|
|
||||||||
J Thorac Cardiovasc Surg 2007;133:1071-1077
© 2007 The American Association for Thoracic Surgery
Cardiothoracic Transplantation |
a Department of Intensive Care, Alfred Hospital, Melbourne, Victoria, Australia.
b Department of Allergy, Immunology, and Respiratory Medicine, The Alfred Hospital, Melbourne, Victoria, Australia.
Received for publication June 29, 2006; revisions received September 28, 2006; accepted for publication October 9, 2006. * Address for reprints: David Pilcher, MD, Intensive Care Unit, The Alfred Hospital, Commercial Rd, Prahran 3181, Victoria, Australia. (Email: d.pilcher{at}alfred.org.au).
| Abstract |
|---|
|
|
|---|
Methods: We retrospectively studied 170 patients who received a single-lung transplant over a 15-year period, 20 (12%) of whom required independent lung ventilation.
Results: Patients requiring independent lung ventilation were similar in age, sex, ischemic time, and donor characteristics to those who required conventional ventilation. Patients receiving independent lung ventilation had a greater degree of preoperative airflow limitation, more hyperinflation, lower postoperative PaO 2/fraction of inspired oxygen ratios, more radiologic mediastinal shift, and more transplant lung infiltrate on the postoperative chest radiograph. Multivariate logistic regression analysis showed that independent lung ventilation was associated with increasing levels of recipient hyperinflation (percentage total lung capacity compared with predicted value; odds ratio, 1.04; 95% confidence interval, 1.01-1.07; P = .032) and reduced early postoperative PaO 2/fraction of inspired oxygen ratio (odds ratio, 0.97; 95% confidence interval, 0.95-0.99; P = .005). Length of ventilation and intensive care unit stay and mortality were higher in the independent lung ventilation group. Among patients who survived to hospital discharge, there were no differences in long-term mortality between the 2 groups.
Conclusions: The need for independent lung ventilation in patients undergoing single-lung transplantation for obstructive lung disease is predicted by the combination of increased hyperinflation measured on recipients preoperative lung function tests and a low PaO 2/fraction of inspired oxygen ratio, indicating graft dysfunction in the immediate postoperative period.
| Introduction |
|---|
|
|
|---|
After transplantation, the new allograft is predisposed to early injury from a variety of sources. These include primary graft dysfunction, injury caused by lung handling, fluid loading in a lung with increased permeability and absent lymphatic drainage, early infection, and, in occasional patients, impaired venous drainage. When early lung allograft injury does occur, it can lead to atelectasis, hypoxia, hypercapnia, and reduced compliance in the transplant lung. The latter increases the proportion of ventilation to the native lung, thereby increasing its dynamic hyperinflation, and can lead to increased intrathoracic pressure, circulatory compromise, and increasing mediastinal shift, with further atelectasis and shunt in the transplant lung. Attempts to alleviate hypoxia, hypercapnia, and transplant lung atelectasis by increasing ventilation or positive end-expiratory pressure (PEEP) often increase dynamic hyperinflation in the native lung and worsen atelectasis and gas exchange in the transplant lung.4
This remains a serious problem, especially in the early postoperative period. Because the ventilation and PEEP requirements of the transplant lung aggravate dynamic hyperinflation in the native (COPD) lung, independent-lung ventilation (ILV) has been proposed as a successful treatment of early postoperative dynamic hyperinflation and graft failure.5,6
The principal aim of our study was to identify preoperative and early postoperative predictors for the need for ILV in patients who received ILV in clinical management. We hypothesized that a number of factors might combine to determine the requirement for ILV: pre-existing dynamic recipient hyperinflation, postoperative hyperinflation of the native lung, size mismatch between the native and transplanted lungs, mediastinal shift caused by hyperinflation of the native lung, and injury to the transplanted lung. Additional aims were to study the early and long-term outcomes (duration of ventilation, length of stay in the intensive care unit [ICU], hospital mortality, and survival after leaving the hospital) of patients requiring ILV compared with those of patients not receiving ILV.
| Materials and Methods |
|---|
|
|
|---|
Preoperative Variables
Recipients age, sex, height, weight, lung function test results, and chest radiographic measurements were recorded. Donor factors included were age, sex, height, PaO
2 (taken on a fraction of inspired oxygen [FIO
2] of 100% and a 5 cm H2O PEEP), and estimated total lung capacity (TLC).7
Total graft ischemic time for the organ was also recorded.
Postoperative Variables
Estimates of lung size and mediastinal shift were taken by using the first postoperative chest radiograph and compared with the most recent preoperative radiograph. The severity of transplant lung infiltrate was graded from 0 (absence of infiltrate) to 4 (dense alveolar consolidation of the entire lung field).8
Gas exchange was assessed by recording the lowest PaO
2/FIO
2 ratio in the first 24 hours or before instituting ILV (whichever came first).
