| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2003;125:1438-1450
© 2003 The American Association for Thoracic Surgery
Cardiopulmonary Support and Physiology |
From Cardiovascular Surgery,a St Vincent Mercy Medical Center, Toledo, Ohio, Saint Luke's Hospital,b Maumee, Ohio; and the Department of Surgery,c Medical College of Ohio, Toledo, Ohio.
Received for publication Feb 12, 2002. Revision requested May 24, 2002; revisions received Aug 21, 2002. Accepted for publication Aug 26, 2002. Address for reprints: Robert H. Habib, PhD, Director, Cardiopulmonary Research, St Vincent Mercy Medical Center, 2213 Cherry St, ACC Building, Suite 309, Toledo, OH 43608 (E-mail: Robert_Habib{at}mhsnr.org).
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 headed by Jonas recently demonstrated, in an intravital animal model, that perfusion of the cerebral microcirculation is not impaired by high hematocrit value (30%) during cardiopulmonary bypass (CPB) compared with moderate (20%) or low (10%) hematocrit value. They also reported evidence of increasingly inadequate cerebral tissue oxygenation and greater white cell-endothelial activation as hematocrit values during CPB decreased. 1 Finally, the authors proposed that their intravital results of increased ischemia and inflammatory tissue injury with greater hemodilution explained earlier findings of improved neurological outcomes with high hematocrit. 2,3
Arguably, the above results may have significant implications to current CPB practice as they contradict prevailing cardiac surgical dogma and current recommendations of textbooks 4,5, that is, that lower hematocrit values (≤20%) minimize microcirculatory disturbances during CPB and hence will improve tissue perfusion and oxygen delivery. Additionally, if these findings in the cerebral microcirculation prevail in other capillary beds, then milder hemodilution may similarly lead to improved post-CPB outcomes of other vital organs.
In adult CPB, variations of body size and pre-CPB hematocrit, coupled with the essentially constant bypass circuit volume, inevitably lead to wide interpatient variability of the nadir on-pump hematocrit. 6 The primary objective of this study was to elucidate the effects of varying "lowest hematocrit" during CPB on postoperative organ dysfunction and patient outcomes. We therefore determined in a large contemporary cardiac surgery series (with CPB) the incidence of serious complications, operative mortality, and resource utilization in terms of varying lowest hematocrit. Also, for isolated coronary artery bypass grafting (CABG), we probed whether and how potential adverse effects of lowest hematocrit levels alter long-term outcomes.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
With approval of the St Vincent Mercy Medical Center (Toledo, Ohio) Human Investigation Committee, we analyzed the clinical outcomes of 5000 consecutive adult cardiac surgery patients from a single institution (1994-2000). Standard CPB techniques were applied with predominant (96%) use of normothermia (body temperature: 35°C to 37°C). 6 Arterial blood flow was determined on the basis of a cardiac index of 2.5 to 3.0 (L · min-1 · m-2), and mean arterial pressure was maintained at greater than 60 mm Hg. All surgeons used standardized clinical care pathway protocols.
Data Collection
Risk factors, CPB variables, complications, and operative mortality (OM; in-hospital death or <30 days out-of-hospital death) data were collected in a dedicated database. Cardiovascular intensive care unit (CVICU) and postoperative length of stays (LOS) were entered for all patients in the same database. Total direct variable costs (Cost) were compiled for patients from 1997 to 2000, and these encompassed every care-related expense throughout the admission period. 7 Cost data are presented in terms of the annual average cost to correct for inflation.
The social security death index database (http://ssdi.geneology.rootsweb.com) was queried in October 2001 for each patient via individual searches to verify accurate out-of-hospital death data. This resulted in survival follow-up data for all patients ranging from up to 8 to 93 months. Our database was updated for any missing late mortality data.
