Preoperative maximal exercise oxygen ... - Wiley Online Library

Blackwell Publishing AsiaMelbourne, AustraliaRESRespirology1323-7799© 2007 The Authors; Journal compilation © 2007 Asian Pacific Society of Respirology2007124505510Original ArticlesOxygen consumption test on major lung resectionAS Bayram et al.

Respirology (2007) 12, 505–510

doi: 10.1111/j.1440-1843.2007.01097.x

ORIGINAL ARTICLE

Preoperative maximal exercise oxygen consumption test predicts postoperative pulmonary morbidity following major lung resection .

Ahmet Sami BAYRAM, TarIk CANDAN AND Cengiz GEBITEKIN Medical Faculty, Department of Thoracic Surgery, Uludag University, Bursa, Turkey

Preoperative maximal exercise oxygen consumption test predicts postoperative pulmonary morbidity following major lung resection BAYRAM AS, CANDAN T, GEBITEKIN C. Respirology 2007; 12: 505–510 Background and objective: Pulmonary resection carries a significant morbidity and mortality. The utility of maximal oxygen uptake test (VO2max) to predict cardiopulmonary complications following major pulmonary resection was evaluated. Methods: Following standard preoperative work-up and VO2max testing, 55 patients (49 male; mean age 59 years, range 20–74) underwent major pulmonary surgery: lobectomy (n = 31), bilobectomy (n = 6) and pneumonectomy (n = 18). An investigator blinded to the preoperative assessment prospectively collected data on postoperative cardiopulmonary complications. Patients were divided into two groups according to preoperative VO2max and also according to FEV1. The frequency of postoperative complications in the groups was compared. Results: Complications were observed in 19 (34.5%) patients, 11 of which were pulmonary (20%). There were two deaths (3.6%), both due to respiratory failure. Preoperative FEV1 failed to predict postoperative respiratory complications. Five of 36 patients with a preoperative FEV1 > 2 L suffered pulmonary complications, compared with six of 19 patients with FEV1 < 2 L. Cardiopulmonary complications were not observed in patients with VO2max > 15 mL/kg/min (n = 27); however, 11 patients with VO2max < 15 mL/kg/min (n = 28) suffered cardiopulmonary complications (P < 0.05). Conclusion: VO2max predicts postoperative pulmonary complications following major lung resection, and the risk of complications increases significantly when the preoperative VO2max is less than 15 mL/kg/min. Key words: exercise test, oxygen consumption, postoperative complication, thoracotomy.

INTRODUCTION Resection remains the best treatment option for patients with non-small cell lung cancer (NSCLC), despite advances in chemotherapy and radiotherapy. As most patients requiring surgery in this setting are smokers or ex-smokers, patients with NSCLC frequently suffer from pulmonary and cardiovascular comorbidities, which place them at a higher risk of postoperative complications.1 Correspondence: Ahmet Sami Bayram, Department of Thoracic Surgery, School of Medicine, Uludag University, Görükle 16059, Bursa, Turkey. Email: asbayram2@ yahoo.com Received 6 March 2006; invited to revise 26 July and 17 August 2006; revised 14 and 18 August 2006; accepted 1 September 2006 (Associate Editor: Kwun Fong).

Static PETs, spirometry and lung volume measurements are standard preoperative assessments for patients requiring lung resection, and even if combined with V/Q lung scanning, these methods are relatively conservative in predicting postoperative lung function. Accurate preoperative risk assessment is required to inform decisions on treatment. As most postoperative complications are cardiopulmonary,2,3 studies have examined preoperative functional cardiopulmonary testing to predict morbidity and mortality post lung resection. There is conflicting evidence of the value of maximal oxygen consumption (VO2max) as assessed by cardiopulmonary stress testing in predicting postoperative complications.4–7 In this prospective study, the value of measuring VO2max preoperatively to predict cardiopulmonary morbidity and mortality in patients following major lung resection was assessed.

