Outcome of infrainguinal arterial reconstruction in women

Outcome of infrainguinal arterial reconstruction in women E. J o h n H a r r i s , Jr., M D , L l o y d M. Taylor, Jr., M D , G r e g o r y L. M o n e t a , M D , and J o h n M. P o r t e r , M D , Stanford, Calif., and Portland, Ore.

Purpose: The outcome of leg bypass in women is unknown. To date, most series of lower extremity bypass have included few women, and the results in women compared with those in men have not been reported. The experience with infrainguinal arterial reconstruction in women treated at the Oregon Health Sciences University has been greater than previously reported. We have reviewed our recent experience with lower extremity bypass to determine whether results in women differ significantly from those in men. Methods: In the past 11 years we have performed 823 infrainguinal arterial reconstructions for lower extremity ischemia in 585 patients, of which 357 procedures were performed in w o m e n and 4 6 6 procedures were performed in men. The mean ages were 65 years for m e n and 68 years for w o m e n . Diabetes was present in 59% o f the m e n and 48% o f the w o m e n . Among men, 84% had a history of tobacco use compared with 66% of the women.

Bypasses in men were performed for limb salvage in 73%, claudication in 22%, and a failing bypass graft in 5% of cases. Bypasses in women were performed for limb salvage in 79%, claudication in 15%, and a failing graft in 6% of cases. Previous revascularizations had been performed in 63% of the men and 71% of the women. Autogenous vein was used in 97% of the bypasses in men and 96% of bypasses in women. Graft distribution among infrainguinal arterial target sites was similar between the groups. Bypasses in men were femoral to above-knee popliteal in 11%, femoral to below-knee popliteal in 40%, femoral to tibial in 38%, popliteal to tibial in 10%, and tibial to tibial in 1% of cases. Bypasses in women were femoral to above-knee popliteal in 14%, femoral to below-knee popliteal in 43%, femoral to tibial in 33%, popliteal to tibial in 9%, and tibial to tibial in 1% of cases. Results: Perioperative 30-day mortality rates were 3.7% overaU, 4.3% in men, and 2.9% in women. Long-term survival at 1, 3, and 5 years in men was 80%, 59%, and 44%, respectively. Long-term survival at 1, 3, and 5 years in women was 83%, 69%, and 44%, respectively. Life-table primary patency rates at 1, 3, and 5 years were 86%, 77%, and 71% for men and 87%, 74%, and 67% for women. Limb-salvage results at i and 5 years were 93% and 91% for men and 96% and 96% for women. Conclusions: These results indicate that long-term graft patency and limb salvage results in women are identical to those obtained in men in this experience with autogenous vein. Infrainguinal arterial reconstruction can be performed in women with mortality rates similar to those of men. In this series long-term survival wassimilar for both men and women. (J VASC SURG 1993;18:627-36.)

W o m e n in western societies live longer than men. This improved survival for w o m e n is not completely explained by demographic, behavioral, or biochemFrom the Divisions of Vascular Surgery, Departments of Surgery, Stanford University School of Medicine, Stanford, and Oregon Health Sciences University, Portland. Presented at the Eighth Annual Meeting of the Western Vascular Society, Sonoma, Calif., Jan. 11, 1993. Reprint requests: E. lohn Harris, Jr., MD, Department of Surgery, MSOB X300, Stanford University Medical Center, Stanford, CA 94305-5408. Copyright © 1993 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery, North American Chapter. 0741-5214/93/$1.00 + .10 24/6/48556

ical differences between the sexes. 1 Despite a lower prevalence and later age at onset, ischemic heart disease is still the leading cause o f death in w o m e n in the United States. 2 Several recent reports have suggested that w o m e n with ischemic heart disease suffer higher morbidity and mortality rates f r o m coronary revascularization, achieve inferior results, and may be denied access to coronary interventional procedures compared with men. 3-s Proper stratification for age, severity o f angina, and delay period from onset o f symptoms to intervention negates m a n y o f these differences, yet m o s t still perceive coronary revascularization in w o m e n as a higher risk proce627

628 Harris et al.

dure with inferior results compared with men. We and others have reported successful performance of infrainguinal arterial reconstruction for severe limb ischemia.911 Although these series have included women, no specific inquiry into the outcome of infrainguinal arterial reconstruction with respect to gender was performed. Some series have reported trends toward inferior results in women compared with men, yet the women were such a small percentage of the study groups and the total enrollment in these series was so small that statistical significance was not reached) 2,1s A recent study retrospectively reviewed the outcome of infrainguinal arterial reconstruction in women compared with a larger contemporaneous cohort of men. Although women constituted 39% of this series, the study follows only 150 bypass grafts in women through 3 years. This report identified comparable operative mortality and limb salvage rates for men and women undergoing infrainguinal arterial reconstruction, but long-term survival and graft patency rates were significantly reduced in women.14 This outcome seemed contrary to our experience with infrainguinal arterial reconstruction in women. In 1990 we reported results of an 8-year series of consecutive infrainguinal arterial reconstructions with reversed vein bypass grafts. 9 Since this report we have maintained prospective follow-up of this patient group, as well as all subsequent patients treated for lower extremity ischemia. This series now covers an 11-year period. These data were reviewed specifically to evaluate significant differences in outcome between genders. METHODS All patients with limb-threatening ischemia or disabling claudication were considered for infrainguinal arterial reconstruction. Patients with chronic neuromuscular impairment and nonambulatory status before the onset of ischemia were not considered for limb-salvage procedures. Failure of previous bypass surgery and presumed lack of adequate autogenous vein were not contraindications to infrainguinal arterial reconstruction. All patients underwent angiography with specific attention to visualization of the distal extremity arteries according to techniques described previously by our radiology group, including selective catheterization, delayed filming, and use of intraarterial vasodilators. 15 Absence of a suitable distal target vessel was defined by the presence of small, unnamed contrast-filled vessels in the foot and distal leg,

JOURNALOF VASCULARSURGERY October 1993

without contrast filling of the named tibial or pedal Vessels.

