You are on page 1of 6

Acta Anaesthesiol Scand 2012; 56: 712717 Printed in Singapore.

All rights reserved

2012 The Authors Acta Anaesthesiologica Scandinavica 2012 The Acta Anaesthesiologica Scandinavica Foundation ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/j.1399-6576.2012.02717.x

Positive uid balance is associated with reduced survival in critically ill patients with cancer
J. P. de Almeida1, H. Palomba2, F. R. B. G. Galas1, J. T. Fukushima1, F. A. Duarte1, D. Nagaoka1, V. Torres2, L. Yu2, J.-L. Vincent3, J. O. C. Auler Jr1 and L. A. Hajjar1
Cancer Institute, Department of Anesthesiology and Critical Care, Intensive Care Unit, School of Medicine, University of Sao Paulo, Sao Paulo, Brazil, 2Cancer Institute, Department of Nephrology, Intensive Care Unit, School of Medicine, University of Sao Paulo, Sao Paulo, Brazil and 3Department of Intensive Care, Erasme Hospital, Universit Libre de Bruxelles, Brussels, Belgium
1

Background: There are no studies that describe the impact of the cumulative uid balance on the outcomes of cancer patients admitted to intensive care units ICUs. The aim of our study was to evaluate the relationship between uid balance and clinical outcomes in these patients. Method: One hundred twenty-two cancer patients were prospectively evaluated for survival during a 30-day period. Univariate (Chi-square, t-test, MannWhitney) and multiple logistic regression analyses were used to identify the admission parameters associated with mortality. Results: The mean cumulative uid balance was signicantly higher in non-survivors than in survivors [1675 ml/24 h (471 2921) vs. 887 ml/24 h (104557), P = 0.017]. We used the area under the curve and the intersection of the sensibility and specicity curves to dene a cumulative uid balance value of 1100 ml/24 h. This value was used in the univariate model. In the multivariate model, the following variables were signicantly

associated with mortality in cancer patients: the Acute Physiology and Chronic Health Evaluation II score at admission [Odds ratio (OR) 1.15; 95% condence interval (CI) (1.051.26), P = 0.003], the Lung Injury Score at admission [OR 2.23; 95% CI (1.293.87), P = 0.004] and a positive uid balance higher than 1100 ml/24 h at ICU [OR 5.14; 95% CI (1.4518.24), P = 0.011]. Conclusion: A cumulative positive uid balance higher than 1100 ml/24 h was independently associated with mortality in patients with cancer. These ndings highlight the importance of improving the evaluation of these patients volemic state and indicate that dened goals should be used to guide uid therapy.
Accepted for publication 11 April 2012 2012 The Authors Acta Anaesthesiologica Scandinavica 2012 The Acta Anaesthesiologica Scandinavica Foundation

he number of cancer patients admitted to intensive care units (ICUs) has increased worldwide. The characteristics and predictors of mortality for these patients have been investigated to identify the benets of intensive care treatment. These investigations allow health-care providers to properly dene and allocate nancial and human resources.13 The prognostic models used for general ICU patients, such as the Acute Physiology and Chronic Health Evaluation (APACHE) II or III, the Simplied Acute Physiology Score (SAPS) II

Author contributions to the manuscript: Ludhmila Hajjar: study design and drafting of the paper. Juliano Almeida, Henrique Palomba: data collection, study design, drafting of the paper. Julia Fukushima: statistical analysis, review of the paper. Danielle Nagaoka, Felipe Duarte, Vernica Torres: data collection. Jean-Louis Vincent: review of the paper, technical support. Luis Yu, Filomena Galas: study design, review of the paper.

and the Mortality Probability Models II, underestimate the risk of dying for cancer patients admitted to ICUs.4 In a prospective study, Lecuyer et al.1 demonstrated that physiologic variables and the number of failed organs at ICU admission were more accurate than cancer characteristics for predicting survival. Additional organ dysfunction and the need for mechanical ventilation, dialysis and vasopressors after 3 days of treatment in the ICU were associated with 100% mortality. In general ICU patients, a positive uid balance has been linked to a higher incidence of complications, such as pulmonary congestion, cardiovascular overload, gastrointestinal dysmotility, coagulation disturbances, immunological dysfunction and increased mortality.57 In critical patients with cancer, uid overload can be more common because of an excess of prescribed uids related