Data Analysis/Statistical Methods
Univariate analyses. Lowest hematocrit is a continuous normally distributed variable, and examining its effects on "dichotomous" outcome measures (eg, mortality [yes/no]) can hinge on how the population is subdivided. 6 Multiple steps were applied to identify outcomes associated to the extent of hemodilution on CPB. First, mean lowest hematocrit values of each complicated subgroup (eg, stroke) were compared with patients free of any complications. Next, incidence of each complication was compared among lowest hematocrit quintile groups (n = 1000 each). Finally, we derived continuous independent variable-outcome relations by consideration of rolling decile subgroups. 6 Here, patient data were first arranged in increasing lowest hematocrit order, and a total of 37 subgroups (75% overlapping ranges) were separately analyzed, such that group 1 included patients 1 to 500 (0%-10%), group 2 included patients 126 to 625 (2.5%-12.5%), and so on. Note, with this method, the smaller subgroup size minimized within-group lowest hematocrit heterogeneity while the 75% overlap smoothed the effects of similar lowest hematocrit values in consecutive subgroups.
Continuous variables were presented as mean ± standard deviation and categorical data were presented as frequency counts (N, %). Nonlinear regression analyses (SPSS, Inc, Chicago, Ill) were used to describe continuous lowest hematocrit-outcome relations. Group comparisons of continuous variables were done using unpaired t test (2 groups) or 1-way analysis of variance (3 or more groups) depending on applicability. 
2-Square tests were used for categorical variables. Long-term survival was assessed via Kaplan-Meier and death hazard analyses. 8
Multivariate analyses. We used multivariate analyses to determine if and how nadir on-pump hematocrit, among other variables, predicted certain outcomes (length of stay as well as early and late mortality). Also, we determined the pre- and intraoperative variables predicting the extent of hemodilution on CPB, or lowest hematocrit (%), via multiple linear regression analysis (SPSS). All multivariate methods were limited to the 3800 patients who underwent isolated CABG to avoid the possible varying effects on outcomes of valve and or combined operations.
First, predictors of operative mortality and perioperative morbidity were derived via multiple logistic regression (SPSS). We defined perioperative morbidity as patients who remained in excess of 48 hours in cardiovascular intensive care (CVICU > 2 days) or patients with postoperative hospital stays greater than 8 days. Cox proportional hazard models were used to define the effect of explanatory variables (including lowest hematocrit) on 0- to 6-year mortality after CABG with CPB. Model selection was first done with backward elimination and variables significant at the P < .05 level were retained in the model as independent predictors. Models were then confirmed using forward selection and stepwise selection. The considered prognostic variables in the multivariate models were the preoperative and intraoperative variables listed in Table 1 in addition to the categorical variables postoperative transfusion and use of intra-aortic balloon pump.
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
). Of these, 3800 underwent isolated CABG (76%). The lowest hematocrit value reached during CPB varied widely among patients (21.4 ± 4.2%, median = 21.2%, range: 9.1%-40.3%). Lowest hematocrit quintiles show that older age, female sex, diabetes, history of cerebrovascular accidents, preoperative renal failure, peripheral vascular disease, non-CABG, reoperation, nonelective status, and increased time on bypass were all more frequent in more hemodiluted patients (Table 1). Alternatively, ejection fraction, chronic obstructive pulmonary disease, left main disease, and three-vessel disease did not vary with this on-pump hematocrit level. Independent predictors of the nadir hematocrit determined via multivariate linear regression included four continuous (pre-CPB hematocrit value, body surface area, time on CPB, and age) and two categorical (female sex and preoperative renal failure) variables (Table 2). Pre-CPB hematocrit value (%) and body surface area (BSA) were the main determinants of the nadir hematocrit on pump as would be expected based on red blood cell mass balance considerations in conjunction with a constant volume of the CPB crystalloid prime. Alternatively the variations in lowest hematocrit value in terms of age and time on CPB are probably indirect, and reflect (1) the on-average lower pre-CPB hematocrit value (39%-35%) and BSA (2.0-1.8 m2) with increasing CABG patient age (60-80 years) and (2) greater blood loss with prolonged pump times.