The techniques used to perform infrainguinal arterial reconstruction in this patient group have been described previously. 16-1s The principal features include a multiteam approach simultaneously to facilitate vein harvests and arterial dissections; careful vein harvest with minimal dissection of the adventifia and gentle distention of the vein with chilled, heparinized autologous blood containing lidocaine and papaverine; anatomic tunneling; distal arterial origins; use of arm vein and vein segment splicing to achieve the maximum number of autogenous reconstructions; and objective evaluation of graft technical adequacy with continuous-wave Doppler insonation, electromagnetic flow measurements, and, when necessary, intraoperative angiography. After infrainguinal arterial reconstruction, all patients were followed up in the vascular clinic and the noninvasive vascular laboratory every 3 months for the first year and every 6 months thereafter until death, loss to follow-up, or the end of the study. Graft patency was monitored by physical exanainarion with assessment of palpable pulses in the graft and distally and maintenance of the initial postoperative improvement in ankle/brachial pressure index, which was greater than 0.30 in all cases. In addition, grafts were surveyed with the duplex scanner at all visits to the noninvasive vascular laboratory, as described previously.19 Graft flow velocities consistently below 45 cm/sec were examined by arteriography. All repairs performed based on these studies were considered end points for primary patency. All patient charts were reviewed for demographic information including age, indication for surgery, smoking history, previous bypass surgery, type of conduit, level of distal anastomosis, and diabetic status. For the purpose of this review, limb-salvage indications for infrainguinal arterial reconstruction were defined as the presence of ischemic rest pain or gangrene together with an ankle/brachial index less than 0.40. Limb salvage was defined as retention of the extremity without need for amputation proximal to the tarsal-metatarsal junction. Primary patency was defined as uninterrupted patency of the original graft without the need for any graft-directed intervention. Patency of the original graft maintained by thrombectomy, angioplasty, patch graft, or lyric therapy was termed secondary patency. Immediate postoperative occlusions that were salvaged were classified as secondarily patent. Occluded grafts that were not salvaged but replaced by a new bypass graft were regarded as occluded. The repeat bypass graft entered

JOURNAL OF VASCULAR SURGERY Volume 18, Number 4

Harris et al.

629

Table I. Comparison of demographic factors

Mean age (yr) Mean follow-up (mo) Incidence of diabetes mellitus (%) Incidence of reported tobacco use (%) 30-Day operative mortality (%) Previous lower limb bypass (%)

Men

Women

p Value

65.4 17 59 84 4.3 63

68.1 20 48 66 2.9 71

<0.05 < 0.05 < 0.005 < 0.005 NS < 0.05

NS, Not significant.

the series as a new graft. These definitions and conventions of reporting are in accordance with the recommendations of the Ad Hoc Committee on Reporting Standards of the Joint Council of the Vascular Societies. 2° Patency, limb salvage, and patient survival data were calculated according to the life-table method. Life tables were compared by the use of the MantelHaenszel log-rank test. 21 Chi-square analysis was used to compare the distribution of smoking and diabetes mellitus, the incidence of previous lower limb revascularization, and 30-day operative mortality rates between men and women. The two-tailed Student t test was used to compare the mean follow-up periods and mean ages between men and women. Significance was designated as p < 0.05. RESULTS From 1980 through 1991, 823 infrainguinal arterial reconstructions were performed in 585 patients. Four hundred sixty-six reconstructions (57%) were performed in 346 men and 357 reconstructions (43%) were performed in 239 women. The mean follow-up period for men was 17 months, with a range from 0 to 106 months. The mean follow-up period for women was 20 months, with a range from 0 to 138 months. Women were followed up for a significantly longer period than were men

(p < 0.05). The mean ages of the two groups were 65 (range 13 to 94) years for men and 68 (range 32 to 94) years for women. In both men and women, 95% of the procedures were performed in patients older than 40 years of age. When evaluated by decade, there were no significant differences in primary patency rates between decades for either gender. The bypass procedures in patients less than 40 years old were performed predominantly for type I diabetes mellims or trauma. Compared with men, women in this series were significantly older (p < 0.05). Three hundred ninety-two bypasses (84%) were performed in male smokers or ex-smokers, and 74 bypasses (16%) were