We analyzed variables as predictors for the need for ILV by comparing preoperative and early postoperative findings in the ILV and non-ILV groups during the first 24 postoperative hours. Outcome parameters were compared between the ILV and non-ILV groups. Outcomes assessed were incidence of ILV, duration of mechanical ventilation, duration of stay in the ICU, ICU mortality, hospital mortality, and long-term survival.
General Method for SLT
Split-lung function studies were performed on all patients (nuclear quantitative perfusion scans). The worst functional side was always selected for transplantation. There were insufficient retrospective data to include this in our analysis. General methods for recipient selection, donor selection, pulmonary allograft procurement, surgical technique for SLT, and postoperative management have been described in detail previously.4,9-11
However, principles included immunosuppressive and antiviral therapy (described previously),12
maintenance of low central venous pressure, minimal PEEP, active airway clearance, maximal bronchodilator therapy, and nitric oxide, where appropriate.
Indications for ILV
There was no protocol for the initiation of ILV. Treating clinicians instigated ILV on the basis of one or a combination of the following: (1) serious hemodynamic compromise attributed to excessive dynamic hyperinflation and (2) serious gas exchange compromise attributed to excessive dynamic hyperinflation.
Transplant lung infiltrate on chest radiography was usually present but was not by itself an indication for ILV. Asymptomatic hyperinflation was not considered an indication for ILV.
Technical Aspects of ILV
Patients who required ILV were intubated with a double-lumen endotracheal tube (Bronchocath; Mallinckrodt, Athlone, Ireland; size range, 35F-41F). The side of the bronchial lumen was positioned bronchoscopically to lie in the main bronchus of the native lung to avoid injury to the bronchial anastomosis. In all patients, asynchronous ILV was used. The native lung was ventilated with low tidal volumes (3-6 mL/kg), a low ventilator rate (3-6 breaths/min), and 0 PEEP or not ventilated at all by using continuous positive airway pressure (5 cm H2O). The transplanted lung was ventilated with higher ventilator rates (10-26 breaths/min) and variable tidal volumes (3.5-10 mL/kg). PEEP (5-15 cm H2O) was adjusted to achieve a maximal improvement in gas exchange and compliance.
Statistical Analysis
All data are expressed as means ± standard deviation. Data were analyzed by using the Student t test and Wilcoxon and
2 tests as appropriate. A stepwise multivariate logistic regression model was used to correct for the effect of multiple variables and their interactions on requiring ILV in the early postoperative period. Variables that were associated with the requirement for ILV at a P value of less than .1 in the univariate analysis were entered into the multivariate model. The initial model was constructed by using preoperative variables. Postoperative variables were selected afterward and then incorporated into the final model. Software used was SAS version 9.1.3 (Enterprise Guide) for Windows (SAS Institute, Inc, Cary, NC).
| Results |
|---|
|
|
|---|
1-antitrypsin deficiency, 8 for obliterative bronchiolitis, 4 for lymphangioleiomyomatosis, 1 for agammaglobulinemia, and 1 for asthma). These 170 patients formed the basis of our analysis. ILV was used in 12% (20/170) of SLTs for obstructive lung disease. One of these 20 patients had obliterative bronchiolitis; the remainder had emphysema caused by smoking. Other than one case, in which the patient was returned to the ICU from the operating theater with a double-lumen endotracheal tube in situ, ILV was instigated by the treating intensivist. ILV was initiated within the first 24 hours in 19 patients. One patient had a pneumothorax in the native lung on the fourth postoperative day, and this patients condition subsequently deteriorated, requiring ILV. There was one instance of ILV use in a patient with restrictive lung disease caused by pulmonary fibrosis (before institution of extracorporeal membrane oxygenation), and this patient was not included in this study. Demographic data for all patients receiving an SLT for COPD are shown in Table 1. The mean age for the whole cohort was typical for this patient group, and there was a female preponderance.
|
|
|
The multivariate analysis (Table 4) demonstrated that the only 2 factors that were significantly associated with the requirement for ILV were TLC percent predicted (P = .032) and the postoperative PaO 2/FIO 2 ratio (P = .005). There was a nonsignificant trend toward a relationship with the degree of the pulmonary infiltrate (P = .06).
|
|
|
| Discussion |
|---|
|
|
|---|
The reported frequency of use of ILV in SLT series for obstructive airways disease also varies widely, from 0%13-17
to 43%,18
with an overall frequency of ILV use of 9%.6,13-21
This compares with 12% in our study. In part, this variability between series relates to the threshold for using ILV and to the availability and frequency of use of techniques such as extracorporeal membrane oxygenation.
A number of other studies have investigated risk factors for ILV, such as severity of airflow obstruction in the native lung, the size of the transplanted lung, and the side of the transplantation. Severity of airflow obstruction will directly affect the degree of dynamic hyperinflation.22
Yonan and colleagues6
found that patients who had native lung hyperinflation had higher pulmonary artery pressures, higher pulmonary artery to wedge pressure gradients, higher residual volumes, and lower FEV1 values than patients who did not have symptomatic postoperative hyperinflation. Pulmonary hypertension and a high transpulmonary gradient were not studied in our patients. Although we found the degree of preoperative airflow limitation (FEV1/forced vital capacity percent predicted) was worse in the ILV group, this was not a significant factor after multivariate adjustment. Because our population size is relatively small, it was unlikely that we would detect many significant factors in multivariate analysis.
