several reasons, 225 patients were excluded for a single reason only. Table I indicates the reasons reported m o s t frequently for this last g r o u p of patients. It also illustrates the a m o u n t of variability b e t w e e n participating facilities. A total of 20 e n r o l l m e n t errors out of the 231 patients enrolled (8.7%) occurred in this s t u d y (Table 2). Thirteen of these cases are considered to be errors because of a possible error in the diagnosis of diabetic gangrene. In eight of these patients, a m p u t a t i o n was due to osteomyelitis w i t h o u t gangrene. In the other five, data on the index event were not sufficient to be certain w h e t h e r g a n g r e n e or osteomyelitis was present. Four patients did not have diabetes, a n d in 3 patients, the index event occurred more t h a n 18 m o n t h s prior to entry. These patients were all retained in the s t u d y a n d initial analyses carried out with patients in the g r o u p to which they were assigned.

Table 2

Enrollment Errors No. patients

Osteomyelitis Questionable gangrene Impaired glucose tolerance without diabetes mellitus Amputation > 18 months before entry TOTAL: NO. ENROLLED: PERCENTAGE

8 5 4 3 20 231 8.7%

173

Amputation in Diabetes: Antiplatelet Agents

Baseline Characteristics A n u m b e r of baseline characteristics were recorded (Table 3). In order to characterize the patient population, the baseline characteristics are presented in two ways. Some are shown as frequencies (number and percent), while others are shown as distribution percentiles. The 50th percentile is the median of the distribution and the 5th and 95th percentiles include 90% of the observations. The patients enrolled in the study had a median age of 60 years and a median duration of diabetes of 10 years. In Group B, the majority had amputations of part or all of the lower extremity below the knee in the past. Only 24 patients were enrolled with gangrene or low Doppler readings (Group A). Sixty-eight percent of the patients were being treated with insulin and the remainder by diet alone. In the past, 55% had been treated with sulfon-

Table 3 Baseline Characteristics Characteristic

N

5%a

50%a

Age

231

47

60

78

Duration of diabetes (Yrs)

231

1

10

31

Site of amputation (Group B) Below knee 1-3 toes Above knee Transmetatarsal

207

Site of gangrene (Group A) 1-3 toes Doppler <0.5

95%°

No. ~

%

103 66 27 11

50 32 13 5

15 9

62 38

162 54 15

70 23 7

156 75

68 32

126 13 43

55 6 19

54 81 96

23 35 42

24

Race White Black Hispanic

231

Current therapy Insulin Diet

231

Previous therapy Sulfonylureas Phenformin Regular aspirin

231

Smoking history Never smoked Ex-smoker Currently smokes Amount smoked (pack Yrs)c

231

History of complications Sensory neuropathy Retinopathy Autonomic neuropathy Renal disease Blindness

231

4

40 140 93 66 30 8

100 61 40 29 13 3

(Continued)

174

]. A. ColweU et al. Table 3

Continued

Characteristic

N

History of vascular disease Hypertension Claudication Heart disease Previous myocardial infarction Congestive heart failure Angina pectoris Cerebrovascular disease Stroke Transient ischemic attack

230

Blood pressure Systolic Diastolic

230

Pulse

5%"

50%" 96 85 67

95%" 42 37 29

43

19

40 30

17 13

30

13 23 7

10 3

110 60

135 80

175 98

227

64

80

100

Percent of desirable bodyweight Serum triglyceride (mg/dl)d Serum cholesterol (mg/dl)r Serum glucose (mg/dl)d

230

79

110

135

135 195 135

62 125 60

139 191 150

335 280 255

Hemoglobin (gin%) Creatinine (mg/dl)

230 230

10.4 0.7

13.2 1.1

16.4 1.9

Urine protein

No. ~

%

175 30 8 8 7

76 13 4 4 3

228

0 1+ 2+ 3+ 4+ "Percentiles. ~Frequendes. ~Exdudes patients who have never smoked. ¢Fasting sample. ¢Random sample.