|
Effects of Lowest Hematocrit on Outcomes
Operative outcomes. Variations in hemodilution levels for patients, which were divided based on incidence of postoperative complications, are summarized in Table 3. Lowest hematocrit for OM (19.0 ± 4.1%) and complicated (19.9 ± 4.0%) patients were significantly lower than for uncomplicated (21.8 ± 4.1%) patients (unpaired t test). Moreover, this trend was true for all types of complications, and more severe hemodilution was associated with greater likelihood of multiple complications.
|
fivefold), tamponade (sixfold), perioperative myocardial infarction (fourfold), cardiac arrest (fourfold), permanent stroke (sixfold), coma (sixfold), prolonged mechanical ventilation (threefold), pulmonary edema (fourfold), septicemia (threefold), use of intra-aortic balloon pump (fourfold), renal failure (fourfold), and multiorgan failure (sevenfold) were all systematically greater as lowest hematocrit value decreased. Note, adverse hemodilution effects were particularly greater for lowest hematocrit value below 22%.
|
|
|
|
|
, bottom).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
). Moreover, this hematocrit-complications association inevitably resulted in significantly and systematically greater intensive care requirements, hospital stays, operative costs, and death with increasing levels of hemodilution, particularly for lowest hematocrit value less than 22% (Figures 3 and 4). Indeed, the observed-to-expected (O/E) CABG mortality decreased systematically and substantially for patient cohorts whose lowest hematocrit was between 15% and 22% and was relatively unchanged for greater values (Figure 3, bottom right). Lowest hematocrit value (%) was also found to be an independent predictor of operative mortality and morbidity (Table 5).
|
). For example, 6-year survival was 80.5% for group I patients in whom the average lowest hematocrit value was a low 16.1%. In contrast, groups IV (lowest hematocrit value: 23.6%) and V (lowest hematocrit: 27.5%) both had a 92.3% survival 6 years after CABG. Granted, systematically younger age of patients for group I (67 years) through group V (60 years) and the fact that operative deaths were also greater in the more hemodiluted patients have contributed to the difference in survival. However, this lowest hematocrit effect on intermediate-term survival that is apparent from 0- to 6-year survival and death hazard analyses in CABG patients (Figure 4) was confirmed to be an independent predictor by multivariate Cox proportional hazard analysis (Table 6).
Other authors 9-11 have suggested a possible link between hemodilution on CPB and worse operative outcomes after CABG but, to our knowledge, no studies have previously reported an association between CPB hemodilution levels and worse longer-term outcomes. Consistent with our data, DeFoe and colleagues 9 reported increased reoperation of bleeding and low cardiac output and in-hospital mortality with increased hemodilution severity. They, however, did not find a systematic increase in stroke. Hardy and coworkers 10 reported increased postoperative renal failure, hemodynamic instability, and death with decreased hematocrit (1) on pump or (2) during the first 24 hours postoperatively. Fang and colleagues 11 used a multivariate analysis that separated low- and high-risk CABG patients and concluded that hematocrit values below 14% in low-risk patients and 17% in high-risk patients are associated with a doubling of the risk of death. A similar variance was also evident by contrasting our LOS (hematocrit and mortality) hematocrit data for young (lower-risk) versus older (higher-risk) patients (Figure 2). These showed that worse outcomes occur at relatively lower hematocrit valuesin young (20%) compared with older (22% to 24%) patients indicating a varying tolerance of hemodilution on CPB.
In this series, we report associations between hemodilution levels and a large number of complications, as well as a number of morbidity and mortality measures. An obvious reason for this is that more outcome variables were included in the analysis. We also contend that this was facilitated by the consideration of lowest hematocrit as a continuous variable as opposed to use of arbitrary hematocrit values to define patient cohorts. Also, via the overlapping deciles analyses one is able to better define the continuous relations between on-pump hemodilutional anemia and specific postoperative complications, resource use, and death.
