Author + information
- Received November 20, 2000
- Revision received April 18, 2002
- Accepted October 25, 2002
- Published online May 7, 2003.
- Niall A Herity, MD*,* (, )
- Sidney Lo, MB, BS*,
- David P Lee, MD*,
- Michael R Ward, MB, PhD*,
- Steven D Filardo, MD, MPH*,
- Paul G Yock, MD, FACC*,
- Peter J Fitzgerald, MD, PhD, FACC*,
- Sharon A Hunt, MD, FACC* and
- Alan C Yeung, MD, FACC*
- ↵*Reprint requests and correspondence:
Dr. Niall A. Herity, Department of Cardiology, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, Northern Ireland.
Objectives We sought to document whether a physiologic change in gender has any effect on coronary arterial size.
Background The coronary arteries are smaller in women, even after correction for body surface area (BSA). These differences may contribute to adverse clinical outcomes after coronary artery bypass graft surgery and myocardial infarction in women. In male and female transsexuals, pharmacologic doses of estrogens and androgens significantly influence vascular diameter. Thus, gender differences in the coronary vasculature may be a reflection of the hormonal environment.
Methods In 86 patients who had undergone orthotopic heart transplantation, serial intravascular ultrasound studies of the proximal left anterior descending coronary artery (LAD) were analyzed. Changes in vessel area (VA) over the first or second post-transplant year were recorded, and comparisons were made between donor hearts that were transplanted in a patient of the same gender and those that were transplanted in a patient of the opposite gender.
Results Vessel area of the proximal LAD increased over time in all patient groups. In hearts transplanted within the same gender and in male donor hearts transplanted to female recipients, the change was small and not significant. However, in hearts transplanted from female donors to male recipients, there was a substantial and highly significant increase in LAD VA (median 16.13 to 17.88 mm2; p = 0.01). This increase was not explained by confounding due to changes in BSA or left ventricular wall thickness.
Conclusions This pattern of arterial remodeling early after heart transplantation supports a link between host gender and coronary arterial size.
Observational evidence suggests that coronary artery disease (CAD) in women is different from that seen in men (1). The relative importance of risk factors is different, the disease presents later in women, and an atypical presentation is more frequent. The first disease manifestation is more likely to be angina pectoris in women, whereas it is acute myocardial infarction (MI) in men (2). Furthermore, morbidity and early mortality rates after MI (3)and after coronary artery bypass graft surgery (CABG) are substantially higher in women (4).
The reasons underlying these adverse clinical outcomes are likely to be multifactorial. Confounding factors may be partly responsible: female patients tend to be older and have a more complicated clinical presentation (5), and compared with men, they are marginally less likely to receive reperfusion therapy for acute MI (6). Anatomic differences in coronary size may also be important. In patients undergoing CABG, the mean luminal diameters of the left coronary arteries were significantly smaller in women (4), and smaller arterial size was an independent predictor of postoperative mortality, irrespective of gender (4). Intravascular ultrasound (IVUS) has also shown that the left main coronary artery (LMCA) and left anterior descending coronary artery (LAD) are significantly larger in men, even after correction for body surface area (BSA) (7).
Gender differences in coronary dimensions that persist after correction for BSA may be due to the influences of sex hormones on coronary arterial structure or tone. The vascular effects of exogenous sex hormone administration have been explored in small cross-sectional studies in men receiving high-dose estrogens (8)and in women receiving high-dose androgens (9). Compared with their respective same-sex controls, brachial artery diameter was significantly smaller in men taking pharmacologic doses of estrogens (8)and significantly larger in women taking high-dose androgens (9).
The present study aimed to further evaluate the relationship between gender and coronary arterial size. We analyzed the changes in the vessel area (VA) of the proximal LAD, as measured by IVUS, that occurred after hearts were transplanted from a donor to a recipient of the opposite gender, compared with those that were transplanted within the same gender. The hypothesis was that hearts transplanted across gender would exhibit changes in coronary arterial size that would differ from those seen in hearts transplanted within the same gender.
Patients who had undergone serial IVUS studies spanning either the first (i.e., within 4 weeks after heart transplantation and at the first annual evaluation) or second (i.e., at the first and second annual evaluations) post-transplant year were eligible for this study. Most were drawn from the IVUS database of the Multicenter Intravascular Ultrasound Transplant Study (10), in which 299 patients at eight heart transplant centers underwent serial annual IVUS examinations after heart transplantation. Additional patients were drawn from the Stanford University database of IVUS recordings from heart transplant patients who were not part of that study.