712
bs_bs_banner

Fluid balance and outcome in cancer patients

to chemotherapy, hypercalcaemia treatment and tumour lysis prevention.8 Fluid overload may worsen lung functioning because of hypoalbuminaemia and increased lung permeability.8 We hypothesized that a positive mean cumulative uid balance in the rst 72 h of an ICU stay predicts ICU mortality in critical patients with cancer.

Material and methods


Following approval by the institutional Medical Ethics Committee, this prospective observational study was performed at the medical-surgical ICU of the Cancer Institute do Estado de Sao Paulo, an oncology reference hospital of the University of Sao Paulo. The closed 58-bed ICU admits 1500 to 2000 patients per year, approximately 60% of which are post-operative and 40% of which are medical patients. There is a global ICU mortality rate of approximately 25%.9 Between July and September 2009, 122 patients were included in this study. Written informed consent was obtained from all patients who participated in the study. We included patients with cancer who were older than 18 years and were admitted to the ICU. Exclusion criteria were palliative care, haemorrhagic shock and end-stage renal disease (i.e., patients who required chronic renal replacement therapy), patients participating in other research and patients with an expected death within 24 h after ICU admission. Baseline characteristics, such as demographic data and the reason for admission, were recorded. The APACHE II was calculated based on the worst clinical and laboratory ndings in the rst 24 h.10 Cancer was classied as a solid tumour or haematologic neoplasm. Information about the cancer status such as the presence of metastasis, the Karnofsky performance status scale results11 and the chemotherapy or radiotherapy treatments performed in the 4 weeks before ICU admission were recorded. The mean uid balance was calculated as the arithmetic mean of the daily uid balance in the rst 72 h of each ICU stay. The uids that were considered included packed red blood cells, starch, albumin, crystalloids and tube feeds. The total uid received, uid losses and diuresis were recorded. Laboratory ndings such as blood gas analyses, lactate levels, haemoglobin levels, white blood cell and platelet counts, creatinine and blood urea nitrogen were recorded during the ICU stay.

Organ failure was assessed daily using the Multiple Organ Dysfunction Scores.12 Sepsis was dened by standard criteria.13 Acute renal failure was dened as an Acute Kidney Injury Network14 classication of Stage 1 or higher. The Lung Injury Score15 (LIS) was used daily to evaluate lung injury. The arterial oxygen pressure to inspiratory oxygen fraction ratio and the need for mechanical ventilation were recorded. In addition, the need for vasopressors, inotropic support, renal replacement therapy, the length of ICU and hospital stay, and mortality were recorded. Two groups were dened according to ICU mortality: survivors and nonsurvivors.

Statistical analysis
We compared baseline characteristics, follow-up measures and clinical outcomes between the groups (survivors and non-survivors). Continuous variables were compared using t-tests or MannWhitney U-tests, and categorical variables were compared using Pearsons Chi-square, Fishers exact or likelihood ratio tests. The normality of the data was assessed with the KolmogorovSmirnov test. The results are expressed as the means with 95% condence intervals (CIs) or medians with interquartile ranges. A stepwise multiple logistic regression analysis was performed to estimate the predictive factors for ICU mortality, including risk factors that were rst estimated in the univariate analysis (P < 0.10). A two-sided P-value less than 0.05 was considered statistically signicant. Statistical analyses were performed using SPSS Version 18.0 (SPSS Inc., Chicago, IL, USA).