Roach and colleagues 12 recently reported that adverse neurologic events of varying severity are unfortunately among the most frequent complications of CPB. Moreover, Newman and coworkers 13 showed a relatively high prevalence and persistence of neurocognitive deficits after CABG. To our knowledge, the substantial dependence of the incidence of stroke on lowest hematocrit is the first such reported association in adult cardiac surgery. This perhaps implicates hemodilution severity as a primary cause of CPB adverse neurologic effects. Importantly, this agrees with experimental findings of reduced cerebral oxygen delivery 1,14-19 and worse neurologic outcomes 2,3 with low hematocrit on CPB. Indeed, in case of normothermic CPB, Liam and colleagues 19 showed using a dog model that both oxygen delivery and whole body oxygen uptake were reduced when hematocrit values fell below 25%, becoming significant at 18% or below. They further concluded that (1) hematocrit levels above 18% were needed to maintain oxygen delivery and consumption, (2) the critical hematocrit value may be higher for bypass compared with nonbypass hemodilution states, and (3) the critical hematocrit value for the body appears to be higher than that required for the brain during warm bypass, presumably because the brain is more effective at increasing its flow in hemodiluted states. Interestingly, the experimental data of Liam and colleagues 19 agrees with our clinical data, which showed little or no adverse effects of hemodilution when hematocrit values did not fall below the 21% to 22% levels (Figure 2). This critical hematocrit falls within the 25% and 18% hematocrit values identified by them for adequate versus compromised oxygenation, respectively.
A number of large studies have convincingly shown that small size and female (who are small relative to males) patients have worse outcomes of CPB including greater OM. 6,9,10,20-22 We believe that the poor outcomes in these two patient subpopulations are directly linked to their relatively more severe hemodilution during CPB. In this series, OM was significantly greater in female and low-BSA patient subpopulations, both of which were characterized by significantly (P < .001) lower nadir hematocrit values on pump: (1) female compared with male patients [OM: 3.3% (39/1175), lowest hematocrit: 18.7 ± 3.9% vs 1.9% (51/2625), 23.1 ± 3.1%] and (2) small-size (BSA ≤ 1.8 m2) compared with larger-size (BSA > 1.8 m2) patients [3.9% (33/846), 18.4 ± 3.2% vs 1.9% (57/2954), 22.7 ± 3.9%].
Prolonged time on cardiopulmonary bypass (>100 minutes) and increased age are also established major determinants of increased morbidity and mortality after CABG, 23a fact confirmed in this series as well. In addition, we have found that severity of hemodilution during cardiopulmonary bypass as defined by the nadir on-pump hematocrit is a strong predictor of increased CABG morbidity (Figures 3 and 4, Table 5) and mortality (Figure 2, Table 5). Yet, the extent to which hemodilution severity alters observed operative outcomes in relation to patient age and duration of bypass is not known. We thus compared the incidence of OM and postoperative hospital stays (POLOS) in (1) age and (2) time on CPB patient decile groups (380 patients each). Here, within each decile group, OM and POLOS were contrasted for patients whose nadir hematocrit dropped to 21% or less versus those with lowest hematocrit values greater than 21% (Figure 5). Irrespective of age and duration of CPB, we found substantial and systematically increased poorer outcomes (OM and POLOS) in the more hemodiluted patients. Moreover, this hemodilution effect on OM tended to be greater in patients who are over 70 years and/or those who remained on bypass for 2 or more hours. Although it remains to be proven by prospective studies, these findings do suggest that the possibility of improved operative outcomes if hemodilution to levels below 21% is avoided during CPB and that older patients and prolonged cardiopulmonary bypass patients might benefit most from such control of on-pump hematocrit levels. Finally, the tendency of greater OM variance with hematocrit value with longer pump runs is suggestive of a dose effect of on-pump hemodilution (ie, adverse effects of the same low [<21%] nadir hematocrit may be increased as the duration of bypass increases).
If the aforementioned associations between low hematocrit levels on CPB and worse outcomes obtained by the reported univariate and multivariate analyses are shown to be causal by prospective randomized studies, then methods aimed at minimizing on-pump hemodilutional anemia (eg, lowest hematocrit ≥ 22%) might lead to significant improvements in patients at risk. Note, hematocrit value during CPB is an intraoperative variable that cardiac surgeons can potentially control so that adverse outcomes are minimized in contrast to preoperative variables over which they have little or no control and, ironically, these have been traditionally the focus of quality improvement.