The size of the donor lung and its relationship to estimated recipient lung volume has been identified by several authors5,17,21
as an important factor. We could not demonstrate any such relationship using radiologic, measured, or estimated measures of lung function.
Several reports have suggested a higher incidence of native lung hyperinflation with left SLTs. Weill and associates20
found radiographic hyperinflation in 37% of left SLTs and 25% of right SLTs, with no difference in the incidence of symptomatic hyperinflation. Angles and colleagues18
did not report radiographic hyperinflation but found symptomatic hyperinflation in 83% of left SLTs and 50% of right SLTs. In our study there was no significant difference in postoperative native lung hyperinflation (as evidenced by mediastinal shift) between left and right SLTs (1.8 ± 1.4 cm vs 1.5 ± 1.2 cm, respectively; P = .158), but ILV was required more than twice as often with left SLTs. However, we have previously reported23
higher long-term mortality (primarily because of airway complications) with left lung transplants compared with that seen with right lung transplants.
A reduced PaO
2/FIO
2 ratio (odds ratio, 0.97; 95% confidence interval, 0.95-0.99, P = .005) in the immediate postoperative period was associated with the requirement for ILV. This is not surprising because impaired gas exchange is likely to have been an important factor in determining the intensivists decision to initiate ILV. A reduced PaO
2/FIO
2 ratio is the basis of the definition for primary graft dysfunction24
and is associated with worse early outcomes after lung transplantation.25,26
It is possible that attempts to improve graft function by providing greater minute ventilation, PEEP, or a longer inspiratory time might worsen dynamic hyperinflation in the native lung, resulting in hemodynamic deterioration. Fluid administration as a treatment for hypotension and low cardiac output might further jeopardize graft function by increasing edema in the transplanted lung.
Other reported treatment options for severe dynamic hyperinflation in association with early graft failure include the use of a bronchial blocker,21
contralateral lobectomy or pneumonectomy,27,28
lung volume reduction surgery of the native lung,29
retransplantation,21
and extracorporeal membrane oxygenation30
(as was required by one of our patients). It is noteworthy that the reported cases of lung volume reduction surgery and contralateral lobectomy or pneumonectomy have not been performed in the setting of early graft injury as an immediate postoperative complication.
The patients who required ILV had an initial hospital mortality of 40%, although it is likely that the mortality within this group would have been much higher without this intervention. Two case series in which data on all patients undergoing SLT and ILV survival were recorded6,19
have reported mortality rates of 25% and 46%, respectively, which is significantly worse than for patients with SLT who required only conventional ventilation. However, the study by Weill and associates,20
which described a mortality of 0%, had only 2 patients who required ILV. However, it is important to note that among those who survive to leave the hospital, satisfactory long-term outcomes can be achieved, as shown in Figure 1. We know of no other reports on the long-term survival of patients who have required ILV.
Potential limitations of our study are that it is a single-center, retrospective, nonrandomized analysis of relatively small numbers of patients. Generalizing our findings to other centers should be done with caution. However, this is the largest case series of which we are aware. It is possible that patient selection for ILV could have been biased by a desire to optimize results. At our center, there has been a trend away from SLT for COPD in recent years (15 SLTs between 2003 and 2005 compared with 45 in the preceding 3 years). This has been driven by a perceived late survival advantage,3
with bilateral transplantation rather than conscious selection of patients at risk of needing ILV because of significant preoperative hyperinflation. We could not include cardiovascular parameters in our study, mainly because there were insufficient data for most patients. The presence of hypotension caused by native lung hyperinflation (with or without a low PaO
2/FIO
2 ratio) has been considered an absolute indication for initiating ILV at our hospital. Our aim was to investigate other parameters that the clinician might use to assess the risk of requiring ILV. The presence of preoperative hyperinflation with TLC percent predicted of greater than 150% should alert those selecting patients for SLT of a significant risk of requiring ILV in the postoperative period. The additional finding of a reduced PaO
2/FIO
2 ratio (<100) in the immediate postoperative period should prompt early consideration of ILV.
| Conclusions |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. Massard, N. Santelmo, P.-E. Falcoz, and R. Kessler Noninfectious complications Lung Transplantation, June 7, 2010; 177 - 193. [Abstract] [Fulltext] [PDF] |
||||
![]() |
J. G.T. Augoustides Independent lung ventilation in adult single-lung transplantation: Is it time for fast-track anesthesia and early tracheal extubation? J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 825 - 825. [Full Text] [PDF] |
||||
![]() |
D. V. Pilcher, G. M. Auzinger, B. Mitra, D. V. Tuxen, R. F. Salamonsen, A. R. Davies, T. J. Williams, and G. I. Snell Reply to the Editor J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 825 - 826. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |