| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2004;109:1489-1496.)
© 2004 American Heart Association, Inc.
Clinical Investigation and Reports |
From the Division of Cardiovascular Surgery, St Vincent Mercy Medical Center, Toledo (A.Z., R.H.H., T.A.S., C.J.R., S.J.D., A.S.); Division of Cardiovascular Surgery, St Lukes Hospital, Maumee (A.Z., R.H.H., T.A.S., C.J.R., S.J.D., A.S.); and Departments of Surgery (A.Z., T.A.S., C.J.R., S.J.D., A.S.) and Medicine (R.H.H.), Medical College of Ohio, Toledo, Ohio.
Correspondence to Dr Robert H. Habib, Director, Cardiopulmonary Research, St. Vincent Mercy Medical Center, 2213 Cherry St, ACC Bldg, Suite 309, Toledo, OH 43608. E-mail Robert_Habib{at}mhsnr.org
Received October 21, 2003; revision received January 9, 2004; accepted January 9, 2004.
| Abstract |
|---|
|
|
|---|
Methods and Results We compared 6-year outcomes in propensity-matched CABG-LITA-LAD patients (925 each) divided into those with
1 radial grafts and those with vein-only grafting. Matched patients had essentially identical demographics, comorbidities, coronary disease, and operative data. Perioperative outcomes, including death (radial, 11 [1.2%]; vein, 10 [1.1%]), were similar for the 2 groups. Cumulative 0- to 6-year survival was better for radial patients (risk ratio, 0.675), particularly after 3 years (P<0.03). Six-year survival in vein (86.8%) and radial (92.1%) patients indicated 67% greater overall vein mortality. Incidence rates of radial and vein repeated catheterization (190 of 925 [20.5%] versus 199 of 925 [21.5%]) and revascularization (8.8% versus 8.5%) were similar. Angiography data in restudied symptomatic patients showed a trend for greater radial patency. Vein failure (66 of 161 [41%]) was significantly worse than radial failure (46 of 157 [29.3%]) in patients receiving both types of grafts (P=0.039).
Conclusions Using radial as a second arterial conduit in CABG-LITA-LAD as opposed to vein grafting improves long-term outcomes as a result of decreased late deaths, especially after the third postoperative year.
Key Words: angiography revascularization coronary disease restenosis
| Introduction |
|---|
|
|
|---|
Surgeons have extrapolated LITA results to other arterial conduits and are using them for revascularization with increasing frequency.610 Right internal thoracic artery (RITA) grafts are used often and are reported to have superior patency compared with vein grafts.68 Recent analyses showed significantly better long-term survival when both ITA conduits are used compared with a single ITA.9,10 Bilateral use of ITA, however, is contraindicated in patients who are at increased risk of sternal wound complications.1113 Moreover, in situ RITA conduits may not reach remote areas of the right coronary and circumflex branches, are more technically demanding (longer operative times) particularly if reoperation is necessary,14 and may not be sufficient for complete arterial revascularization. Also, RITA use as a free graft diminishes its long-term patency.15,16
Acar and colleagues17 reintroduced the radial artery as a viable conduit for CABG. Improved harvesting techniques,18 together with potential benefits of calcium channel blockers,19 have led to radial becoming an increasingly popular alternative/adjunct conduit.20,21 Early and intermediate-term angiographic data have so far indicated relatively better radial versus vein patency,2225 yet, to date, improved long-term outcomes with radial use have not been shown, particularly in conjunction with the potentially dominant survival benefit of the LITA-LAD graft. In the present study, we analyzed a large CABG-LITA-LAD experience with the primary aim of investigating whether use of radial versus vein as the second conduit of choice provides improved perioperative and long-term outcomes.
| Methods |
|---|
|
|
|---|
99%) was used unless aorta quality was suboptimal. Cardiopulmonary bypass (92%) and sequential grafting were used to a similar extent in both groups. Radial grafting was rarely used in emergency and peripheral vascular disease patients. Radial use increased from <20% (1996) to >50% (2002; detailed surgeon data in provided in the online Data Supplement).