Intravascular ultrasound recordings that included adequate images of the proximal segment of the LAD (proximal to the first diagonal branch) were identified. Only disease-free vessels were included to avoid the potential confounding effects of vascular remodeling in response to atherosclerosis. All female heart transplant recipients were included, as long as they met the aforementioned criteria. In addition, all male recipients of a female donor heart were included, as were selected male recipients of a male donor heart (the largest patient group). Selection of patients for the male-to-male group followed the identification of the other three patient groups and was performed by choosing patients of similar donor age, similar recipient age, and similar dates of IVUS studies to all other patient groups. Furthermore, we ensured that the distribution of serial IVUS studies between the first and second post-transplant year was approximately equal in this group, as in the other patient groups.
Intravascular ultrasound imaging and analysis
The imaging protocol has been described (10). In brief, after coronary angiography and myocardial biopsy, intravenous heparin (5,000 to 10,000 U), sublingual nitroglycerin, and intracoronary nitroglycerin (200 μg) were administered. An angioplasty guide wire (0.014 in.) was advanced to the distal LAD, over which a 4.3F, 30-MHz IVUS catheter (Cardiovascular Imaging Systems, Sunnyvale, California) was advanced to the distal segment. Ultrasound time-gain compensation settings were adjusted to allow maximal gray-scale differentiation for the cross-sectional images within the target segment. In addition to selection of specific sites for follow-up comparisons, a continuous manual pullback recording of the artery was made until the guiding catheter artifact appeared. Images were recorded on S-VHS videotape and analyzed off-line. To minimize interobserver variability, one observer who was blinded to the donor–recipient gender profile performed all of the IVUS measurements. Quantitative cross-sectional analysis was performed using commercially available software (TapeMeasure, Indec Corp., Mountain View, California). Tapes from serial annual studies were reviewed side-by-side, and appropriate matching single segments of the proximal LAD were identified, most of which were immediately proximal to the first diagonal branch or immediately distal to the bifurcation of the LMCA. For each cross section, the VA was measured as the area bound by the external elastic lamina at end systole. Given that patients with any noticeable intimal thickening in the target segment were excluded, a three-layered vessel wall appearance was not the norm, and intimal area and lumen area were therefore not examined as separate variables for this study.
Demographic data and immunosuppressant drug doses were taken from the patients’ transplant, and medical records and echocardiographic data from medical and echocardiographic records. Immunosuppressant drug doses were recorded at the time of the follow-up IVUS study. End-diastolic dimensions of the interventricular septum were obtained from reports of echocardiographic examinations performed within three months of the initial and follow-up IVUS studies.
Data are expressed as the median and interquartile ranges or mean and standard error or standard deviation, as appropriate. Differences between groups were assessed by Kruskal-Wallis one-way analysis of variance on ranks, with post hoc testing by Dunn’s test (SigmaStat, Jandel Corp., San Rafael, California). Alternatively, we used the Mann-Whitney Utest when only two groups were involved. Differences within groups were analyzed using the Wilcoxon signed-rank test.
Eighty-six pairs of IVUS studies were identified from 86 different patients, spanning either the first (n = 41) or second (n = 45) year after heart transplantation. Transplants within the same gender accounted for 48 patients (32 male to male, 16 female to female), and there were 38 cross-gender transplants (17 male to female, 21 female to male). From these paired studies, 86 disease-free, matched segments of the proximal LAD were selected for analysis. Twenty patients were excluded from the study because of intimal thickening or an incomplete IVUS study (6 each from the female-to-female and female-to-male groups, 5 from the male-to-male group, and 3 from the male-to-female group).
Clinical and echocardiographic characteristics are shown in Table 1. Donor age and recipient age were very similar across the four groups. The BSA of male donors was greater than that of female donors, and the initial recipient BSA of male recipients was greater than that of female recipients. The donor BSA and initial recipient BSA were well-matched within the four patient subgroups. The median number of rejection episodes per patient was similar between the groups, as was the percentage requiring increased immunosuppressive drug therapy (Table 2). Hypertension was uncommon among both donors and recipients at the time of transplantation, although a substantial percentage of patients received antihypertensive agents during the study period (Table 2).
