The nadir hematocrit (HCT) on cardiopulmonary bypass (CPB) is a recognized independent risk factor for major morbidity and mortality in cardiac surgery. The main interpretation is that low levels of HCT on CPB result in a poor oxygen delivery and dysoxia of end organs. Hyperlactatemia (HL) is a marker of dysoxic metabolism, and is associated with bad outcomes in cardiac surgery. This study explores the relationship between nadir HCT on CPB and early postoperative HL.
Measurements and Main Results
Nadir HCT on CPB and other potential confounders were explored for association with blood lactate levels at the arrival in the Intensive Care Unit (ICU), and with the presence of moderate (2.1 – 6.0 mMol/L) or severe (> 6.0 mMol/L) HL. Nadir HCT on CPB demonstrated a significant negative association with blood lactate levels at the arrival in the ICU. After adjustment for the other confounders, the nadir HCT on CPB remained independently associated with moderate (odds ratio 0.96, 95% confidence interval 0.94-0.99) and severe HL (odds ratio 0.91, 95% confidence interval 0.86-0.97). Moderate and severe HL were significantly associated with increased morbidity and mortality.
Citation: Ranucci M, Carboni G, Cotza M, Bianchi P, Di Dedda U, Aloisio T, et al. (2015) Hemodilution on Cardiopulmonary Bypass as a Determinant of Early Postoperative Hyperlactatemia. PLoS ONE 10(5): e0126939. https://doi.org/10.1371/journal.pone.0126939
Academic Editor: Chiara Lazzeri, Azienda Ospedaliero-Universitaria Careggi, ITALY
Received: February 23, 2015; Accepted: April 9, 2015; Published: May 18, 2015
Copyright: © 2015 Ranucci et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: Data are stored in the institutional database of the IRCCS Policlinico San Donato. The dataset on which the analysis is based is an SSPS file that is available on request to the Local Ethics Committee of the IRCCS Policlinico San Donato, due to ethical restrictions and patient confidentiality requirements. To request the data, please contact: Ms Elisabetta Riva: email@example.com or Marco Ranucci (author): firstname.lastname@example.org.
Funding: This work was funded by internal research funds of the IRCCS Policlinico San Donato - Research Line: Anesthesia, ICU, and extracorporeal technologies.
Competing interests: The authors have declared that no competing interests exist.
The nadir hematocrit (HCT) on cardiopulmonary bypass (CPB) is widely recognized as a risk factor for postoperative acute kidney injury [1–4], stroke [5,6], and mortality . The commonly accepted interpretation for this association is that running CPB at a very low HCT value may determine a poor oxygen delivery, leading to end organ dysoxia and consequent organ failure [7–9].
A prolonged condition of poor oxygen delivery triggers the anaerobic energy production, leading to increased levels of blood lactates . Hyperlactatemia (HL) during CPB is associated with a low oxygen delivery  and with bad outcomes in adult  and pediatric patients . HL immediately following heart surgery is a marker of an impaired hemodynamic condition and is associated with an increased morbidity and mortality [14–16].
It is therefore reasonable to hypothesize that the nadir HCT on CPB may be a determinant of early postoperative HL through the exposure of the patient to a poor oxygen delivery during CPB, consequently linking hemodilution, inadequate oxygen delivery, and early postoperative HL to major morbidity and mortality.
Presently, there is a gap of knowledge with respect to the association of the nadir HCT on CPB with early postoperative HL. The present study aims to verify the hypothesis that low values of nadir HCT on CPB are associated with increased levels of blood lactates at the arrival in the ICU, therefore confirming the dysoxic interpretation of the link between nadir HCT on CPB and bad outcomes.
This is a retrospective study, based on our institutional database. The Local Ethics Committee (IRCCS San Raffaele Hospital) approved this study, waiveing the need for an informed consent from the patients. At the hospital admission, all the patients gave written approval to the treatment of their data in an anonymous form and for scientific purposes.
We analyzed data routinely collected in our institutional database from January 1st, 2010, through December 31st, 2013. Routine inclusion of blood gas analysis data at the arrival in ICU (including blood lactate values) was available in this period. The database includes all the patients receiving a cardiac surgery operation, with the exclusion of transplant operations (not performed at our Institution). The initial patient population included 5,645 patients. Patients aged < 18 years and patients receiving an off-pump procedure were excluded from the study, reaching a study population of 3,851 patients.