Radial Harvesting
Hand collateral circulation and palmar arch status were assessed by modified Allen test and Doppler ultrasonography, respectively. Intraoperative plethysmography and oximetry were used before vessel removal. Radial harvesting was abandoned after exploration in case of small-vessel caliber and/or significant calcification. Harvesting was done as a pedicle without electrocautery at the same time of the LITA dissection and immersed in diluted papaverine. Mechanical dilatation was not used to avoid intimal injury. Intravenous calcium channel blockers (diltiazem) were administered intraoperatively and continued through day 1. Oral calcium channel blockers were prescribed for 2 months.
Propensity Matching
Study patients were drawn from observational data, and radial and vein groups were hence characterized by significant demographic and risk factor differences (Table 1). Because such differences substantially influence outcomes,26 we used propensity matching to overcome such confounding effects, with radial considered treatment.26,27 Briefly, the probability that a patient received a radial graft was defined by a propensity score via logistic multivariate modeling applied to all patients. Most risk factors, demographics, and operative variables were entered into the model regardless of significance level. Month of surgery was entered to account for increasing radial use over time. Highly redundant variables were avoided.
|
As expected, radial use propensity score distributions derived for radial versus vein patients were distinct (Figure 1). One-to-one propensity matching of radial patients with the closest unique vein match was obtained by a computer algorithm allowing a maximal difference of ±1%. Patient matching was done on the basis of propensity scores only (standard) and by propensity scores in addition to gender and triple-vessel disease status (yes/no) restrictions (restricted).
|
Follow-Up
Long-term survival data were secured from our service follow-up and verified from individual queries of the US Social Security Death Index database (June 2003). Allowing for a 3-month lag, this corresponded to between 3 (minimum) and 87 (maximum) months of follow-up. Post-CABG recatheterization data were obtained from cardiac surgery and cardiac catheterization databases and from referring cardiologists and primary care physicians. All angiography reports were retrieved and reviewed, and graft data were entered into a dedicated database. A coronary graft was considered to have failed in case of complete occlusion, stenosis of
90%, or presence of extensive conduit narrowing or "string sign." Cumulative patency calculations in restudied patients were done with the following guidelines: (1) Time of graft failure is the date of recatheterization when failure was documented; (2) grafts were presumed patent unless a subsequent recatheterization showed otherwise; and (3) patent grafts are censored at death.
Statistical Analysis
Continuous data were expressed as mean±SD. Univariate comparisons were done with
2 or Fishers exact test for categorical variables and the unpaired t test or the nonparametric Mann-Whitney rank-sum test for continuous variables. Kaplan-Meier plots were derived for survival, recatheterization-free survival, revascularization-free survival, and graft patency (all compared via the log-rank test). Survival data were verified by bootstrapping (see online Data Supplement).28 Effects of explanatory variables on survival were derived by Cox proportional-hazard analysis. Model selection was done with backward elimination (Wald statistic confirmed by forward and stepwise selection), and variables with P<0.05 were retained (SPSS version 10.0, SPSS Inc).
| Results |
|---|
|
|
|---|
97%) had multivessel coronary disease. Overall radial and vein patient data are compared in Table 1. Briefly, vein patients were older (mean age, 67 versus 61 years), were more often female (37% versus 23%), were smaller, and had more comorbidities. Consequently, radial patients were associated with generally better operative outcomes (Table 1). Also, radial showed superior 0- to 6-year (P<0.001; Figure 2), recatheterization-free, and revascularization-free survival (both P<0.001; Figure I in the online Data Supplement).
|
To account for confounding effects of treatment group differences, 925 of 1292 radial patients (72%) were propensity matched to the same number of unique vein patients (Figure 1). Matched groups had essentially identical demographics, risk factors, and operative data (Table 1). Comparisons of key continuous variables (age, follow-up, perfusion, ejection fraction, body surface area, and mass index) for matched patients are provided in the online Data Supplement (Figure II). All subsequent comparisons are done on matched groups.