Changes in VA of proximal LAD with time
At the initial study, the median VA of the proximal LAD was larger in male than in female donor hearts, although the difference was not statistically significant (Table 3). Over the duration of the study, the VA tended to increase in all patient groups, although the magnitude of this change differed substantially between groups. In hearts that were transplanted to a recipient of the same gender, there were small changes in VA over time (Table 3, Fig. 1), and, in hearts transplanted from male donors to female recipients, a small and non-significant increase in VA was also observed. In contrast, in female hearts transplanted to male recipients, a striking and highly significant increase in VA occurred (mean 11.0 ± [SEM] 2.4% [SD 14.0%]) (Table 3, Fig. 2). An increase in VA was seen in 14 of 21 patients studied (7 of 11 and 7 of 10 patients from the first and second post-transplant years, respectively) and was of a similar magnitude when comparing the first and second post-transplant years (8.1 ± [SEM] 4.5% [SD 15.5%] and 11.3 ± 3.9% [SD 14.0%], respectively).
Changes in BSA over time
Body surface area increased between the initial and follow-up evaluations in all patient groups—a change that was more pronounced in female transplant recipients. In the group that had the greatest LAD enlargement, the female donor–male recipient group, only a small increase in BSA was observed (Table 1). By linear regression analysis, we tested whether the change in VA was dependent on the change in recipient BSA and found no evidence of a relationship (R2= 0.04 , p = 0.55).
Changes in interventricular septal thickness with time
Serial echocardiographic measurements of interventricular septal thickness were available for 77 of the 86 patients (Table 1). The initial septal thickness in the male-to-male group was somewhat greater than that in the other groups and remained so throughout the follow-up study. There were no major increases in interventricular septal thickness between the initial and follow-up studies. In the female donor–male recipient group, the median initial and follow-up septal thicknesses were virtually identical.
This study shows that in female hearts transplanted into male recipients, the VA of the proximal LAD increased significantly over the early transplant years—a change not seen in hearts transplanted from donors to recipients of the same gender, nor in male donor hearts transplanted into female recipients. Concomitant serial echocardiographic studies showed that this increase in vessel size was not a response to cardiac hypertrophy, and serial BSA measurements indicated that it was not accounted for by changes in body size.
Gender, sex hormones, and arterial size
Cardiac surgery studies have shown that men have larger coronary arteries than do women (4)and that smaller coronary arterial size is an independent risk factor for mortality after bypass surgery (4,11). Furthermore, IVUS showed that the LMCA and LAD are significantly larger in men, even after correction for BSA (7)and, hence, that the observed differences between men and women are not explained simply by different body size. Although we found that the VA of the proximal LAD was somewhat larger in men than in women, this difference did not reach statistical significance. We also evaluated whether the change in VA during the study was dependent on the change in BSA over the same time and found no evidence of a relationship.
The study of sex hormones in relation to vascular biology has concentrated on the impact of estrogens (and to a lesser degree, androgens) on arterial reactivity, rather than on arterial size. In post-menopausal women, intra-arterial infusion of 17β-estradiol potentiated endothelium-dependent vasodilation to acetylcholine in both the forearm (12)and coronary beds (13), although these effects were not reproduced during long-term estrogen administration (14). The effects of estrogen on vascular reactivity in men have also been documented. In two studies of transsexuals taking high-dose estrogen therapy, flow-mediated and nitroglycerin-mediated brachial artery dilation were enhanced when compared with male controls (8,15). It is interesting that in both studies the baseline brachial artery diameter was smaller in men taking estrogen than in control males, consistent with a link between estrogen and arterial size. This putative link is strengthened by a separate observation that long-term, high-dose androgen therapy in transsexual females was associated with a significantly larger baseline brachial artery diameter than that of control females (9)and that androgen-deprivation therapy in men was associated with a slightly smaller baseline brachial diameter than that of their matched controls (16). Although these studies were not designed to detect links between sex hormone status and arterial size, their findings suggest that androgen or estrogen exposure may have opposing vascular remodeling effects.
Thus, the distinct remodeling response seen in the female-to-male group in the present study may be explained on an endocrine basis. It assumes that these hearts moved from a high-estrogen, low-androgen environment (healthy females in their mid-20s) to a high-androgen, low-estrogen environment (middle-aged men) and that either estrogens act to reduce coronary arterial size or that androgens increase coronary arterial size. The fact that coronary arteries did not get smaller when hearts moved from healthy males in their 20s (a putative high-androgen, low-estrogen environment) to women in middle age (a low-androgen environment) may indicate that estrogens have little direct impact on coronary size, that most female transplant recipients were postmenopausal (age range 43 to 55 years) and therefore did not have high circulating estrogen levels, or alternatively, that androgen-induced changes in coronary artery structure are irreversible, irrespective of what endocrine environment they subsequently move to.
