010 cypel normothermic evlp in clinical tx

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 364;15 nejm.org april 14, 2011 1431

original article

Normothermic Ex Vivo Lung Perfusion in Clinical Lung Transplantation

Marcelo Cypel, M.D., Jonathan C. Yeung, M.D., Mingyao Liu, M.D., Masaki Anraku, M.D., Fengshi Chen, M.D., Ph.D., Wojtek Karolak, M.D.,

Masaaki Sato, M.D., Ph.D., Jane Laratta, R.N., Sassan Azad, C.R.A., Mindy Madonik, C.C.P., Chung-Wai Chow, M.D., Cecilia Chaparro, M.D., Michael Hutcheon, M.D., Lianne G. Singer, M.D., Arthur S. Slutsky, M.D.,

Kazuhiro Yasufuku, M.D., Ph.D., Marc de Perrot, M.D., Andrew F. Pierre, M.D., Thomas K. Waddell, M.D., Ph.D., and Shaf Keshavjee, M.D.

From the Toronto Lung Transplant Pro-gram (M.C., J.C.Y., M.L., M.A., F.C., W.K., M.S., J.L., S.A., M.M., C.-W.C., C.C., M.H., L.G.S., K.Y., M.P., A.F.P., T.K.W., S.K.) and the Interdepartmental Division of Critical Care Medicine (A.S.S.), University of To-ronto; the McEwen Centre for Regenera-tive Medicine, Toronto General Research Institute (M.C., M.L., T.K.W., S.K.); and the Keenan Research Centre, Li Ka Shing Knowledge Institute, St. Michael’s Hospi-tal (A.S.S.) — all in Toronto. Address re-print requests to Dr. Keshavjee at Toron-to General Hospital, 200 Elizabeth St., 9N946, Toronto, ON M5G 2C4, Canada, or at [email protected].

N Engl J Med 2011;364:1431-40.Copyright © 2011 Massachusetts Medical Society.

A bs tr ac t

Background

More than 80% of donor lungs are potentially injured and therefore not considered suitable for transplantation. With the use of normothermic ex vivo lung perfusion (EVLP), the retrieved donor lung can be perfused in an ex vivo circuit, providing an opportunity to reassess its function before transplantation. In this study, we exam-ined the feasibility of transplanting high-risk donor lungs that have undergone EVLP.

Methods

In this prospective, nonrandomized clinical trial, we subjected lungs considered to be high risk for transplantation to 4 hours of EVLP. High-risk donor lungs were defined by specific criteria, including pulmonary edema and a ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (PO2:FiO2) less than 300 mm Hg. Lungs with acceptable function were subsequently transplanted. Lungs that were transplanted without EVLP during the same period were used as controls. The pri-mary end point was primary graft dysfunction 72 hours after transplantation. Sec-ondary end points were 30-day mortality, bronchial complications, duration of me-chanical ventilation, and length of stay in the intensive care unit and hospital.

Results

During the study period, 136 lungs were transplanted. Lungs from 23 donors met the inclusion criteria for EVLP; in 20 of these lungs, physiological function remained sta-ble during EVLP and the median PO2:FiO2 ratio increased from 335 mm Hg in the donor lung to 414 and 443 mm Hg at 1 hour and 4 hours of perfusion, respectively (P <0.001). These 20 lungs were transplanted; the other 116 lungs constituted the con-trol group. The incidence of primary graft dysfunction 72 hours after transplantation was 15% in the EVLP group and 30% in the control group (P = 0.11). No significant differences were observed for any secondary end points, and no severe adverse events were directly attributable to EVLP.

Conclusions

Transplantation of high-risk donor lungs that were physiologically stable during 4 hours of ex vivo perfusion led to results similar to those obtained with conventionally selected lungs. (Funded by Vitrolife; ClinicalTrials.gov number, NCT01190059.)


n engl j med 364;15 nejm.org april 14, 20111432

Lung transplantation is lifesaving for patients with end-stage lung diseases. However, the number of patients waiting for

a lung transplant greatly exceeds the number of available donors. On average, only 15% of lungs from multiorgan donors are used for transplan-tation1; the rest are considered unsuitable owing to the lung injury that occurs after brain death and to complications associated with treatment in the intensive care unit (ICU) (e.g., barotrauma and pulmonary edema). Although nonstandard donor lungs (i.e., lungs with suboptimal gas-exchange function or infiltrates visible on chest radio-graphs)2 have been used successfully,3,4 increased primary graft dysfunction — an acute lung injury that appears within 72 hours after transplantation — has been reported in some studies.2,5,6 Such injury affects early outcomes and is associated with an increased risk of chronic graft dysfunc-tion.7-9 Thus, clinicians tend to be highly conser-vative when selecting donors,10 and because of the relatively small number of organs that are deemed to be acceptable, mortality is high among patients awaiting transplantation.11,12

Increased use of available lungs is a promising means of augmenting the number of lung trans-plants. Although management strategies in the multiorgan donor are important in preventing lung deterioration,13 organs are often retrieved before the lungs can recover from brain death or related injuries.14 The use of static hypothermia is widely accepted for preserving lung viability after removal,15,16 but the inhibition of cellular metab-olism as a result of hypothermia17 makes it dif-ficult to repair lungs or assess them (i.e., retest the organ) during the preservation period. Ex vivo lung perfusion (EVLP) with the use of acellular normothermic perfusate after organ retrieval is a potential method for assessing lung viability. With EVLP, lungs are perfused and ventilated ex vivo at body temperature to mimic physiologic conditions. Preclinical studies have shown that normal and injured donor lungs that were main-tained for up to 12 hours in the EVLP system had excellent, sustained lung function after trans-plantation.18,19 In this study, we examined the feasibility of assessing the physiological integrity of high-risk lungs before implantation by means of EVLP.