in male nonsmokers. Two hundred thirty-nine bypasses (66%) were performed in female smokers or ex-smokers, and 118 bypasses (33%) were performed in female nonsmokers. A smoking history was identified significantly more often in men compared with women (p < 0.005). Diabetes mellitus was present in 59% of the men and 48% of the women. Diabetes mellitus was significantly more frequent in men compared with women (p < 0.005). Indications for bypass in men were limb salvage in 73%, disabling claudication in 22%, and failing bypass graft in 5% of cases. Indications for bypass in women were limb salvage in 79%, disabling claudication in 15%, and failing bypass graft in 6% of cases. In men, 451 infrainguinal reconstructions (97%) were performed with autogenous vein and 15 (3%) were performed with prosthetic materials. O f the autogenous grafts, 370 (82%) were reversed saphenous vein, 29 (6%) were reversed arm vein, and 52 (12%) were combinations of lesser saphenous and arm vein segments. In women, 341 infrainguinal reconstructions (96%) were performed with autogenous vein and 16 (4%) were performed with prosthetic materials. Of the autogenous reconstructions in women, 279 (82%) were performed with reversed saphenous vein, 27 (8%) with reversed arm vein, and 35 (10%) with combined lesser saphenous and arm vein segments. Previous lower limb revascularization had been performed in 63% of men and 71% of women (p < 0.05). These were predominantly contralateral infrainguinal reconstructions or inflow reconstructions, with reoperative ipsilateral reconstructions identified in 11% of the men and 10% of the women. Graft distribution among infrainguinal arterial target sites was similar between the two groups. In men, 50 bypasses (11%) were femoral to above-knee (AK) popliteal, 187 bypasses (40%) were femoral to below-knee (BK) popliteal, 178 bypasses (38%) were femoral to tibial, 45 bypasses (10%) were popliteal to tibial, and 6 bypasses (1%) were tibial to tibial. In women, 52 bypasses (14%) were femoral to AK

630

JOURNALOF VASCULARSURGERY October 1993

Harris et al.

100

100

"'"1"3"7"'"'T 87

.q

~

8o

,2 ~

162

7o

'~..$..7....T ''''"'''"

.h

eo

b~

40 M E N (n = 346) ....... W O M E N (n = 239)

0

T

T

~

m

lz

p

--

,1

T

r...:iZ....: ~ " " "7; ......... "~ 7 " 1

T 4o

N O N - D I A B E T I C M E N ( n = 189) D I A B E T I C M E N ( n -- 2 7 7 )

2o

ns

17 T

r'7i ..... ".

60

r'"ml

2o

T ,,

..°.~...°..°Oo

80

p = 0.025 I

I

I

I

i

i

0

1

2

3

4

5

0

I

I

I

I

I

1

2

3

4

5

YEARS

YEARS

Fig. 1. Long-term patient survival.

Fig. 3. Primary graft patency of male patients with diabetes versus those without diabetes.

I00 ! ... A8.~....T... F.?...."a:...7.1......T.....F.....T..... 1......"a-

100

200 .L124 .1. 6s .L 80

i " T" 71 ~

T " T

37

60

~s _L

1 .L

80

r;=l

b, 40 ......

20

0

4o

M E N (n = 466) W O M E N ( n = 357)

p = ns

......

20

I

[

I

I

I

I

0

1

2

3

4

5

YEARS

0

I

0

I

I

1

MEN (n = 466) W O M E N (n = 357)

I

I

I

2

I

3

p = ns

r

I

4

I

5

YEARS

Fig. 2. Primary graft patency of men versus women.

Fig. 4. Limb salvage in men versus women.

popliteal, 152 bypasses (43%) were femoral to BK popliteal, 117 bypasses (33%) were femoral to tibial, 33 bypasses (9%) were popliteal to tibial, and 3 bypasses (1%) were tibial to tibial. Perioperative 30-day mortality rates were 3.7% for all patients, 2.9% for women, and 4.3% for men (difference not significant) (Table I). Long-term survival in men was 80% at 1 year, 59% at 3 years, and 44% at 5 years. Long-term survival in women was 83% at I year, 69% at 3 years, and 44% at 5 years (Fig. 1). There was no significant difference in survival life-tables between men and women by the log-rank test. Primary graft patency rates for men at 1, 3, and 5 years were 86%, 77%, and 71%, respectively. Primary graft patency rates for women at 1, 3, and 5 years were 87%, 74%, and 67%, respectively. There was no significant difference in primary patency life tables between men and women by the log-rank test

(Fig. 2). Life-table primary graft patency rates for various subsets of men and women were also compared. Primary graft patency rates for male smokers at 1, 3, and 5 years were 86%, 77%, and 70%, respectively. Primary graft patency rates for female smokers at 1, 3, and 5 years were 88%, 74%, and 66%, respectively. There was no significant difference in primary patency rates between male and female smokers. There was no significant difference in life-table primary patency rates between male and female nonsmokers. Primary graft patency rates for men and women with diabetes mellitus were also compared, and there was no significant difference. Life-table primary patency rates for men with diabetes mellitus at 1, 3, and 5 years were 83%, 72%, and 63%, and for women with diabetes mellitus they were 90%, 78%, and 70%, respectively. There was no significant difference between life-table primary patency rates between women with and without dia-

JOURNAL OF VASCULAR SURGERY Volume 18, Number 4

Harris et al.