Results
Three hundred thirty-seven cancer patients admitted to the ICU were assessed for eligibility (Fig. 1). After exclusions for various medical reasons and lack of consent, 122 patients were enrolled in the study. The main causes of ICU admission were severe sepsis or septic shock and post-operative patients of major oncologic surgery (Table 1). Twenty-ve patients died during the study period (20.5%). Compared with survivors, non-survivors were more severely ill at admission, had a higher prevalence of metastatic disease and had a lower performance status score. Non-survivors were also mainly admitted for medical causes, had a higher LIS and more often required mechanical ventilation (Table 1).

713

J. P. Almeida et al.
337 patients assessed

215 excluded 88 discharged within 24 h 37 enrolled in other studies 26 receiving palliative care 24 refused consent 22 died within 24 h 12 had haemorrhagic shock 6 underwent renal replacement therapy 122 patients enrolled

97 survivors

25 non-survivors

Fig. 1. Study ow.

The mean uid balance was signicantly higher in non-survivors than in survivors [1675 (4712921) vs. 887 (1041557) ml/24 h, P = 0.017; Table 2]. We used the area under the curve and the intersection of the sensibility and specicity curves to dene an accumulated uid balance value (1100 ml/24 h) for the univariate model (Fig. 2). For all patients, the prescribed uid was crystalloid solution (lactated Ringers solution or 0.9% saline). In the multivariate model, the following variables were signicantly associated with ICU mortality in cancer patients: the APACHE II score at ICU admission [(OR (1.15); 95% CI (1.051.26), P = 0.003], the LIS at ICU admission [OR (2.23); 95% CI (1.293.87), P = 0.004] and a positive uid balance that was higher than 1100 ml/24 h at ICU admission [OR (5.14); 95% CI (1.4518.24), P = 0.011; Table 3].

ICU outcomes
Compared with survivors, non-survivors presented a higher need for vasopressor agents [34 (35.1%) vs. 19 (76%), P < 0.001] and renal replacement therapy [5 (20%) vs. 1 (1%), P < 0.001], and non-survivors had a higher incidence of acute renal failure [33 (34%) vs. 18 (72%), P < 0.001]. The need for invasive mechanical ventilation was also higher among nonsurvivors [14 (56%) vs. 5 (5.2%), P < 0.001], and nonsurvivors presented an increased length of ICU stay [8 (413) vs. 5 (47) days, P = 0.015; Table 4].

Discussion
Our study reports predictive factors for critically ill cancer patients admitted to the ICU. The mean uid

balance within 72 h of ICU admission was independently associated with ICU mortality in patients with cancer. These ndings suggest that a liberal uid replacement strategy may be detrimental to critically ill patients with cancer. An approach that focuses on guiding uid therapy and avoiding overhydration may improve the clinical outcome for these patients, particularly for patients with acute lung injury. Compared with survivors, nonsurvivors presented greater acute lung injury severity, which was characterized by a higher LIS score at ICU admission. The higher uid retention in the non-survivor group may have contributed to the worsening of respiratory function and may justify the higher incidence of invasive mechanical ventilation in this group compared with the survivor group. The detrimental consequences of a positive uid balance on acute lung injury (ALI) or acute respiratory distress syndrome (ARDS) have been demonstrated in previous studies.5,1619 In patients with severe ALI/ARDS, a restrictive uid strategy is associated with better outcomes, such as improvements in oxygenation, a reduced duration of mechanical ventilation and a shorter length of ICU stay.19 Therefore, in the present study, a positive uid balance may have been a determinant of ALI in these patients, which can lead to hypoxemia, organ failure and death. In this observational study, it was not possible to determine whether the positive uid balance in non-survivors was the cause or result of greater illness severity. However, the nding of an independent association between uid balance and mortality highlights concerns about suitable uid management in these patients. The best uid replacement strategy may require an individualized approach based on an effective haemodynamic assessment and an effort to minimize pulmonary oedema, particularly in patients with established ALI. Patients who develop ALI/ARDS without shock or renal failure will likely benet from a restrictive uid strategy with a zero or negative uid balance. On the contrary, patients who are haemodynamically unstable will likely benet from a treatment based on haemodynamic goals, vasoactive agents and uid replacement.17,2024 The amount of uid the patient tolerates may depend on the severity of the ALI, which is described by the LIS. In other words, patients with higher LIS scores will most likely tolerate lower uid balances than patients with lower LIS scores.