Me thods

Study DesignIn this single-institution, prospective, nonrandom-ized trial, we compared outcomes in recipients of high-risk donor lungs that had been subjected to normothermic EVLP20 and contemporaneous re-cipients of conventionally assessed lung trans-plants. Donor lungs that met the entry criteria (i.e., were at high risk for nonuse) were retrieved, delivered to our center by means of standard cold-storage transport in a low-potassium dex-tran solution (Perfadex, Vitrolife), and perfused in the EVLP system for 4 hours. Lungs were con-sidered suitable for transplantation if during EVLP the PO2:FiO2 ratio — that is, the partial pressure of oxygen ex vivo (PO2) to the fraction of inspired oxygen (FiO2) — was 350 mm Hg or more and if deterioration from baseline levels of all three physiological measurements (pulmo-nary vascular resistance, dynamic compliance, and peak inspiratory pressure) was less than 15% while the lungs were ventilated with the use of a tidal volume of 7 ml per kilogram of donor body weight and a rate of 7 breaths per minute during the perfusion period.

Recipients provided written informed consent to participate in this study in accord with the protocol, which was approved by the University Health Network research ethics board. Recipients were chosen sequentially, independently of their assignment to either the EVLP group or the con-trol group, and were selected on the basis of blood type, size of the organ (total lung capacity), and waiting-list status, in keeping with our usual practice. Group assignments were based solely on whether the donor lung met the entry criteria; if it did, it was implanted in the intended recipi-ent after EVLP. There was one exception, since one recipient had not provided consent for participa-tion (>90% of patients on the waiting list provid-ed consent during the pretransplantation visits). In that case, the next recipient was chosen. A separate consent form was used for the three initial patients in the safety and feasibility study. Recipients were informed before transplantation that they would be receiving EVLP lungs. After the lungs were transplanted, standard care was provided in both groups, including fluid manage-

Normothermic Ex Vivo Perfusion in Lung Tr ansplants


ment, antibiotic prophylaxis, immunosuppression, and surveillance bronchoscopy.21

The study was conducted in accordance with the protocol. The sponsor (Vitrolife) was not involved in the conduct of the study, analysis or storage of the data, or preparation of the manu-script.

Inclusion and Exclusion Criteria

DonorsHigh-risk donor lungs were defined as those meeting any one of the following criteria: best ra-tio of the partial pressure of arterial oxygen (PaO2) to FiO2 of less than 300 mm Hg; pulmonary ede-ma, defined as bilateral interstitial infiltrates with-out evidence of infection, detected on the last chest radiograph by the lung-transplantation physician assessing the donor; poor lung deflation or infla-tion during direct intraoperative visual examina-tion at the donor site; blood transfusions exceed-ing 10 units; and donation after cardiac death, as defined by Maastricht category III (donor without a heartbeat and with cardiocirculatory death im-minent after withdrawal of treatment) or category IV (cardiocirculatory death in a brain-dead do-nor).22 Donor lungs with established pneumonia, severe mechanical lung injury (i.e., contusions in more than one lobe), or gross gastric aspiration were excluded. Donor age was neither an exclusion nor an inclusion criterion.

A previously described donor score based on age, status with respect to smoking, chest radio-graphic findings, bronchopulmonary secretions, and arterial blood gas measurements was used to compare the severity of donor risk factors be-tween groups.23 The score ranges from 0 to 18, with higher scores indicating more risk factors; a maximum score of 7 was considered to be the cutoff level for transplant acceptability.23

RecipientsAll patients on our waiting list for single or bilat-eral transplantation or re-transplantation were eli-gible. Candidates for combined heart–lung trans-plantation were excluded.

Study Logistics

Donor lungs were offered to our lung-transplanta-tion program through our provincial organ pro-

curement organization (Trillium Gift of Life Net-work). Assessment of potential donor lungs was based on the usual constellation of clinical fac-tors, including history, PaO2:FiO2, bronchoscopic findings, radiologic assessment, and direct ex-amination of the lung during procurement. As is our practice, we pretreated all potential lung do-nors with glucocorticoids (methylprednisolone, 15 mg per kilogram of body weight given intrave-nously every 24 hours); according to the ICU policy of the donor hospital, antibiotics were given to donors if a specific organism had been identified on bronchoalveolar lavage. Donor lungs were in-cluded in the EVLP group only after a complete assessment by our team was carried out at the donor hospital before harvesting. Once accepted for EVLP, donor lungs were transported to our center, where a team of trained nurses, perfusion-ists, and transplantation surgeons had prepared the EVLP system in a sterile operating room. All lung-transplantation surgeons were aware of the design and objectives of the study.

A safety and logistic feasibility study was first performed in three patients who underwent bilat-eral lung transplantation, with the use of stan-dard criteria for donor lungs. One lung was trans-planted according to conventional practice, and the other was transplanted after 1 hour of EVLP (see Fig. 1A in the Supplementary Appendix, avail-able with the full text of this article at NEJM.org). Post-transplantation chest radiographs and bron-choscopic inspection of the anastomosis in the lungs transplanted according to conventional practice and the lungs transplanted after EVLP were similar (Fig. 1B and 1C in the Supplemen-tary Appendix).

Ex Vivo Lung Perfusion

TechniqueThe acellular EVLP technique has been described in detail elsewhere (Fig. 2 in the Supplementary Ap-pendix, and Video 1, available at NEJM.org).18,19,24 For details concerning the EVLP technique and the composition of the perfusate, see the descrip-tion and Table 1 in the Supplementary Appendix. After 4 hours of EVLP, the lungs were cooled to 10°C over a period of 10 minutes. Thereafter, per-fusion and ventilation were stopped (with FiO2 changed to 0.5 for the purpose of lung storage),

A video showing acellular ex vivo lung perfusion is available at NEJM.org



and the trachea was clamped at full inspiration to maintain the lungs in a state of inflation. The lungs were then stored at 4°C in Perfadex until transplantation.