In the female donor–male recipient group, the magnitude of vascular enlargement was similar in the first and second post-transplant years. If these changes are attributed to endocrine influences, our findings suggest that the effects are sustained rather than transient. Outward vascular remodeling patterns in response to transplant CAD have previously been shown to be more pronounced when they occur late than when they occur early post transplantation (17). Long-term remodeling responses in cross-gender heart transplants are not addressed in the present study and may be a focus for future research.
Donor and recipient gender and outcome after heart transplantation
A variety of studies addressing the question of gender and clinical outcomes after heart transplantation have shown that survival is influenced by donor gender and donor–recipient gender matching but not by recipient gender. Specifically, reduced long-term survival rates have been consistently reported in patients who receive female compared with male donor hearts (18–21)as well as in gender-mismatched compared with gender-matched transplants (18,20,21). Among gender-mismatched transplants, the worst survival is seen when female donor hearts are transplanted to male recipients (18,20,21)—an observation not readily explained by differences in the number of rejection episodes (18,20). The present study was not designed to study clinical outcomes in different heart transplant populations, but it is interesting that we observed considerable early outward remodeling in a subgroup that consistently has a worse long-term prognosis than all others. It may be speculated that outward coronary arterial remodeling is not invariably a beneficial or compensatory phenomenon and that, in particular circumstances, it is an early marker of vascular injury or endothelial dysfunction (22). We excluded patients with evidence of intimal thickening on IVUS at either the initial or follow-up study; hence, we were unable to detect whether the pattern of arterial enlargement observed translated to any difference in the rate of development or progression of transplant CAD.
Influences on vascular remodeling patterns
Arterial remodeling occurs in response to diverse physiologic or pathophysiologic stimuli (23). The former includes growth, exercise, and increased metabolic demand, whereas the most common known pathologic stimulus is coronary atherosclerosis. Although gender differences seen in the present study may intuitively suggest a hormonal influence, other factors need to be considered.
All of the recipient patients studied were adults with a minimum age of 39 years (Table 1). It is therefore unlikely that the transplanted hearts were exposed to either physiologic growth stimuli or excessive exercise training stimuli. Body surface area is one of several variables considered when transplant donors are matched to potential recipients, although some variation is frequent. In the present study, the female donor–male recipient group had a greater donor–recipient mismatch in median BSA than the other patient groups (Table 1). However, the magnitude of the difference was small, and the change in recipient BSA between the first and second IVUS studies in this group was less than that found in some other patient groups. The possibility that donor–recipient BSA mismatch accounted for the increase in coronary artery size observed is not excluded by the present study.
Compensatory remodeling in response to transplant CAD is unlikely to account for the changes we observed. Although a range of remodeling responses—from overcompensation to no compensation or shrinkage—has been described in response to transplant CAD (24), an integral feature of the present study was the exclusion of all vessel segments that had IVUS evidence of intimal thickening either at baseline or follow-up. Intravascular ultrasound is a highly sensitive technique for such intimal thickening and has become the accepted standard for detection of transplant CAD in vivo. Therefore, it is unlikely that undetected transplant CAD accounted for the vascular remodeling seen in these patients.
Immunosuppressant drugs and sex hormone levels
Of the immunosuppressant agents taken by heart transplant recipients, both cyclosporin A (CsA) and prednisone may modulate plasma sex hormone levels. Cyclosporin A causes dose-dependent hirsutism, but in men and women taking CsA for rheumatoid arthritis (25)or after renal transplantation (26), circulating testosterone levels were not changed, and the androgenic effect of CsA was better explained by accelerated peripheral testosterone metabolism (25). However, in post-menopausal women with primary biliary cirrhosis, plasma estradiol levels were significantly suppressed by treatment with CsA (27). Oral prednisone therapy is also associated with suppression of estrogen levels in women (28)and with suppression of androgen levels in men (29,30).
In the heart transplant population, iatrogenic suppression of sex hormones is well recognized and is thought to contribute to accelerated bone turnover in the early post-transplant years (31). Whether such hypogonadism contributes to changes in vascular structure is not known, although the similar drug doses across the four groups of patients, coupled with disparate remodeling behavior in one patient group compared with all others, suggest that a ubiquitous influence like immunosuppressant therapy is unlikely to adequately explain the present observations.
