| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2006;29:S115-S125
© 2006 Elsevier Science NL
a Department of Surgery, Division of Cardiothoracic Surgery, David Geffen School of Medicine at University of California Los Angeles, 10833 Le Conte Avenue, 62-258 CHS, Los Angeles, CA 90095-1701, USA
b Option on Bioengineering, California Institute of Technology, Pasadena, CA, USA
Received 17 February 2006; accepted 27 February 2006.
* Corresponding author. Tel.: +1 310 206 1027; fax: +1 310 825 5895. (Email: gbuckberg{at}mednet.ucla.edu).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
Key Words: Heart anatomy • Ventricular septum • Helical heart • Ventricular myocardial band • Sonomicrometry
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
Septal studies are traditionally separated into anatomic and physiologic frameworks. However, insight into the interaction of the different muscular components that spatially form the ventricular mid-wall is needed to determine mechanisms of form and function to explain sequential cardiac events. The helical cardiac configuration was initially described over 500 years ago [2], together with extensive recent anatomic studies that separate cardiac muscle into superficial sinospiral and deep bulbospiral structures [3]. Unfortunately, the classic reports by Pettigrew et al. [4] did not distinguish how the oblique septal fibers (Fig. 1A) that interact to form the ventricular vortex, can coordinate to produce the twisting and untwisting action described so elegantly in (a) the visual images from Mall's anatomic report in 1911 [3] (Fig. 1B) and (b) more recently in vivo by MRI analyses [5,6].
|
|
|
Connection of physiology to function requires analysis of contractile shortening of the left and right sides of the thick septum, with findings based upon an anatomic framework. This study tests a hypothesis that the septum reflects the underlying structure of the free left ventricular wall, because it is comprised of the same descending and ascending segments of the apical loop configuration described by Torrent-Guasp et al. [10,11]. Accumulated information shall satisfy the objectives of (a) identifying the dominant angle of maximal muscular contraction on thick septal left and right septal sides, (b) defining a similar sequential shortening pathway as exists within the thick left ventricular free wall, and (c) clarifying a comparable response to positive and negative inotropic influences. The validity of findings may fulfill the existent gap in knowledge about septal anatomy and function.
|
2. Materials and methods |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
Ten early studies were done without bypass, and 30 subsequent studies were done using extracorporeal circulation to allow direct septal visualization.
Forty Yorkshire–Duroc pigs (29.5–38.4 kg) were premedicated with ketamine hydrochloride (15 mg/kg) and diazepam (0.5 mg/kg) intramuscularly and anesthetized with inhaled isoflurane 1.5% (MAC 1%). Support with a volume-controlled ventilator (Servo 900C, Siemens-Elema, Sweden) was started after tracheotomy and endotracheal intubation. The left extrathoracic mammary artery was cannulated for arterial pressure measurement and blood gas analysis. Arterial blood gases were measured to keep oxygen tension, carbon dioxide tension and pH values in normal range. A balloon-tipped pulmonary artery catheter (Model 132F5, Baxter Healthcare Corp., Irvine, CA, USA) was inserted through the right subclavian vein to measure cardiac output (thermodilution technique) and pulmonary artery pressure. Each pig underwent median sternotomy, and exposure of the heart, following a pericardial incision.
Thirty pigs underwent systemic heparinization (300 units/kg), and a 12 F arterial cannula (Medtronic Inc., Minneapolis, MN, USA) was inserted into the superficial femoral artery, and two 17 F venous cannulas (Medtronic Inc.) were inserted into the superior and inferior vena cava through the superficial jugular vein and femoral vein, respectively. Extracorporeal circulation was instituted using a membrane oxygenator (Affinity NT 541, Medtronic Inc.) and an extracorporeal pump (Sarns, Ann Arbor, MI, USA) with the circuit primed with 1000 ml Plasma-Lyte solutions (Baxter Healthcare Corp.), 700 ml stored porcine packed blood, and calcium chloride for normocalcemia (1.0–1.2 mmol/l). Cardiopulmonary bypass (CPB) was started at 300 mmHg oxygen tension, 50–70 mmHg arterial pressure, and flow adjustment to keep 
70% mixed venous oxygen saturation and 35–37 °C rectal temperature.