Ex Vivo Functional AssessmentFor the functional assessment ex vivo, tidal vol-ume was set at 10 ml per kilogram of donor body weight and 10 breaths per minute, with FiO2 at 1.0. Lung function was evaluated hourly during EVLP according to the following calculations: PO2 = [left atrial PO2 – pulmonary-artery PO2 (in mm Hg)], and pulmonary vascular resistance = [(pulmonary-artery pressure – left atrial pressure) × 80] ÷ pulmonary-artery f low (in dynes · sec-onds · cm–5), dynamic compliance (in milliliters per centimeter of water), and peak inspiratory pressure (in centimeters of water). Radiography of the ex vivo lung and flexible bronchoscopy were performed at 1 hour and 3 hours of EVLP.

Study End Points

The primary end point was primary graft dysfunc-tion of grade 2 (PaO2:FiO2 of 200 to 300 mm Hg) or grade 3 (PaO2:FiO2 <200 mm Hg), according to the International Society for Heart and Lung Transplantation classification (ISHLT),25 at 72 hours after transplantation. Grades 0 and 1 repre-sent good graft function (PaO2:FiO2 >300 mm Hg) without abnormalities on chest radiographs (grade 0) or with radiographic abnormalities (grade 1). Secondary end points were PaO2:FiO2 at the time of arrival in the ICU and at 24 hours, and 48 hours after lung transplantation; the need for extracor-poreal membrane oxygenation; bronchial compli-cations requiring intervention; duration of me-chanical ventilation; length of stays in the ICU and hospital; and mortality at 30 days.

Statistical Analysis

All statistics were calculated with GraphPad Prism 5, with results expressed as medians and ranges. A nonparametric Mann–Whitney test was per-formed to compare numerical data; Fisher’s exact test was used for categorical data. Percentage-point differences (with 95% confidence intervals) are provided for study end points expressed as pro-portions. A post hoc analysis was performed to compare the effects of EVLP stratified according to the two subgroups of donors — those without a heartbeat and those who were brain-dead. For differences in lung function at several time points over the 4-hour period of EVLP, repeated-measures

analysis of variance was used. P values of less than 0.05 were considered to indicate statistical significance.

R esult s

Donor Lungs

From September 2008 through January 2010, lungs from 306 multiorgan donors were offered to our program, and 136 lung transplantations were performed. Lungs from 23 donors were identified as high-risk on the basis of entry crite-ria and were evaluated for 4 hours while being perfused in the EVLP system (Table 2 in the Sup-plementary Appendix). Twenty of these lungs had physiologically stable pulmonary function and were accepted for transplantation (Fig. 1); 9 were from donors without a heartbeat (Maastricht cat-egory III, 45%) and 11 were from brain-dead do-nors (Maastricht category IV, 55%).22

Donor age and status with respect to smoking did not differ significantly between the EVLP group and the control group, but there were sig-nificant between-group differences in donor lung characteristics at baseline (Table 1). Donor lungs in the EVLP group had significantly worse gas exchange and more radiographic and broncho-scopic abnormalities (e.g., mucopurulent or bloody airway secretions and evidence of charcoal aspi-ration). Within the EVLP group, lungs from do-nors without a heartbeat had higher PaO2:FiO2 values and lower donor scores than did lungs from brain-dead donors.

The median time from harvest to implantation for the EVLP group was 653 minutes (range, 267 to 1021) versus 370 minutes (range, 163 to 662) for the control group (P<0.001). The three preservation periods (i.e., cold ischemic time 1, EVLP time, and cold ischemic time 2) for the EVLP group are shown in Figure 3 in the Supplementary Appendix.

Lung Function during EVLP

There was significant improvement in gas exchange during EVLP in the 20 lungs used for transplan-tation (Fig. 2). The median PaO2:FiO2 while the lungs were still in the donors was 335 mm Hg; after 1 hour and 4 hours of EVLP, the PO2:FiO2 increased to 414 and 443 mm Hg, respectively (P<0.001). In addition, pulmonary vascular resis-tance, dynamic compliance, and peak inspiratory pressure were stable throughout the 4 hours of EVLP in all 20 lungs used for lung transplantation. Lung function worsened in 3 of the 23 donor



lungs, with the EVLP PO2:FiO2 decreasing to less than 350 mm Hg in some cases, and these lungs were not transplanted (Fig. 4 in the Supplementa-ry Appendix). The 20 lungs selected for transplan-tation showed stable or improved radiographic findings during EVLP (e.g., reduced pulmonary edema or decreased consolidation and atelectasis between 1 hour and 3 hours of EVLP) (Fig. 5 in the Supplementary Appendix).

Clinical Outcome in the EVLP Group

Primary graft dysfunction was defined as impaired gas exchange meeting ISHLT criteria for grade 2 or 3 dysfunction (PaO2:FiO2 <300 mm Hg)25 after lung transplantation, in the absence of other causes of impaired gas exchange. The incidence of pri-mary graft dysfunction 72 hours after lung trans-plantation tended to be lower in the recipients of EVLP lungs than in the controls (15% vs. 30.1%; 95% confidence interval [CI], −2.6% to 32.8%; P = 0.11) (Table 2). None of recipients in the EVLP group had severe primary graft dysfunction at 72 hours (PaO2:FiO2 <200 mm Hg) (Fig. 6 in the

Supplementary Appendix), as compared with 9.4% of controls (P = 0.36). A post hoc analysis of the effects of EVLP stratified according to donor sub-group (i.e., donors without a heartbeat and brain-dead donors) (Table 2) showed no significant dif-ference in primary graft dysfunction, although the statistical power of the study to detect such a difference was low.

