Ann Thorac Surg 1995;59:1423-1428
© 1995 The Society of Thoracic Surgeons
Experimental Supplemental Vein Grafting and Hypoperfusion Syndrome
Masaki Otaki, MD,
Robert M. Lust, PhD,
You Su Sun, MD,
Terry O. Norton, MD,
David T. Rock, MD,
Paul A. Spence, MD,
W. Randolph Chitwood, Jr, MD
Departments of Surgery and Physiology, East Carolina University School of Medicine, Greenville, North Carolina
 |
Abstract
|
|---|
An additional saphenous vein graft (SVG) sometimes is required to the same coronary system if acute internal thoracic artery (ITA) graft flow is inadequate. These experiments were conducted to determine the consequences produced by ITA--SVG dual grafting. Fourteen dogs each received two coronary grafts (without bypass, using local occlusion) to the proximal circumflex coronary artery, using the ITA and an SVG, and then the circumflex artery was ligated proximally. Simultaneous flow in both grafts was determined at rest and after pharmacologic (adenosine, phenylephrine) or physiologic (cardiac pacing) stimulation. Serial angiography was performed during the first 4 weeks after grafting to determine patency patterns of the ITAs and SVGs. In the resting heart, flow was 7.5 ± 1.6 mL/min (17.5%) in the ITA graft and 35.3 ± 5.2 mL/min (82.5%) in the SVG (mean ± standard deviation [% total distal perfusion]), and the combined flow was not significantly different from the original native flow. Intravenous adenosine (0.2 mg • kg-1 • min-1) preferentially increased both the total ITA flow and its fractional contribution to total distal perfusion (18.4 ± 3.2 [31.1%]; p < 0.05 versus rest). Saphenous vein graft flow was not changed significantly (40.3 ± 6.0 mL/min), in part due to a modest decrease in arterial pressure. In contrast, intravenous phenylephrine (0.003 mg • kg-1 • min-1) decreased both absolute ITA flow and its relative contribution to distal perfusion (6.1 ± 1.1 [10.9%]; p < 0.05 versus rest), despite an increased systemic perfusion pressure, which increased SVG flow significantly (50.1 ± 4.8 [89.1%]; p < 0.05 versus rest). Increasing ventricular rate by 25% from 120 to 150 beats/min increased both ITA and SVG flow significantly (ITA, 10.8 ± 2.5 mL/min; SVG, 42.6 ± 6.2 mL/min; both p < 0.05 versus rest) but did not change the flow balance between the grafts. Six animals were studied chronically, and serial angiography demonstrated the following results: all vein grafts retained patency throughout, but 2 ITA grafts were occluded within 7 days, 2 additional grafts were occluded within 2 weeks, and the remaining 2 grafts were occluded within 3 weeks after grafting. These data demonstrate that the ITA graft did not survive in this canine dual grafting model, and suggest that the necessary acute perfusion benefit of supplemental vein grafting for hypoperfusion syndrome may eliminate any long-term patency advantages of the ITA graft.
|
Introduction
|
|---|
The benefits of using the internal thoracic artery (ITA) graft in coronary artery bypass grafting have been documented [1, 2]. Hypoperfusion syndrome describes conditions of inadequate myocardial perfusion, documented by intraoperative electrocardiographic or wall motion changes (echocardiogram), acutely following coronary artery bypass grafting, most commonly in regions supplied by an ITA graft. Hypoperfusion syndrome has been related to ITA spasms, flow-limiting technical problems at the distal anastomosis, or a disproportion between ITA flow and myocardial oxygen demands, and may be more common during reoperations when an ITA graft is used to replace a stenotic vein graft [3–5]. In these situations, immediate supplemental saphenous vein (SV) grafting is indicated to increase perfusion before discontinuation of cardiopulmonary bypass [3–5].