A solid-stated pressure transducer-tipped catheter (Model MPC-500, Millar Instruments Inc., Houston, TX, USA) was inserted via the left ventricular apex for left ventricular hemodynamic measurements. Arterial pressure (AP), left ventricular pressure (LVP), dP/dt, and sonomicrometer crystals data were digitally processed by specific hardware and software (Sonometrics, London, Ont., Canada). Regional shortening was measured with pairs of 2 mm ultrasonic microtransducer crystals (Sonometrics).
Percentage of segmental shortening (SS%) was calculated as follows
|
|
Velocity of sound through cardiac tissue was fixed to 1590 m/s. Sonomicrometer measurements were recorded with a sampling rate of 195.8 samples/s, a transmitter spacing of 652 µs, transmit inhibit delay of 1.18 µs, and transmit pulse length of 375 ns. Synchronicity between myocardial shortening was compared to left ventricular performance with 5 ms precision. All studies were performed and analyzed by the same surgeon.
2.1 Septal approach
Figs. 3–7
define the anatomic considerations underlying this effort to record sonomicrometer crystal tracings from the left and right ventricular aspects of the septum. The bias was that (a) unfolding the right ventricular free wall (Figs. 3 and 4) will expose the helical septum constructed from overlap of the oblique left-sided descending segment and a right-sided ascending segment of the helical heart apical loop; (b) the dominant shortening direction should conform to vectors of crystal placement confirmed in left ventricular free wall (Figs. 5–7) [20]; and (c) sequential performance of this muscular septal infrastructure will parallel the sequential shortening sequence in the LV free wall caused by positive and negative inotropic interventions.
|
|
|
|
A basic requirement is satisfactory recording of sequential shortening of the descending and ascending segments of the left ventricular free wall. The logic is related to (a) prior LV free wall studies [20] establishing the correct angulation of crystal placement (Figs. 5 and 6), and (b) testing this bias by comparison of free wall records to septal tracings. Failure to make this correlation invalidates the hypothesis of this experiment.
2.2 Without CPB
Two methods were tested and involved either (a) approaching the septum surface on the outside surface of the right and left ventricular sides, or (b) transmurally via septal muscle penetration by an external blind approach (via the free LV and RV walls).
Initial studies used the free wall of the left and right ventricles to advance crystals into the respective cavity (confirmed by back bleeding) with subsequent septal surface penetration on either side. Free wall angulation positions were used as guidelines for satisfactory positions. Despite good initial recordings, these site locations were transient. Consequently, either crystal fixation by suture or impaction against the underlying septum (as with free wall endocardial placement) was needed to sustain the position.
Secondary efforts used transmural crystal placement by advancing a longer insertion sheath to cross the septum flowing free wall insertion, and then withdrawing it to endocardially attach crystals on the right and left ventricular septal sides. Constant recordings could not be obtained and postmortem studies showed improper positioning. These inconsistent findings led to subsequent direct septal vision studies.
2.3 With CPB
The first study addressed the left- and right-sided septums by direct vision via a 4 cm ventriculotomy on the left and right sides. The beating heart was used and crystals were placed on left and right septal endocardium to avoid transmural changes due to multiple septal punctures. Placement angulation was either in an oblique pathway along the free wall fiber direction pathway (Fig. 6) or in a transverse direction and across the expected maximum shortening direction as reported previously [20].
Each crystal was fixed with a prolene stay suture, but the direction of suture placement effected results. Initially satisfactory recordings failed after transverse suture placement, as this positioning compressed underlying oblique fibers, whereas oblique direction suture placement allowed consistent and satisfactory left- and right-sided septal tracings.
Despite valid recordings in the open ventricle, consistent irregular left-sided recordings occurred after discontinuation of bypass for subsequent measurements. The need for repeat left ventriculotomy for replacement of crystals created an unsatisfactory solution because new septal damage occurred after many repetitive septal punctures.