With regard to secondary end points, there were no significant differences between the EVLP and control groups in the occurrence of primary graft dysfunction at the other three time points (at ICU arrival and 24 and 48 hours after trans-plantation) (Table 2). None of the patients with EVLP lung transplants had gas-exchange abnor-malities sufficient to warrant extracorporeal mem-brane oxygenation. The incidence of bronchial complications requiring intervention (e.g., dilation) was similar in the EVLP and control groups (5% and 4%, respectively; P = 1.0), and no significant differences between the groups were observed in the median duration of post-transplantation me-chanical ventilation (2 days for both groups,

306 Consecutive offers of lungs were received from deceased donors

Lungs from 172 donors were rejected for transplantation

Lungs from 111 donors were accepted for transplantation

Lungs from 23 donors underwent4 hours of EVLP

116 Patients underwent trans-plantation (control group)

Lungs from 20 donors wereaccepted for transplantation

(PO2:FiO2 ≥350 mm Hg at 2, 3,or 4 hours of EVLP)

Lungs from 3 donors were rejected for transplantation(PO2:FiO2 <350 mm Hg or

deterioration of lung function[PVR, PIP, or Cdyn] or both)

20 Patients underwent transplan-tation (EVLP group)

11 Were from brain-dead donors9 Were from donors without

a heartbeat

Figure 1. Disposition of Donor Lungs.

Cdyn denotes dynamic compliance, EVLP ex vivo lung perfusion, PO2:FiO2 ratio of the partial pressure of oxygen to the fraction of inspired oxygen, PIP peak inspiratory pressure, and PVR pulmonary vascular resistance.



P = 0.15), length of stay in the ICU (4 days for both, P = 0.68), and hospital length of stay (23 and 27 days, respectively; P = 0.39). Two of 20 patients (10%) died within 30 days after transplantation in the EVLP group, as compared with 6 of 116 in

the control group (5.2%, P = 0.33). Two of the pa-tients who received EVLP lungs died; one patient died on day 7 after lung transplantation, owing to gram-negative sepsis unrelated to the donor, and the other died on day 16 from massive retroperi-

Table 1. Donor, Recipient, and Transplantation Characteristics of Ex Vivo Lung Perfusion (EVLP) Lungs and Control Lungs.*

CharacteristicEVLP Lungs

(N = 20)Control Lungs

(N = 116) P Value†

Lungs from Donors without a Heartbeat

(N = 9)

Lungs from Brain-Dead Donors

(N = 11)Total

(N = 20)

Donor

Age (yr) 0.07

Median 40 38 39 45

Range 16–68 17–69 16–69 6–79

Best PaO2:FiO2 (mm Hg) while in donor <0.001

Median 420 275 335 459

Range 195–532 160–469 160–532 290–590

Abnormal chest radiograph (%) 33 91 70 45 0.05

Abnormal bronchoscopic results (%) 89 91 90 52 0.001

Smoking history >10 packs/day (%) 33 9 20 23 0.78

Positive culture on bronchoalveolar lavage (%) 78 82 80 60 0.04

Donor score‡ <0.001

Median 5 8 8 3

Range 1–9 2–10 1–10 1–9

Recipient

Age (yr) 0.81

Median 52 57 56 56

Range 28–69 36–66 28–69 19–73

Diagnosis of pulmonary fibrosis (%) 33 36 35 37 0.65

UNOS lung-allocation score§ 0.22

Median 34 37 35 33

Range 31–57 29–72 29–72 26–71

Transplantation

Bilateral transplantation (%) 89 63 75 85 0.36

Retransplantation (%) 0 9 5 4.3 0.81

Total preservation time (min) <0.001

Median 632 703 653 370

Range 267–760 467–1021 267–1021 163–662

* PaO2:FiO2 denotes ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen, and UNOS United Network for Organ Sharing.

† P values are for comparisons of all EVLP lungs with control lungs and were calculated with the use of Fisher’s exact test for discrete vari-ables and the Mann–Whitney test for continuous variables.

‡ The donor score ranges from 0 to 18 and is based on age, status with respect to a history of smoking, PaO2:FiO2, chest radiographs, and bronchoscopic findings; higher scores indicate the presence of more risk factors.23

§ The UNOS lung-allocation score ranges from 0 to 100, with higher numbers indicating worse clinical status and greater potential trans-plantation benefit.



toneal bleeding due to anticoagulation therapy for atrial fibrillation. Both causes of death were con-firmed by autopsy. No serious adverse events were directly related to EVLP (Table 3 in the Supplemen-tary Appendix). The survival rate at 1 year was 80% in the EVLP group and 83.6% in control group (P = 0.54); 15 of the 20 recipients of EVLP lungs (75%) and 94 of the 116 controls (81%) survived, with median follow-up times of 561 days (range, 7 to 821) and 542 days (range, 9 to 828), respectively.

Discussion

Lung transplantation is limited by the shortage of available organs and by complications that occur after transplantation, such as primary graft dys-function. The condition of the transplanted lung during the perioperative period strongly influenc-

es both short-term and long-term outcomes.7,9,26-28 In our study, we found that it was feasible to im-prove the outcome of transplantation if lungs from donors without a heartbeat or brain-dead donors were subjected to 4 hours of normothermic ex vivo perfusion during the organ-preservation phase.