714

Fluid balance and outcome in cancer patients


Table 1
Baseline characteristics of survivor and non-survivor patients. Variables Male* Age, years APACHE II MODS Type of neoplasm* Solid Haematologic Karnofsky scale Metastasis* Chemotherapy in the last 30 days Cause of ICU admission Sepsis* Respiratory failure* Neurological Cardiovascular* Surgical* Acute renal failure at ICU admission Lactate levels at admission (mM) Baseline serum creatinine Mechanical ventilation at ICU admission Vasopressor at ICU admission LIS at admission All patients (n = 122) 68 63 14 1 106 16 70 43 20 47 35 3 29 61 22 2 0.84 6 44 0.5 (55.7%) (6165) (1020) (03) (86.9%) (13.1%) (70100) (35.2%) (16.3%) (38.5%) (28.7%) (2.5%) (23.8%) (50.0%) (18.0%) (13) (0.661.06) (4.9%) (36.1%) (01) Survivors (n = 97) 57 63 13 1 87 10 80 28 13 31 21 2 20 58 15 2 0.84 2 32 0 (58.8%) (6065) (1018) (02) (89.7%) (10.3%) (70100) (28.9%) (13.4%) (32.0%) (21.6%) (2.1%) (20.6%) (59.8%) (15.5%) (13) (0.661.03) (2.1%) (33.0%) (01) Non-survivors (n = 25) 11 63 19 3 19 6 70 15 7 16 14 1 9 3 7 3 0.88 4 12 1 (44.0%) (5967) (1724) (16) (76.0%) (24.0%) (5078) (60.0%) (28%) (64.0%) (56.0%) (4.0%) (36.0%) (12.0%) (28.0%) (23) (0.61.26) (16.0%) (48.0%) (03) P 0.185 0.850 < 0.001 < 0.001 0.095 0.026 0.004 0.125 0.003 0.001 0.501 0.107 < 0.001 0.154 0.063 0.091 0.016 0.163 0.010

*n (%), Pearsons Chi-square test. Mean (95% condence interval), Students t-test. Median (interquartile range), MannWhitney test. n (%), Fishers exact test. APACHE, Acute Physiology and Chronic Health Evaluation; MODS, Multiple Organ Dysfunction Scores; LIS, Lung Injury Score; ICU, intensive care unit.

Table 2
Mean daily uid balance in the rst 72 h after intensive care unit admission between survivor and non-survivor patients. Variables Mean uid balance (ml/24 h)* < 1100 ml/24 h > 1100 ml/24 h All patients (n = 122) 956 (1471712) 69 (56.6%) 53 (43.4%) Survivors (n = 97) 887 (1041557) 60 (61.9%) 37 (38.1%) Non-survivors (n = 25) 1675 (4712921) 9 (36.0%) 16 (64.0%) P 0.017 0.020

*Median (interquartile range), MannWhitney test.

Table 3
Mortality predictors in cancer patients admitted to the intensive care unit: stepwise multiple logistic regression. Variable APACHE II LIS score Mean uid balance (> 1100 ml/24 h) Odds ratio (95% CI) 1.15 (1.051.26) 2.23 (1.293.87) 5.14 (1.4518.24) P 0.003 0.004 0.011

CI, condence intervals; APACHE, Acute Physiology and Chronic Health Evaluation; LIS, Lung Injury Score.