ylureas a n d 6% w i t h p h e n f o r m i n at some time. Only 42% were smokers at the time of entry a n d 23% h a d never smoked. M a n y patients had a history of complications at entry: sensory n e u r o p a t h y in 61%, retinopathy in 40%, autonomic n e u r o p a t h y in 29%, a n d history of renal disease in 13%. Hypertension h a d existed in 42% of the patients, and 37% gave a history of claudication. Twenty-nine percent of patients had a history of heart disease (previous myocarfli~l infarction, congestive heart failure, or angina pectoris). Thirteen percent h a d a history of cerebrovascular disease. Median blood pressure was 135/80 a n d weight was 110% of desirable level. Laboratory values revealed m e d i a n serum cholesterol a n d fasting serum triglyceride levels in the normal range (191 a n d 139 mg/dl, respectively) a n d a slightly elevated fasting serum

Amputation in Diabetes: Antiplatelet Agents

175

glucose level of 150 mg/dl. M e d i a n e n t r y h e m o g l o b i n w a s 13.2 g m % . S e r u m creatinine w a s 1.1 mg/dl, a n d 24% h a d s o m e d e g r e e of proteinuria u p o n enrollment. Distribution of baseline characteristics b y t r e a t m e n t g r o u p s is s h o w n in Tables 4 a n d 5. T r e a t m e n t g r o u p s w e r e c o m p a r a b l e in regard to age, site of a m p u t a t i o n , weight, history of p r e v i o u s myocardial infarction, congestive h e a r t failure, angina, renal disease, a u t o n o m i c n e u r o p a t h y , a n d s e r u m cholesterol a n d creatinine levels. Equal n u m b e r s of patients (55%) h a d t a k e n oral h y p o g l y c e m i c a g e n t s prior to entry. Regular aspirin use w a s also c o m p a r a b l e in the t w o g r o u p s b y history.

Table 4

Baseline Characteristics b y T r e a t m e n t G r o u p s (History)

No. Age (years) <56 56--60 61-65 >65 Mean (yr.) Site of amputation (Group B) Below knee 1-3 toes Above knee Transmetatarsal Site of gangrene (Group A) 1-3 toes Doppler <0.5 Race White Black Hispanic % Desirable body weight <91 91-110 111-130 >130 Mean % Large vessels Hypertension Mild Moderate Severe Claudication Previous myocardial infarction Congestive heart failure Cerebrovascular disease" Stroke ~ Transient ischemic attack Angina

39 29 19 23

Aspirin + Dipyridamole % N = 110 35 26 17 21 59.4

Placebo No. 39 26 34 22 N -- 108

47 32 15 5

56 34 12 6

55 45

9 4

74 24 3

81 28 12

12 37 45 6

15 44 47 15

N = 11 6 5

N = 121 48

26 20 7

43 24 18 6

33 16 17 19 16 5

14

12 36 39 12 110.1

N = 110

36 19 18 21

67 23 10 N = 121

110.0 53

69 31 N = 121

N = 109 13 40 49 7

52 31 11 6 N = 13

N = 110 81 26 3

32 21 28 18 59.9

N = 99 47 32 15 5

% N = 121

49 24 22 9 15 5

13

36 13 18 6

11 19 6 40 20 18 7

7 2 16

6 2 13

(Continued)

176

J.A. Colwell et al.

Table 4

Continued Aspirin + Dipyridamole % No.

Small vessels Retinopathy Blindness Renal disease Neuropathy Sensory Autonomic Duration of DM (years) <6 6--15 16-25 >25 Mean (years)

Placebo No.