To confirm that donor lung function was ad-equate for transplantation, we used a combina-tion of physiological variables that addressed gas exchange (PO2:FiO2 >350 mm Hg), pulmonary mechanics (stable peak inspiratory pressure and dynamic compliance), and pulmonary vascula-ture (stable pulmonary vascular resistance). One could argue that the advantage of EVLP is that it can be used to identify lungs that are not suitable for transplantation. Indeed, two of the three lungs that were rejected after EVLP appeared to function reasonably well on the basis of assess-

*

PaO

2:Fi

O2

Rat

io (m

m H

g)600

400

300

500

200

0Donor 2 31 4

Hours of EVLP

A

Pulm

onar

y V

ascu

lar

Res

ista

nce

(dyn

es·s

ec·c

m−5

)

600

400

200

02 31 4

Hours of EVLP

BD

ynam

ic C

ompl

ianc

e(m

l/cm

of H

2O)

150

100

50

02 31 4

Hours of EVLP

C

Peak

Insp

irat

ory

Pres

sure

(cm

of H

2O)

30

10

20

02 31 4

Hours of EVLP

D

Figure 2. Function during Ex Vivo Lung Perfusion (EVLP) in Donor Lungs.

Twenty high-risk lungs were accepted for transplantation. The gas-exchange function is represented by the best ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (PaO2:FiO2) achieved with the donor lung in vivo and the ratio of the partial pressure of oxygen (PO2) (left atrial PO2 − pulmonary-artery PO2) to FiO2 during4 hours of EVLP. Oxygenation improved significantly during EVLP. Panel A shows median PaO2:FiO2 values in the donor lungs in vivo and during EVLP; the values in vivo and at 1 hour and 4 hours of perfusion were 335, 414, and 443 mm Hg, respectively (P<0.001 by repeated-measures analysis of variance). Panels B, C, and D show pulmonary vascular resis-tance, dynamic compliance, and peak inspiratory pressure, respectively, all of which were stable during EVLP.



ments carried out while the lungs were still in vivo (i.e., the PaO2:FiO2 was adequate and the chest radiograph was normal), but because they were obtained from donors without a heartbeat, they were included in the study (see cases 17 and 18 in Table 2 in the Supplementary Appendix). How-ever, these lungs were not transplanted because pulmonary vascular resistance, peak inspiratory

pressure, or dynamic compliance deteriorated dur-ing EVLP. Unfortunately, the study design does not allow us to conclude that this decision was correct, since there was no group of matched controls in whom such lungs were transplanted.

We chose primary graft dysfunction at 72 hours as the primary end point on the basis of evidence suggesting that it is the best determi-

Table 2. Outcomes in the EVLP and Control Groups.*

End PointEVLP Lungs

(N = 20)Control Lungs

(N = 116)Absolute

Difference† P Value‡

Donors without a Heartbeat

(N = 9)

Brain-Dead Donors (N = 11)

Total(N = 20)

percentage points (95% CI)

Primary end point§

PGD grade 2 or 3 at 72 hr (%) 11 18 15 30 15 (−3 to 33) 0.11

Secondary end points§

PGD grade 2 or 3 at ICU arrival (%) 33 18 25 30 5 (−15 to 26) 0.30

PGD grade 2 or 3 at 24 hr (%) 11 18 15 36 21 (3 to 39) 0.07

PGD grade 2 or 3 at 48 hr (%) 33 27 30 35 5 (−17 to 27) 0.46

ECMO (%) 0 0 0 4 0.37

PaO2:FiO2 on arrival in ICU (mm Hg) 0.51

Median 420 423 422 372

Range 85–518 86–538 85–538 49–591

Mechanical ventilation after transplan-tation (days)

0.15

Median 2 2 2 2

Range 1–27 1–101 1–101 1–43

ICU stay after transplantation (days) 0.68

Median 4 5 4 4

Range 1–34 1–101 1–101 1–103

Hospital stay after transplantation (days)

0.39

Median 19 34 23 27

Range 7–54 11–101 7–101 9–156

Airway complications (%)¶ 11 0 5 4 −1 (−10 to 10) 1.0

Mortality at 30 days (%) 22 0 10 5 −5 (−19 to 9) 0.33

* ECMO denotes extracorporeal membrane oxygenation, ELVP ex vivo lung perfusion, ICU intensive care unit, and PGD primary graft dysfunction.† The differences between all EVLP lungs and control lungs are shown in percentage points (with 95% confidence intervals [CI]) for each

study end point.‡ P values were calculated with the use of Fisher’s exact test for discrete variables and the Mann–Whitney test for continuous variables and

are for the comparison between all EVLP lungs and control lungs.§ Primary graft dysfunction was defined as a ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (PaO2:FiO2)

of less than 300 mm Hg, according to the International Society for Heart and Lung Transplantation classification.25 Grade 0 indicatesPaO2:FiO2 ≥300 mm Hg with clear chest radiographs, grade 1 PaO2:FiO2 ≥300 mm Hg with infiltration on chest radiographs, grade 2 PaO2:FiO2 ≥200 but <300 mm Hg, and grade 3 PaO2:FiO2 <200 mm Hg.

¶ Airway complications were defined as those requiring interventions, such as bronchial dilation.



nant of early outcomes: 30-day mortality was 36% among patients with grade 3 primary graft dys-function at 72 hours versus 5% among patients without grade 3 dysfunction.9,26 In the patients in our study who received lungs assessed during EVLP, the incidence of primary graft dysfunction was low, and severe dysfunction (grade 3) at 72 hours was absent, despite the fact that the func-tion of the EVLP lungs was significantly more impaired at baseline (on the basis of the donor score) than that of the control lungs. In light of the confidence intervals for the differences be-tween the two groups (95% CI, −3% to 33%), we are 97.5% confident that the incidence of pri-mary graft dysfunction in the EVLP group was not more than 3% higher than that in the con-trol group. Other measures of the outcome in recipients, such as the number of days of me-chanical ventilation and the duration of the ICU and hospital stays, were also acceptable and were similar in the two groups. Mortality at 30 days was 10% in the EVLP group — twice that in the control group; however, this difference repre-sented only one excess death, and the causes of death, as assessed on autopsy, were not directly related to graft dysfunction.