When an ITA is grafted to an incompletely occluded coronary artery, some controversy exists regarding the effects of the residual native flow on ITA flow characteristics and patency [5–7]. Previous experimental animal studies have demonstrated that when the ITA was grafted to a nonstenotic circumflex artery (CFX), the ITA graft retained patency despite significant residual native flow and demonstrated dynamic interaction with the native artery in supplying distal myocardial perfusion [6, 7]. Because a vein graft has flow-regulating characteristics quite different from those of a native coronary artery, it remains unclear how dual flow sources terminating in the coronary system will interact, especially when the potential for flow-limiting ITA graft-specific vasoconstriction in response to systemic pressors also has been demonstrated [8, 9]. Therefore, these experiments were designed to determine the effects of ITA-SV dual grafting on flow responses in each graft individually, on the relative contribution of each graft to meet distal perfusion requirements, and on the continued angiographic patency of each graft over time.
|
Material and Methods
|
|---|
Surgical Preparation
The guiding principles for the Care and Use of Laboratory Animals of the American Physiology Society were followed carefully during the conduct of these experiments. Fourteen dogs weighing 17 to 25 kg (mean, 21 ± 2.7 kg) were premedicated with sodium thiopental (15 mg/kg intravenously) and intubated. Anesthesia was maintained subsequently by inhalation using a mixture of nitrous oxide, oxygen, and halothane. An arterial sheath was placed in the femoral artery percutaneously, and arterial blood pressure (Millar Instruments, Houston, TX) and a surface electrocardiogram (lead II) were monitored continuously. A left fifth intercostal thoracotomy was performed and a pedicled ITA graft was rotated from the left chest wall from its origin proximally to the level of the eight intercostal space distally. A segment of the SV graft from the dorsum of the left hind leg was harvested and prepared for use as a bypass graft. The pericardium then was incised and a small margin of the left atrial appendage was ligated and retracted to expose the atrioventricular groove. A section of the CFX between the posterior descending branch and the first large marginal branch was exposed and prepared for grafting. The SV--aorta proximal anastomosis was constructed using 7-0 suture and partial occlusion of the descending aorta. The coronary anastomosis was performed using a brief local occlusion of the coronary artery, without cardiopulmonary bypass or shunt, as described previously [6–8]. Briefly, snares were placed around the vessel at each end of the dissected coronary segment. Lidocaine (1.5 mg/kg) and esmolol (500 µg/kg bolus, 50 µg • kg-1 • min-1 infusion) were given intravenously. The snares were tightened, and the vessel was elevated somewhat, thereby minimizing movement artifacts associated with ventricular contraction and stabilizing the anastomotic field. The SV--CFX distal anastomosis was constructed using 7-0 suture in a continuous manner. After 20 minutes of reperfusion, the snares were tightened again and the ITA--CFX anastomosis was constructed just proximally to the SV--CFX anastomosis, using the same technique. During the ITA--CFX anastomosis, distal CFX perfusion was maintained through the newly completed SV--CFX graft. After both grafts were placed on the CFX, the CFX was ligated proximally (Fig 1
), rendering distal perfusion completely dependent on the two grafts.
View larger version (34K):
[in this window]
[in a new window]
|
Fig 1. . Diagram of the complete preparation showing the approximate location of both bypass grafts. (AO = ascending aorta; CFX = circumflex artery; ITA = internal thoracic artery; LAD = left anterior descending coronary artery; LM = left main trunk coronary artery; SV = saphenous vein.)
|
|
Measurements
All flow measurements were made using calibrated electromagnetic probes (Zepeda Instruments, Seattle, WA). Flows in the ITA graft and the SV graft were measured and arterial pressure was recorded after intravenous administration of adenosine (0.2 mg • kg-1 • min-1), after intravenous administration of phenylephrine (0.003 mg • kg-1 • min-1), and in response to ventricular pacing at a heart rate of 150 beats/min (compared with a control rate of 120 beats/min). All treatments were continued for 10 minutes before measurements were obtained, and 20 minutes between each intervention was provided, enabling complete recovery of resting parameters.
Angiographic Evaluations
In 6 animals these surgical procedures were performed in a sterile operative field with aseptic instruments to enable subsequent postoperative angiographic evaluations. After the intraoperative flow studies had been completed, the thoracotomy was repaired in layers, and the animals were allowed to recover from the anesthesia. Standard postoperative antibiotic (penicillin, 106 units per day subcutaneously for 5 days) was administered, and the animals were maintained on aspirin (325 mg/day) for the duration of the study.