631

Table II. Comparison of primary patency rates and survival life tables Group Total grafts Men Women Smokers Men Women Diabetes meUitus Men Women Women Non-DM DM Men Non-DM DM BK-POP Men Women FEM-TIB Men Women Limb salvage Men Women Survival Men Women

p Value

1 Year (%)

3 Years (%)

5 Years (%)

NS

86 87

77 74

71 67

NS

86 88

77 74

70 66

NS

83 90

72 78

63 71

NS

85 90

71 78

66 71

< 0.025

90 83

85 72

85 63

NS

90 92

79 83

71 73

NS

83 77

77 64

70 59

NS

93 96

91 96

91 96

NS

80 83

59 69

44 44

NS, Not significant; DM, diabetes mellims; BK-POP, below-knee femoropopliteal bypass; FEM-T/B, femorotibial bypass.

betes mellitus. Women without diabetes mellims showed primary patency rates of 85%, 71%, and 66% at 1, 3, and 5 years. There was a significant difference (p < 0.025) in life-table primary patency rates between men with and without diabetes mellims. Men without diabetes mellitus showed primary patency rates of 90%, 85%, and 85% at 1, 3, and 5 years (Fig. 3). There were no significant differences between life-table primary patency rates for smokers compared with nonsmokers in either sex. Similarly, patients undergoing multiple revascularization procedures showed no significant difference in life-table primary patency rates compared with patients undergoing initial revascularization, either among or between genders. There were no significant differences between men and women when life-table primary patency rates of each level of distal anastomoses were compared by the log-rank test. For the AK femoropopliteal bypass, primary patency rates at 1, 3, and 5 years were 92%, 92%, and 92% in men and 88%, 67%, and 67% in women. Above-knee femoropopliteal grafts were at risk in 50 men and 52 women. For the BK femoropopliteal bypass, primary patency rates at 1, 3, and 5 years were 92%, 83%, and 73% for men and 90%, 78%, and 71% for women.

Below-knee femoropopliteal grafts were at risk in 152 men and 187 women. For the femorotibial bypass, primary patency rates at 1, 3, and 5 years were 83%, 77%, and 70% for men and 77%, 64%, and 59% for women. Femorotibial grafts were at risk in 178 men and 117 women. Limb-salvage analysis by life-table and log-rank tests showed no significant differences by gender. For men, limb salvage at 1, 3, and 5 years was 93%, 91%, and 91%. For women, limb salvage at 1, 3, and 5 years was 96%, 96%, and 96% (Fig. 4). The comparative life-table data from this series are summarized in Table II. DISCUSSION

In this series of continuous, unselected patients, women underwent infraingninal arterial reconstruction with results similar to those of men. These results confirmed our anecdotal observations that women were not at higher risk for infrainguinal arterial reconstruction than men and their results were not dissimilar to those of men. This report differs from the recent report of Magnant et al.,l~ which concluded that gender substantially influenced infrainguinal graft patency and, from the trends in earlier series of infrainguinal revascularizations, suggested

632

Harris et aL

results for women to be inferior compared with those of men. 12'is The series ofMagnant et al. is remarkable for the diminished primary graft patency rates of both men and women. Primary graft patency rates were calculated by life-table methods and at 1 and 3 years were 73% and 70% for men and 59% and 54% for women. Prosthetic bypass grafts did not influence these results because autogenous grafts were used in more than 90% of the series. Three-year limb salvage rates were not significantly different between men and women, suggesting improved secondary graft patency rates for women. Interestingly, secondary graft patency rates of the women in their series at 1 and 3 years were also significantly inferior to those of men. In the Framingham study, tobacco abuse in men and the presence of diabetes mellitus in women were the most influential risk factors for the development of any of the cardiovascular end points examined, including congestive heart failure, intermittent clandication, and coronary artery disease. 22 The finding of increased susceptibility to cardiovascular disease in female diabetics is not unique to the Framingham study. 23 Similar findings were reported in the Joslin Clinic experience24 and the Rancho Bernardo study. 2s The Rancho Bernardo study group suggested the higher risks associated with diabetes in women were largely a function of the superior survival of nondiabetic women compared with men. In the report of Magnant et al.,14 diabetes increased the mortality rate of women far more than that of men and decreased long-term survival compared with nondiabetic women. An interesting finding in their series was the 72% 3-year survival rate of men and the 79% 3-year survival rate of nondiabetic women, clearly better than our observed survival statistics. The 54% 3-year survival rate of diabetic women in their series was more in keeping with our survival data. Several investigations have suggested an insulinandrogen interaction as the basis of diabetes influencing cardiovascular morbidity and mortality rates in women. Women with non-insulin-dependent diabetes often show a male or central pattern of obesity that is associated with insulin resistance and hyperinsulinemia. Nondiabetic women with this central pattern of obesity often show androgen excess and some degree of insulin resistance and hyperinsulinemiaY 27 Central obesity in women is also associated with less favorable high-density lipoprotein levels and a greater risk of coronary artery disease. 28 In the Cleveland Clinic review of 2445

JOURNAL OF VASCULAR SURGERY October 1993

consecutive women undergoing coronary artery bypass grafting, body surface area was the most important predictor of operative death. When matched for age, severity of angina, and extent of coronary atherosclerosis, women still had twice the operative deaths of men. When these patients were stratified for body surface area, gender was not an important predictor of operative risk. 6 In neither the series ofMagnant et al.14 nor this series were patients with diabetes mellitus categorized as having type I or having type II diabetes. Patients with type I diabetes would be expected to have smaller body surface areas, whereas patients with type II diabetes would more likely exhibit larger body surface areas. Perhaps stratification by body surface area would provide more insight into the gender difference in the results of bypass identified in the series of Magnant et al. In this series of consecutive patients, gender did not influence either the mortality rate or success of the bypass procedure. Furthermore, diabetic women had results similar to those of nondiabetic women. These results were obtained in a population in which 71% of the women had undergone previous attempts at infrainguinal revascularization. Women in this series were significantly older and followed up for a significantly longer period than were men. It is obvious that we did not exclude women from consideration for revascularization because of a perception of inferior results and this series included a population of putatively high-risk women. The observation that most previous studies suggesting inferior results in women had a female population less than 20% of the series, whereas women represent 43% of our series, suggests, as in the coronary artery revascularization series, that selection bias may influence results. The series of Magnant et aL 14 included 39% women, yet tiffs current series followed up more than two times as many patients, men and women, as their series and it included 2 more years of follow-up. Many have suggested that women, by virtue of decreased arterial diameter, should be expected to have diminished bypass graft patency compared with men, although smaller vessel size has been documented only in the aorta and iliac and coronary a r t e r i e s . 29-31 This report refutes this notion because the life-tahle primary patency rates for BK femoropopliteal bypass and femorotibial bypass in women were not significantly different from those in men. These findings were confirmed in another report by Shah et alY evaluating the outcome of tibial artery bypass in diabetic patients. In this series, 33% of the patients were women and the primary 5-year graft