To our knowledge, the use of EVLP in humans has not previously been assessed prospectively in a systematic fashion. Clinical experience with EVLP is limited to six lung transplants that un-derwent short-term (1-hour), blood-based EVLP.27

Our study involved a normothermic, acellular, blood-free perfusate; protective perfusion and ven-tilation strategies; and a prolonged perfusion time. This approach was associated with gradual im-provement in the function of most of the lungs that were subjected to EVLP.

The few series of cases of lung transplanta-tion involving lungs from donors without a heart-beat (Maastricht category III) have indicated good

early outcomes28-32; however, one recent study showed increased rates of primary graft dysfunc-tion, as well as increased in-hospital mortality.33 Therefore, most lung-transplantation centers do not use lungs from donors without a heartbeat. Indeed, we were hesitant to transplant such lungs, even if they seemed otherwise adequate, until we were able to assess lung function using EVLP.

The limitations of this study are inherent in studies with a small number of cases and lack of randomization. Thus, our study does not provide definitive evidence that the lungs subjected to EVLP would have performed well if they had been directly transplanted, without the use of EVLP. Although a randomized clinical trial of EVLP would be ideal, overcoming the ethical dif-ficulty of directly transplanting so-called ques-tionable organs will present a challenge.

In conclusion, our study shows that the use of extended normothermic EVLP allows an objective assessment of high-risk donor lungs. When these lungs are transplanted, acceptable rates of pri-mary graft dysfunction are achieved, and the early outcomes are similar to those with conven-tionally selected and transplanted lungs.

Supported by Vitrolife.Disclosure forms provided by the authors are available with

the full text of this article at NEJM.org.We thank all members of the Toronto Lung Transplant Pro-

gram as well as the patients; the Toronto General Hospital per-fusion department and Paul Chartrand for assistance in the EVLP procedure and logistics; Kevin E. Thorpe (biostatistician–trialist, Knowledge Translation Program, Keenan Research Cen-tre, Li Ka Shing Knowledge Institute of St. Michael’s Hospital, and assistant professor, Dalla Lana School of Public Health, University of Toronto) for his support during this study; Trillium Gift of Life Network, Ontario, for collaboration in obtaining consent for research on the donor lungs used in this study; As-tellas Pharma Canada and the McEwen Centre for Regenerative Medicine for providing support to accomplish our preclinical translational research projects that ultimately led to the comple-tion of this trial; and Vitrolife for supplying the XVIVO system disposables and Steen Solution.

References1. Punch JD, Hayes DH, LaPorte FB, Mc-Bride V, Seely MS. Organ donation and utilization in the United States, 1996-2005. Am J Transplant 2007;7:1327-38.2. Botha P, Trivedi D, Weir CJ, et al. Ex-tended donor criteria in lung transplanta-tion: impact on organ allocation. J Thorac Cardiovasc Surg 2006;131:1154-60.3. Aigner C, Winkler G, Jaksch P, et al. Extended donor criteria for lung trans-plantation — a clinical reality. Eur J Car-diothorac Surg 2005;27:757-61.4. Van Raemdonck D, Neyrinck A,

Verleden GM, et al. Lung donor selection and management. Proc Am Thorac Soc 2009;6:28-38.5. de Perrot M, Liu M, Waddell T, Kes-havjee S. Ischemia-reperfusion-induced lung injury. Am J Respir Cell Mol Biol 2003;167:490-511.6. Pierre AF, Sekine Y, Hutcheon MA, Waddell TK, Keshavjee SH. Marginal do-nor lungs: a reassessment. J Thorac Car-diovasc Surg 2002;123:421-7.7. Bharat A, Narayanan K, Street T, et al. Early posttransplant inflammation pro-

motes the development of alloimmunity and chronic human lung allograft rejec-tion. Transplantation 2007;83:150-8.8. Bharat A, Kuo E, Steward N, et al. Im-munological link between primary graft dysfunction and chronic lung allograft rejection. Ann Thorac Surg 2008;86:189-95.9. Lee JC, Christie JD, Keshavjee S. Pri-mary graft dysfunction: definition, risk factors, short- and long-term outcomes. Semin Respir Crit Care Med 2010;31:161-71.



10. Kawut SM, Reyentovich A, Wilt JS, etal. Outcomes of extended donor lung re-cipients after lung transplantation. Trans-plantation 2005;79:310-6.11. De Meester J, Smits J, Persijn G, Hav-erich A. Listing for lung transplantation: life expectancy and transplant effect, strat-ified by type of end-stage lung disease, the Eurotransplant experience. J Heart Lung Transplant 2001;20:518-24.12. Lederer DJ, Arcasoy SM, Wilt JS,D’Ovidio F, Sonett JR, Kawut SM. Six-minute-walk distance predicts waiting list survival in idiopathic pulmonary fi-brosis. Am J Respir Crit Care Med 2006; 174:659-64.13. Angel LF, Levine DJ, Restrepo MI, etal. Impact of a lung transplantation do-nor-management protocol on lung dona-tion and recipient outcomes. Am J Respir Crit Care Med 2006;174:710-6.14. Avlonitis VS, Wigfield CH, Kirby JA,Dark JH. The hemodynamic mechanisms of lung injury and systemic inflammatory response following brain death in the transplant donor. Am J Transplant 2005; 5:684-93.15. Keshavjee SH, Yamazaki F, CardosoPF, McRitchie DI, Patterson GA, Cooper JD. A method for safe twelve-hour pulmo-nary preservation. J Thorac Cardiovasc Surg 1989;98:529-34.16. Fischer S, Matte-Martyn A, De PerrotM, et al. Low-potassium dextran preserva-tion solution improves lung function after human lung transplantation. J Thorac Cardiovasc Surg 2001;121:594-6.17. Southard JH, Belzer FO. Organ pres-ervation. Annu Rev Med 1995;46:235-47.