Hemodilution during CPB results from the mixing of pump crystalloid and colloid prime solution with the patient's blood, and these two relative volumes, along with pre-CPB hematocrit, will largely determine the nadir hematocrit (Table 2). In that sense, lowest hematocrit value during CPB is then a potentially modifiable risk factor. Changes to several areas of practice in CPB patients can alter the extent of hemodilution experienced in a given patient. These include controlling preoperative blood loss during routine phlebotomy and cardiac catheterization; possible redesign and use of variable-size, or multiple sizes of, CPB circuits to be used according to patient BSA; minimizing of the tubing size (length and diameter) connecting the patient and pump; more timely return of collected cells to the circulating volumes; more strict control of intraoperative blood loss and fluid administration; use of retrograde autologous priming of the CPB circuit, which has been shown to reduce hemodilution and transfusion requirements 24,25; and, lastly, freer use of blood transfusions during CPB so as to maintain hematocrit at predetermined levels. The latter is perhaps the most controversial of the above recommendations. Yet, as Figure 3 shows, a large fraction of the more hemodiluted patients are transfused anyway. Moreover, at a minimum, our data raises the prospect that adverse transfusion effects in CPB patients derive, in a significant way, from the associated on-pump hemodilution as opposed to being a consequence of transfusion per se. Obviously, this question will only be resolved by appropriate prospective studies.
In conclusion, we found strong systematic associations between severity of hemodilution on CPB and serious complications affecting most major vital organs; particularly when the nadir hematocrit drops below 22%. Expectedly, these associations then lead to (1) increased resource utilization such as CVICU/hospital stays and operative costs and, more importantly, (2) substantially greater operative mortality. We also found evidence of potential long-term effects of on-pump hemodilutional injury to vital organs as manifested by the worse 0- to 6-year survival due to a sustained increase in death hazard after CPB in the more severely hemodiluted patients. We believe these data represent compelling evidence of substantially greater adverse outcomes when hematocrit value during CPB is allowed to fall below a critical value (22%) and hence warrant future prospective confirmatory investigation that should also aim to separate the correlated effects of substantial hemodilution on bypass and transfusion of blood products. We further speculate that these adverse effects might have as common mechanisms (1) ischemic vital organ tissue injury within vital organs due to inadequate oxygen delivery at low hematocrits, and (2) systemic inflammatory injury due to increased white cell-endothelial activation in microcirculatory beds of vital organs. These hemodilution effects during CPB per se have been convincingly demonstrated in recent experimental studies by the Boston group 1-3 and by Cook, Daly, and coworkers. 14-16,19
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. A.J.M. Huybregts, R. de Vroege, and W. van Oeveren A New System for Right Atrial Cooling Ann. Thorac. Surg., April 1, 2008; 85(4): 1421 - 1424. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ranucci, A. Pazzaglia, C. Bianchini, G. Bozzetti, and G. Isgro Body Size, Gender, and Transfusions as Determinants of Outcome After Coronary Operations Ann. Thorac. Surg., February 1, 2008; 85(2): 481 - 486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Karkouti, D. N. Wijeysundera, W. S. Beattie, and for the Reducing Bleeding in Cardiac Surgery (RBC) Risk Associated With Preoperative Anemia in Cardiac Surgery: A Multicenter Cohort Study Circulation, January 29, 2008; 117(4): 478 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ranucci Perioperative Renal Failure: Hypoperfusion During Cardiopulmonary Bypass? Seminars in Cardiothoracic and Vascular Anesthesia, December 1, 2007; 11(4): 265 - 268. [Abstract] [PDF]
