Early Institution of Extracorporeal Membrane Oxygenation for Primary Graft Dysfunction After Lung Transplantation Improves Outcome

CLINICAL LUNG AND HEART/LUNG TRANSPLANTATION

Early Institution of Extracorporeal Membrane Oxygenation for Primary Graft Dysfunction After Lung Transplantation Improves Outcome Christopher H. Wigfield, MD, FRCS,a Joshua D. Lindsey, MBS,a Thomas G. Steffens, CCP,a Niloo M. Edwards, MD, FACS,a and Robert B. Love, MD, FACSb Background: Primary graft dysfunction (PGD) after lung transplantation (LTx) carries a significant mortality and clinical management is controversial. Extracorporeal membrane oxygenation (ECMO) has been used infrequently for recovery from acute lung injury (ALI) in this setting. We reviewed our experience with ECMO after primary LTx. Methods: The present study is a retrospective analysis of all LTx patients between 1991 and 2004. Twenty-two patients sustained severe PGD with subsequent placement on ECMO. We analyzed indications and 30-day, 1-year and 3-year mortality. Complications and incidence of multiple-organ failure (MOF) were determined. Critical appraisal of the evidence available to date was performed. Results: A total of 297 LTxs were performed during the study period, with 97.5%, 88.6% and 73.8% survival at 30 days, 1 year and 3 years, respectively. Twenty-two patients (7.9%) had severe allograft dysfunction leading to ECMO support. Twelve patients received single-lung (SLTx), 8 double-lung (BLTx), 1 single-lung/kidney (SLKTx) and 1 heart/lung (HLTx) transplantation. Thirty-day, 1-year and 3-year survival of LTx recipients with ECMO support post-operatively were 74.6%, 54% and 36%, respectively. MOF was the predominant cause of death (58.3%) in patients on ECMO support for PGD. Conclusions: Our data suggest that, in addition to prolonged ventilation and pharmacologic support, ECMO should be considered as a bridge to recovery from PGD in lung transplantation. Early institution of ECMO may lead to diminished mortality in the setting of ALI despite the high incidence of MOF. Late institution of ECMO was associated with 100% mortality in this investigation. J Heart Lung Transplant 2007;26:331– 8. Copyright © 2007 by the International Society for Heart and Lung Transplantation.

Severe primary graft dysfunction (PGD) remains the primary cause of 30-day mortality and amelioration would improve both early and late lung transplantation (LTx) outcomes.1 Improvements in the procurement protocols, surgical procedure and post-operative intensive care have not resulted in a notable reduction of PGD after LTx.2 Treatment remains primarily supportive for acute lung injury (ALI),3 as defined by ventilator requirements and oxygenation indices in the absence of cardiogenic and infective etiology of acute respiratory

From the aDepartment of Cardiothoracic Surgery, University of Wisconsin, Madison, Wisconsin; and bDepartment of Cardiothoracic Surgery, Loyola University Health System, Maywood, Illinois. Submitted April 25, 2006; revised November 3, 2006; accepted December 12, 2006. Reprint requests: Christopher H. Wigfield, MD, Department of Cardiothoracic Surgery and Transplantation, Freeman Hospital, High Heaton, Newcastle upon Tyne NE7 7DN, UK. Telephone: 0044 191 244 8393. E-mail: [email protected]. Copyright © 2007 by the International Society for Heart and Lung Transplantation. 1053-2498/07/$–see front matter. doi:10.1016/ j.healun.2006.12.010

failure (onset ⬍72 hours). Although ALI in this context is essentially reversible, it may progress and result in profound hypoxemia, refractory acidosis and eventually multiple-organ failure (MOF). Recent reports provided encouraging results with extracorporeal membrane oxygenation (ECMO) in selected patients in adult respiratory distress syndrome (ARDS) and suggest that ECMO may be an underutilized modality in PGD.4 Criteria and timing for institution of ECMO under these circumstances are not well defined and information regarding outcomes in this setting is scarce. We therefore retrospectively examined our experience with early ECMO institution in PGD after LTx. Our emphasis was on reviewing survival after early ECMO institution and the cause of death in patients not successfully weaned from ECMO to long-term recovery. METHODS A single-center analysis of 297 consecutive lung transplants (LTx) at the University of Wisconsin Hospital was performed. Retrospective analysis of UNOS data sheets and clinical case documentation was conducted for all 331

