Acta Anaesthesiol Scand 2004; 48: 704—710 Printed in Denmark. All rights reserved

Copyright # Acta Anaesthesiol Scand 2004 ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/j.1399-6576.2004.00388.x

Increased glomerular permeability and pulmonary dysfunction following major surgery: correlation of microalbuminuria and PaO2/FiO2 ratio T. SZAKMA´NY and Z. MOLNA´R Department of Anaesthesia and Intensive Care, Faculty of General Medicine, University of Pe´cs, Hungary

Background: The aim of our trial was to evaluate the ability of microalbuminuria as an indicator of outcome and to investigate its relationship with the postoperative respiratory dysfunction in the initial postoperative period in a high-risk patient group. Methods: In our prospective, observational study patients were consecutively recruited following elective oesophagectomy, total gastrectomy, Whipple-resection of the pancreas and liver resection due to tumour removal. Microalbuminuria (expressed as urine albumin:creatinine ratio, M:Cr) was measured before (tp), and after surgery (t0, t6, t24, t48, t72). Multiple Organ Dysfunction Scores were monitored on ICU admission than daily (t1, t2, t3). For statistical analysis, Wilcoxon’s ranksum test, Mann—Whitney’s U-test, receiver operating characteristic curve analysis and Spearman’s rho test were used as appropriate. Results: One hundred and forty patients (118 survivors and 22 non-survivors) were recruited. Significantly higher Multiple Organ Dysfunction Scores were observed in non-survivors throughout the study period (P < 0.001). Microalbuminuria (Cr) increased significantly (P < 0.01) on admission to the ICU (t0) compared with the preoperative levels, but levels returned to normal within 6 h and remained so for the rest of the study.

There was a significant difference between survivors and nonsurvivors at t0 (P < 0.01). However the ROC curve indicated that M:Cr is not a reliable descriptor of outcome. Comparison of Cr values with the PaO2/FiO2 ratio showed an inverse relationship on admission, which remained so for t24 and t48. Conclusion: M:Cr measured on admission to the ICU was significantly higher in non-survivors than in survivors, and also showed an inverse relationship with the PaO2/FiO2 ratio following extended abdominal surgery. However, on admission, M:Cr did not discriminate survivors from non-survivors. Further studies are required to evaluate the prognostic value of this test for postoperative patients with risk of respiratory failure.

M

coworkers found that in trauma patients increased M:Cr was associated with changes in the PaO2/FiO2 ratio, suggesting that this index may also represent the changes of capillary permeability in the lungs (5). Following elective major abdominal surgery respiratory dysfunction is the leading cause of postoperative morbidity (10). In our previous observational study we found that the PaO2/FiO2 ratio representing the respiratory function measured on the first postoperative day significantly differentiated survivors from non-survivors (11). The PaO2/FiO2 ratio has also been proved to be a sensitive marker of respiratory dysfunction after trauma (12). In a very recent clinical trial, Gosling et al. showed that M:Cr measured within 15 min was able to

(urinary albumin/creatinin ratio; M:Cr) has been suggested to reflect the changes in the vascular permeability, as it is thought that this represents the renal manifestation of generalized increase in systemic vascular permeability (1—3). Several studies have described a rapid increase in M:Cr during trauma, pancreatitis, surgery and ischemiareperfusion (2, 4—6). We have shown earlier that in a heterogeneous ICU patient population microalbuminuria is an early, sensitive and specific marker of the severity of critical illness (7). It has been shown by others that increased urinary albumin excretion occurs within minutes or hours after surgical trauma, which has recently been supported by our data (8, 9). Furthermore, Gosling and

704

ICROALBUMINURIA

Accepted for publication 28 January 48 2004

Key words: Major surgery; microalbuminuria; PaO2/FiO2 ratio; respiratory dysfunction. #

Acta Anaesthesiologica Scandinavica 48 (2004)

Microalbuminuria and pulmonary dysfunction

differentiate between survivors and non-survivors in a mixed ICU population (13). It has also been suggested that the predictive value of M:Cr and its relationship to the PaO2/FiO2 ratio should be evaluated in postoperative patients (14). Therefore the aim of our trial was to evaluate the ability of M:Cr as an indicator of outcome and to investigate its relationship with the postoperative respiratory dysfunction as monitored by the PaO2/ FiO2 ratio in the initial postoperative period in a highrisk patient group.