JOURNAL OF VASCULAR SURGERY Volume 18, Number 4

p a t e n c y rate for w o m e n was significantly b e t t e r t h a n for m e n (p = 0.05). A l s o in c o n t r a s t t o t h e F r a m i n g h a m study, cigarette s m o k i n g d i d n o t adversely affect p r i m a r y p a t e n c y rates o f infrainguinal r e c o n s t r u c t i o n s in either m e n o r w o m e n in this series. P e r h a p s a larger p o p u l a t i o n o f n o n s m o k e r s w o u l d have l e d to differe n t results. This i n f o r m a t i o n clearly does n o t alter o u r s t r o n g c o m m i t m e n t t o educate o u r patients a b o u t s m o k i n g cessation. I n conclusion, this s t u d y reinforces o u r percept i o n t h a t patients p r e v i o u s l y c o n s i d e r e d h i g h risk for arterial r e c o n s t r u c t i o n can u n d e r g o bypass s u r g e r y w i t h l o w o e p r a t i v e m o r t a l i t y rates a n d acceptable graft p a t e n c y a n d limb-salvage results. I n o u r experience g e n d e r differences d o n o t exist a n d a p p e a r irrelevant in c o n s i d e r a t i o n o f patients for infraing u i n a l arterial r e c o n s t r u c t i o n . O n c e again o u r results c o n f i r m t h a t l o n g - t e r m survival in a p o p u l a t i o n o f patients r e q u i r i n g i n f r a i n g u i n a l arterial reconstruct i o n is significantly r e d u c e d f r o m t h a t o f a g e - m a t c h e d c o n t r o l subjects w i t h o u t s y m p t o m a t i c l i m b ischemia, c o n f i r m i n g an i m p o r t a n t message o f t h e F r a m i n g h a m s t u d y (i.e., s y m p t o m a t i c atherosclerotic vascular disease is associated w i t h a c o n s i d e r a b l y decreased life expectancy). REFERENCES

1. Wingard DL. The sex differential in morbidity, mortality, and lifestyle. Annu Rev Public Health 1984;5:433-58. 2. Cunningham MA, Lee TH, Cook EF, et al. The effect of gender on the probability of myocardial infarction among emergency department patients with acute chest pain. J Gen Intern Med 1989;4:392-8. 3. Ayanian JZ, Epstein AM. Differences in the use of procedures between women and men hospitalized for coronary heart disease. N Engl J Med 1991;325:221-5. 4. Steingart RM, Packer M, Hamm P, et al. Sex differences in the management of coronary heart disease. N Engl J Med 1991;325:226-30. 5. Maynard C, Litwin PE, Martin JS, et al. Gender differences in the treatment and outcome of acute myocardial infarction. Arch Intern Med 1992;152:972-6. 6. Loop FD, Golding LR, MacMillan JP, et al. Coronary artery surgery in women compared with men: analyses of risks and long term results. J Am Coil Cardiol 1983;1:383-90. 7. Khan SS, Nessim S, Gray R, et al. Increased mortality of women in coronary artery bypass surgery: evidence for referral bias. Ann Intern Med 1990;112:561-7. 8. Kahn JK, Rutherford BD, McConahay DR, et al. Comparison of procedural results and risks of coronary angioplasty in men and women for conditions other than acute myocardial infarction. Am J Cardiol 1992;69:1241-2. 9. Taylor LM Jr, Edwards JM, Porter ~M. Present status of reversed vein bypass grafting: five-year results of a modern series. J VAsc SURG 1990;11:193-206.

Harris et al.