332

Wigfield et al.

primary single and bilateral LTxs. Patients with severe PGD, Grade 3 according to ISHLT guidelines,5 and subsequent ECMO were compared with an unaffected LTx cohort. For the evaluation of outcomes the patient cohort was divided into two groups: early institution of ECMO (⬍24 hours after LTx) and late institution of ECMO (⬎24 hours after LTx). ECMO support principles, cannulation methods, surgical techniques, anti-coagulation protocols and intensive-care records were evaluated. Ventilatory management during ECMO support and subsequent weaning strategies were assessed. We also reviewed ECMO management parameters, including oxygen saturation (SVO2) and systemic oxygenation (PaO2). The incidence of disseminated intravascular coagulation (DIC), renal failure, vascular complications and rates of infection during ECMO were established. The incidence of MOF, defined as presence of greater than two organ system dysfunctions, was assessed and reviewed as a potential cause for death after ECMO. Outcome analysis focused on death as the primary end-point at 30 days, 3 months, 1 year and 3 years, utilizing Kaplan–Meier methods. Donor selection criteria and recipient matching procedure were in accordance with international guidelines.6 Allograft preservation involved procurement with 4 to 6 liters of cold University of Wisconsin (UW) solution (50 to 100 ml/kg body weight) in all cases with pre-treatment utilizing either 1 mg prostacyclin or phentolamine (10 mg Regitine), administered prior to pulmonary flush directly via pulmonary artery (PA). Prior to en-bloc retrieval, the donor trachea was stapled with lungs inflated at a fraction of inspired oxygen (FIO2) of 1.0 and allografts were transported in UW solution at 2° to 4°C. A retrograde flush was administered utilizing 25 to 30 ml/kg of UW solution. Ischemic times for the donor allografts were recorded and the use of cardiopulmonary bypass was reviewed. Cardiopulmonary bypass was employed to facilitate implantation of lung allografts when there was either excessive PA pressure, hemodynamic instability after clamping of the PA, or when persistent hypoxemia occurred with single-lung ventilation during transplantation. Criteria for implementing ECMO included progressive hypoxemia despite optimized ventilatory support, high peak airway pressures, administration of nitrous oxide (NO), falling venous saturation (SVO2) and vasopressor requirements. The decision for ECMO institution was made for each case individually considering the following patient parameters: intra- or post-operative oxygenation indices (PaO2/FIO2); assessment of static lung compliance (tidal volume/peak airway pressure); respiratory status after optimized ventilatory settings; and the required amount of hemodynamic support needed, including pulmonary vasodilators.

The Journal of Heart and Lung Transplantation April 2007

Conventional post-operative respiratory support was predominantly pressure-controlled ventilation with optimized positive end-expiratory pressure (PEEP, maximum 10 to 12 cm H2O), with a single endotracheal tube in all patients. Fluid management was utilized to achieve negative volume balance to augment clearance of pulmonary edema. This included judicious diuretic administration and continuous volume monitoring. Contraindications for ECMO included severe biventricular cardiac failure as a result of PGD, such that venoarterial ECMO would not provide sufficient support, in addition to clinical evidence of DIC prior to ECMO initiation. All patients received venoarterial (V-A) ECMO, either through the femoral artery and vein, or central cannulation. Three patients had central V-A cannulation and the remaining 19 patients received peripheral V-A cannulation. Two patients initially on peripheral V-A ECMO required a change to venovenous (V-V) ECMO after limb-threatening vascular complications were observed. ECMO Principles and Techniques The vascular access was determined based on timing of institution of ECMO with respect to cardiopulmonary bypass (CPB). Patients with immediate primary graft dysfunction, as evident in the operating room after lung transplantation utilizing CPB, received central cannulation. This was facilitated using venous drainage from the right atrium and arterial return via the ascending aorta. The majority of patients in this cohort received ECMO via surgical cutdown of the femoral artery and vein while in the intensive-care unit. Femoral cannulae used were 14F to 21F for arterial and 17F to 29F for venous cannulation. The ECMO circuit oxygenators employed in this case series were Carmeda-coated Maxima-Plus PRF (Medtronic, Inc., Anaheim, CA) and Affinity NT (Medtronic) membranes in addition to the Biomedicus BP-80 Centrifugal Pump (Medtronic, Inc., Minneapolis, MN). ECMO flow rates were maintained at 2 to 2.5 liters/min/m2 and the sweep rate adjusted according to patient PCO2. Anti-coagulation with heparin was commenced typically within 24 hours to achieve activated clotting times of 180 ⫾ 20 seconds, unless specific contraindications, such as coagulopathy with hemorrhage, was prohibitive. Coagulation parameters were monitored as required in addition to repeated platelet counts. Platelet administration was considered in patients with active bleeding or for ⬍100,000 platelets/ml. Adequate oxygenation was assessed continuously by systemic oxygen saturation (⬎95%) as well as monitoring of mixed venous blood gas (SVO2), with a target of ⬎65%. Ventilatory management included reduction of tidal volumes and peak airway pressures to minimize volutrauma and barotrauma to lung allografts. Inspired