N = 110 48 6 16

N = 121 44 5 15

45 2 14

65 28

69 35

24 46 17 13

39 49 20 13

N = 110 26 51 19 14

78 32

Previous therapy Sulfonylureas Phenformin Regular aspirin use

60 6 21

Smoking Present smoker Present nonsmoker Never smoked

52 58 23

121 57 29

N = 121 32 41 16 il 12.0

13.6

Present therapy Insulin Diet

37 2 12 N=

N = 110 71 31

%

N = 110

N = 121 71 29

78 43

55 5 19

66 7 22

47 53 21

44 77 31

N = 110

65 35 N=

N = 110

121 55 6 18

N = 121 36 64 26

Note: Differences are not statistically significant (p > 0.05) except where noted. "p = 0.01 (Fisher's Exact Test [2-tailed]) [45]. bp = 0.03. The t r e a t m e n t g r o u p w a s at h i g h e r risk r e g a r d i n g a history of cerebrovascular disease (p = 0.01) d i a g n o s e d as stroke (p = 0.03). M o r e t h a n twice as m a n y p a t i e n t s in the t r e a t m e n t g r o u p t h a n in the placebo g r o u p h a d a history of stroke or t r a n s i e n t ischemic attack. Further, this g r o u p h a d a slightly higher risk w i t h r e g a r d to the following variables, t h o u g h n o n e of these differences w e r e statistically significant: mild h y p e r t e n s i o n , r e t i n o p a t h y , s e n s o r y n e u r o p a t h y , d u r a t i o n of diabetes, t r e a t m e n t w i t h insulin, a n d s m o k i n g history. Fasting s e r u m glucose a n d triglyceride levels w e r e also slightly higher in the treated g r o u p (Table 5). The placebo g r o u p h a d slightly m o r e patients w i t h claudication.

DISCUSSION It w a s p l a n n e d at the outset that the m a j o r vascular o u t c o m e variables of the s t u d y s h o u l d be easily recognizable a n d subject to m i n i m a l o b s e r v e r error. Therefore, a m p u t a t i o n of the o p p o s i t e extremity for g a n g r e n e a n d vascular d e a t h w e r e c h o s e n as p r i m a r y o u t c o m e variables.

177

Amputation in Diabetes: Antiplatelet Agents Table 5

Baseline Characteristics by T r e a t m e n t G r o u p (Laboratory) Aspirin + Dipyridamole No.

Fasting serum glucose (mg/dl) > 250 201-250 151-200 < 151 Mean Fasting serum triglycerides (mg/dl) > 250 201-250 151-200 151 Mean Serum cholesterol (mg/dl) > 250 201-250 151-200 < 151 Mean Serum creatinine (rag/dl) > 2.0 1.6--2.0 1.1-1.5 < 1.0 Mean

Placebo

%

No.

10 15 30 45

4 9 17 38

N = 67 7 10 20 30

N = 68

163 N = 67 9 9 12 37

5 7 9 47

160 N = 94 13 25 45 18

195 N = 109

13 25 45 18 193 N = 121

4 15 39 43 1.19

7 10 13 69 150 N = 101

9 29 46 17

4 16 42 47

6 13 25 56 148 N = 68

13 13 18 55

8 27 43 16

%

4 11 55 51

3 9 45 42 1.18

O t h e r o u t c o m e variables of obvious interest were described prospectively, and carefully d o c u m e n t e d . These included same side amputations, myocardial infarction (with central ECG interpretation), cerebrovascular disease, lower extremity vascular bypass procedures, and serial Doppler readings of the leg vessels. Finally, r e t i n o p a t h y and renal failure were followed regularly by clinical a n d laboratory procedures, as were the more subjective o u t c o m e variables of claudication a n d angina. As is true in m a n y cooperative studies [39], we were overly ambitious in expectations of r e c r u i t m e n t in one of the groups (Group A: patients with fresh gangrene). We f o u n d that these patients were usually surgical candidates on admission to the hospital a n d neither patients nor physicians were seriously interested in e n r o l l m e n t in a double blind placebo controlled study. Most of these patients w e r e soon eligible for e n r o l l m e n t after amputation, and could be included in G r o u p B. Therefore, w e included all patients in the long-term study, a n d expected to enroll the majority of the patients following a m p u tation for gangrene. Estimates of major vascular outcome variables were made, including vascular d e a t h s and opposite side amputations, and a sample size was t h e n calculated. We were successful in recruiting this sample, which w o u l d , hopefully, give us sufficient p o w e r for testing the primary h y p o t h e s e s that antiplatelet t h e r a p y might protect these patients from s u b s e q u e n t vascular d e a t h a n d / o r a m p u t a t i o n of the opposite leg.