Data collection and definitions
For each patient, the following data were collected and available:
Preoperative: demographics; left ventricular ejection fraction (%); preoperative HCT (%); congestive heart failure; cardiogenic shock; active endocarditis; unstable angina; preoperative intra-aortic balloon pump; serum creatinine value (mg/dL); serum bilirubin value (mg/dL); chronic dialysis; chronic obstructive pulmonary disease; diabetes (on medication); previous cerebrovascular accident; previous cardiac surgery; non-elective procedures. Operative: type of operation (other than isolated coronary surgery; combined surgery [coronary + valve surgery or double/triple valve surgery], mitral valve surgery; aortic valve surgery; ascending aorta surgery); cardiopulmonary bypass (CPB) duration (minutes); nadir HCT on CPB; nadir temperature (°C) on CPB. Postoperative: blood lactate values (mMol/L) at the arrival in the ICU; stroke; acute kidney injury (peak postoperative serum creatinine double the baseline value); bloodstream infections (with positive cultures); operative (in-hospital or within 30 days after discharge) mortality.
For the purposes of the present analysis, patients were attributed to three groups according to the blood lactate values at the arrival in the ICU: no HL (lactate value ≤ 2.0 mMol/L), moderate HL (lactate value 2.1–6.0 mMol/L) and severe HL (lactate value > 6.0 mMol/L).
Surgery and CPB
Patients were generally treated under moderate hypothermia (32°C—34°C) unless for specific procedures. Roller or centrifugal pumps were used, the CPB circuit was primed with colloid solutions at variable volumes, ranging from 800 to 1,200 mL. Pump flow was set between 2.0 and 2.8 L. min-1 m-2, and adjusted according to the temperature and the HCT value. The heart was arrested using antegrade cold crystalloid cardioplegia or cold blood cardioplegia according to the surgeon’s preference. Anticoagulation was achieved with unfractionated heparin according to our standard protocols (loading doses 300 IU/kg to reach a target activated clotting time of 450–480 seconds; additional doses of 80 IU to maintain this value), and heparin reversal was achieved with adequate doses of protamine sulfate. All the patients received tranexamic acid at a dose of 15 mg/kg before CPB and 15 mg/kg after protamine administration. Cell-saver was used during the operation in selected cases.
All data are presented as number with percentage for categorical variables, mean with standard deviation for normally distributed continuous variables, and median with interquartile range for continuous, non-normally distributed variables. Normality of distribution was checked with the Kolgomorov-Smirnov test.
The association between continuous variables was tested using polynomial regression analyses, testing different equations, and the best model was identified based on the R2 value.
Differences between groups were tested with the Pearson’s Chi square test, the Student’s t test, and the Mann-Whitney test when appropriate.
The variables being significantly associated with moderate and severe HL were entered into two multivariable logistic regression analyses (one for moderate HL and one for severe HL), producing odds ratios with 95% confidence interval. A maximum of one variable per each 10 events was admitted to the model; however, due to the expected large number of HL events, no over-fitting of the model was anticipated. Multi-collinearity among the independent variables in multivariable logistic regression analyses was checked using collinearity statistics with measurement of condition indices and Eigenvalues. A condition index greater than 30 was considered indicative for multi-collinearity.
All tests were two-sided. A p-value < 0.05 was considered significant for all statistical tests. Statistical calculations were performed using a computerized statistical program (SPSS 13.0, Chicago, IL).
Demographics and general characteristics of the patient population are depicted in Table 1. Overall, moderate HL at the arrival in the ICU was observed in 837 (21.7%) patients and severe HL in 153 (4.0%). The values of nadir HCT on CPB were 27.4±3.9 in the no-HL group, 26.5±4.1 in the moderate HL group, and 25.4±4.3 in the severe HL group (P = 0.001 for between-groups difference).
The univariate association between nadir HCT on CPB and levels of lactates at the arrival in the ICU was defined by a quadratic equation (Fig 1) with higher values of lactates observed for lower values of nadir HCT.
Dashed lines: 95% confidence interval.