© 2007 The Authors Journal compilation © 2007 Asian Pacific Society of Respirology

506

METHODS Study subjects Patients referred for major lung resection had a standard assessment for operability, including pulmonary function tests, resting and postexercise arterial blood gas analysis (ABG), and a 6-min walk test. All patients with a history of ischaemic heart disease were seen by a cardiologist to optimize cardiovascular function. Eighty-three consecutive patients deemed fit to undergo, and listed for major lung resection (lobectomy, bilobectomy or pneumonectomy), were recruited prospectively into the trial. Written, informed consent was obtained from each patient and was approved by the local ethics committee. Patients were excluded from analysis if the lesion was determined to be unresectable at thoracotomy, or if they underwent lesser resections (wedge excision or segmentectomy) or more extensive resections (chest wall, diagphragmatic or angioplastic resections) than was originally planned. No patient was excluded on the basis of their VO2max result.

Study design All patients underwent preoperative measurement of FEV1 and VO2max. Patients who met the study criteria were followed postoperatively and complications recorded. The patients were grouped by VO2max and by FEV1 to determine whether either of these variables could be used to predict the risk of postoperative complications.

Procedure The maximal exercise O2 consumption test was carried out using an Ergomedic Monark 824E bicycle (Varberg, Sweden). Oxygen consumption and CO2 production were measured using a metabolic analyser (SensorMedics 2900C; Yorba Linda, CA). Before each test, room temperature, barometric pressure and humidity were recorded. During the test, the patient’s SaO2, blood pressure and ECG were continuously and non-invasively recorded. The test was begun with a load of 0.5 kg on the bicycle, and this was increased every 3 min in 0.5-kg increments. The test was ended according to the patient’s tolerance, development of angina or changes in ECG trace. Patients were divided into two groups according to O2 consumption. All lung resections were carried out via a posterior, full muscle-sparing, mini-thoracotomy. Postoperative analgesia was achieved via an epidural catheter. Patient-controlled i.v. analgesia or paravertebral nerve block was added as required. Fifty-five patients underwent lung resections that qualified for inclusion in the study: 52 for NSCLC and three for bronchiectasis. Postoperative morbidity and mortality data were prospectively collected for each patient by an investigator blinded to the VO2max results. On each day of

AS Bayram et al.

their hospital admission, each patient was clinically examined and CXR and laboratory tests studied. Death within 30 days of surgery was accepted as surgical mortality. Assessments aimed to identify postoperative complications of respiratory failure (mechanical ventilation > 48 h or re-intubation), pneumonia (infiltrate on CXR, pyrexia > 38°C and leukocytosis > 12.0 cells/L), atelectasis (requiring bronchoscopic intervention), broncho-pleural fistula (confirmed on bronchoscopy), prolonged air leak (longer than 7 days) and cardiac complications (arrhythmias requiring intervention or cardiac ischaemia).

Data analysis The data were analysed to determine whether VO2max was a better test for predicting postoperative cardiovascular complications than FEV1. Patients were divided into two groups (group V1 and group V2) on the basis of a VO2max of greater or less than 15 mL/ kg/min. The division into these groups was based on the published literature that suggests a VO2max of 15 mL/kg/min is considered indicative of a patient having a standard risk for complications.8 Patients were also divided into two groups (groups F1 and F2) on the basis of an FEV1 of greater or less than 2 L. The published literature suggests that an FEV1 of 1–2 L has a low risk for the patient having thoracic surgery.9 Comparison was made using the chi-squared test and Fisher’s exact t-test; significance was accepted at P < 0.05.