Angiographic studies were performed 7 days, 14 days, and 21 days after grafting. The animals were sedated and placed in a right lateral position. A sterile field was prepared, and percutaneous access of the arterial system was obtained via the right or left femoral artery. The graft was imaged by injection of a contrast material (Mallinkrodt Medical, Inc, St. Louis, MO) and angiographic images were recorded on videotape.
Statistics
All data are expressed as the mean ± the standard deviation of the mean. Differences between a resting value and a value in response to pharmacologic or physiologic stimulation were assessed using Student's t test for paired differences, and corresponding probability values less than 0.05 were considered statistically significant.
|
Results
|
|---|
Under resting conditions, flow and each vessel's contribution to total distal perfusion, respectively, were 7.5 ± 1.6 mL/min and 17.5% in the ITA graft; and 35.3 ± 5.2 mL/min and 82.5% in the SV graft. Intravenous adenosine preferentially increased the ITA flow and its contribution total distal perfusion (18.4 ± 3.2 mL/min and 31.1%, p < 0.05), despite a significant decrease in mean arterial pressure (87.5 ± 6.1 versus 73.1 ± 6.5 mm Hg; p < 0.05). Saphenous vein graft flow was not different from rest (40.6 ± 5.8 mL/min; p = not). In contrast, phenylephrine decreased both absolute ITA flow and its relative contribution to total distal perfusion (6.1 ± 1.1 mL/min, 10.9%; p < 0.05 versus rest), despite an increased systemic perfusion pressure (86.0 ± 6.5 versus 107.5 ± 8.0 mm Hg; p < 0.05). Reciprocally, absolute flow in the SV graft was increased, as was its relative contribution to total distal perfusion (50.1 ± 4.8 mL/min and 89.1%; p < 0.05 versus rest). Increasing ventricular rate by 25%, from 120 to 150 beats/min, increased flow in both the ITA and SV grafts significantly (ITA, 10.8 ± 2.5 mL/min; SV, 42.6 ± 6.2 mL/min; both p < 0.05 versus rest) but did not change the ITA graft fractional contribution to distal perfusion (20.0%), and the flow balance between the grafts was not different from resting control (Fig 2).
View larger version (27K):
[in this window]
[in a new window]
|
Fig 2. . Measured flow in each graft (A) and the fractional contribution provided by each graft to distal perfusion (B) in the resting heart, after intravenous administration of adenosine (ADO) or phenylephrine (PE), and with ventricular pacing. Values shown are mean ± standard deviation (ITA = internal thoracic artery; SV = saphenous vein; *p < 0.05.)
|
|
Serial angiography in 6 dogs demonstrated the following results: all vein grafts retained patency throughout, but 2 ITA grafts were occluded within 7 days, 2 additional grafts were occluded within 14 days, and the remaining 2 grafts were occluded within 21 days after grafting. Figure 3 demonstrates that both grafts were widely patent angiographically in the first week after grafting, but Figure 4 (angiography in the following week) illustrates the fully patent vein graft in contrast to the angiographically occluded ITA graft. Figure 5 demonstrates the ITA graft patency in competition with the SV graft, in contrast to patency of the ITA graft completed by the normal CFX from a previous study.
View larger version (101K):
[in this window]
[in a new window]
|
Fig 3. . Angiography performed in the first week after grafting demonstrated that both grafts were fully patent and runoff through the distal circumflex artery (CFX) was evident. (ITA = internal thoracic artery; SV = saphenous vein.)
|
|
View larger version (159K):
[in this window]
[in a new window]
|
Fig 4. . Angiography of the same vessels shown in Figure 3 performed 1 week later showed a fully patent vein graft (SV), but the internal thoracic artery graft no longer could be imaged. Subsequent physical examination at autopsy revealed a microscopically intact lumen less than 0.5 mm in diameter in a vessel that was fibrotic along much of its length. (CFX = circumflex artery.)