Methods Patients The study protocol was approved by the regional Ethics Committee, which waived the need for written informed consent. All patients undergoing elective oesophagectomy, total gastrectomy, Whippleresection of the pancreas and liver resection due to tumour removal were entered into the study between January 1998 and January 2001. Those whose tumour proved to be inoperable were excluded from the study, as they were not admitted to our 20-bed teaching hospital intensive/high-dependency care unit. Patients with chronic renal failure requiring renal replacement therapy as indicated by the APACHE II guidelines were also excluded, as renal insufficiency and oliguria may significantly affect the renal albumin excretion (15).

Surgery All operations were performed by the same two surgeons. Tumours were removed via laparotomy or thoracolaparotomy.

Anaesthetic management All patients received routine anaesthetic management and monitoring including: premedication with benzodiazepine; induction with propofol; muscle relaxation with atracurium; and maintenance with isoflurane, fentanyl and epidural analgesia in the postoperative period. Full blood count, arterial blood gases, cardiorespiratory parameters and blood loss were checked and recorded hourly during surgery.

Clinical management Patients received routine intensive monitoring and therapy. Regarding respiratory support if invasive mechanical ventilation was needed, lung protective ventilation (VT6 ml kg1) with the open-lung concept was applied with PEEP 5—25 cmH2O in pressure

controlled/supported mode (Servo 300, Siemens, Solana Sweden). PEEP levels and breathing frequency were adjusted to arterial blood gas results.

Multiple organ dysfunction score (MODS) In the ICU, the progression of organ dysfunction was monitored by the MODS, as designed by Marshall et al., daily during the first three postoperative days (t1, 2, 3; 16).

Urine sampling and microalbuminuria assay Urine samples (10 ml) were taken from an indwelling bladder catheter after anaesthesia was commenced (tp), on arrival in the ICU (t0), then at 6, 24, 48 and 72 h (t6, 24, 48, 72) postoperatively. Urine albumin and creatinine concentrations were determined by immunoturbidimetry (Turbox1, Orion Diagnostica, Finland). In order to correct for intraindividual variation in the urine flow rate, microalbumine levels were quantified to the urine creatinine concentration (analyzed by the Jaffe method), and expressed as the microalbumin:creatinine (M:Cr) ratio (17). The assigned laboratory reference range was 225 mmHg had a 81% sensitivity and 60% specificity for survival. Therefore in this trial in a post hoc analysis we defined a ‘normal’ group as having a PaO2/FiO2 ratio at t0 higher than 225 mmHg, whereas in the ‘low’ group it was below 225 mmHg, which in the MOD scoring system represents two points, indicating moderate respiratory dysfunction. As the admission M:Cr and the PaO2/FiO2 ratio showed significant negative correlation we tried to assess whether M:Cr can predict the postoperative respiratory dysfunction. M:Cr as measured on ICU admission was significantly higher in the ‘low’ group [29.55 (13.90—76.42) vs. 18.86 (6.57—42.99) mg mmol1; ‘low’ vs. ‘normal’ group, respectively], however, the ROC curve indicated that it is not a reliable descriptor of respiratory failure (Fig. 3).

Discussion In this prospective study patients following extended abdominal tumour surgery demonstrated raised M:Cr

Table 1 Perioperative data of the patients.

Age (years) Sex (M/F) Time of surgery (min) APACHE II (first 24 h) Serum creatinine (mmol l1) Oesophagectomy Total gastrectomy Whipple-resection Liver resection

Survivors (n ¼ 118)

Non-survivors (n ¼ 22)

P

57 (50—67) 88/30 240 (180 360) 3 (2—6) 67 (48—83) 84 22 8 4

63 (54—72) 18/4 260 (236 375) 4 (2—7) 72 (59—81) 17 5 0 0

NS NS NS NS NS

Data are presented as medians and interquartile ranges in (). For statistical analysis the Mann—Whitney U-test and the Chi-square test were used, respectively.

706

Microalbuminuria and pulmonary dysfunction Table 2

110

Course of the multiple organ dysfunction score in survivors and non-survivors.