Coronary Grafts
Coronary vessel disease and grafting data for both groups are detailed in Table I (online Data Supplement). Briefly, the overall number of completed grafts was similar for radial (3020 grafts: 946 LITA, 19 RITA, 1180 radial, 875 SVG) and vein (2997 grafts: 942 LITA, 24 RITA, and 2031 SVG), with similar distributions of graft target location and stenosis (Table II and Figure III). Sequential radial grafts were used in 200 patients (21.6%). All arterial grafting was achieved in 311 patients (33.6%) with a total of 468 radial grafts (1.5 per patient). The remaining 614 radial patients received a total of 712 radial grafts (1.16 per patient) and 875 vein grafts (1.43 per patient).
Patient Outcomes
Postoperative outcomes in matched patients, including rates of transfusion and use of intra-aortic balloon pump, are summarized in Table 1. Similar operative mortality, defined as in-hospital or out-of-hospital death within 30 days, rates were documented (11 radial [1.2%] versus 10 vein [1.1%]). Most postoperative complications, length of stays, and readmission rates were also similar.
Long-term follow-up was similar for radial (1364±687 days; range, 1 to 2702 days; median, 1380 days) and vein (1325±771 days; range, 3 to 2703 days; median, 1382 days) patient groups. A total of 47 late deaths (5.1%) were documented in the radial group occurring at 816±691 days. In vein patients, there were 72 late deaths (7.8%) occurring at 875±708 days. Cumulative 0- to 6-year survival was significantly better in the radial group (P<0.03; Figure 3, top). Here, estimated survival was 96.8% versus 96.1% at 1 year and 92.1% versus 86.8% after 6 years for radial and vein, respectively. Thus, the corresponding 0- to 6-year (13.2% versus 7.9%) and 1- to 6-year (9.3% versus 4.7%) mortality rates were 67% and 98% greater, respectively, in vein patients. Generally, survival was similar for the 2 groups up to about 6 months; after that, the incidence of late death was relatively greater in vein patients, particularly after the third post-CABG year (Figure 3, bottom).
|
Importantly, these survival results were closely reproduced via 3 complementary analyses. First, we obtained essentially identical survival trends using "restricted" matching groups (618 each; Figure 3, top). Second, analysis of 1000 bootstrap populations resulted in similar differences in averaged survival curves with relatively narrow confidence bounds (Figure 3, middle). Finally, radial use was an independent predictor of decreased 0- to 6-year mortality (risk ratio, 0.675; P=0.04; Table 2), along with older age, diabetes, and ejection fraction <40%. The radial survival benefit was also true, albeit differently, when patients were subdivided on the basis of specific risk factors. In women, triple-vessel disease patients, younger patients (<65 years of age), and diabetic patients (Figure 4, left), improved survival is evident relatively early. Alternatively, their counterparts exhibited this benefit only after the third year after surgery (Figure 4, right).
|
|
One or more repeated angiographies were documented in a similar proportion of radial (190 of 925 [20.5%]) and vein (199 of 925 [21.5%]) patients. The average time to recatheterization did not differ between the 2 groups (radial: 668±518 days; range, 4 to 2142 days; median, 558 days; vein: 649±539 days; range, 3 to 2267 days; median, 532 days). A total of 264 catheterizations were performed in 190 radial patients, compared with 272 in 199 vein patients (both 1.4 per patient). Need for revascularization was also similar (radial, 81 [8.8%]; vein, 79 [8.5%]): repeated CABG in 4 versus 3, angioplasty in 15 versus 14, and stenting in 75 versus 71. Actuarial analysis showed relatively better catheterization-free and revascularization-free survival trends in radial patients (Figure 5, top), but these differences can be attributed almost entirely to fewer late deaths rather than to less frequent intervention (Figure 5, middle).
|
Graft Patency
For the 190 radial patients with
1 repeated angiography, a total of 599 (242 radial, 161 vein, 191 LITA, 5 RITA) grafts were restudied (3.2 per patient). Similarly, for the 199 vein patients, a total of 637 (427 vein, 203 LITA, 7 RITA) grafts were restudied (3.2 per patient). The corresponding cumulative patency rates for all the restudied LITA, radial, and vein grafts are shown in Figure 5 (bottom). Briefly, LITA patency was superior to that of radial and vein grafts (both P<0.001). The 6-year radial graft patency trend was systematically greater than that of vein grafts, but these cumulative differences did not reach statistical significance by Kaplan-Meier analysis (P=0.11, log-rank test).