The Journal of Heart and Lung Transplantation Volume 26, Number 4

oxygen concentrations (FIO2) were weaned empirically to reduce the potential of oxygen toxicity. All ECMO protocols were implemented 24 hours/day by certified cardiovascular perfusionists under the supervision of senior clinical staff. An aggressive approach to nutritional status optimization was implemented, including enteral feeding within 48 hours after ECMO implementation when feasible; total parenteral alimentation was avoided whenever possible. Statistical Analysis Standard descriptive statistics were employed for the patient demographics and comparison of early and late ECMO group differences. A Kaplan–Meier analysis of survival at 3 months, 1 year, 3 years and 5 years was performed. Statistical assessment was done using EXCEL 2002 (Microsoft, Redmond, WA). RESULTS Patient Demographics Twenty-two patients, 7.9%, had severe PGD (ISHLT Grade 3) requiring ECMO support due to refractory hypoxemia after lung transplantation (male-to-female ratio: 1.2:1). The primary pulmonary diagnoses of recipients leading to end-stage respiratory failure prior to transplantation are shown in Tables 1 and 2. Eleven of these patients were single-lung transplant (SLTx) recipients, 8 bilateral lung transplants (BLTx), 1 heart–lung transplant (HLTx), 1 patient was an SLTx who was placed on ECMO as a bridge to bilateral re-transplantation (BLTx, Patient 8, see Table 1), and 1 patient was the recipient of a combined SLTx and renal transplantation (Patient 22, see Table 2). Seventeen patients received ECMO support within 24 hours after lung transplantation (early ECMO cohort) and five recipients had ALI resulting in ECMO institution ⬎24 hours after completion of lung transplantation (late ECMO cohort). The median of time for institution of ECMO in the late group was 11 days (range 2 to 81 days) post-operatively for potentially reversible respiratory failure. Median age at time of LTx was 48 and 47 years for the early and late ECMO groups, respectively (range 24 to 67 years). Patient characteristics and clinical information is summarized in Tables 1 and 2. Outcomes A total of 297 consecutive lung transplants were performed between July 1991 and July 2004. Of these, 286 were primary SLTx or BLTx recipients. Twenty-two of the 286 patients (7.9%) had severe allograft dysfunction, meeting the criteria for ALI subsequently requiring ECMO support. Cardiogenic, infective or other nonrespiratory causes for respiratory failure were ruled out in all patients prior to commencing ECMO support.

Wigfield et al.