633

10. Kent KC, Whittemore AD, Mannick JA. Short-term and mid-term results of an all autogenous tissue policy for infrainguinal reconstruction. J V3sc SURG 1989;9:107-14. 11. Leather RP, Shah DM, Chang BB, et al. Resurrection of the in-situ saphenous vein bypass. Ann Surg 1988;208:435-42. 12. Ramsburgh SR, Lindenauer SM, Weber TR, et al. Femoropopliteal bypass for limb salvage. Surgery 1977;81:453-8. 13. King BR, Myers KA, Scott DF, et al. Femoropopliteal vein grafts for intermittent claudication. Br J Surg 1980;67:48992. 14. Magnant JG, Cronenwert JL, Walsh DB, et al. Surgical treatment of infrainguinal arterial occlusive disease in women. J VASe SURG (in press). 15. Kozak BE, Bedell JE, Rosch J. Small vessel leg angiography for distal vessel bypass grafts. J VASC SURG 1988;8:711-5. 16. Taylor LM Jr, Edwards JM, Brant B, et al. Autogenous reversed vein bypass for lower extremity ischemia in patients with absent or inadequate greater saphenous vein. Am ~ Surg 1987;153:505-10. 17. Taylor LM Jr, Phinney ES, Porter JM. Present status of reversed vein bypass for lower extremity ischemia. J VAsc SURG 1986;3:288-97. 18. Taylor LM Jr, Porter JM. Technique of reversed vein bypass to distal leg arteries. In: Bergan JJ, Yao JST, eds. Techniques in arterial surgery. Orlando: WB Saunders, 1989:109-22. 19. Mills JL, Harris EJ Jr, Taylor LM Jr, et al. The importance of routine surveillance of distal bypass grafts with duplex scanning: a study of 379 reversed vein grafts. J VAsc SURG 1990; 12:379-89. 20. Ad Hoc Committee on Reporting Standards, Society for Vascular Surgery/North American Chapter, International Society for Cardiovascular Surgery. Standards for reports dealing with lower extremity ischemia. J Vasc SURG 1986; 4:80-94. 21. Matthews DE, Farewell VT. The log-rank or MantelHaenszel test for the comparison of survival curves. In: Matthews DE, Farewell VT, eds. Using and understanding medical statistics. New York: Karger, 1985:79-86. 22. KannelWB, McGeeDL. Diabetes and cardiovascular disease: the Framingham Study. JAMA 1979;241:2035-8. 23. Kessler II. Mortality experiences in diabetic patients. Am Med 1971;51:715-24. 24. Barrett-Connor EL, Cohn BA, Wingard DL, et al. Why is diabetes mellitus a stronger risk factor for fatal ischemic heart disease in women than in men? JAMA 1991;265:627-31. 25. Krotkiewski M, Bjorntorp P, Sjostrom L, et al. Impact of obesity on metabolism in men and women: importance of regional adipose tissue distribution. J Clin Invest 1983;72: 1150-62. 26. Evans DJ, Hoffman RG, Kalkhoff RK, et aL Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolic aberrations in premenopause women. J Clin Endocrinol Metab 1983;57:304-10. 27. Seidell JC, Cigolini M, Deurenberg P, et al. Fat distribution, androgens, and metabolism in nonobese women. Am J Clin Nutr 1989;50:269-73. 28. Malley M. The ratio of waist-to-hip circumference, plasma insulin level~and glucose intolerance as independent predictors of the HDL 2 cholesterol levels in older adults. N Engl J Med 1990;322:229-34. 29. Cronenwett JL, Davis JT Jr, Gooch JB, et al. Aortoiliac occlusive disease in women. Surgery 1980;88:775-83.

634

Harris et al.

30. Tyras DH, Barner FIB, Kaiser GC, et al. Myocardial revascularization in women. Ann Thorac Surg 1978;25:449-53. 31. Kennedy JW, Kaiser GC, Fisher LD, et al. Multivariate discriminant analysis of the cfinical and angiographic predictors of operative mortality from the Collaborative Study in Coronary Artery Surgery (CASS). J Thorac Cardiovasc Surg 1980;80:876-87.

JOURNAL OF VASCULARSURGERY October 1993

32. Shah DM, Chang BB, Fitzgerald KM, et al. Durability of the tibial artery bypass in diabetic patients. Am J Surg 1988;156: 133-5.

Submitted Feb. 23, 1993; accepted May 10, 1993.

DISCUSSION Dr. Linda M. Reilly (San Francisco, Calif.). You have reported a large series ofinfrainguinal revascularizations in which yon detected no difference between men and women in terms of mortality, primary graft patency, or limb salvage rates. Although I am convinced that there are certain political groups that would cheer this demonstration o f gender equality, I am afraid that I have some reservations. It also occurred to me that I might have been asked to discuss this study because I am one o f the few members o f this society for whom it might be safe to suggest that women are not equal to men, at least with regard to this particular issue. I am puzzled by the lack of an influence of smoking on graft patency rates among your patients. In your patient group, smoking had absolutely no effect, either within genders or across genders. This contradicts a lot o f published data about the adverse effects of smoking on durability of revascularizations in the carotid location, the infrainguinal location, and aortoiliac reconstructions. I note that only 16% of your male patients and about 33% of your female patients were nonsmokers; I wonder if the substantially greater volume o f smokers in this study actually results in a sample size of nonsmokers that is too small to show an effect. You did not provide me with the life tables comparing nonsmokers with the smoking group, and therefore I cannot assess the impact of sample size. Perhaps you can comment on this lack of an effect of smoking, and if you have the additional raw fife-table data could you present it. Second, published data have reported substantially lower graft patency rates, as well as lower rates o f limb salvage, among diabetics undergoing infrainguinal revascularization. However, in your study diabetes had no impact on graft patency among women. Diabetes did adversely affect patency rates among men, making it even more confusing that it did not affect women, because, as you point out, it is well known that the impact of diabetes on the development and progression of cardiovascular disease is substantially greater in women than in men. I wonder if you could comment on the reason for this equally perplexing lack of an effect. Your male patients had a significantly greater prevalence o f diabetes and a higher prevalence of smoking. It seems to me that the women should have had a better outcome than the men; I wonder if in this particular case equivalency is actually doing worse.