178

J.A. Colwell et al. Platelet aggregation was monitored at ~ach hospital. Because the effect of aspirin to inhibit the platelet release reaction can easily be seen by the technician performing the platelet aggregation curves, we set up a scheme to keep personnel masked as to which sample was under study. Platelet aggregation responses in drug and control group were to be analyzed centrally, with data used to determine if the treatment group had appropriate responses to therapy or if the placebo group had responses compatible with aspirin effect. Peak plasma levels of dipyridamole and salicylate were checked in order to give some index of drug absorption and distribution in the drug treatment group and of compliance in the placebo group. Indices of metabolic control in this study are serial levels of fasting serum glucose, triglycerides, and serum cholesterol. At the inception of the study, hemoglobin Ale was not recognized to be a useful index of metabolic control in diabetic patients. We were well into the study when HgbAlc methods were developed. Further, HDL cholesterol was found to be a probable risk factor for diabetic vascular disease after the study began. Since baseline data were lacking, these determinations were not added during routine visits. Exclusion criteria in this study were such that enrollment was 41% of all patients who were reported as screened. Since it is likely that a group of patients was excluded based on chart reviews, this figure is probably too high. It is apparent that this fact, plus the heterogeneous nature of the disease process, limits any broad generalizations from the study to a general population of diabetic subjects. On the other hand, results from the study will be applicable to a select group of noninsulin-dependent males, with recent gangrene or amputation for gangrene of the lower extremity. The reasons for patient exclusion were most frequently due to an assessment that the patient might be uncooperative or unreliable, or that he simply did not wish to participate in the study. Because of some evidence that therapy with oral antidiabetic agents may be associated with premature deaths from cardiovascular disease [40], such patients were excluded from this study. Results will thus be applicable to patients treated with insulin or diet alone during the course of the study. It is apparent from the baseline characteristics that these patients represent a group with evidence of advanced generalized vascular disease (Tables 3 and 4). This was not surprising, and is in accord with other published studies [6-19]. It clearly raises the question of the suitability of such patients for an antiplatelet trial, since such secondary intervention could be too late in the course of vascular disease. However, we felt that this concern was more than balanced by the view that this population would proceed at a rather predictable rate with major vascular end points and that a protective effect of antiplatelet agents, if present, might be seen in a relatively short time span. Further, the presence of many findings of vascular disease at baseline would allow multivariate analysis upon completion of the study in order to identify potentially high risk patients in the placebo group. Also, the probability of multiple sites of preenrollment amputations would give an opportunity to relate this index of severity of vascular disease to subsequent events. As the study progressed, we were encouraged to learn that evidence of platelet activation was reported in several studies of patients with peripheral vascular disease [41-43]. In addition, one report suggested that aspirin and

Amputation in Diabetes: Antiplatelet Agents

179

dipyridamole could reverse symptoms associated with pregangrene or gangrene in patients with hypersensitive platelets and spontaneous platelet aggregation [44]. As shown in Tables 4 and 5, the treatment and placebo groups are generally comparable regarding baseline characteristics. One variable, a history of cerebrovascular disease was significantly more common in the treatment group, and several other variables were more common though not significantly so. These factors could put the treated patients at a higher inherent risk for vascular disease when compared to controls, and could contribute to a poor response to antiplatelet drugs. However, if a protective effect of antiplatelet therapy on major vascular outcome variables is seen in the study, it is not likely to be due to decreased vascular disease or its risk factors in the treatment group. Furthermore, use of these variables as covariates in the Cox model will adjust for recognized imbalance [15]. Therefore, this study should give definitive information about the efficacy of antiplatelet agents in diabetic men with recent gangrene or amputation for gangrene, in preventing vascular death and subsequent amputation of the opposite extremity. It may also provide useful data about the role of these agents in preventing such events as same side amputations, myocardial infarction, cerebrovascular accident, retinopathy, renal failure, and the need for arterial bypass in the lower extremity. Finally, it is anticipated that useful data on natural history and on the significance of presumed risk factors in this selected group of patients should emerge from this study. ACKNOWLEDGMENT This work was supported by the Veterans AdministrationCooperativeStudies Program of the Medical Research Service. We would like to gratefullyacknowledgethe help of the housestaffand attendingphysiciansat participating V.A. Medical Centers and affiliatedinstitutions for their assistance in patient recruitment. We would also like to thank Dr. Peter J. Savageof Allen Park, Michigan, Dr. Philip L. Hooper of Albuquerque, New Mexico,and Dr. G. S. Hicksof Jackson, Mississippifor helping with the follow-upof patients who moved during the study. Finally,the expert secretarial help at the PerryPoint Centerand of DonnaAncellat the CharlestonV.A. MedicalCenteris gratefully acknowledged.

REFERENCES 1. Garcia MJ, McNamara PM, Gordon T, Kannel WB: Morbidity and mortality in diabetes in the Framingham population. Six-year follow-up study. Diabetes 23:105-111, 1974 2. Robertson WB, Strong JP: Atherosclerosis in persons with hypertension and diabetes mellitus. Lab Invest 18:538-551, 1968 3. Bell ET: A post-mortem study of vascular disease in diabetes. AMA Arch Pathol 5 3 : ~ J.55, 1952 4. Warren S, LeCompte PM, Legg MA: The Pathology of Diabetes Mellitus, 4th Edition. Philadelphia: Lea and Febiger, 1966 5. Kessler II: Mortality experience of diabetic patients. Am J Med 51:715-724, 1971 6. Levin CM, Dealey FN: The surgical diabetic, a five-year survey. Ann Surg 102:1029-1039, 1935

180

J.A. Colwell et al. 7. Pearse HE, Zeigler HR: Is the conservative treatment of infection or gangrene in diabetic patients worthwhile? Surgery 8:72-78, 1940 8. MendelbergA, SheinfieldW: Diabetic amputations; amputation oflower extremity in diabetes; analysis of 128 cases. Am J Surg 71:70-75, 1944 9. Silbert S: Amputation of the lower extremity in diabetes mellitus. Diabetes 1:297-299, 1952 10. Smith BD: A twenty-year follow-up in fifty below-knee amputations for gangrene in diabetes. Surg Gynec Obstet 103:625-630, 1956 11. Hoar CS, Tortes J: Evaluation of below-the-knee amputation in the treatment of diabetic gangrene. N Engl J Med 266:440--443, 1962 12. Cameron HC, Lennard-Jones JE, Robinson FMD" Amputations in the diabetic: Outcome and survival. Lancet 2:605-607, 1964 13. Warren R, Kihn RB: A survey of lower extremity amputations for ischemia. Surgery 63:107-120, 1968 14. Whitehouse FW, Jurgensen C, Black MA: The later life of the diabetic amputee. Another look at fate of the second leg. Diabetes 17:520-521, 1968 15. Haimovici H: Peripheral arterial disease in diabetes meilitus. In: Diabetes Mellitus, Theory and Practice, EUenberg M, Rifkin H Eds., New York: McGraw-Hill Book Company, 1970 16. Ecker ML, Jacobs BS: Lower extremity amputation in diabetic patients. Diabetes 19:189-195, 1970 17. Kahn O, Wagner W, Bessman AN: Mortality of diabetics treated surgically for lower limb infection and/or gangrene. Diabetes 23:287-292, 1974 18. Roon AJ, Moore WS, Goldstone J: Below-knee amputation: A modern approach. Am J Surg 134:153--158, 1977 19. Goldner MG: The fate of the second leg in the diabetic amputee. Diabetes 9:100-103, 1960 20. Strandness DW, Priest RW, Gibbons GE: Combined clinical and pathologic study of diabetic peripheral arterial disease. Diabetes 13:336-372, 1964 21. Goldenberg S, Alex M, Joshi RA, Blumenthal HT: Nonatheromatous peripheral vascular disease of the lower extremity in diabetes mellitus. Diabetes 8:261-273, 1959 22. Ross R, Glomset JA: The pathogenesis of atherosclerosis. N Engl J Med 295:369-377, 420-425, 1976 23. Mustard JF, Packham MA: The role of blood and platelets in atherosclerosis and the complications of atherosclerosis. Thromb Diath Haemorrh 33:4~ ~56, 1975 24. Stemerman MB: Atherosclerosis: The etiologic role of blood elements and cellular changes. Cardiovasc Med 3:17-36, 1978 25. Colwell JA, Lopes-Virella M, Halushka PV: Pathogenesis of atherosclerosis in diabetes mellitus. Diabetes Care 4:121-133, 1981 26. The Canadian Cooperative Study Group: A randomized trial of aspirin and sulfinpyrazone in threatened stroke. N Engl J Med 299:53-59, 1978 27. Harter HR, Burch JW, Majerus PW, Stanford N, Delunez JA, Anderson CB, Weerts CA: Prevention of thrombosis in patients on hemodialysis by low-dose aspirin. N Engl J Med 301:557-579, 1979 28. Elwood PC, Cochrane AL, Burr ML, Sweetnam PM, Williams G, Welsby E, Hughes SJ, Renton R: A randomized controlled trial of acetyl salicylic acid in the secondary prevention of mortality from myocardial infarction. Br Med J 1:436 ~0, 1974 29. Sagel J, Colwell JA, Crook L, Laimins M: Increased platelet aggregation in early diabetes mellitus. Ann Intern Med 82:733-738, 1975 30. Harker LA, Slichter SJ: Studies of platelet and fibrinogen kinetics in patients with prosthetic heart valves. N Engl J Med 283:1302-1305, 1970