100

Survivors

Non-survivors

P

t0 t1 t2 t3

1 (0—3) 1 (0—3) 1 (1—3) 1 (0—3)

3 (2—4) 3 (2—4) 4 (3—6) 5 (3—7)

0.001 0.001 0.001 0.001

Data are presented as median (interquartile range). For statistical analysis the Mann—Whitney U-test was used. MODS ¼ multiple organ dysfunction score.

levels comparable to, or greater than, those reported in previous studies (2, 6, 7, 14). Microalbuminuria was significantly higher in non-survivors compared with survivors as early as on admission to the ICU. Moreover, M:Cr inversely correlated with the PaO2/FiO2 ratio, and patients with moderate to severe respiratory dysfunction had significantly increased microalbuminuria. Many conditions such as trauma burns and surgery trigger the inflammatory response which involves a wide variety of mediators (18). Endothelial cells play a pivotal role in this process. Activated polymorphonuclear leukocytes can accumulate at the site of inflammation, causing direct and indirect damages at the endothelial cells by the obstruction of the capillaries or by the release of inflammatory mediators to 700

PaO2/FiO2 ratio (mmHg)

600

*

*

*

*

500 400 300 t0

200

t24 100

t48 t72

0 Survivors

∗†

90 Microalbuminuria (mg mmol–1)

MODS



Non-survivors

Fig. 1. Changes in PaO2/FiO2 ratio from ICU admission (tp) to 72 h (t72) postoperatively in survivors and non-survivors. Data are presented as boxes and whisker plots. The boxes enclose the interquartile range and median (middle line in each box); the whiskers enclose the minimum and maximum. For statistical analysis Wilcoxon’s test was used. *P < 0.001 vs. t0. Differences between the two groups at each assessment point were tested with the Mann—Whitney U-test.

80 70 60

tp 50

t0

40

t6

30

t24

20

t48

10

t72

0 Survivors

Non-survivors

Fig. 2. Progress of Microalbuminuria (M:Cr) preoperatively (tp) to 72 h (t72) postoperatively in survivors and non-survivors. Data are presented as boxes and whisker plots. The boxes enclose the interquartile range and median (middle line in each box); the whiskers enclose the minimum and maximum. For statistical analysis Wilcoxon’s test was used. *P < 0.001 vs. t0. Differences between the two groups at each assessment point were tested with the Mann—Whitney U-test. †P < 0.05

the systemic circulation (19). Whatever the exact pathway would be, the net effect is an increase in the endothelial permeability. The exact method proposed to detect these changes involves the use of radiolabelled albumin, which is technically complex, invasive, expensive and hardly available at the routine clinical setting (1). It has been suggested that the kidneys are uniquely placed to mirror changes in capillary permeability, as they receive about 25% of the cardiac output and small changes in glomerular permeability lead to large changes in microalbuminuria (20). Indeed, increased renal albumin excretion was found in several acute conditions such as pancreatitis and trauma (4, 5). An early report from Gosling et al. found that the urinary concentration of total proteins increased in proportion to the magnitude of the surgical insult (2). Studies with similar results, including our recent report, following oesophagectomy have confirmed their data (8, 9). Our present results are in accordance with these findings. Following extended abdominal tumour surgery the observed values were in a higher range, as seen by De Gaudio and coworkers (14). They found significantly increased M:Cr in patients progressing to sepsis; however, as there were no deaths in their series they could not evaluate the prognostic value of M:Cr (14). We found that significantly higher M:Cr was associated with adverse outcome and also

707

T. Szakma´ny and Z. Molna´r Table 3 Correlation coefficients for Microalbuminuria (M:Cr) with PaO2/ FiO2 ratio in patients following extended abdominal surgery. t0 t24 t48 t72

PaO2/FiO2 correlation coefficient

P

0.218 0.193 0.238 0.057

0.012 0.038 0.025 0.644

that lower PaO2/FiO2 ratio values were seen in patients with increased glomerular permeability. We observed higher M:Cr values preoperatively compared with the normal value of 3 mg mmol1 which was obtained in healthy subjects, however, this difference was not statistically significant. It has been shown that many conditions can cause an increase in albumin excretion by the kidneys, such as high blood pressure, smoking and age (>40 years), as we have also seen in our patients (20, 21). Indeed, as perioperative serum creatinine and blood urea nitrogen values were in normal range in all of our patients this finding may support our opinion that the 3 mg mmol1 cut-off point may not be suitable in the critically ill patient population, however, it has to be determined by screening large numbers of ICU patients (22). Authors in most cases emphasize that microalbuminuria measured within 6—48 h after the initiating insult is of particular value (1—8). Previously, we 1.00

Sensitivity

0.75

0.50 AUC=0.615

0.25

0.00 0.00

0.25

0.50

0.75

1.00

1 - Specificity Fig. 3. Respiratory dysfunction (PaO 2 /FiO 2 < 225 mmHg) receiver operating characteristic curves, with area under the curve for Microalbuminuria (M:Cr) ¼ 0.615.

708

have shown in a heterogeneous critical care population that M:Cr measured 6 h after admission to the ICU significantly differentiated survivors from nonsurvivors (7). In a recent pilot study it was found that decreasing microalbuminuria had a good sensitivity and specificity to predict the absence of acute respiratory failure and multiple system organ failure during the first 48 h in medical ICU patients (23). Gosling et al. showed that on admission M:Cr had excellent predictive value for mortality in a mixed ICU population (13). In the present trial increased urinary albumin excretion was seen in our patients on admission, which normalized within 6 h and remained so for the rest of the observational period. Although the increase in the glomerular permeability was associated with worse respiratory parameters, the ROC data indicate that M:Cr is not a reliable predictor of neither the respiratory dysfunction nor the outcome. The lack of consistency between our results and those of Gosling et al. may be partially explained by the different study population. In this study, a relatively homogenous patient population was examined and more accurate urine sampling was achieved in relation to the initiating insult. On the other hand, Gosling et al. proposed an association between microalbuminuria and systemic vascular endothelial damage. They suggested that microalbuminuria in the critically ill patient is an indicator of the severity of the capillary leakage, which may lead to multiple system organ failure (1, 3—5). However, this statement is lacking of clear evidence. For instance, in our recently published trial we have shown that in patients with developed septic shock, M:Cr do not correlate with the PaO2/ FiO2 ratio or with the extravascular lung water (24). Also De Gaudio and coworkers were unable to find correlation between M:Cr and oxygenation in postoperative septic patients (14). There might be explanations other than systemic capillary leakage for the correlation between the M:Cr and PaO2/FiO2 ratios observed in this study. As the increase in urinary albumin excretion diminished as early as 6 h postoperatively, it is likely that this phenomenon represents an acute response to the surgical insult rather than a sustained elevation in the systemic endothelial permeability. Moreover, the PaO2/FiO2 ratio was used to assess the degree of respiratory failure, and endothelial permeability is not the only factor influencing this parameter. As pneumonia was the cause of the multiple organ failure in those who did not survive their illness, they might have had impaired defence mechanisms such as bronchoalveolar cleaning at the time of ICU admission, compared with the survivors. There is a possibility that those with

Microalbuminuria and pulmonary dysfunction

preoperatively unnoticed, compromised respiratory function reflect the acute insult with a higher degree of albumin excretion. The link between acute lung injury and renovascular permeability may be explained by tissue hypoxia, complement and neutrophil and macrophage activation, which are not only key features of acute respiratory failure but are also seen in other organs in patients following extended tissue damage (25). Significantly higher MOD scores were observed in the non-survivor group throughout the study, which were attributable to the lower PaO2/FiO2 ratio. Furthermore, the MOD scores were predictive for outcome on the first postoperative day, and in accord with previous reports the MODS seems to be a reliable tool for monitoring the severity of illness (26). In conclusion, microalbuminuria measured on admission to the ICU was significantly higher in the non-survivors compared with the survivors, and also showed an inverse relationship with the PaO2/FiO2 ratio following extended abdominal surgery. The fast onset and magnitude of the renal permeability as a response to remote injury and the correlation with respiratory dysfunction suggest, that measurement of urinary albumin excretion on ICU admission following major abdominal surgery may have a role in the early identification of patients at risk of developing multiple system organ failure. Further studies are required to evaluate the prognostic value of this easyto-apply bedside test for postoperative patients with risk of respiratory failure who would benefit from more aggressive therapeutic interventions.

Acknowledgements The authors would like to thank the staff on the ICU at Medical University of Pe´cs and at the Department of Clinical Biochemistry, especially Tama´s K’szegi MD, PhD, for their kind and generous help throughout the study, without which this work would have not been completed. This work was partially funded by NKFP 1 A/0026 research grant, Ministry of Education, Hungary.

References 1. Fleck A, Raines G, Hawker F et al. Increased vascular permeability: a major cause of hypoalbuminemia in disease and injury. Lancet 1985; 1: 781—4. 2. Gosling P, Shearman CP, Gwynn BR, Simms MH, Bainbridge ET. Microproteinuria: response to operation. Br Med J 1988; 296: 3338—9. 3. Gosling P, Bascom JU, Zikria BA. Capillary leak, oedema and organ failure: breaking the triad. Care Crit Ill 1996; 12: 191—7. 4. Gosling P, Shearman CP. Can we predict disease severity of acute pancreatitis? Ann Clin Biochem 1992; 29: 9—10.

5. Gosling P, Sanghera K, Dickson G. Generalized vascular permeability and pulmonary function in patients following serious trauma. J Trauma 1994; 36: 477—81. 6. Smith CT, Gosling P, Sanghera K, Green MA, Paterson IS, Shearman CP. Microproteinuria predicts the severity of systemic effects of reperfusion injury following infra aortic aneurysm surgery. Ann Vasc Surg 1994; 8: 1—5. 7. MacKinnon KL, Molnar Z, Lowe D, Watson ID, Shearer E. Use of microalbuminuria as a predictor of outcome in critically ill patients. Br J Anaesth 2000; 84: 239—41. 8. De Gaudio AR, Piazza E, Barneschi MG, Ginanni R, Martinelli P, Novelli GP. Peri-operative assessment of glomerular permeability. Anaesthesia 1999; 50: 810—2. 9. Molnar Z, Szakmany T, Koszegi T, Tekeres M. Microalbuminuria and serum procalcitonin levels following oesophagectomies. Eur J Anaesth 2000; 17: 464—5. 10. Svanes K, Stangeland L, Viste A, Varhaug JE, Gronbech JE, Soreide O. Morbidity, ability to swallow, and survival, after oesophagectomy for cancer of the oesophagus and cardia. Eur J Surg 1995; 161: 669—75. 11. Szakmany T, Molnar Z. Clinical rather than biochemical variability may predict mortality following oesophagectomy. Br J Anaesth 2001; 2: 320. 12. Navarrete-Navarro P, Ruiz-Bailen M, Rivera-Fernandez R, Guerrero-Lopez F, Pola-Gallego-de-Guzman MD, VazquezMata G. Acute respiratory distress syndrome in trauma patients: ICU mortality and prediction factors. Intensive Care Med 2000; 26: 1624—9. 13. Gosling P, Brudney S, McGrath L, Riseboro S, Manji M. Mortality prediction at admission to intensive care: a comparison of microalbuminuria with acute physiology scores after 24 hours. Crit Care Med 2003; 31: 98—103. 14. De Gaudio AR, Adembri C, Grechi S, Novelli GP. Microalbuminuria as an early index of impairment of glomerular permeability in postoperative septic patients. Intensive Care Med 2000; 26: 1364—8. 15. Knaus WA, Draper EA, Waner DP, Zimmerman JE, APACHE II. a severity of disease classification system. Crit Care Med 1985; 13: 818—29. 16. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CR, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995; 23: 1638—52. 17. Watts GF, Morris RW, Khan K, Polak A. Urinary albumin excretion in healthy adult subjects: reference values and some factors affecting their interpretation. Clin Chem Acta 1988; 172: 191—8. 18. Beal AL, Cerra FB. Multiple organ failure syndrome in the 1990s: Systemic inflammatory response and organ dysfunction. JAMA 1994; 271: 226—33. 19. Adams DH, Nash GB. Disturbance of leucocyte circulation and adhesion to the endothelium as factors in circulatory pathology. Br J Anaesth 1996; 77: 17—31. 20. Gosling P. Microalbuminuria: marker of systemic disease. Br J Hosp Medical, 1995: 285—90. 21. Mogensen CE. Microalbuminuria and hypertension with focus on type 1 and type 2 diabetes. J Intern Med 2003; 7: 45—66. 22. Molnar Z, Szakmany T, Heigl P. Microalbuminuria in severe sepsis: timing is difficult, ‘normal’ levels are uncertain. Intensive Care Med 2003; 29: 1395. 23. Abid O, Sun Q, Sugimoto K, Mercan D, Vincent J-L. Predictive value of microalbuminuria in medical ICU patients. Chest 2001; 120: 1984—8. 24. Molnar Z, Szakmany T, Heigl P. Microalbuminuria does not reflect increased systemic capillary permeability in septic shock. Intensive Care Med 2003; 29: 391—5.

709

T. Szakma´ny and Z. Molna´r 25. Nuytinck JKS, Goris RJA, Weerts JGE et al. Acute generalised microvascular injury by activated complement and hypoxia: The basis of the adult respiratory distress syndrome and multiple organ failure? Br J Exp Path 1986; 67: 537. 26. Jacobs S, Zuleika M, Mphansa T. The Multiple Organ Dysfunction Score as a descriptor of patient outcome in septic shock compared with two other scoring systems. Crit Care Med 1999; 27: 741—4.

710

Address: Dr Tamas Szakmany Department of Anaesthesia and Intensive Care Faculty of General Medicine University of Pe´cs Ifju´sa´g u. 13 7643, Hungary e-mail: [email protected]