A second study used a brief interval of hyperkalemic cardioplegia to expedite ventriculotomy, crystal placement, and closure. These studies were also unsuccessful because initial assessment of correct function was delayed until ventricular closure, which often resulted in dislocation of crystals.
The third study was successful, as only a 4 cm right ventriculotomy was used to approach both sides of the septum (Figs. 5 and 6). Left-sided septum crystals were placed via a transmural puncture (similar to transmural epicardial procedure used to place free wall endocardial crystals) with a specially designed PVC tube, and right-sided endocardial crystals were inserted just beneath the RV septal surface. All data were collected following right ventriculotomy closure and discontinuation of bypass.
2.4 Experimental protocol
Each pig underwent placement of two pairs of sonomicrometer crystals on the epicardial and endocardial sides of the left ventricular free wall (Fig. 5). These baseline free wall tracings were then compared to septal crystal tracings. On the free wall epicardial side, crystals were placed beneath a 1 mm epicardial incision. Endocardial crystals were then placed by inserting a specific crystal introducer (1 mm diameter PVC tube) beyond the 1 mm epicardium incision into the ventricular cavity (identified by pulsatile bleeding), the sensor cord was then pulled against the endocardial surface with subsequent epicardial fixation by a 4/0 Prolene suture. Postmortem examination confirmed LV surface and septum position.
Local heart coordinates were used to select highest point at aortic annulus, and lowest at the apex. A North and South fashion was used to define the crystal placement angles for most powerful segmental shortening as described previously [20]. Epicardial crystal angles were between 140° and 150°, compared to between 80° and 90° (Fig. 5) for endocardial sites.
After free wall crystal placement, CPB was initiated and total bypass was achieved by snaring the inferior and superior cava veins. A beating heart was always used, and the 3–4 cm right ventriculotomy was made parallel to the left anterior descending artery. Following placement of free wall traction sutures, the RV septum underwent placement of 4/0 Prolene stay sutures for crystal placement.
The ascending segment forms the right ventricular side of the septum [10,21] and three crystals were placed in two angulation directions. One orientation site followed the specific oblique ascending segment fiber orientation, while the other direction was perpendicular to the free wall ascending segment fiber direction (Fig. 6).
The descending segment forms predominantly the left ventricular side of the septum, and was approached via the right ventricular side of the septum. Following the 1 mm incision, a 1 mm diameter PVC crystal introducer tube was pushed into the left ventricular cavity. Pulsatile bleeding confirmed the transmural position, and crystals were brought to septal endocardial wall by pulling on the electrical cord wire to place the sensor against the LV septum. As with the right side, placement was done in a direction along the descending segment fibers, and then perpendicular to them (Figs. 6 and 7). The right ventricle was closed with a 2/0 Prolene suture and CPB was slowly discontinued.
Baseline hemodynamic measurements and crystal recordings were performed 15 min after weaning from CPB. Responses to positive and negative inotropic agents were subsequently recorded after administering (a) dopamine at 10 µg/(kg min) and (b) a 50 mg esmolol bolus injection after the dopamine effect subsided.
2.5 Statistical analysis
Hemodynamic and sonomicrometer data were compared by t-test between pharmacologic interventions and are reported as mean ± standard deviation (mean ± SD). p-values <0.05 were considered statistically significant.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
Baseline LV free wall SS% were 21 ± 3% and 10 ± 3% for the descending and ascending segments, respectively, and SS% of the septal descending and ascending segments were 13 ± 2% and 12 ± 2%, respectively (Table 1 ). These tracings displayed the expected 75 ± 11 ms time-delay between the start of descending and ascending segment shortening. Similarly, a 86 ± 21 ms hiatus existed between the end of ascending segment shortening and the earlier completion of descending segment shortening (Fig. 8 ; Table 2 ). As described below, a similar pattern of delays between the initiation and termination of contraction of the left and right sides of the septum exited, but timing and magnitude depended upon the positioning of the paired crystals.