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig 5. . Cumulative patency of the internal thoracic artery graft in competition with the saphenous vein (SV) graft in contrast to patency in competition with a nonstenotic circumflex artery (CFX). Data for native artery patency from [7].
|
|
|
Comment
|
|---|
The association between use of the ITA graft as a coronary artery bypass graft and both superior patency rate and long-term survival has been documented [1, 2]. However, the potential for a relative flow deficit (hypoperfusion syndrome) using ITA grafts is a persistent concern. Hypoperfusion has been observed during reperfusion while the patient remains on cardiopulmonary bypass, or just after weaning from cardiopulmonary bypass, and has been documented by intraoperative electrocardiographic changes or wall motion abnormalities by echocardiography [3, 4]. Hypoperfusion syndrome may result from intraoperative spasm of the ITA graft, technically inadequate, flow-limiting distal anastomoses, anatomically limited flow capacity of the ITA graft, and hypertrophy of the underlying myocardium. In addition, Navia and associates [5] reported on a clinical series of reoperative coronary grafting in which the mortality rate was more than three times higher when stenotic vein grafts were replaced with an ITA alone, compared with when the stenotic vein graft was left intact or when a supplemental vein graft was placed to diagonal branch. They noted that the increase in mortality was attributed to a hypoperfusion syndrome, especially when revascularization was attempted using the ITA alone to a target myocardium previously supplied by a vein graft [5]. In any situation in which mechanical performance is judged to be graft flow--limited, the addition of a supplemental saphenous vein graft to the myocardium suspected of hypoperfusion before cardiopulmonary bypass is discontinued has been recommended [3–5].
The acute necessity for the supplemental vein graft is clear, but the dynamic flow characteristics and the long-term consequences of this dual grafting strategy are less clear. Previous experiments from our laboratory have documented the effects of competing flow from a normal nonstenotic coronary artery (CFX) on ITA flow. These experimental results suggested that ITA graft flow was not overwhelmed by flow from the normal coronary vessel in the dog, and that relative ITA graft flow in competition with native coronary flow was proportional to the relative diameter of the ITA graft compared with the native coronary artery [6]. Chronic experimental studies using a similar model [7] also demonstrated continued patency of the ITA graft despite significant residual flow through the native coronary artery, and that dynamic reactivity of the vessel was preserved. It was hypothesized that retention of dynamic regulatory capacity in the ITA graft allowed the ITA graft to interact with native artery much more dynamically to meet distal perfusion demands and may have been a key factor in preserving the ITA patency, even under competitive flow from a fully patent native CFX. These experimental studies are consistent with clinical results by Urschel and associates [10], who analyzed 1,000 patients who had undergone intraoperative transluminal balloon angioplasty as an adjunct to coronary artery bypass and found that postoperative ITA graft status was not influenced by the severity of proximal stenosis in the grafted coronary artery.
Although the previous data suggest that ITA flow may not be affected significantly by competing flow from native sources, it is unknown what effects competing flow from relatively nondynamic conduit vein grafts might have on ITA flow. Twenty years ago, Flemma and associates [11] reported intraoperative results from a clinical study of ITA-vein double grafting to the left anterior descending coronary artery. They noted that the SV graft limited the ITA graft severely, and that the vein graft flow was 2.7 times higher than the ITA flow, placing the ITA contribution to total distal perfusion at approximately 25%. Acute flow measurements in our present study demonstrated that the vein graft provided the CFX regions with more than 80% of the total distal perfusion and limited the ITA graft contribution to less than 20%. These experimental findings are consistent with and extend the clinical results reported by Flemma and associates [11]. The observation that the ITA fares poorly in competition with SV grafts is quite different from results previously observed with ITA-native coronary flow competition [6, 8]. The present data suggest that flow competition from nonreactive, relatively passive vein grafts produces more severe limitations on ITA graft flow than does flow competition between two dynamically reactive arteries (native coronary artery and ITA), at least acutely. Alternatively, the data suggest that a minimum absolute flow or a minimum contribution to distal perfusion by the ITA may be required to preserve ITA patency.
Increased myocardial oxygen demands secondary to pharmacologic or physiologic stimulation also can change the flow balance between the ITA and SV grafts, and potentially exacerbate flow disparities. In a previous report [8], the ITA graft increased flow in response to increased coronary flow requirements with adenosine-induced coronary vasodilation or cardiac pacing. In contrast, the ITA graft decreased flow in response to phenylephrine-induced vasoconstriction, despite increased myocardial oxygen consumption. Both the adenosine and phenylephrine effects of this study were attributed to direct vasodilation and vasoconstriction, respectively, of the ITA graft. Adenosine is a well-known mediator of coronary vasoreactivity, and produces general vasodilation in the peripheral circulation through direct actions on vascular smooth muscle and also through interaction with vascular endothelium and endothelium-derived relaxation factors [12]. In the present study, adenosine decreased the perfusion pressure, which partially limited vein graft flow, but also directly dilated the ITA graft and increased the ITA contribution to distal perfusion from 17.5% to 33% (p < 0.05).