Amputation in Diabetes: Antiplatelet Agents

181

31. Honour AJ, Hockaday TDR, Mann JI: The synergistic effect of aspirin and dipyridamole upon platelet thrombi in living blood vessels. Br J Exp Patho158:268-272, 1977 32. Remein QR, Wilkerson HL: The efficiency of screening tests for diabetes. J Chron Dis 13:6--21, 1961 33. Special Report: Nomenclature and criteria for diagnosis of ischemic heart disease. Circulation 59:607-609, 1979 34. Cox DR: Regression models and life-tables. ] R Statist Soc B 34:187-202, 1972 35. Born GVR: Aggregation of blood platelets by adenosine diphosphate and its reversal. Nature 194:927-920, 1963 36. Manual of Lipid Operations. Lipid Research Clinics Program, Vol. 1. Lipid and Lipoprotein Analysis, Department of Health, Education, and Welfare, Publ. No. (NIH) 75-628, 1974 37. Miles CI, Schenk GH: Fluorescence of acetylsalicylic acid in solution and its measurement in presence of salicylic acid. Analyt Chem 42:656-659, 1970 38. Zak SB, Tallan HH, Quinn GP, Fratta I, Greengard P: The determination and physiological distribution of dipyridamole and its glucuronides in biological material. J Pharm Exptl Therap 141:392-398, 1963 39. Collins JF, Bingham SF, Weiss DG, Williford WO, Kuhn RK: Some adaptive strategies for inadequate sample acquisition in Veterans Administration cooperative trials. Controlled Clin Trials 1:227-248, 1980 40. Klimt CR, Knatterud GL, Meinert CL et al.: The University Group Diabetes Program (UGDP): A study of the effects of hypoglycemic agents on vascular complications in patients with adult-onset diabetes. II. Mortality results. Diabetes 19 (Suppl):789-830, 1970 41. Bacle G, Bogaerts H, Clement DL, Pannier R, Barbier F: Platelet activation during treadmill exercise in patients with chronic peripheral arterial disease. Thromb Haemostas 46:141, (Abstract) 1981 42. Kazmier F], Fuster V, Chesebro JH, O'FaUon WM, Palumbo PJ: Serial platelet survival half-life in peripheral vascular disease and diabetes meUitus. Thromb Haemostas 46:288, (Abstract) 1981 43. Lowe GDO, Reavey MM, Johnston RV, Forbes CD, Prentice CRM: Increased platelet aggregates in vascular and non-vascular illness: Correlation with plasma fibrinogen and effect of ancrod. Thromb Res 14:377-386, 1979 44. Morris-Jones W, Preston FE, Greaney M, Chatterjee DK: Gangrene of the toes with palpable peripheral pulses. Response to platelet suppressive therapy. Ann Surg 193:462--466, 1981 45. Bishop YMM, Fienberg SE, Holland PW: Discrete Mutivariate Analysis: Theory and Practice. Cambridge: MIT, Press 1975

APPENDIX Veterans Administration Cooperative Study Group Office of the Chairman John A. Colwell, M.D., Ph.D., Chairman William Facteau, Administration Assistant Charleston V.A. Medical Center, Charleston, SC

182

J.A. ColweU et al.