333

The mean ischemic time for allografts transplanted requiring ECMO was 435.9 minutes (all ECMO) and 452.7 minutes for early ECMO recipients (p ⫽ not significant [NS]; Table 3). Lung allograft ischemia duration was similar when compared with all other LTx patients not having PGD (422.9 minutes, p ⫽ NS). The range of ECMO duration varied from 5 to 100 hours, with a median of 47 hours for the entire ECMO cohort reviewed. The mean support duration for patients with early ECMO and late ECMO institution was 58.6 hours and 51 hours, respectively (p ⫽ NS). Late ECMO was associated with 100% mortality in this study, with a mean survival of 33 days. The mean survival time of patients dying with MOF was 62 days (early and late ECMO groups). Early ECMO support for PGD resulted in 67% and 49% survival at 1 and 5 years, respectively. Two of the five patients in the late ECMO cohort were weaned off of ECMO successfully with satisfactory oxygenation but did not recover from MOF or sepsis. The remaining three patients died while on ECMO or had ECMO withdrawn as all measures proved futile. Kaplan–Meier survival plots with sub-group analysis of ECMO recipient cohorts is shown in Figure 1. The comparative survival curve for all LTx recipients without severe PGD is provided with resulting survival at 30 days and 1 and 5 years of follow-up. When survival was plotted as conditional on 1-year survival in the two groups, 66% of patients with ECMO support for severe PGD after LTx were alive at 4 years, compared with 70% of unaffected lung transplant recipients (data not shown). DISCUSSION The renewed interest in the application of ECMO for potentially reversible causes of respiratory failure has coincided with the refined technology and innovative ventilator strategies developed in the 1990s. The primary principle in this approach is to provide membrane oxygenation as a temporary replacement measure to allow for intrinsic mechanisms to repair the diffuse alveolar damage in a non-proliferative stage of acute lung injury. Although conceptually appealing, ECMO after lung transplantation has been viewed as a last resort for patients with edema from severe ischemia– reperfusion injury. A number of case reports have described successful management of patients utilizing ECMO after LTx.7–12 ECMO has also been used as a bridge to primary or re-do lung transplantation in a number of ventilator-dependent patients, with variable success.9 –11 Anderson et al compared the historically early experiences with ECMO for PGD with more recent experiences and provided evidence for the utility of ECMO in supportive critical care.13 In addition, sev-

334

Wigfield et al.

The Journal of Heart and Lung Transplantation April 2007

Table 1. Patient Demographics and Cohort Analysis

Pt. no.

Age/ gender

Diagnosis

Transplant SLTx

Status (days): Alive/ Expired

CPB use (yes/no)

ECMO start post-Tx (h)

ECMO time (h)

Yes

1

45

3E

1

67 F

PF

2

46 M

PHTN secondary to cardiac

H/LTx

Yes

12

71

4E

3

43 F

CF

BLTx

Yes

3

104

338 A

4 5 6

55 M 58 M 53 F

COPD PF COPD

SLTx SLTx SLTx (ASD closure)

Yes Yes Yes

1 0 20

98 (weaned) 27 38 (weaned)

54 E 56 E 166 E

7 8

56 M 56 F

COPD PF

SLTx SLTx (retransplant)

No Yes

2 1

9 10

61 F 48 F

COPD COPD

SLTx SLTx

Yes Yes

4 23

11

30 M

CF

BLTx

Yes

NA

37

651 E

12 13 14 15

33 F 38 M 28 F 33 F

CF CF CF PF

BLTx BLTx BLTx BLTx

Yes No Yes Yes

0 12 2 0

76 49 58 52

964 A 1,042 A 1,295 A 1,348 A

16 17

24 M 51 M

CF COPD

BLT SLTx

Yes No

9 6

27 63

1,506 A 1,034 A

37 79

93 96 (weaned)

535 A 562 A

601 A 582 E

Complications during ECMO Disseminated coagulopathy, hemorrhage, massive transfusions Coagulopathy mediastinal bleeding ASD closure hepatorenal syndrome Membrane failure, bleeding at cannulation site, hemodialysis Renal failure, CVVH Bleeding Bleeding at cannulation site, CMV pneumonitis, line sepsis, tracheostomy, cerebellar hematoma None Compartment syndrome, left leg amputation, cholestasis, stent, hemopericardium— pericardial window, delayed wound healing Cardiac pacer Rejection, left pneumothorax, left bullectomy, CMV pneumonitis Gram-negative sepsis

None None Membrane failure Bleeding, hemofiltration None None

Cause of death or outcome Exsanguination

Hemorrhage, MOF

Alive

MOF MOF Cardiac arrest

Alive Alive

Alive Infection/Grade III rejection

Re-transplant, renal failure, pan-resistant Pseudomonas, pneumonia, MOF Alive Alive Alive Alive Alive Alive

Early ECMO (institution ⬍24 hours after lung transplantation). For abbreviations see Table 2.

eral centers have documented improved outcomes after recent refinements of both surgical technique as well as ECMO technology and timing.13–16 Temporary extracorporeal maintenance of systemic oxygenation allows for advanced ventilatory strategies to influence recovery from alveolar volutrauma and inflammatory parenchymal lung injury.

Surgical Approaches Both venovenous and venoarterial perfusion circuits have been successfully employed in the management of ARDS and after lung transplantation for PGD.17 Venovenous ECMO has been the preferred option for PGD after LTx in some centers.14 Clinical advantages of venovenous ECMO include relative technical ease of

The Journal of Heart and Lung Transplantation Volume 26, Number 4

Wigfield et al.

335

Table 2. Patient Demographics and Cohort Analysis

CPB use No

ECMO start post-Tx 6 days

ECMO time (h) 17

Status (days) Alive/Expired 9E

SLTx

Yes

2 days

81 (weaned)

64 E

CF

BLTx

Yes

11 days

38

13 E

47 (M)

COPD

SLKTx

No

16 days

38

19 E

61 (M)

COPD

SLTx

No

85 days

81 (weaned)

102 E

Age (gender) 33 (F)

Diagnosis PF

Transplant SLTx

19

59 (M)

COPD

20

24 (M)

21

22

Pt. no. 18

Complications during ECMO Anoxic brain injury, renal failure, compartment syndrome fasciotomy bleeding Bronchial stent placement, supporative cholangitis CVVH for renal failure, Gram-negative sepsis, tracheostomy Enterococcus infection, Pseudomonas and MRSA, pneumonia, coagulopathy Enterovirus infection, right lung consolidation, pulmonary hemorrhage Renal failure, Aspergillus sepsis, hemodialysis

Cause of death MOF

MOF

Bilateral pneumonia

MOF

Sepsis

Late ECMO (institution ⬎ 24 hours after lung transplantation). PF, fibrotic lung diseases; COPD, chronic obstructive and emphysematous pulmonary diseases; PHTN, pulmonary hypertension; CF, cystic fibrosis; Tx, transplant; SLTx, single-lung transplant; BLTx, bilateral lung transplant; H/LTx, heart–lung transplant; SLKTx, single-lung and kidney transplant; MO, multiple-organ failure; NA, data not available.

cannulation, increased aortic oxygen saturation, and reduced risk of systemic embolization. This approach has significant disadvantages, including right heart failure and relative coronary artery hypoxemia. These problems when encountered render the patient more vulnerable to hemodynamic instability in the immediate post-transplant period. The venoarterial approach, when instituted early, can provide a higher oxygen delivery capacity and additional circulatory support to facilitate early recovery from the ALI process. Reducing pulmonary vascular flow potentially modulates the endothelial activation and aggravation of pulmonary edema secondary to reperfusion injury. In addition, reduced pulmonary vascular resistance with the use of inhaled nitrous oxide may be more effective with this approach. Concerns raised by Glassman et al regarding bronchial anastomotic healing without bronchial artery perfusion may be counterbalanced with this approach.18 The clinical advantages of the venovenous and venoarterial approaches for ECMO in this setting are summarized in Table 4. Peripheral cannulation is feasible in the ICU setting for PGD after lung transplantation. Central CPB cannulation can be utilized to continue support with ECMO in patients with fulminant allograft failure immediately after reperfusion. This was required in three patients in

our series due to profound hypoxemia after weaning CPB support. This approach requires delayed primary chest closure after discontinuation of ECMO and potentially increases the infectious risks. In recipients with progressive intra-operative hypoxemia or with marked pulmonary artery pressure rise during implantation after cross-clamping, CPB is required to facilitate the procedure. CPB has been implicated as a potentially aggravating factor in reperfusion injury after ischemia of allografts in lung transplantation due to the systemic inflammatory response elicited.19 The frequent need for CPB in the early ECMO group (13 of the 17 patients) may have influenced outcomes, theoretically worsening existing pulmonary reperfusion response and PGD. The ventilatory strategy during ECMO requires an individually tailored approach to optimize lung recovery by minimizing airway pressures and inspired oxygen concentration and allowing for “best PEEP” principles with limited hypercapnia. Pulmonary vasodilators and nitric oxide may provide additional pharmacologic support. Three of seven deaths in the early ECMO group and two of the five in the late ECMO group were in patients who underwent single-lung transplantation for chronic obstructive pulmonary disease (COPD). Despite refined ventilator techniques to optimize recovery of lung

336

Wigfield et al.

The Journal of Heart and Lung Transplantation April 2007

Table 3. Mean ECMO Data Comparison for Early and Late ECMO Cohortsa Early ECMO (n ⫽ 10)

ECMO parameter Pre-ECMO ventilator settings Rate (Hz) FIO2 (%) PIP (mm Hg) PEEP (cm H2O) Pre-ECMO hemodynamics Blood pressure Systolic (mm Hg) Diastolic (mm Hg) Mean arterial (mm Hg) Pulmonary artery pressure Systolic (mm Hg) Diastolic (mm Hg) Mean (mm Hg) Cardiac index (liters/min/m2) Pump flow (liters/min) Hour 4 Hour 24 Hour 24 ventilator settings Rate (Hz) FIO2 (%) PIP (mm Hg) PEEP (cm H2O) Mean airway pressure

Late ECMO (n ⫽ 5)

15 (10–32) 99.3 (92–100) 38.5 (25–48) 10.1 (5–15)

17 (14–20) 100 (100) 31 (26–39) 12 (12)

84 (60–103) 54 (40–77) 70 (54–99)

93 (72–114) 44 (32–56) 60 (45–75)

43 (24–69) 26 (12–50) 29 (19–42) 2.8 (1.9–4.6)

46 (37–55) 31 (30–33) 38 (32–44) 4.3 (2–6.7)

4.1 (3.3–5) 3.7 (3–4.2)

3.7 (3.2–4.4) 4.4 (4.3–4.4)

7.3 (6–10) 50.9 (40–80) 30.4 (20–39) 6.6 (5–10) 11.2 (8–16)

12 (10–14) 45 (40–50) 27.5 (26–29) 5 (5) 11.5 (10–13)

FIO2, fraction of inspired oxygen; PIP, positive inspiratory pressure; PEEP, positive end-expiratory pressure. a Data collected and reported to the extracorporeal life support organization (ELSO) starting in the Year 2000. Numbers are rounded to nearest tenth decimal place, with the exception of pH. Ranges are included within parentheses.

allografts, separate lung ventilation may be beneficial in these patients. We recorded improved systemic oxygenation in both the early and late groups during ECMO, compared with baseline blood gas analysis (see Table 5). Permissive hypercapnia did not result in respiratory acidosis in either group during ECMO. ECMO with 100

patient survival (%)

90 80 70 60 50 40

All Lung Tx

30

Early

20

Late

10

ALL ECMO

0 0

1

2

3

4

5

6

7

8

9

10

11

12

13

years post transplant

Figure 1. Kaplan–Meier survival curves. Extracorporeal membrane oxygenation support cohorts after lung transplantation with severe PGD compared with all lung transplant recipients without severe PGD.

Table 4. ECMO Cannulation for PGD After Lung Transplantation: Physiologic Considerations and Potential Complications of Venovenous and Venoarterial ECMO Support Issue and Factor Cannulation Technique Flow rates Oxygenation O2 delivery PaO2 range Pulmonary physiology Pulmonary perfusion Pulmonary emboli Left to right shunt Right to left shunt Systemic adverse effects CNS/CVS emboli Cardiovascular effects Pre-load effect

Cardiac effect Circulatory support

Venovenous ECMO

Venoarterial ECMO

One or two sites Up to 130 ml/kg/ min

Central or femoral Up to 100 ml/kg/min

Acceptable PaO2: 40–80 mm Hg

Excellent PaO2: 80–150 mm Hg

Not affected, possible

Decreased flow, none

Essentially none

Decreased aortic saturation

Unlikely

Possible

Increased venous return

Decreased pre-load, increased afterload Partial/complete CPB

Modestly improved CO Minimal cardiac effect

O2

Potential for reduced inotropes

continued ventilation allows for uncomplicated CO2 elimination and correction of acid– base imbalances. Refined ventilatory strategies, including inverse ratio ventilation (IRV), pressure-limited ventilation (PLV), prone positioning and high-frequency ventilation (HFV), have had a noticeable impact on oxygenation and survival in patients with ARDS and require a formal evaluation in this setting. We observed complications in 15 of 22 patients. Clinically significant complications included ventilatorassociated pneumonia, sepsis and sepsis syndrome, vascular complications requiring surgical intervention and coagulopathies. Coagulopathies and bleeding problems requiring invasive measures were observed in six patients. Three cannulation site–related vascular problems resulted in transfer to venovenous support in two of these and a below-knee amputation in one survivor after early ECMO. Cannulation options for both venovenous and venoarterial approaches have to be considered individually for each ECMO candidate and improved outcomes have recently been reported with venovenous support. Hartwig et al compared venovenous to veno-arterial ECMO in their cohort of ECMO patients after PGD and documented better outcomes after improving their anti-microbial, specifically their anti-fungal, regimen.14

The Journal of Heart and Lung Transplantation Volume 26, Number 4

Wigfield et al.

337

Table 5. Arterial Blood Gas Comparison: Pre- and Post-ECMO Institution Data for Early and Late ECMO Groups Early ECMO group (n ⫽ 17) Arterial blood gases pH PaCO2 PaO2 HCO3 O2 saturation (%)

Pre-ECMO 7.32 (6.89–7.45) 53.6 (35–123) 46.6 (29–98) 25.8 (18.9–30.6) 76 (57–94)

Late ECMO group (n ⫽ 5)

Post-ECMO 7.40 (6.95–7.57) 44.7 (35–134) 285.0 (74–446) 25.6 (19.6–31.4) 96.5 (93–98)

Pre-ECMO 7.25 (7.25–7.44) 63.5 (45–90) 52.5 (39–62) 26.1 (19.9–26.1) 75.1 (64.5–87)

Post-ECMO 7.44 (7.35–7.57) 34.2 (25–45) 360.8 (273–468) 23.4 (16–30.5) 97.3 (95.5–98)

Values represent the arterial blood gas (ABG) values up to ECMO initiation and the preceding ABG value after ECMO initiation. Mean values are shown (range, low– high). Numbers are rounded to nearest tenth decimal place, with the exception of pH.

Sepsis and septis syndrome occurred in six patients and was associated with MOF in four of five patients in the late ECMO group. Sepsis therefore constitutes a further limiting factor in ECMO, complicating the potential resolution of diffuse alveolar damage. Other clinically significant complications include ventilatoryassociated pneumonia and renal failure requiring temporary hemofiltration (CVVH) or hemodialysis. Complications affecting more than one organ system were seen in 14 of 22 patients and MOF (more than two organ systems) was the predominant cause of death (58.3%) after ECMO for PGD. MOF was also the predominant cause of death in patients receiving early ECMO for PGD (57.1%; see Tables 1 and 2). MOF is clearly associated with ECMO and carries a poor prognosis in this setting. Remarkably, the incidence of central nervous system (CNS) complications in this cohort utilizing venoarterial ECMO was very low (2 of 22 patients). One patient had a clinically devastating global CNS anoxia, and a second patient had a clinically compensated dense cerebellar hemorrhage (see Tables 1 and 2). Others investigators have reported a higher incidence of CNS complications than reported herein.14 All except one patient in the late ECMO group were supported prior to 1997 and historic bias is possible. Membrane failure requiring discontinuation of ECMO support was noted in two patients but they were weaned off ECMO successfully. ECMO system requirements and continuous monitoring have considerable resource implications as well as the need for set up on very short notice. ECLS protocol improvements during the study period may also have contributed to the historic bias in our study. Limitations of this study include the retrospective nature of analysis, the undetermined influence of intraoperative CPB on PGF, and protocol changes made after the initial experience with ECMO in this case series. A concerted, prospective, multicenter study would allow for a more conclusive evaluation of the benefits of ECMO in PGD after lung transplantation. In conclusion, ECMO should be considered early in patients likely to require prolonged ventilation and