We also found that higher APACHE II and LIS were associated with ICU mortality. Previous studies identied the severity of acute illness at ICU admission and the presence of ALI as predictors of

negative outcomes in patients with cancer. In a prospective study with 28 ICUs and 717 patients, Soares et al.2 reported that the severity of organ failure at ICU admission, a low performance status and the need for mechanical ventilation were independently associated with mortality. In a substudy of a prospective multi-centre cohort with 3147 patients (of which 473 had malignancies), Taccone et al.3 reported that higher SAPS II, ALI and ARDS scores, a greater need for mechanical ventilation and sepsis were associated with higher mortality in patients with cancer.2,3 Our study has some limitations. The study was performed in a single centre with a small sample of patients, and the generalizability of our ndings may be limited. In fact, it is unclear whether uid

715

J. P. Almeida et al.
1 0.8 Sensivity 0.6 0.4 0.2 0 1 0.8 0.6 0.4 0.2 0 4000 2000

AUC = 0.66 (0.520.79)


0 0.2 0.4 0.6 1-Specicity 0.8 1

Sensivity/Specicity

0 2000 4000 6000 8000 Fluid balance (ml)

Fig. 2. (A) Receiver-operating characteristics curve for the mean uid balance (ml/24 h) as a predictor of mortality in cancer patients admitted to the intensive care unit. (B) The intersection of sensibility and specicity to dene the optimal trade-off between the measures. AUC, area under the curve.

Table 4
Intensive care unit (ICU) patient outcomes. Variables Dobutamine* Vasopressors Acute renal failure RRT* Invasive mechanical ventilation* ICU days of endotracheal intubation (median) ICU length of stay (days) *n (%), Fishers exact test. n (%), Pearsons Chi-square test. Median (interquartile range), MannWhitney test. RRT, renal replacement therapy. All patients 21 53 51 6 19 1 5 (17.2%) (43.4%) (41.8%) (4.9%) (15.6%) (07) (49) Survivors 17 34 33 1 5 1 5 (17.5%) (35.1%) (34%) (1%) (5.2%) (01) (47) Non-survivors 4 19 18 5 14 3 8 (16%) (76%) (72%) (20%) (56%) (09) (413) P 1.000 < 0.001 0.001 0.001 < 0.001 0.129 0.015

balance is the cause of the problem or the marker of severe illness, as non-survivors were more severely ill at admission. Further studies are necessary to dene strategies that reduce the cumulative uid balance in critically ill patients with cancer, which will translate to better clinical outcomes.

cally ill patients with cancer, these ndings highlight the importance of volemic status evaluations and uid therapy that is guided by dened goals. Conict of interest: The authors conrm that there is no conict of interest. Financial disclosures: None reported.

Conclusions
Our results show that a positive uid balance in the rst 72 h of the ICU stay is associated with reduced survival in patients with cancer. Furthermore, in critically ill patients with cancer, a liberal uid replacement strategy may be detrimental, and an approach that focuses on avoiding uid retention may improve clinical outcomes. For criti-

References
1. Lecuyer L, Chevret S, Thiery G, Darmon M, Schlemmer B, Azoulay E. The ICU trial: a new admission policy for cancer patients requiring mechanical ventilation. Crit Care Med 2007; 35: 80814. 2. Soares M, Caruso P, Silva E, Teles JM, Lobo SM, Friedman G, Dal Pizzol F, Mello PV, Bozza FA, Silva UV, Torelly AP, Knibel MF, Rezende E, Netto JJ, Piras C, Castro A, Ferreira BS, Ra-Neto A, Olmedo PB, Salluh JI, Brazilian Research in Intensive Care Network (BRICNet). Characteristics and out-