|
|
|
|
After baseline measurements, dopamine was administered at 10 µg/(kg min) which resulted in an increase of heart rate from 82 ± 12 beats/min to 105 ± 9 beats/min (p < 0.05; Table 1; Fig. 10 ). The time-delay between the start of contraction in the descending and ascending segments decreased from 75 ± 11 ms to 54 ± 6 ms (p < 0.05), and the duration hiatus between the descending and ascending segment at end of contraction was shortened from 86 ± 21 ms to 68 ± 10 ms (p < 0.05; Table 2; Fig. 10). Left ventricular free wall SS% increased to 24 ± 2% and 14 ± 3% at the descending and ascending segments, respectively, and significantly exceeded baseline SS% (p < 0.05). Similarly, septal contractility increased to 16 ± 2% and 15 ± 2% at the descending and ascending segments, respectively (p < 0.05 vs baseline; Table 1).
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
The study was biased, since the experimental design was geared to evaluate: (a) if the septum had a spatial configuration that matched the helical ascending and descending segments of the apical loop that was presumed to form it, (b) if the major septal shortening sequence pattern existed within fiber orientation pathways that followed those preciously established in the left ventricular free wall [20], and (c) if septal muscle displayed a similar response to positive and negative inotropic drugs to confirm the contractile nature of active diastolic filling from isovolumetric contraction (rather than conventional isovolumetric relaxation). Each goal was accomplished to provide a novel septal structure and function relationship.
Analysis of dominant shortening characteristics along the fiber orientation within the ventricular myocardial band model of myocardial structure requires evaluation of the validity of the rationale behind the experimental design. The helical rope like model consists of a single muscle band that extends from below the pulmonary artery to form a basal loop that wraps transversely around the right and left ventricles, undergoes a spiral ventricular myocardial fold to become a helix composed an oblique descending segment, apical vortex and reciprocal oblique ascending segment to finish at the trigones of the aorta.
The septum is thus surrounded by the right ventricular segment of the basal loop, and when exposed (Fig. 11 ), is made from three muscular strata [10] that include: (a) the right stratum represented by oblique fibers of right ventricular free wall, which turn towards inside of the ventricle and cover subendocardially the right ventricular side of the septum after reaching the anterior interventricular surface, (b) the middle stratum consisting of oblique fibers of ascending segment, and (c) the left stratum composed of oblique fibers of the descending segment.
|
Septal transverse or oblique directions were tested to determine timing and extent of maximum shortening. The oblique direction was essential (Figs. 5, 6 and 8) and confirmed the hypothesis. Furthermore, tracing distortion occurred when stay sutures were placed in transverse rather than oblique position, thereby confirming the importance of the oblique angulation. Conversely, failure to reproduce free wall findings would invalidate the hypotheses. Furthermore, understanding this design bias clarifies why (a) sonomicrometer crystals were used to address how ascending and descending segments formed the left and right sides of septum, and (b) studies on and off bypass were done with use of the beating heart, cardioplegia, and biventricular and right ventricular approaches, to evolve the confirmatory data that comprises this report.
The heart undergoes sequential activation of the ventricular band to produce four cardiac movements which include: (a) narrowing by contraction of outer basal loop and internal surface of the descending segment, (b) shortening by co-contraction of the predominant descending and then ascending segments of the apical loop, (c) lengthening by unopposed contraction of the ascending segment of the apical loop, and (d) relaxation without contraction of any part of the myocardial band. Theoretical form animation of this sequence is seen on the web site http://www.gharib.caltech.edu/, under helical heart. These four movements explain both phases of cardiac function, including co-contraction for systolic ejection, and unopposed ascending contraction to allow the active muscular early suction phase for rapid ventricular filling; an action that differs from the conventional acceptance of elastic recoil from stored potential energy during systole to explain isovolumetric relaxation [5,24].
Reproduction of left ventricular free wall studies was essential to match with the septum, and current findings confirmed previous sequential movements causing systolic ejection and early suction phase of the diastolic cardiac function [20]. This is shown in Fig. 8, demonstrating (a) simultaneous early shortening of descending segment of the free wall and septum, (b) later origin of ascending segment shortening in free wall and septum, (c) co-contraction of descending and ascending segments of free wall and septum during ejection, (d) ongoing ascending shortening in free wall and septum during rapid fall in left ventricular pressure, just before when active suction occurs, and (e) similar delay between end of descending segment shortening and later completion of ascending shortening.