Among the variables considered, 19 factors were significantly associated with moderate HL, severe HL, or both (Table 2).
Factors being associated with moderate or severe HL at the univariate analysis were entered into two distinct multivariable logistic regression analyses having moderate or severe HL as dependent variables. CPB duration was not included in the model, due to severe multi-collinearity with the other independent variables (ejection fraction, serum creatinine, congestive heart failure, preoperative IABP, active endocarditis, redo surgery, non-elective surgery, non-isolated coronary surgery, nadir temperature on CPB, and nadir HCT on CPB) with a condition index of 84.9.
After correction for the potential confounders (Table 3), the nadir HCT on CPB remained an independent risk factor for both moderate and severe HL. The relative risk of HL increased by 4% per each unit decrease of the nadir HCT on CPB for moderate HL, and by 9% for severe HL.
The unadjusted relationship between nadir HCT on CPB and moderate or severe HL is defined by logistic regressions shown in Fig 2. Within the considered range of nadir HCT values, the relationship is almost linear for moderate HL, with an absolute risk of moderate HL of 25% at a nadir HCT on CPB of 25%, increasing to 31% for a nadir HCT value of 20% (relative risk increase 20%). Conversely, the absolute risk of severe HL is 4.3% at a nadir HCT on CPB of 25%, increasing to 8.7% at a nadir HCT on CPB of 20% (relative risk increase 100%).
Dashed lines: 95% confidence interval.
The clinical outcome of the patients according to the presence of moderate or severe HL is shown in Table 4. Major morbidity and operative mortality were significantly worse in both moderate and severe HL groups.
This study demonstrates, in a large series of patients who underwent cardiac surgery with CPB, that the nadir HCT on CPB is independently associated with both moderate and severe HL. The impact of the nadir HCT on CPB as a risk factor for HL is more pronounced for severe HL: the same decrease of nadir HCT on CPB induces an increase in the relative risk of HL that is 4 times higher for severe HL than for moderate HL.
HL is the consequence of lactic acidosis, through the buffering of lactic acid by anions. There are two different species of L-lactic acidosis. Type A recognizes a hypoxic nature, is a marker of shock, and is found during septic shock, mesenteric ischemia, hypoxemia, hypovolemic or cardiogenic shock, poisoning by carbon monoxide or cyanide . Type B is non-hypoxic and may be related to medications, thiamine deficiency, and intoxications. In the setting of cardiac surgery, type A is the dominant pattern [14–18]. However, the temporal aspects of postoperative HL are important to correctly interpret this finding. HL during CPB is generally associated with the exposure to an inadequate oxygen delivery, in particular during the rewarming phase [12,13]. HL at the arrival in the ICU (otherwise defined “early HL”) may reflect an inadequate oxygen delivery during CPB or after weaning from CPB, or both. It is difficult or even impossible, with an isolated blood lactate measure, to draw conclusions about the exact onset of HL, because in presence of an organ dysoxia blood lactate formation is rapidly triggered; however, lactate clearance takes more time, is a liver-dependent phenomenon, and may be hampered itself by a condition of low hepatic blood flow . In our study, we measured blood lactate at the arrival in the ICU and we addressed moderate and severe early HL. It is likely that some patients may have started lactate formation during CPB, while others may not. However, the finding that the nadir HCT on CPB is independently associated especially with severe HD is suggestive for a hemodilution-related (low oxygen content) type A HL at least in a percentage of our patients. Conversely, other patients may have faced a condition of low cardiac output only once weaned from CPB, leading to a different kind of type A HL (low oxygen transport).
There are few studies linking oxygen delivery on CPB with early postoperative HL. However, Demers and associates  could notice that low levels of hemoglobin on CPB were associated with the onset of HL during CPB. More importantly, Abraham and associates  observed that postoperative HL in children undergoing atrial septal repair was associated with lower rates of pump flow and oxygen delivery.