I suppose my problem is that there are too many contradictions between the results o f your study and those of other published data. This makes me wonder if there is some inadvertent bias in your patient group, which brings me to my final point. In your methodology you state that patients who underwent a second bypass after failure of the first bypass were reentered into the study group as an old patient with a new bypass. I believe I have correctly stated your method. However, because you did not provide a breakdown of how many patients had undergone two, three, or more bypasses, I cannot tell exactly how many patients were reentered in this maimer. At most it could amount to 40% of your patient group. This poses a problem because of the large number of patients involved. Every time you reenter a new bypass on an old patient you are essentially double-counting all the demographic and risk factors that are pertinent to that patient. What this means is that patients who do well never get a chance to be counted more than once, but patients who do poorly will be counted every time they reenter the study. I wonder if this has inadvertently weighted the data toward those factors that are present in patients whose grafts fail. This could be avoided by reporting patency results by patient rather than by operation. I believe the most definitive way to present your data would be to take those patients whose initial operation was done by your group and present the follow-up on those patients only until they reached one of your end points or the end of the follow-up interval. Have you analyzed your data in that way and, if so, could you provide us with those results? Having said all that, and in spite of the fact that you are from Stanford, I actually befieve that you are probably correct. When properly analyzed and treated, women may well have the same results as men. However, I do not believe that you can conclusively state that the data in your study resolve the issue. Dr. E. John Harris (Stanford, Calif.). First, regarding the issue of smoking and the perplexing ability o f smoking to enhance the patency rate in these patients, we agree that perhaps a larger series o f nonsmokers might shed some more light on this issue. The life-tables that were compared did have a significant amount of patients such that at the 5-year end point an SE o f less than 10% was realized, which I believe is the accepted standard for reporting. As far as the role o f diabetes in women, I do not think

JOURNAL OF VASCULAR SURGERY Volume 18, Number 4

this issue is clear. If you look carefully at the cardiac data and some of the catheterization studies, it appears that it is really not diabetes per se, but rather body surface area that influences patient outcomes. In fact, if you categorize patients by body surface area, all differences in gender and diabetes factor out. I believe if we look at these data critically, perhaps we should have calculated individual body surface area and compared these values for outcome analysis; that would be more valid than using diabetes alone. Type I or II diabetes was not factored out in this series. It just was the presence or absence of diabetes mellitus. As far as the third point about our reporting standards, yes, I agree. If you saw how many patients we had in this series as opposed to how many bypass procedures were performed, a predominant number of patients did undergo multiple revascularizations. You can argue that by following these patients with postoperative graft surveillance, we are taking out a lot of patent grafts that might go further, weakening our primary patency value, which does indeed happen. When comparing our data, with a significant number of reoperafive cases, with recent series in the literature, I note a similar degree of disparity between patient number and bypasses performed. The series from Dartmouth used multiple procedures in a smaller patient subset, and they showed absolutely opposite results. So again, in spite of the limitations of the reporting standards, we have adhered to the Joint Societies' recommendations for reposing, allowing us to compare our data with that already published with use of similar standards. Dr. Wesley S. Moore (Los Angeles, Calif.). One of several pet peeves I have has to do with the issue of obtaining an accurate smoking history from a patient. I am sure that all of you have shared the experience of asking a patient, "Are you a smoker," to which the patient replies, '~qo." Few people will then ask the next question: 'q/Vhen did you quit?" The answer to that question ranges from yesterday to perhaps 1 or 2 years ago. As a result, misinformation is recorded in the chart and the same misinformation is abstracted at the time of chart review. This was highlighted most recently by a report of a prospective randomized trial on femoropopliteal reconstruction conducted in the United Kingdom. In a recent commentary, Greenhalgh (Wiseman S, Powell J, Greenhalgh R, et al. Eur J Vase Surg 1990;4:57-61) pointed out that they would not accept smoking history alone but required chemical data. They used thiocyanate analysis to determine whether a patient was smoking. When the investigators correlated graft patency data as a function of cigarette smoking history, there appeared to be no difference between the smokers and the nonsmokers. However, when they compared patency data of smokers versus nonsmokers based on validation by thiocyanate analysis, nonsmokers clearly fared better than smokers. As a result of this observation, I suggest that rather than reporting categories of smokers and nonsmokers, we should report patients with peripheral vascular disease as being either smokers or liars. Could you tell us how you obtained the smoking history in this report?

Harris et al.

635

Dr. Harris. I reviewed two places. The vas~xflar laboratory had a separate history sheet that would ask the patient to record a yes or no answer on whether they were smokers or had hypertension and what medicines they took. That was reviewed as one arm of the study; the other was looking into the patient charts, which do suffer all the bias you just described. No, we did not specifically test patients; in fact, I lumped together both patients who were currently smoking and those who said that they had smoked but currently were not smoking, assuming that most of these patients were currently smoking. Dr. N o r m a n H e ~ z e r (Cleveland, Ohio). About I year ago I had the opportunity to prepare a book chapter on peripheral vascular disease for a National Institutes of Health conference concerning cardiovascular health and disease in women. I conducted a pretty extensive review of the literature, as I am sure you have done, and I found that very little has been reported regarding gender differences and surgical outcome, particularly for lower extremity bypass. There is some evidence that recurrent carotid artery stenosis is more common in women, but that is also a matter of debate. This whole topic may be one of those things that we think we know about but really do not. In an attempt to obtain some data for my chapter, I reviewed the Cleveland Vascular Registry for the past 15 years with respect to surgical resuks in women versus men. At that time this registry included a total of 6855 patients who had undergone femoropopliteal or femorotibial revascularization, 34% of whom were women. In fact, women had significantly higher risks for early graft thrombosis (9.8% vs 7.8%), amputation (8.2% vs 6.2%), and operative death (3.9% vs 2.5%). Despite their statistical significance within nearly 7000 observations, however, one still wonders what the clinical relevance of these findings may be. I also found in my literature review that, because surgical results are so rarely reported according to gender, it is impossible to perform metaanalysis for distal revascularization. You see, any single series of femoropopliteal or distal bypasses usually has sufficiently few women in it that their outcomes are not described separately. If there is one thing that we as vascular surgeons can do during the next few years, irrespective of the number of women in each of our series, it would be to report their outcome specifically so that metaanalysis eventually would be possible. I think this is an important issue, and I hope you do tOO.