Phenylephrine is an alpha-adrenergic agonist that produces vasoconstriction in the larger epicardial coronary vessels and increases heart rate and myocardial oxygen consumption [9]. When phenylephrine was infused, the effects of vasoconstriction dominated the response in the ITA graft, reducing perfusion through this conduit, while the increased perfusion pressure systemically and tachycardia-induced coronary dilatation locally increased flow through the vein graft. Thus phenylephrine increased vein graft flow but decreased ITA flow to less than 10% of the total distal perfusion. This effect was directly related to vasoconstriction of the ITA, because cardiac pacing alone produced comparable flow increases in both grafts and did not change the ITA contribution to distal perfusion.
Late angiographic studies demonstrated that all ITA grafts were occluded within 21 days after grafting, but all vein grafts retained patency throughout. However, the angiographic demonstration of true ITA graft failure is difficult to distinguish from a functional, nonperfused ITA (``string sign''). Therefore, angiographically occluded ITA grafts were harvested at autopsy and examined physically in 4 dogs. Three of the ITA grafts had a definable lumen but displayed fibrotic narrowing along the length of the graft. In one graft, an organized intraluminal thrombus was defined. However, angiographically occluded ITA graft may demonstrate reversible patency with progression of native disease or occlusion of the supplemental vein graft [13–15]. The time course, minimal residual ITA flow, and anatomic characteristics of the native circulation between the two grafts all may play significant roles in the ultimate patency of the ITA in this dual grafting setting.
Within the context of the present study, and previous reports using native coronary arteries, the effects of competing flow from other sources on ITA patency are probably more complex than originally anticipated. At a minimum, it appears that competition between two relatively equally reactive vessels is better suited to preserve long-term patency than is competition between one highly reactive and one relatively nonreactive vessel. Although the relative size of the ITA with respect to the coronary circulation in the dog is relatively large, compared with humans, this only serves to emphasize the potential for early ITA graft failure when redundant vein grafts are placed. Despite the relatively large size of the ITA graft with respect to the total distal perfusion, it was still not enough to maintain patency against flow provided from the additional vein graft. Although this report discusses the effects of dual grafting as used clinically to address hypoperfusion, it should be reiterated that hypoperfusion syndrome per se has not been demonstrated using this model, again, probably due to the relative size of the canine ITA. In humans with coronary artery disease, it is more likely that some degree of stenosis will be present in the native circulation between the insertion of the two grafts, which could reduce the flow-limiting impact of the SV on the ITA graft and preserve long-term patency. The possibility that the sequence of distal graft anastomoses contributed to the observed results was considered. As reported, the SV graft insertion always was distal to the ITA graft. In a separate series of controls, the sequence of grafts was reversed. The ITA anastomosis was constructed first, and placed distally to the subsequent SV graft anastomosis. However, similar flow patterns were observed, and failed patency in the ITA persisted [16].
In summary, these experimental results suggest that in circumstances in which the flow capacity of the ITA is suspect, it may be preferable to choose a vein graft alone, rather than a vein graft in addition to the ITA graft, because the long term patency of the ITA in this dual grafting may not be realized. If the ITA graft can autoregulate to contribute less than 20% of the total distal perfusion, long-term patency is less likely, and the necessary acute perfusion benefit of supplemental vein grafting for hypoperfusion syndrome may eliminate any long-term patency advantages of the ITA graft.
|
Acknowledgments
|
|---|
We recognize the technical assistance of Kathy Dennis in completing these studies, the editorial assistance of Laurie Rouse in preparing the manuscript, and the generous donation of suture used in the anastomoses by Davis + Geck, Inc.
|
Footnotes
|
|---|
Presented at the Forty-first Annual Meeting of the Southern Thoracic Surgical Association, Marco Island, FL, Nov 10–12, 1994.