The notion that severe hemodilution on CPB may be deleterious, and that every effort should be applied to avoid it, is nowadays generally recognized. However it is still unclear whether the relationship between the nadir HCT on CPB and bad outcomes is directly causative or rather an epiphenomenon, a reflection of other pathological conditions (like preoperative anemia), or even simply the trigger for allogeneic blood product transfusions-related adverse effects. The present study suggests, in a large series of patients, that even after correction for potential confounders hemodilution on CPB accounts for a certain percentage of patients experiencing moderate or severe early postoperative HL. The dysoxic chain represented by hemodilution / poor oxygen delivery / organ ischemia / hyperlactatemia / bad outcomes could explain many of the complications observed in presence of very low values of HCT on CPB.
This study has limitations. The most important is the absence of additional available data during CPB, like the pump flow and the oxygen delivery. This would have been useful to better elucidate the link between oxygen delivery and postoperative HL. As a matter of fact, it is our routine policy is to adjust the pump flow according to the HCT in order to maintain an adequate oxygen delivery. Our data suggest that, despite this strategy, very low levels of HCT are determinants of anaerobic production of lactates. A second potentially useful lacking datum is the time of exposure to the nadir HCT while on CPB. If the hemodilution is leading to a poor oxygen delivery, then the longer is the exposure, the more critical will be the organ ischemia and the higher will be the value of postoperative lactates. Further studies are certainly warranted to elucidate the role of time within the context of the relationship between hemodilution, HL, and outcome.
In conclusion, our study confirms the role of early postoperative HL as a predictive marker of bad outcomes in cardiac surgery, and highlights the role of hemodilution on CPB as an independent determinant of moderate and severe early postoperative HL. Hemodilution on CPB is far from being the only or the major factor leading to postoperative HL: other mechanisms, occurring before, during, and after CPB may be advocated as determinants of type A postoperative HL. However, in order to limit the modifiable mechanism linking hemodilution, HL, and bad outcomes, possible interventions include (i) the prevention of severe hemodilution by reduction of the CPB circuit priming volume and (ii) the increase of pump flow to compensate severe hemodilution and guarantee an adequate oxygen delivery. Further prospective studies exploring these strategies are needed.
Marco Ranucci declares that he had full access to all the data in the study and that he takes responsibility for the integrity of the data and the accuracy of the data analysis, including and especially any adverse effects. Marco Ranucci assumes full responsibility for the integrity of the submission as a whole, from inception to published article.
Marco Ranucci did the statistical analysis and wrote the draft paper; Giovanni Carboni and Mauro Cotza were responsible for the study design, data interpretation, and article supervision; Paolo Bianchi, Tommaso Aloisio and Umberto Di Dedda contributed to data analysis and interpretation, and to the manuscript preparation.
The Surgical and Clinical Outcome Research (SCORE) Group is composed, for this study, by Marco Ranucci, Giovanni Carboni, Mauro Cotza, Paolo Bianchi, Umberto Di Dedda and Tommaso Aloisio.
Conceived and designed the experiments: MR GC MC. Performed the experiments: MR GC MC. Analyzed the data: MR PB TA UDD. Wrote the paper: MR MC GC.
- 1. Ranucci M, Pavesi M, Mazza E, Bertucci C, Frigiola A, Menicanti L, et al. Risk factors for renal dysfunction after coronary surgery: the role of cardiopulmonary bypass technique. Perfusion. 1994; 9: 319–326. pmid:7833539
- 2. Fang WC, Helm RE, Krieger KH, Rosengart TK, DuBois WJ, Sason C, et al. Impact of minimum hematocrit during cardiopulmonary bypass on mortality in patients undergoing coronary artery surgery. Circulation. 1997; 96 (suppl II): 194–199. pmid:9386097