Documented graft failure was lowest for LITA conduits (24 of 394 [6.1%]; P<0.001 compared with both radial and vein grafts), which were used exclusively as LITA-LAD grafts. A total of 77 of 242 restudied radial grafts failed (31.8%). Of these, 31 of 85 (36.5%) occurred in 54 restudied patients with arterial grafts only compared with 46 of 157 failures (29.3%) in 137 radial patients with
1 additional vein graft. Overall, 216 of 588 vein grafts (36.7%) were found to have failed. Vein graft failure rates derived from the radial (66 of 161 [41%]) and vein (150 of 427 [35.1%]) groups were similar (P=0.22). Note, however, that when compared in the same patients, ie, radial patients with additional vein grafts, the documented 41% vein graft failure was significantly greater than the 29.3% radial graft failure (P=0.039).
| Discussion |
|---|
|
|
|---|
Multiple groups have reported better CABG survival when both ITA conduits are used concomitantly compared with the use of a single ITA with additional vein grafts.710 However, these promising data have been clouded by contradictory reports31,32 and by the fact that not all included patients received LITA-LAD grafting, which may have a dominant effect on survival. Caputo and colleagues33 compared use of radial versus RITA as the second arterial conduit in combination with LITA and found better perioperative outcomes and improved event-free survival at 18 months using radial. However, their study included a relatively small number of unmatched patients (336 RITA, 325 radial) unevenly distributed over a 5-year period and drawn from an experience with a relatively low rate of multiple arterial conduit use (15%). Nevertheless, if independently confirmed, their findings are significant, given the many potential advantages to using radial artery conduits. Indeed, compared with RITA, radials (1) are larger and easier to work with; (2) are more straightforward to prepare; (3) are harvested concomitantly with LITA, reducing operative time; (4) are more amenable to revascularize remote branches than the RITA pedicle graft; and (5) avoid increased risk of sternal wound infections caused by bilateral ITA dissection.1113
This study examined the potential outcome benefits of using radial as a second arterial graft as opposed to vein graftingstrictly in conjunction with the LITA-LAD graftin well-matched patient groups. Perioperative outcomes, including operative deaths, and resource use were mostly similar (Table 1), yet univariate analyses showed a trend for lower vein harvest site wound infection rates in radial patients, a finding similar to that in a recent report.34 More importantly, we found significantly better 0- to 6-year CABG survival when radial is used as a second arterial conduit as opposed to vein (Figure 3). The separation in the observed survival trends started at 6 to 12 months after CABG and became more pronounced after the third year. Such a trend toward a greater difference in survival with increasing time after CABG is analogous to that reported for CABG with 2 ITA conduits compared with a single ITA with additional vein grafts.9
A limitation of our analysis was that it was performed on a retrospective observational CABG series. In addition, the question of whether radial grafting improves CABG outcome is perhaps best addressed via specifically designed prospective multicenter randomized trials. Fortunately, advances in statistical methods have overcome many of the known shortcomings of observational data analyses.27,28 If used properly, these methods may adequately answer such important clinical questions.9 Here, we concomitantly used several available methods to increase the confidence in and applicability of our findings. The main approach was to propensity match radial and vein patients in whom radial use was considered a treatment. To match patients, we applied a strict matching criteria of ±1% maximum propensity score difference, and matching was done on a 1-to-1 basis, resulting in the same number of unique patients in the 2 comparison groups (avoiding population-size bias) rather than the more frequently used decile- or quintile-based propensity matching regardless of group size.9 This approach successfully identified 2 closely matched groups (Table 1 and Figure II). Next, we applied a restricted form of the same 1-to-1 matching in which patients were always matched to the same gender and coronary vessel disease type. Analyses of the restricted-matched patient groups produced survival data that were essentially identical to those from standard matching findings (Figures 3 and 4
).
Another approach to test the effect of radial versus vein grafting on survival was done via Cox multivariate modeling applied to all matched patients. Here, too, radial grafting was determined to be an independent predictor of better survival. Finally, we used 1000-times bootstrapping to determine the reproducibility and confidence bounds of the derived radial and vein survival trends.26,28 Results of this computer-intensive method agreed closely with the above-described findings, and the computed survival trends were characterized by narrow confidence bounds, indicating good reproducibility (Figure 3, middle). We contend that convergence of all the aforementioned methods significantly advances our main finding of superior survival when radial grafting is preferred in CABG-LITA-LAD.
Improved long-term CABG outcome when 2 arterial conduits are used is presumably linked to superior patency of arterial grafts. A number of recent studies have demonstrated superior patency relative to vein of all types of arterial conduits, including radial.110,2225 Ideally, graft patency is best characterized by reexamination of all or a representative, randomly drawn sample of patients. The available angiography data were instead derived from patients restudied because of recurrent cardiac symptoms and/or positive stress tests. These data represented a substantial number of grafts derived from
20% of patients. Overall, we find more vein than radial failures. When we consider patients with both radial and vein grafts (ie, patient is own control), we documented significantly better radial patency (111 of 157 [71%] versus 95 of 161 [59%]; P=0.039). Also, using actuarial analysis, we found that the cumulative patency derived for 588 restudied vein grafts was systematically worse between 1 and 6 years compared with radial, even if this finding is not statistically significant (Figure 5; P=0.11). The lack of significance may stem in part from the fact that the patency data were derived from symptomatic patients exclusively who reflected a worst-case scenario.35
| Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. A. Schwann, A. Zacharias, C. J. Riordan, S. J. Durham, A. S. Shah, and R. H. Habib Does radial use as a second arterial conduit for coronary artery bypass grafting improve long-term outcomes in diabetics? Eur. J. Cardiothorac. Surg., May 1, 2008; 33(5): 914 - 923. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Damgaard, J. T. Lund, N. B. Lilleor, M. J. Perko, K. Sander, B. Dimo, M. B. Jensen, J. K. Madsen, H. Kelbaek, and D. A. Steinbruchel Comparable three months' outcome of total arterial revascularization versus conventional coronary surgery: Copenhagen Arterial Revascularization Randomized Patency and Outcome trial. J. Thorac. Cardiovasc. Surg., May 1, 2008; 135(5): 1069 - 1075. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Bleiziffer, I. Hettich, B. Eisenhauer, D. Ruzicka, M. Wottke, J. Hausleiter, S. Martinoff, M. Morgenstern, and R. Lange Patency rates of endoscopically harvested radial arteries one year after coronary artery bypass grafting J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 649 - 656. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. A. Schwann, A. Zacharias, C. J. Riordan, S. J. Durham, A. S. Shah, and R. H. Habib Survival and Graft Patency After Coronary Artery Bypass Grafting With Coronary Endarterectomy: Role of Arterial Versus Vein Conduits Ann. Thorac. Surg., July 1, 2007; 84(1): 25 - 31. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. H. Habib, A. Zacharias, and T. A. Schwann Long-term Mortality Associated With Aprotinin Following Coronary Artery Bypass Graft Surgery JAMA, June 13, 2007; 297(22): 2476 - 2477. [Full Text] [PDF] |
||||
![]() |
J. W. Fischer Tenascin-C: A key molecule in graft stenosis Cardiovasc Res, June 1, 2007; 74(3): 335 - 336. [Full Text] [PDF] |
||||
![]() |