A predictable limitation of a study of this size is its potential lack of power to detect subtle differences between study groups or within groups over time. This is because of the highly select nature of the study population: cross-gender heart transplants comprise only a small percentage of the total transplant population, and from that small cohort, only a minority have sequential IVUS data from the early post-operative years. On the other hand, these constraints on study size mean that observed differences had to be of considerable magnitude to attain statistical significance and, hence, were likely to be biologically important when they arose. Secondly, we chose to use interventricular septal thickness as a surrogate marker of left ventricular (LV) mass, with specific reference to the territory of the artery under investigation—the LAD. This is not the same as formal measurement of LV mass, although septal thickness is an integral component of LV mass calculation (32).
Plasma concentrations of sex hormone levels were not assayed. Thus, although it may be assumed that a group of young female donors (median age 24 years [range 17 to 38]) is likely to exhibit, on average, a pre-menopausal estrogen profile, this is not proven conclusively by the present study. Similarly, it is assumed but not proven that male recipients exhibited physiologic androgen levels. To confidently ascribe the observations made to an endocrine cause, it would be necessary to demonstrate expected hormone profiles in the population under investigation.
Finally, post-transplant coronary arteries are a unique group of vessels for the study of arterial remodeling, in that they are denervated, have undergone an ischemic insult, and are exposed to complex immunologic and pharmacologic stimuli. Although we have taken measures to exclude confounding as the principal mechanism underlying our observations, we recognize that they may not translate completely to the much larger non-transplant population.
In this serial IVUS study, the proximal LAD of female hearts transplanted into male recipients underwent significant enlargement early after transplantation, an effect not seen in any other patient group. The findings were not explained by confounding due to increased BSA nor by LV hypertrophy and support a link between host gender and coronary arterial size, possibly mediated by sex hormone exposure.
We appreciate the help and permission given by the following institutions in the collection of clinical and echocardiographic data: Stanford University (J. Miller, RN; P. Gamberg, RN; H. Luikart, RN; M. A. Merlo, RN; D. Liang, MD), Brigham and Women’s Hospital (R. Prpic, MD), University of California, Los Angeles (M. Brown, RN), Indiana University Hospital, Ochsner Cardiac Transplantation, and St. Louis University.
☆ Dr. Herity was supported by the Northern Ireland Council for Postgraduate Medical and Dental Education and by the Queen’s University of Belfast. Dr. Ward was supported by the National Heart Foundation of Australia.
This study was presented in part at the 49th Annual Scientific Sessions of the American College of Cardiology in March 2000 in Anaheim, California.
- body surface area
- coronary artery bypass graft surgery
- coronary artery disease
- cyclosporin A
- intravascular ultrasound
- left anterior descending coronary artery
- left main coronary artery
- left ventricular
- myocardial infarction
- vessel area
- Received November 20, 2000.
- Revision received April 18, 2002.
- Accepted October 25, 2002.
- American College of Cardiology Foundation
- Knopp R.H.
- O’Connor G.T.,
- Morton J.R.,
- Diehl M.J.,
- et al.
- New G.,
- Timmins K.L.,
- Duffy S.J.,
- et al.
- McCredie R.J.,
- McCrohon J.A.,
- Turner L.,
- Griffiths K.A.,
- Handelsman D.J.,
- Celermajer D.S.
- the Multicenter Intravascular Ultrasound Transplant Study Group,
- Yeung A.C.,
- Davis S.F.,
- Hauptman P.J.,
- et al.
- Gilligan D.M.,
- Badar D.M.,
- Panza J.A.,
- Quyyumi A.A.,
- Cannon R.O.
- Gilligan D.M.,
- Quyyumi A.A.,
- Cannon R.O.
- McCrohon J.A.,
- Walters W.A.,
- Robinson J.T.,
- et al.
- Herman S.M.,
- Robinson J.T.,
- McCredie R.J.,
- Adams M.R.,
- Boyer M.J.,
- Celermajer D.S.
- Pethig K.,
- Heublein B.,
- Meliss R.R.,
- Haverich A.
- Ward M.R.,
- Pasterkamp G.,
- Yeung A.C.,
- Borst C.
- Lim T.T.,
- Liang D.H.,
- Botas J.,
- et al.
- Cutolo M.,
- Giusti M.,
- Villaggio B.,
- et al.