716

Fluid balance and outcome in cancer patients


comes of patients with cancer requiring admission to intensive care units: a prospective multicenter study. Crit Care Med 2010; 38: 915. Taccone FS, Artigas AA, Sprung CL, Moreno R, Sakr Y, Vincent JL. Characteristics and outcomes of cancer patients in European ICUs. Crit Care 2009; 13: R15. Soares M, Fontes F, Dantas J, Gadelha D, Cariello P, Nardes F, Amorim C, Toscano L, Rocco JR. Performance of six severity-of-illness scores in cancer patients requiring admission to the intensive care unit: a prospective observational study. Crit Care 2004; 8: R194203. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, deBoisblanc B, Connors AF Jr, Hite RD, Harabin AL. Comparison of two uid-management strategies in acute lung injury. N Engl J Med 2006; 354: 256475. Payen D, de Pont AC, Sakr Y, Spies C, Reinhart K, Vincent JL, Sepsis Occurrence in Acutely Ill Patients (SOAP) Investigators. A positive uid balance is associated with a worse outcome in patients with acute renal failure. Crit Care 2008; 12: R74. Shum HP, Lee FM, Chan KC, Yan WW. Interaction between uid balance and disease severity on patient outcome in the critically ill. J Crit Care 2011; 26: 6139. Mank A, Semin-Goossens A, Lelie J, Bakker P, Vos R. Monitoring hyperhydration during high-dose chemotherapy: body weight or uid balance? Acta Haematol 2003; 109: 1638. Hajjar LA, Nakamura RE, de Almeida JP, Fukushima JT, Hoff PM, Vincent JL, Auler JO Jr, Galas FR. Lactate and base decit are predictors of mortality in critically ill patients with cancer. Clinics (Sao Paulo) 2011; 66: 203742. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classication system. Crit Care Med 1985; 13: 81829. Schag CC, Heinrich RL, Ganz PA. Karnofsky performance status revisited: reliability, validity, and guidelines. J Clin Oncol 1984; 2: 18793. Cook R, Cook D, Tilley J, Lee K, Marshall J, Canadian Critical Care Trials Group. Multiple organ dysfunction: baseline and serial component scores. Crit Care Med 2001; 29: 204650. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G, SCCM/ ESICM/ACCP/ATS/SIS. 2001 SCCM/ESICM/ACCP/ ATS/SIS International Sepsis Denitions Conference. Crit Care Med 2003; 31: 12506. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A, Acute Kidney Injury Network. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007; 11: R31. 15. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded denition of the adult respiratory distress syndrome. Am Rev Respir Dis 1988; 138: 7203. 16. Rosenberg AL, Dechert RE, Park PK, Bartlett RH, NIH NHLBI ARDS Network. Review of a large clinical series: association of cumulative uid balance on outcome in acute lung injury: a retrospective review of the ARDSnet tidal volume study cohort. J Intensive Care Med 2009; 24: 3546. 17. Roch A, Guervilly C, Papazian L. Fluid management in acute lung injury and ARDS. Ann Intensive Care 2011; 1: 16. 18. Martin GS, Mangialardi RJ, Wheeler AP, Dupont WD, Morris JA, Bernard GR. Albumin and furosemide therapy in hypoproteinemic patients with acute lung injury. Crit Care Med 2002; 30: 217582. 19. Schuller D, Mitchell JP, Calandrino FS, Schuster DP. Fluid balance during pulmonary edema. Is uid gain a marker or a cause of poor outcome? Chest 1991; 100: 106875. 20. Vincent JL, Weil MH. Fluid challenge revisited. Crit Care Med 2006; 34: 13337. 21. Santry HP, Alam HB. Fluid resuscitation: past, present, and the future. Shock 2010; 33: 22941. 22. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, Reinhart K, Angus DC, Brun-Buisson C, Beale R, Calandra T, Dhainaut JF, Gerlach H, Harvey M, Marini JJ, Marshall J, Ranieri M, Ramsay G, Sevransky J, Thompson BT, Townsend S, Vender JS, Zimmerman JL, Vincent JL. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Intensive Care Med 2008; 34: 1760. 23. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M, Early Goal-Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345: 136877. 24. Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds RM, Bennett ED. Early goal-directed therapy after major surgery reduces complications and duration of hospital stay. A randomised, controlled trial [ISRCTN38797445]. Crit Care 2005; 9: R687693.

3. 4.

5.

6.

7. 8.

9.

10. 11. 12. 13.

14.

Address: Juliano Pinheiro de Almeida Department of Anesthesia and Critical Care Instituto do Cancer do Estado de Sao Paulo Av. Dr. Arnaldo, 251 Sao Paulo, 01246000 Brazil e-mail: juliano.almeida@usp.br

717

You might also like