S. Na, Y. J. Oh, Y. H. Shim, Y. W. Hong, S. O. Bang, and Y. L. Kwak Effects of milrinone on blood flow of the Y-graft composed with the radial and the internal thoracic artery in patients with coronary artery disease. Eur. J. Cardiothorac. Surg., August 1, 2006; 30(2): 324 - 328. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M. Wildhirt, B. Voss, F. von Canal, M. Benz, J. B. Grammer, R. Bauernschmitt, P. Tassani, and R. Lange Graft function, histopathology and morphometry of radial arteries used as conduits for myocardial revascularization in patients beyond age 70. Eur. J. Cardiothorac. Surg., August 1, 2006; 30(2): 333 - 340. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. G. Nezic, A. M. Knezevic, P. S. Milojevic, B. P. Dukanovic, M. D. Jovic, M. D. Borzanovic, and A. N. Neskovic The fate of the radial artery conduit in coronary artery bypass grafting surgery. Eur. J. Cardiothorac. Surg., August 1, 2006; 30(2): 341 - 346. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Di Lazzaro, T. Ragni, G. Di Manici, G. Bardelli, U. Da Col, F. Grasselli, A. Antoniella, W. Papa, F. Crusco, and A. Giovagnoni Noninvasive midterm follow-up of radial artery bypass grafts with 16-slice computed tomography. Ann. Thorac. Surg., July 1, 2006; 82(1): 44 - 50. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. G. Nezic, A. M. Knezevic, M. Dj. Jovic, and S. D. Borovic Angiographic Patency of the Radial Artery Conduit in Coronary Artery Bypass Grafting Ann. Thorac. Surg., April 1, 2006; 81(4): 1551 - 1551. [Full Text] [PDF] |
||||
![]() |
L. R. Sajja and G. Mannam Reply Ann. Thorac. Surg., April 1, 2006; 81(4): 1551 - 1552. [Full Text] [PDF] |
||||
![]() |
Vascular-wall remodeling of 3 human bypass vessels: organ culture and smooth muscle cell properties. J. Thorac. Cardiovasc. Surg., March 1, 2006; 131(3): 651 - 658. |
||||
![]() |
I. Parvaiz, J. T. Lund, and H. Kelbaek The Arterial Sling Operation: One-Year Follow-Up Ann. Thorac. Surg., October 1, 2005; 80(4): 1375 - 1380. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Knobloch, A. Lichtenberg, S. Tomaszek, C. Hagl, N. Khaladj, U. Klima, and A. Haverich Long-Term Physical Activity and Neurologic Function After Harvesting of the Radial Artery as T-Graft or Free Graft in Coronary Revascularization Ann. Thorac. Surg., September 1, 2005; 80(3): 918 - 921. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Zacharias, R. H. Habib, T. A. Schwann, and C. J. Riordan Arterial versus vein graft patency in coronary artery bypass grafting patients with ischemia-directed repeat angiography Eur. J. Cardiothorac. Surg., September 1, 2005; 28(3): 510 - 511. [Full Text] [PDF] |
||||
![]() |
D. Glineur, P. Noirhomme, J. Reisch, G. El Khoury, P. Astarci, and C. Hanet Resistance to Flow of Arterial Y-Grafts 6 Months After Coronary Artery Bypass Surgery Circulation, August 30, 2005; 112(9_suppl): I-281 - I-285. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Kobayashi, T. Tashiro, M. Ochi, H. Yaku, G. Watanabe, T. Satoh, O. Tagusari, H. Nakajima, S. Kitamura, and for the Japanese Off-Pump Coronary Revascularizati Early Outcome of a Randomized Comparison of Off-Pump and On-Pump Multiple Arterial Coronary Revascularization Circulation, August 30, 2005; 112(9_suppl): I-338 - I-343. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. P. Georghiou, B. A. Vidne, and J. Dunning Does the radial artery provide better long-term patency than the saphenous vein? Interactive CardioVascular and Thoracic Surgery, August 1, 2005; 4(4): 304 - 310. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Matsuura, J. Kobayashi, O. Tagusari, K. Bando, K. Niwaya, H. Nakajima, T. Yagihara, and S. Kitamura Off-Pump Coronary Artery Bypass Grafting Using Only Arterial Grafts in Elderly Patients Ann. Thorac. Surg., July 1, 2005; 80(1): 144 - 148. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Zacharias, R. H. Habib, T. A. Schwann, C. J. Riordan, S. J. Durham, and A. Shah Radial artery conduits in coronary artery bypass grafting: Current perspective J. Thorac. Cardiovasc. Surg., July 1, 2005; 130(1): 232 - 233. [Full Text] [PDF] |
||||
![]() |
|