Dr. George Andros (Burbank, Calif.). Dr. Reilly's comment is particularly important if you believe the statistics dealing with the rate of increase in new smokers under the age of 25 years; the rate for women exceeds the rate for men. In addition, I noted that at least 20% of your patients underwent "other autogenous bypasses." Some of these were probably arm veins and I have been often asked if arm veins in women were as usable as they are in men. I have no hard data on that question, but it is my clinical impression that as sources of vein bypasses arms are politically correct for both genders and are equivalent.

636

Harris et al.

Dr. Eugene Strandness (Seattle, Wash.). One point that is often forgotten, at least in the diabetic patient, is that the gender difference disappears when you have type II diabetes. With this disease, the prevalence is equal among men and women. One of the key questions is why is this true? It turns out that women with type II diabetes have higher low-density lipoprotein cholesterol levels than do men. It is also well known that women have smaller arteries than do men. So I suspect that the diabetic population are an entirely different group, and I do not think they should be mixed together. They all have type II diabetes as far as I am concerned. You could not have had any patients with type I diabetes in this study. Dr. Harris. There were some patients in their teens in the series that may have had type I diabetes. The only evidence I found in the literature for difference in size of artery was with the coronary and iliac arteries and the aorta, but I guess it gets back to the issue of how can you normalize all of these? With type II diabetes, there is a change in the body fat distribution. There is a change in androgen levels, and all of these things are important; we do not know how these changes affect arterial reconstruction. These patients should not be turned away because one perceives that they are going to have an inferior result, on the basis of sex or diabetic status. These patients had a high incidence of reoperation. They came to us for limb salvage, and they had acceptable results. Dr. Eugene F. Bernstein (La Jolla, Calif.). Have you separately compared the outcome of patients who presumably are all male at the Veterans Hospital with your University Hospital male and female population and are there significant differences in those groups? Have you tried to explain the difference between your results and those of the Albany-Dartmouth groups? Dr. Harris. The Albany group says women do much better. They had an 85% primary patency rate at 5 years compared with 60% for men. In the Albany group they showed that it was a small subset of patients, and they did not show more information than that. Just in the body of their discussion they said that women had an 85% primary patency rate at 5 years compared with men. The focus of their study was looking at diabetic versus nondiabefic patients in femorotibial in situ bypass, and they found that diabetes made no difference in their series. They did not subdivide the women into diabetic women versus nondiabefic women. They said that, overall, women had better results at 5 years than men. The Dartmouth group said women had worse results as far as patency rates and survival, but their limb-salvage and operative mortality rates were the same as those for men. That is somewhat

JOURNALOF VASCULARSURGERY October 1993

confusing considering that their primary and secondary patency rates were inferior to those of the men. It is difficult for me to understand how their limbsalvage rate can be similar to that for men when their bypass grafts were doing so poorly. Dr. Bernstein. What were the results with the group at the Veterans Administration Hospital? Dr. Harris. I have not specifically looked at that. I do not know if Dr. Moneta knows about that. This series does not include any patients from the Portland Veterans Affairs Medical Center. Dr. Lloyd M. Taylor, Jr. (Portland, Ore.) When viewing the results of these series of bypass grafts, it is important to understand the nature of the patient population being described. What you have presented are the results of an analysis of a very large number of bypass grafts, more than 800, in a very large number of patients, more than 500, almost all of whom were operated on for severe limb-salvage indications and, in fact, almost 65% of whom were reoperative cases. This is a very skewed distribution of severity of disease in the patients who were seen at the University Hospital. One of the characteristics of the patients undergoing limbsalvage procedures that is very different from the population treated primarily for arterial disease is the numbers of men and women treated are about the same, whereas in patients treated for claudication or coronary artery disease, a much yotmger group of less severely affected patients, the numbers of women have traditionally been much lower. This is the only series presented to date in which the numbers of men and women are about the same. The conclusions drawn from this study are based on a much more equal proportion of men and women treated by the same method as the Dartmouth group, which had fewer than 30% women. Some of the other studies that have been reported have numbers as low as 15% to 20% women. I think our population at the Veterans Administration Hospital is much more disease orientcd. The numbers of patients who are undergoing repeat operations are in the 20% to 25% range instead of 65% to 70%. They are really not comparable. This study is just another description of a group of patients that can be placed in context only when you recognize the patient group that you are talking about. Most of the superior bypass graft results that are attributed to diabetics and diabetic women are the result of an artifact in the life table that have to do with editing out patients with patent grafts who die at a much faster rate if they have diabetes and if they are women or if they are older and have diabetes. So it is not improved graft patency. It is a much worse survival rate that makes the patency rate look better.