These interactions produce the following physiologic results relative to septal contributions to cardiac action. The co-contraction following shortening of both the descending and ascending segments of the septum allows the whole septum to contribute to longitudinal shortening of the ventricle during systolic ejection, which is associated with clockwise twisting on MRI analysis [5,6]. The ascending segment continues to shorten after the descending segment stops to allow suction for rapid filling, thereby using the descending segment as a fulcrum to lengthen during the rapid fall in the ventricular pressure curve [25]. The suction phase of the cardiac cycle is characterized by reciprocal twisting to accentuate rapid filling [25].
The observed role of the septum in cardiac anatomic morphology now interfaces with its dominant role in heart physiology and function. Reproduction of prior free wall findings related to ejection and suction following positive and negative inotropic drug administration [20] confirmed the calcium related muscular components of septal action. Dopamine produced tachycardia, increased SS%, and decreased the time-delay between descending and ascending sequences of shortening. Conversely, beta blockade with esmolol decreased heart rate and SS%, increased the time-delay between the beginning of the contraction of the descending and ascending segments, and preferentially delayed relaxation of the septal endocardial descending segment to impede rapid filling.
Ventricular suction normally occurs during the duration hiatus between the end of shortening of the descending and ascending segments, so that restriction of this hiatus by a negative inotropic agent at a lower heart rate would interfere with suction. Indirect evidence of this comes from a slower down slope of the left ventricular pressure curve and less negative dP/dt [20]. The dynamic effect of less vigorous relaxation and this observation serves as a keynote into diastolic dysfunction; more filling pressure is needed when a negative inotropic influence impairs suction.
Dell’Italia [26] in 1991 observed that despite markedly different muscle mass and chamber geometry, both ventricles are bound together by remarkable spiral muscle bundles that encircle them in a complex interlacing fashion that includes the septum to form a highly interdependent functional unit. The septum, together with the left ventricle free wall, is the basic structural components of the left ventricle. The anatomic components of the right ventricle functional performance are the external free wall and internal septum [27,16]. Forces are transmitted from one ventricle to another through the septum, independently of the neural, humoral and circulatory effects [16]. These effects are immediate, on beat-to-beat basis, and they are known as ventricular interdependence [16]. The septum is a key element for ventricular interaction, now shown to have oblique fiber orientation in the helical heart model, and this central structure predominantly contributes to the normal shortening and lengthening of the ventricles during cardiac cycle that produces systolic ejection and also contributes to the early suction phase of rapid ventricular filling [28,29].
4.1 Fiber orientation and septal function
During normal cardiac action, MRI studies confirm that the predominant heart motions include twisting for clockwise ejection and reciprocal twisting for suction [5,6]. These motions are determined by fiber orientation. Sallin [30] and Ingels [24] report that ejection fraction is 60% with oblique orientation, and reduced to 30% after transverse or horizontal fiber orientation. The sonomicrometer studies evaluated maximum shortening and confirm the oblique maximum direction of fiber orientation in septum and free wall of the normal ventricle. While crystal shortening reflects local dimensional changes in only the limited region studied, this action likely underlies the transmural twisting of ventricular movements responsible for ejection and suction.
Recorded oblique septum and free wall observations differ from prior right ventricular sonomicrometer studies [20] showing a transverse, rather than oblique orientation of basal loop fibers that wrap around the right ventricle. Functional constriction or narrowing follows this horizontal orientation, accounting for isometric cardiac compression that narrows the mitral valve annulus during this pre-ejection interval [31]; a cocking or counterclockwise motion [32] event that results from the horizontal basal loops shortening during this early isometric phase of the cardiac cycle. In contrast, longitudinal septal motion predominantly contributes to ventricular shortening during ejection and lengthening during suction, so that oblique fiber direction accounts for clockwise twisting during right ventricular ejection. This differs from the left ventricle, where systolic clockwise twisting occurs in both the septum and left ventricle free wall to generate high systemic pressure during normal left-sided ejection.