pharmacologic support after lung transplantation in the presence of severe PGD. We have provided evidence justifying ECMO as a bridge to recovery from acute allograft lung injury in the absence of cardiogenic and infectious etiology for allograft failure. Our findings clearly indicate improved outcomes with utilization of ECMO support within 24 hours after lung transplantation in patients with PGD. Later institution of ECMO has proven futile in our experience. The primary cause of death observed in these patients has been MOF, despite reliable improvement in oxygenation status and weaning off ECMO in the majority of these patients. REFERENCES 1. Boucek MM, Edwards LB, Keck BM, et al. Registry of the International Society for Heart and Lung Transplantation: eighth official pediatric report—2005. J Heart Lung Transplant 2005;24:968 – 82. 2. Kuntz X, Hadjiliadis X, Sager X, et al. Changes in incidence and 30-day mortality of primary graft dysfunction. J Heart Lung Transplant 2006;25:47–57. 3. Artigas A, Bernard GR, Carlet J, et al. The American–European Consensus Conference on ARDS. Part 2. Ventilatory, pharmacologic, supportive therapy, study design strategies and issues related to recovery and remodeling. Intensive Care Med 1998;24:378 –98. 4. Oto T, Rosenfeldt F, Rowland M, et al. Extracorporeal membrane oxygenation after lung transplantation: evolving technique improves outcomes. Ann Thorac Surg 2004;78:1230 –5. 5. Christie JD, Carby M, Bag R, et al. Report of the ISHLT Working Group on Primary Lung Graft Dysfunction. Part II: Definition. A consensus statement of the International Society for Heart and Lung Transplantation. J Heart Lung Transplant 2005;24:1454 –9. 6. Orens JB, Boehler A, de Perrot M, et al. A review of lung transplant donor acceptability criteria. J Heart Lung Transplant 2003;22:1183–200. 7. Ball JW Jr, Noon GP, Short HD, Scheinin SA. Extracorporeal membrane oxygenation for early graft dysfunction in lung transplantation: a case report. J Heart Lung Transplant 1997;16:468 –71. 8. Badesch DB, Zamora MR, Jones S, et al. Independent ventilation and ECMO for severe unilateral pulmonary edema after SLT for primary pulmonary hypertension. Chest 1995;107:1766 –70. 9. Haverich A, Hirt SW, Wahlers T, et al. Functional results after lung retransplantation. J Heart Lung Transplant 1994;13:48 –54. 10. Jurmann MJ, Schaefers HJ, Demertzis S, et al. Emergency lung transplantation after extracorporeal membrane oxygenation. ASAIO J 1993;39:M448 –52. 11. Demertzis S, Haverich A, Ziemer G, et al. Successful lung transplantation for posttraumatic adult respiratory distress syn-

338

12.

13.

14.

15.

Wigfield et al.

drome after extracorporeal membrane oxygenation support. J Heart Lung Transplant 1992;11:1005–7. Jurmann MJ, Haverich A, Demertzis S, et al. Extracorporeal membrane oxygenation as a bridge to lung transplantation. Eur J Cardio-Thorac Surg 1991;5:94 –7. Anderson HL III, Delius RE, Sinard JM, et al. Early experience with adult extracorporeal membrane oxygenation in the modern era. Ann Thorac Surg 1992;53:553– 63. Hartwig MG, Appel JZ III, Cantu E III, et al. Improved results treating lung allograft failure with venovenous extracorporeal membrane oxygenation. Ann Thorac Surg 2005;80:1872– 80. Meyers BF, Sundt TM III, Henry S, et al. Selective use of extracorporeal membrane oxygenation is warranted after lung transplantation. J Thorac Cardiovasc Surg 2000;120:20 – 6.

The Journal of Heart and Lung Transplantation April 2007

16. Dahlberg PS, Prekker ME, Herrington CS, et al. Medium-term results of extracorporeal membrane oxygenation for severe acute lung injury after lung transplantation. J Heart Lung Transplant 2004;23:979 – 84. 17. Pierre AF, Keshavjee S. Lung transplantation: donor and recipient critical care aspects. Curr Opin Crit Care 2005;11:339 – 44. 18. Glassman LR, Keenan RJ, Fabrizio MC, et al. Extracorporeal membrane oxygenation as an adjunct treatment for primary graft failure in adult lung transplant recipients. J Thorac Cardiovasc Surg 1995;110:723– 6. 19. Keshavjee S, Davis RD, Zamora MR, et al. A randomized, placebocontrolled trial of complement inhibition in ischemia–reperfusion injury after lung transplantation in human beings. J Thorac Cardiovasc Surg 2005;129:423– 8.