RESULTS Twenty-eight patients were excluded from the study due to inoperability or resections that were smaller or greater than the planned major resection. A total of 55 patients met the inclusion criteria: 49 (89%) were male and 6 (11%) female; mean age was 59 years (range 20–74). Lobectomy was performed in 31 patients (56%), bilobectomy in 6 (11%) and pneumonectomy in 18 (33%). Complications occurred in 19 patients (34.5%). Pulmonary complications (respiratory failure, pneumonia, atelectasis and bronchopleural fistula) occurred in 11 patients (20%). The mean VO2max for the entire group was 15.12 mL/kg/min (range 2.7–32.5). The mean preoperative FEV1 was 2.21 L (range 0.79–3.46). The postoperative mortality was 3.6% (2/55). Based on preoperative VO2max, the 55 patients were classified into two groups: group V1 (n = 27) and group V2 (n = 28). Group demographic and operative data are given in Table 1. Group V1 had a mean VO2max of 19.15 mL/kg/min (range 15.6–32.5) and group V2 a mean VO2max of 11.95 mL/kg/min (range 7.2–14.8). Both patients who died were in group V2.


507

Oxygen consumption test on major lung resection Table 1

Operative and demographic characteristics of subgroups used to evaluate preoperative testing

Type of resection Lobectomy Bilobectomy Pneumonectomy Median age in years(range)

Group V1

Group V2

Group F1

Group F2

13 3 11 62 (34–74)

18 3 7 54 (20–71)

18 3 15 55 (20–70)

13 3 3 65 (32–74)

Group V1, preoperative VO2max > 15 mL/kg/min; group V2, preoperative VO2max < 15 mL/kg/min; group F1, FEV1 > 2 L; group F2, FEV1 < 2 L.

Table 2 Postoperative mortality and morbidity in group V1 (VO2max > 15 mL/kg/min) compared with group V2 (VO2max < 15 mL/kg/min) Complications Mortality Respiratory failure Pulmonary Pneumonia Atelectasis Bronchopleural fistula Prolonged air leak Arrhythmia

Group V1 n (%)

Group V2 n (%)

Total n (%)

2 (7.1) 1 (3.5) 11 (40) 3 (11) 6 (21) 2 (7.1) 12 (42) 1 (3.5)

2 (3.6) 1 (1.8) 11 (20) 3 (5.5) 6 (11) 2 (3.6) 19 (34) 3 (5.5)

0 0 0

7 (26) 2 (7.4)

Statistical significance

P < 0.05

NS NS


Table 3

Postoperative mortality and morbidity in group F1 (FEV1 > 2 L) compared with group F2 (FEV1 < 2 L)

Complications Mortality Respiratory failure Pulmonary complications Pneumonia Atelectasis Bronchopleural fistula Prolonged air leak Arrhythmia

Group F1 (n = 36) n (%)

Group F2 (n = 19) n (%)

Total (n = 55) n (%)

Statistical significance

1 (2.7) 1 (2.7)

1 (5.2) 0

2 (3.6) 1 (1.8)

NS

1 (2.7) 4 (11) 0 11 (30) 3 (8.3)

2 (10.5) 2 (10.5) 2 (10.5) 8 (42) 0

3 (5.5) 6 (11) 2 (3.6) 19 (34) 3 (5.5)

NS

NS NS


Complications overall were significantly less in group V1 than in group V2 (7/27 (25%) vs 12/28 (43%) patients; P >0.05) (Table 2), and pulmonary complications were also significantly less in Group V1 than in Group V2 (0/27(0%) vs 11/28 (39%), P < 0.05). There was no significant difference in the average length of postoperative stay (6 days vs 7 days). The 55 patients were also divided into two groups using pre operative FEV1. Group F1 contained 36 patients (65%) with an FEV1 > 2 L, and group F2 comprised 19 patients (35%) with an FEV1 < 2 L. Group demographics and operative data are given in Table 1. Group F1 had a mean preoperative FEV1 of 2.57 L (range 2.01–3.46), and group F2 had a mean preoperative FEV1 of 1.57 L (range 0.79–1.97). Of the

two patient deaths, one was from group F1 and the other from group F2. There was no significant difference in the proportion of each group that experienced general complications (10/36 (28%) vs 9/19 (47%)), although a greater proportion occurred in group F2 (Table 3). Of group F1 patients, 5/36 (14%) experienced pulmonary complications compared with 6/19 (32%) in group F2. Comparison of all complications and pulmonary complications between the two groups revealed no significant differences. There was no significant difference in length of stay between the two groups (5 days vs 8 days). Patients who suffered pulmonary complications had a VO2max that was significantly less compared


508

AS Bayram et al.

with the mean VO2max of those who did not suffer pulmonary complications (8.8 mL/kg/min vs 16.0 mL/kg/min, P < 0.05). There was no significant difference between the mean FEV1 of the two groups of patients.

DISCUSSION The causes of morbidity and mortality after thoracic surgery are predominantly pulmonary and cardiac, and preoperative testing focuses on this as well as looking for evidence of metastases. Historically, spirometry and informal exercise tests have determined who undergoes lung resection. A patient is deemed operable if the predicted postoperative FEV1 (ppoFEV1) is greater than 800 mL.10 Standard spirometry is a poor predictor of a patient’s ability to cope with lung resection postoperatively. Patients with COPD may be able to meet ventilatory demands during resting conditions but exhibit severe limitation on exercise11 or, conversely, following pulmonary rehabilitation, may increase their VO2max and exercise capacity with little effect on spirometry, lung volumes or gas exchange.12 ppoFEV1 is useful in defining patients at low risk of complications after pulmonary resection, but is a poor predictor in high-risk patients. ppoFEV1 correlates well with actual postoperative FEV1 6 months after surgery for patients undergoing lobectomy but not for patients who have a pneumonectomy, where ppoFEV1 underestimates the final value.13 There is a poor relationship between dyspnoea and spirometry, but a correlation does exist between dyspnoea and exercise studies. Exercise is an interactive process of pulmonary and cardiac function, together with tissue utilization of oxygen.14 Preoperative testing using an exercise stress may well be a more accurate model of the postoperative stresses that a lung resection patient will have to undergo. In this study of 55 patients following major lung resection, a mortality rate of 3.6% and all-cause morbidity of 34.5% was observed. Pulmonary complications occurred in 20% of patients. A cut-off value for VO2max of 15 mL/kg/min divided the patients into two groups that experienced a significantly different rate of postoperative pulmonary complications. The two deaths were both in the patients with

VO2max < 15 mL/kg/min. Grouping patients by FEV1 using an arbitrary cut-off of 2.0 L was not predictive of postoperative pulmonary complications. The cardiac complications that occurred in the study were limited to arrhythmias and affected three patients (5.5%); neither VO2max nor FEV1 predicted their occurrence. A number of studies have found no predictive value for VO2max or an inconsistent correlation with postoperative outcome (Table 4). Colman et al. found a relationship between VO2max and the development of postoperative complications in 47 patients following lung resection, although surgical complications such as gastrointestinal haemorrhage, wound infections and excessive operative blood loss were included in their study.4 A further study of 17 patients with mild to moderate COPD undergoing lung resection found no relationship between postoperative cardiopulmonary complications and preoperative VO2max.5 Markos et al. found an inconsistent correlation between VO2max and postoperative complications; in the 55 patients studied, a VO2max of 15 mL/kg/min and were offered resection; of those patients, eight accepted surgery. There was no postoperative mortality, but two patients experienced cardiopulmonary complications.24 Studies among this high-risk group of patients have refuted a link between preoperative VO2max and postoperative complications. In Pate et al.’s study of 12 high-risk patients with FEV1 < 2 L who underwent a range of resections for lung cancer, seven patients experienced complications. The mean VO2max of the patients who experienced complications was 14.1 mL/kg/min, compared with 13.7 mL/kg/min for those who did not.18 In a prospective study of 40 patients undergoing lung resection with a ppoFEV1 of