18. Cypel M, Yeung JC, Hirayama S, et al.Technique for prolonged normothermic ex vivo lung perfusion. J Heart Lung Transplant 2008;27:1319-25.19. Cypel M, Rubacha M, Yeung J, et al.Normothermic ex vivo perfusion prevents lung injury compared to extended cold preservation for transplantation. Am J Transplant 2009;9:2262-9.20. de Perrot M, Bonser RS, Dark J, et al.Report of the ISHLT Working Group on Primary Lung Graft Dysfunction part III: donor-related risk factors and markers. J Heart Lung Transplant 2005;24:1460-7.21. de Perrot M, Chaparro C, McRae K, etal. Twenty-year experience of lung trans-plantation at a single center: influence of recipient diagnosis on long-term survival. J Thorac Cardiovasc Surg 2004;127:1493-501.22. Kootstra G, Daemen JH, Oomen AP.Categories of non-heart-beating donors. Transplant Proc 1995;27:2893-4.23. Oto T, Levvey BJ, Whitford H, et al.Feasibility and utility of a lung donor score: correlation with early post-trans-plant outcomes. Ann Thorac Surg 2007; 83:257-63.24. Cypel M, Liu M, Rubacha M, et al.Functional repair of human donor lungs by IL-10 gene therapy. Sci Transl Med 2009;1(4):4ra9.25. Christie JD, Carby M, Bag R, Corris P,Hertz M, Weill D. Report of the ISHLT Working Group on Primary Lung Graft Dysfunction part II: definition: a consen-sus statement of the International Society for Heart and Lung Transplantation. J Heart Lung Transplant 2005;24:1454-9.

26. Christie JD, Bellamy S, Ware LB, et al.Construct validity of the definition of pri-mary graft dysfunction after lung trans-plantation. J Heart Lung Transplant 2010; 29:1231-9.27. Ingemansson R, Eyjolfsson A, MaredL, et al. Clinical transplantation of ini-tially rejected donor lungs after recondi-tioning ex vivo. Ann Thorac Surg 2009; 87:255-60.28. Erasmus ME, Verschuuren EA, Ni-jkamp DM, Vermeyden JW, van der Bij W. Lung transplantation from nonheparin-ized category III non-heart-beating do-nors: a single-centre report. Transplanta-tion 2010;89:452-7.29. Cypel M, Sato M, Yildirim E, et al. Ini-tial experience with lung donation after cardiocirculatory death in Canada. J Heart Lung Transplant 2009;28:753-8.30. Snell GI, Levvey BJ, Oto T, et al. Earlylung transplantation success utilizing controlled donation after cardiac death donors. Am J Transplant 2008;8:1282-9.31. Mason DP, Thuita L, Alster JM, et al.Should lung transplantation be performed using donation after cardiac death? The United States experience. J Thorac Cardio-vasc Surg 2008;136:1061-6.32. De Oliveira NC, Osaki S, Maloney JD,et al. Lung transplantation with donation after cardiac death donors: long-term follow-up in a single center. J Thorac Car-diovasc Surg 2010;139:1306-15.33. Puri V, Scavuzzo M, Guthrie T, et al.Lung transplantation and donation after cardiac death: a single center experience. Ann Thorac Surg 2009;88:1609-14.Copyright © 2011 Massachusetts Medical Society.

SPECIALTIES AND TOPICS AT NEJM.ORG

Specialty pages at the Journal’s Web site (NEJM.org) feature articles in cardiology, endocrinology, genetics, infectious disease, nephrology, pediatrics, and many other medical specialties. These pages, along with collections of articles on clinical and nonclinical topics, offer links to interactive and multimedia content and feature

recently published articles as well as material from the NEJM archive (1812–1989).

Supplementary Appendix

This appendix has been provided by the authors to give readers additional information about their work.

Supplement to: Cypel M, Yeung JC, Liu M, et al. Normothermic ex vivo lung perfusion in clinical lung transplan-tation. N Engl J Med 2011:1431-40.

Supplementary Material Table of contents

Supplementary methods. Ex vivo lung perfusion technique and perfusate

composition. Supplementary figure 1. Safety and logistic feasibility study first performed in

3 patients. Supplementary figure 2. Schematic of the EVLP System. Supplementary figure 3. Subdivisions of total preservation time in EVLP

donor lungs. Supplementary figure 4. Lung function during EVLP in 3 lungs that did not

meet criteria for transplantation. Supplementary figure 5. Ex vivo lung x-ray during EVLP. Supplementary figure 6. Distribution of primary graft dysfunction grades in

EVLP group. Supplementary table 1. Components of Steen Solution™ (Vitrolife) Supplementary table 2. Donor lungs in EVLP arm Supplementary table 3. Adverse events in EVLP group

Supp. Methods – Perfusate Composition and Ex vivo Lung Perfusion Technique

Perfusate

The circuit was primed with 2 L of Steen Solution™ (XVIVO, Vitrolife), a buffered dextran

containing extracellular-type solution with an optimized colloid osmotic pressure developed

specifically for EVLP (Supplementary table 1). In addition, 500 mg of Solumedrol (Sandoz,USA),

500 mg of Primaxin (Merck, USA) and 3000 IU of Heparin (Organon, Canada) were added to

the perfusate. Every hour after the initiation of EVLP, 500 ml of circulated perfusate was

removed and replenished with 500 ml of fresh solution.