|
|
|
|
|
|
G. J. Murphy, B. C. Reeves, C. A. Rogers, S. I.A. Rizvi, L. Culliford, and G. D. Angelini Increased Mortality, Postoperative Morbidity, and Cost After Red Blood Cell Transfusion in Patients Having Cardiac Surgery Circulation, November 27, 2007; 116(22): 2544 - 2552. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. K. Shoemaker Hemodilution Impairs Cerebral Autoregulation, Demonstrating the Complexity of Integrative Physiology Anesth. Analg., November 1, 2007; 105(5): 1179 - 1181. [Full Text] [PDF]
|
|
|
|
|
|
R. H. Habib, A. Zacharias, and T. A. Schwann Minimally Invasive Closed Circuit Versus Standard Cardiopulmonary Bypass: Is It Renoprotective? Ann. Thorac. Surg., October 1, 2007; 84(4): 1426 - 1427. [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Habib and A. Zacharias Body Size and the Early Mortality Gender Gap in Coronary Artery Bypass Grafting Surgery J. Am. Coll. Cardiol., September 11, 2007; 50(11): 1095 - 1095. [Full Text] [PDF]
|
|
|
|
|
|
K. H. Humphries, M. Gao, A. Pu, S. V. Lichtenstein, and C. R. Thompson Reply J. Am. Coll. Cardiol., September 11, 2007; 50(11): 1095 - 1096. [Full Text] [PDF]
|
|
|
|
 |
|
 F. Pappalardo, C. Corno, A. Franco, G. Giardina, A.M. Scandroglio, G. Landoni, G. Crescenzi, and A. Zangrillo Reduction of hemodilution in small adults undergoing open heart surgery: a prospective, randomized trial. Perfusion, September 1, 2007; 22(5): 317 - 322. [Abstract] [PDF]
|
|
|
|
 |
|
 G. M. T. Hare, C. D. Mazer, J. S. Hutchison, A. T. McLaren, E. Liu, A. Rassouli, J. Ai, R. E. Shaye, J. A. Lockwood, C. E. Hawkins, et al. Severe hemodilutional anemia increases cerebral tissue injury following acute neurotrauma J Appl Physiol, September 1, 2007; 103(3): 1021 - 1029. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kulier, J. Levin, R. Moser, G. Rumpold-Seitlinger, I. C. Tudor, S. A. Snyder-Ramos, P. Moehnle, D. T. Mangano, and for the Investigators of the Multicenter Study of Impact of Preoperative Anemia on Outcome in Patients Undergoing Coronary Artery Bypass Graft Surgery Circulation, July 31, 2007; 116(5): 471 - 479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Perthel, L. El-Ayoubi, A. Bendisch, J. Laas, and M. Gerigk Clinical advantages of using mini-bypass systems in terms of blood product use, postoperative bleeding and air entrainment: an in vivo clinical perspective Eur. J. Cardiothorac. Surg., June 1, 2007; 31(6): 1070 - 1075. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A.J.M. Huybregts, A. M. Morariu, G. Rakhorst, S. R. Spiegelenberg, H. W.A. Romijn, R. de Vroege, and W. van Oeveren Attenuated Renal and Intestinal Injury After Use of a Mini-Cardiopulmonary Bypass System Ann. Thorac. Surg., May 1, 2007; 83(5): 1760 - 1766. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
The Society of Thoracic Surgeons Blood Conservatio, V. A. Ferraris, S. P. Ferraris, S. P. Saha, E. A. Hessel II, C. K. Haan, B. D. Royston, C. R. Bridges, R. S.D. Higgins, G. Despotis, et al. Perioperative Blood Transfusion and Blood Conservation in Cardiac Surgery: The Society of Thoracic Surgeons and The Society of Cardiovascular Anesthesiologists Clinical Practice Guideline Ann. Thorac. Surg., May 1, 2007; 83(5_Supplement): S27 - S86. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ranucci, C. Bellucci, D. Conti, A. Cazzaniga, and B. Maugeri Determinants of Early Discharge From the Intensive Care Unit After Cardiac Operations Ann. Thorac. Surg., March 1, 2007; 83(3): 1089 - 1095. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