Physiologic consequences follow restriction of the oblique pattern responsible for twisting to the septum, since the right-sided basal loop has predominantly transverse fibers. This transverse fiber configuration during basal loop contraction leads to circumferential compression, which adequately maintains right-sided ejection with normal pulmonary artery pressure. However, loss of septal twisting will accentuate right ventricular failure if there is pulmonary hypertension. These observations are linked to the importance of functional oblique septal fibers, and imply that the septum should be considered the ‘lion of right ventricular function’.
Fiber orientation and functional observations have clear impact during cardiac surgery, as septal hypokinesia or akinesia is a common finding [33–35] which may reflect stunning caused by impaired myocardial protection [36]. Iatrogenic loss of the septal twisting responsible for generating adequate RV function with pulmonary hypertension thereby provides insight into causative reasons for clinical difficulty in correcting right heart failure. This new knowledge about septal fiber orientation and its interaction with right heart failure may focus future investigations upon evaluation of septal function, and underscore development of protective methods that avoid this complication. Iatrogenic septal damage is consistent with Klima et al.'s [17] and Agarwal et al.'s [19] observations that define how septal paralysis impairs right heart function.
Subsequent systolic function can also be markedly altered by changes in septal anatomy that result from diastolic septum stretching. For example, distension will distort fiber orientation thereby creating a more horizontal pattern [37,16] that differs from the normal curvilinear septal shape that protrudes into the right ventricle. Septal stretch alters the normal oblique septal fiber orientation toward a more transverse or horizontal position, and then changes left and right heart hemodynamics; this event creates ‘an architectural disadvantage’ that relates to functional alterations produced by septal displacement.
This anatomic underpinning is introduced to explain ventricular interdependence, whereby myocardial factors determine how one ventricle affects the performance of the other ventricle. Bernheim [38] in 1910 described right ventricular compression after left ventricular hypertrophy, and Dexter [39] in 1956 described a ‘reverse Bernheim event’ after large atrial septal defects and right-sided volume overload. These changes in contralateral ventricular function are now amended by adverse septal stretch following left-sided dilation from ischemic [40], valvular [41], and non-ischemic cardiomyopathy [42], or from right septal stretch following extensive left ventricular decompression following LVAD insertion [43] or pulmonary hypertension [44].
|
5. Conclusions |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
The predominantly oblique architecture of septal muscle likely governs the twisting cardiac action during ventricular ejection, compared to the transverse fiber structure of basal segment of the free right ventricular wall. These normal functional events were related to reciprocal oblique fiber orientation comprising the septum, and mechanisms of impaired septal performance were suggested when pathologic lesions disrupted the fiber orientation of septal architecture. These observations further suggest that the septum might be considered the ‘lion of right ventricular function’ in patients with pulmonary hypertension.
|
References |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W.-J. Kim, B. H. Lee, Y. J. Kim, J. H. Kang, Y. J. Jung, J.-M. Song, D.-H. Kang, and J.-K. Song Apical Rotation Assessed by Speckle-Tracking Echocardiography as an Index of Global Left Ventricular Contractility Circ Cardiovasc Imaging, March 1, 2009; 2(2): 123 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. J. Cho, P. Jiamsripong, A. M. Calleja, M. S. Alharthi, E. M. McMahon, B. K. Khandheria, and M. Belohlavek Right ventricular free wall circumferential strain reflects graded elevation in acute right ventricular afterload Am J Physiol Heart Circ Physiol, February 1, 2009; 296(2): H413 - H420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Saleh, O. J. Liakopoulos, and G. D. Buckberg The septal motor of biventricular function Eur. J. Cardiothorac. Surg., April 1, 2006; 29(Suppl_1): S126 - S138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. D. Buckberg and the RESTORE Group The ventricular septum: the lion of right ventricular function, and its impact on right ventricular restoration Eur. J. Cardiothorac. Surg., April 1, 2006; 29(Suppl_1): S272 - S278. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