Participating V.A. Medical Centers Augusta, GA: C. L. Lutcher, M.D., Lisa Pitts Boston, MA (Jamaica Plains): Clark Sawin, M.D., Cynthia Silbert, M.D., Donald Francis Boston, MA (Outpatient Clinic): Stephen Podolsky, M.D., Michael Blaha Cincinnati, OH: Paul Chandler, M.D. (1976-77), Darryl J. Sutorius, M.D. (1977-82), William H. Pearce, M.D. (1982-83), Pat Struharik Hines, IL" Carlos Abraira, M.D., LaVerne Hodges Houston, TX: John P. Comstock, M.D., Beverly Poole Lexington, KY: James W. Anderson, M.D. Long Beach, CA: Leona Miller, M.D. (1976-77), Phillip M. Evanski, M.D. (1977-78), Robert Bielen, M.D. (1978-83), Xenia Kozbur Los Angeles (Wadsworth), CA" Seymour Levin, M.D., Evon Fanous Memphis, TN: Solomon S. Solomon M.D., Harriett Ricks Minneapolis, MN" Frank Q. Nuttall, M.D., Joanne Tallman

V.A. Cooperative Study Program Biostatistics and Research Data Processing Stephen F. Bingham, Ph.D. (Study Biostatistidan) Robert Kuhn, Ph.D. (Statistidan, Chief) Hong Jen-Yu (Statistician/Programmer) Roderic D. Gillis (Computer Systems Analyst) Dorothy Morson (Study Encoder)

Clinical Research Pharmacy Coordinating Center Clair M. Haakenson, R.Ph., M.S. (Clinical Research Pharmacist) Roy W. Fetter, Rudolpho Jaurequi (Research Pharmacy Technicians) Sandra Buchanan (Administrative Assistant)

Human Rights Committee Colleen Crigler (Chairman), Perry Point, MD Samuel L. Caesar, Baltimore, MD Susan K. Gauvey, Baltimore, MD Rev. Maurice Moore, Ashton, MD William Beard, Baltimore, MD Ronald S. Lipman, Ph.D., Baltimore, MD Lino Covi, Baltimore, MD

Central Administration, Cooperative Studies Program James A. Hagans, M.D., Ph.D. (Chief) Ping C. Huang, Ph.D. (Staff Assistant) Priscilla Craig (Technical Assistant for Travel and Budget) Freda Cherry (Administrative Aide for Budget Control) P. Dennis GiUiland (Technical Assistant for Planning and Evaluation) C. James Klett, Ph.D. (Chief, Cooperative Studies Program Coordinating Center, Perry Point, MD)

Amputation in Diabetes: Antiplatelet Agents

183

Bertha D. Carter (Administrative Officer, Cooperative Studies Program Coordinating Center, Perry Point, MD Mike R. Sather, R.Ph., M.S. (Chief, Cooperative Studies Program Clinical Research Pharmacy Coordinating Center, Albuquerque, NM)

Study Committees Operations Committee H. James Day, M.D. (Chairman), Philadelphia, PA Curtis Meinert, Ph.D. (Biostatistician). Baltimore, MD Robert W. Barnes, M.D., Richmond, VA John K. Davidson, M.D., Atlanta, GA

Executive Committee John A. ColweIl, M.D., Ph.D. (Chairman), Charleston, SC Carlos Abraira, M.D., Hines, IL James W. Anderson, M.D., Lexington, KY Stephen F. Bingham, Ph.D. (Biostatistician), Perry Point, MD Hau C. Kwaan, M.D., Chicago, IL

Enrollment Committee Stephen Podolsky, M.D. (Chairman), Boston, MA John A. Colwell, M.D., Ph.D., Charleston, SC John P. Comstock, M.D., Houston, TX

Major End Points Committee Carlos Abraira, M.D. (Chairman), Chicago, IL James Anderson, M.D., Lexington, KY John A. Colwell, M.D., Ph.D., Charleston, SC James Anderson, M.D., Lexington, KY John P. Comstock, M.D., Houston, TX Frank Q. Nuttall, M.D., Ph.D., Minneapolis, MN

Laboratories and Consultants Biochemical Reference Maria Lopes-Virella, M.D. John A. ColweU, M.D., Ph.D., Larry Long, Charlene Daniel, Charleston, SC

Egg Consultants Peter Gazes, M.D., Charleston, SC Arthur Fromm, M.D., Minneapolis, MN

Hematology Reference Hau C. Kwaan, M.D., David Eggena, Chicago, IL