Technique

After the lungs were transferred to the XVIVOTM chamber (Vitrolife), the left atrial (LA) cannula

was connected to the circuit. Flow was initiated slowly in a retrograde fashion to de-air

through the pulmonary artery (PA) cannula. The PA cannula was then connected to the circuit

and anterograde flow was started at 150 ml/min with the perfusate at room temperature. The

temperature of the perfusate was then gradually increased to 37°C. When 32°C was reached

(usually over 30 min), ventilation was started and the perfusate flow rate was gradually

increased. The flow of gas used to deoxygenate and provide carbon dioxide to the inflow

perfusate via a gas exchange membrane was then initiated at 1L/min. We used 40% of the

estimated donor cardiac output (CO) as the target maximum maintenance perfusate flow rate

to perfuse both lungs. A positive LA pressure was maintained between 3 and 5 mmHg by

adjusting the height of the hard-shell reservoir. A protective mode of mechanical ventilation

was applied using a tidal volume of 7 ml/kg (based on donor ideal body weight), at 7 breaths

per min, positive end-expiratory pressure (PEEP) of 5 cmH2O and an inspired oxygen fraction

(FiO2) of 21%. The lungs were recruited with inspiratory holds to a peak airway pressure

(PawP) of 20 cmH2O every hour. The pH, pCO2, electrolytes and glucose were maintained at

physiologic levels in the perfusate.

At the end of 4h of EVLP, the lung block was cooled down in the circuit to 10°C over 10

minutes. Thereafter perfusion and ventilation were stopped (FiO2 was increased to 50% for

lung storage), and the trachea was clamped to maintain the lungs in an inflated state. The

lungs were then statically preserved at 4°C in Perfadex® until transplantation.

Supplementary Material - Legends

Supplementary Figure 1. Safety and logistic feasibility study was first performed in 3 patients who underwent bilateral LTx using standard criteria donor lungs, where one lung was transplanted as per current practice and the other lung transplanted following 1h of EVLP (A). No differences were observed between the perfused (left) and non-perfused (right) lung in CXR performed at ICU arrival (B). Six weeks after transplantation flexible bronchoscopy demonstrated normal bronchial healing of bronchial anastomosis (C). CSP: cold static preservation; EVLP: ex vivo lung perfusion.

Supplementary Figure 2. (A) Schematic of the EVLP System. The lungs are placed within the XVIVOTM

chamber (Vitrolife AB, Sweden). The perfusate leaves the lungs via the LA cannula and enters the reservoir. From there, the perfusate is pumped using a centrifugal pump into the oxygenator and heat exchanger where it is deoxygenated by a gas mixture (86% N2, 8% CO2 and 6% O2) and warmed to normothermia. The perfusate then passes through a leukocyte filter before reentering the lungs via the PA cannula for oxygenation by the lung. (B) Human donor lung in the EVLP system prior to transplantation.

Supplementary Figure 3. Subdivisions of total preservation time in EVLP donor lungs. Supplementary Figure 4. Lung function during EVLP in 3 lungs that did not meet criteria for

transplantation. Case 17, deterioration in all 4 functional parameters and EVLP delta P/F <350mmHg. Case 18, increase in PawP and EVLP delta P/F <350mmHg. Case 19 - poor deflation likely related to some degree of emphysema in donor lungs.

Supplementary Figure 5. Illustration of a lung ex vivo x-ray. After 3 h of EVLP, there is marked

improvement in lung aeration and decreased radiologic evidence of pulmonary edema (upper panel) and improvement in a right lower lobe focal non-pneumonic consolidation/atelectasis after EVLP (lower panel).

Supplementary Figure 6. Distribution of primary graft dysfunction grades within 72h in the EVLP

group. Note that at 72 h after transplantation 100% of the patients were free of PGD III (P/F < 200mmHg).

Supplementary Table 1. Components of Steen Solution™ (Vitrolife). Supplementary Table 2. Specifications of donor lungs in EVLP arm Supplementary Table 3. Adverse events in EVLP group

A

B

C

EVLP lung CSP lung

CSP lung EVLP lung

Supplementary Figure 1


A

B

Preservation time in EVLP group

Cold isch

emic

time 1

Prepara

tion

EVLP

Cold isch

emic

time 2

0

1

2

3

4

5

Hou

rs


PVR

1h 2h 3h 4h100

200

300

400

500Case 17

Case 19Case 18

dyne

s.se

c.cm

-5

Delta P/F

1h 2h 3h 4h100

200

300

400

500

mm

Hg

PIP

1h 2h 3h 4h18

20

22

24

26

28

Hours of EVLP

cmH

20

CDyn

1h 2h 3h 4h0

10

20

30

40

50

Hours of EVLP

ml/c

mH

2O


A 1h EVLP 3h EVLP

B


ICU arriv

alT 24

hT 48

hT 72

h

1

2

3

0/

Post LTx

PGD

sco

res


COMPOSITION Calcium chloride Magnesium chloride Sodium chloride Dextran 40 Potassium chloride Sodium dihydrogen phosphateGlucose Sodium bicarbonate Water Human serum albumin

PRODUCT PROPERTIES SPECIFICATION RESULT

pH (at +20ºC and ambient atmosphere) 7.4 0.3 7.35Osmolality [mOsm/kg] 295 20 296

Supplementary Table 1. Components and properties of Steen solution. This is a buffered extracellular solution containing dextran with an optimal colloid osmotic pressure developed specifically for extra-corporeal perfusion of lungs.

Documents

010 cypel normothermic evlp in clinical tx