- 3. Swaminathan M, Phillips-Bute BG, Conlon PJ, Smith PK, Newman MF, Stafford-Smith M. The association of lowest hematocrit during cardiopulmonary bypass with acute renal injury after coronary artery bypass surgery. Ann Thorac Surg. 2003; 76: 784–792. pmid:12963200
- 4. Karkouti K, Beattie WS, Wijeysundera DN, Rao V, Chan C, Dattilo KM, et al. Hemodilution during cardiopulmonary bypass is an independent risk factor for acute renal failure in adult cardiac surgery. J Thorac Cardiovasc Surg. 2005; 129:391–400. pmid:15678051
- 5. Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ, Shah A. Adverse effects of low hematocrit during cardiopulmonary bypass in the adult: should current practice be changed? J Thorac Cardiovasc Surg. 2003; 125:1438–1450. pmid:12830066
- 6. Karkouti K, Djaiani G, Borger MA, Beattie WS, Fedorko L, Wijeysundera D, et al. Low hematocrit during cardiopulmonary bypass is associated with increased risk of perioperative stroke in cardiac surgery. Ann Thorac Surg. 2005; 80: 1381–1387. pmid:16181875
- 7. Ranucci M, Romitti F, Isgrò G, Cotza M, Brozzi S, Boncilli A, et al. Oxygen delivery during cardiopulmonary bypass and acute renal failure after coronary operations. Ann Thorac Surg. 2005; 80: 2213–2220. pmid:16305874
- 8. de Somer F, Mulholland JW, Bryan MR, Aloisio T, Van Nooten GJ, Ranucci M. O2 delivery and CO2 production during cardiopulmonary bypass as determinants of acute kidney injury: time for a goal-directed perfusion management? Crit Care. 2011; 15: R192. pmid:21831302
- 9. Bennett MJ, Rajakaruna C, Bazerbashi S, Webb G, Gomez-Cano M, Lloyd C. Oxygen delivery during cardiopulmonary bypass (and renal outcome) using two systems of extracorporeal circulation: a retrospective review. Interact Cardiovasc Thorac Surg. 2013; 16:760–764. pmid:23427316
- 10. Berend K, de Vries AP, Gans RO. Physiological approach to assessment of acid-base disturbances. N Engl J Med. 2014; 371:1434–1445. pmid:25295502
- 11. Ranucci M, Isgrò G, Romitti F, Mele S, Biagioli B, Giomarelli P. Anaerobic metabolism during cardiopulmonary bypass: predictive value of carbon dioxide derived parameters. Ann Thorac Surg. 2006; 81: 2189–2195. pmid:16731152
- 12. Demers P, Elkouri S, Martineau R, Couturier A, Cartier R. Outcome with high blood lactate levels during cardiopulmonary bypass in adult cardiac surgery. Ann Thorac Surg. 2000; 70: 2082–2086. pmid:11156124
- 13. Ranucci M, Isgrò G, Carlucci C, De La Torre T, Enginoli S, Frigiola A. Central venous oxygen saturation and blood lactate levels during cardiopulmonary bypass are associated with outcome after pediatric cardiac surgery. Crit Care. 2010; 14:R149. pmid:20684758
- 14. Maillet J-M, Le Besnerais P, Cantoni M, Nataf P, Ruffenach A, Lessana A, et al. Frequency, risk factors, and outcome of hyperlactatemia after cardiac surgery. Chest. 2003; 123: 1361–1366. pmid:12740248
- 15. Ranucci M, Ballotta A, Castelvecchio S, Baryshnikova E, Brozzi S, Boncilli A. Intensive care unit admission parameters improve the accuracy of operative mortality predictive models in cardiac surgery. PLoS One 2010; 5:e13551. pmid:21042411
- 16. Hajjar LA, Almeida JP, Fukushima JT, Rhodes A, Vincent JL, Osawa EA, et al. High lactate levels are predictors of major complications after cardiac surgery. J Thorac Cardiovasc Surg. 2013; 146:455–460. pmid:23507124
- 17. Boldt J, Piper S, Murray P, Lehmann A. Severe lactic acidosis after cardiac surgery: sign of perfusion deficits. J Cardiothorac Vasc Anesth 1999; 13: 220–224. pmid:10230962
- 18. Raper RF, Cameron G, Walker D, Bowey CJ. Type B lactic acidosis following cardiopulmonary bypass. Crit Care Med. 1997; 25: 46–51. pmid:8989175
- 19. Takala J, Uusaro A, Parviainen I, Ruokonen E. Lactate metabolism and regional lactate exchange after cardiac surgery. New Horiz. 1996; 4: 483–492. pmid:8968981
- 20. Abraham BP, Prodhan P, Jaquiss RD, Bhutta AT, Gossett JM, Imamura M, et al. Cardiopulmonary bypass flow rate: A risk factor for hyperlactatemia after surgical repair of secundum atrial septal defect in children. J Thorac Cardiovasc Surg. 2010; 139:170–173. pmid:20106363