Address reprint requests to Dr Lust, Cardiothoracic Research, Department of Surgery, East Carolina University School of Medicine, Brody 4S-22, Greenville, NC 27858-4354.
|
References
|
|---|
- Ivert T, Huttunen K, Björk VO. Angiographic studies of internal mammary artery grafts 11 years after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1988;96: 1–12.[Abstract]
- Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PL. Long-term (5 to 12 years) serial studies of internal mammary artery and saphenous vein coronary grafts. J Thorac Cardiovasc Surg 1985;89:248–58.[Abstract]
- Jones EL, Lattauf OM, Weitraub WS. Catastrophic consequences of internal mammary artery hypoperfusion. J Thorac Cardiovasc Surg 1989;98:902–7.[Abstract]
- Tector AJ, Amundsen S, Schmahl TM, Kress DC, Peter M. Total revascularization with T grafts. Ann Thorac Surg 1994;57:33–9.[Abstract]
- Navia D, Cosgrove DM III, Lytle BW, et al. Is the internal thoracic artery the conduit of choice to replace a stenotic vein graft? Ann Thorac Surg 1994;57:40–4.[Abstract]
- Spence PA, Lust RM, Zeri RS, et al. Competitive flow from a fully patent coronary artery does not limit acute mammary graft flow. Ann Thorac Surg 1992;54:21–6.[Abstract]
- Lust RM, Zeri RS, Spence PA, et al. Effects of chronic native flow competition on internal thoracic artery grafts. Ann Thorac Surg 1994;57:45–50.[Abstract]
- Otaki M, Lust RM, Sun YS, et al. Myocardial perfusion in canine internal thoracic artery bypass graft-dependent regions. Ann Thorac Surg (in press).
- Jett GK, Arcidi JM Jr, Dorsey LMA, Hatcher CR Jr, Guyton RA. Vasoactive drug effects on blood flow in internal mammary artery and saphenous vein grafts. J Thorac Cardiovasc Surg 1987;94:2–11.[Abstract]
- Urschel HC, Razzuk MA, Miller E, Chung SY. Operative transluminal balloon angioplasty. J Thorac Cardiovasc Surg 1990;99:581–9.[Abstract]
- Flemma RJ, Singh HM, Tector AJ, et al. Comparative hemodynamic properties of vein and mammary artery in coronary bypass operation. Ann Thorac Surg 1975;20:619–27.[Abstract]
- Berne R. The role of adenosine in the regulation of coronary blood flow. Circ Res 1980;47:907–13.
- Kitamura S, Kawachi K, Seki T, Sawabata N, Morita R, Kawata T. Angiographic demonstration of no-flow anatomical patency of internal thoracic--coronary artery bypass grafts. Ann Thorac Surg 1992;53:156–9.[Abstract]
- Aris A, Borras X, Ramio J. Patency of internal mammary artery grafts in no-flow situations. Ann Thorac Surg 1987;93:62–4.
- Shammas RL, Mehta PM, Jolly SR, Lust RM, Zeri RS, Spence PA. Reversibility of the ``string sign'' of the left internal mammary artery graft. Cathet Cardiovasc Diagn 1993;30:236–9.[Medline]
- Otaki M, Lust RM, Sun YS, et al. Does proximal or distal placement of a supplemental vein graft have different effects on ITA flow or distal perfusion? [Abstract]. J Invest Surg 1994;7:363.
This article has been cited by other articles:
|
|
|
|
 
G. Bolotin, A. P. Kypson, L. W. Nifong, and W. R. Chitwood Jr
A technique for evaluating competitive flow for intraoperative decision making in coronary artery surgery
Ann. Thorac. Surg.,
December 1, 2003;
76(6):
2118 - 2120.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Louagie, M. Buche, P. Eucher, and J.-C. Schoevaerdts
Intraoperative flow measurements in gastroepiploic grafts using pulsed Doppler
Eur. J. Cardiothorac. Surg.,
March 1, 1999;
15(3):
240 - 246.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Pagni, E. Salloum, J. Storey, W. Montgomery, P. Cerrito, D. Van Himbergen, L. A. Gray Jr., and P. A. Spence
Double grafting of the left anterior descending artery: is the distance between the internal mammary artery and supplemental vein graft anastomoses relevant in graft survival?
Eur. J. Cardiothorac. Surg.,
January 1, 1998;
13(1):
36 - 41.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
