Pulmonary Retransplantation in the Nordic Countries Sandra Lindstedt, MD, PhD, G€ oran Dellgren, MD, PhD, Martin Iversen, MD, Gerdt C. Riise, MD, Øystein Bjørtuft, MD, Pekka H€ amm€ ainen, MD, Ingrid Skog, MD, Arnt Fiane, MD, and Per Wierup, MD, PhD, for the Nordic Thoracic Transplantation Study Group (NTTSG)
Background. The increasing demand for pulmonary retransplantation (re-LTx) raises ethical issues on the correct allocation of the scarce donor pool. Thus, we performed a thorough review of the current results for reLTx in the Nordic countries. Methods. Seventy-five patients with a median age of 50 years (range, 22 to 64 years) underwent re-LTx from 1992 until June 2013, of which 53 had single re-LTx, 21 had double re-LTx, and 1 patient underwent a heart-lung retransplantation. Primary graft dysfunction (PGD) was the primary indication in 9 cases, bronchiolitis obliterans syndrome (BOS) in 62 cases, and airway complications in 4 cases. Results. Patients who underwent re-LTx in the period 1992 to 1999 (n [ 16) had a 1-year survival of 37.5% (95% confidence interval [CI], 19.9 to 70.6), whereas patients who underwent re-LTx in the period 2000 to 2013 (n [ 64) had a 1-year survival of 81.0% (95% CI, 71.5 to 91.8). Corresponding 5-year survival was 25.0% (95% CI, 10.7 to 58.4) in the early era group (1992 to 1999) and 57.2% (95% CI, 44.3 to 73.7) in the more recent era group (2000 to 2013; p [ 0.0151). Patients with BOS who underwent re-LTx in the period 1992 to 1999 (n [ 13) had a 1-year survival of 38.5% (95% CI, 19.3 to 76.5), whereas patients with BOS
who underwent re-LTx in the period 2000 to 2013 (n [ 49) had a 1-year survival of 85.4% (95% CI, 75.9 to 96.0). Corresponding 5-year survival was 23.1% (95% CI, 8.6 to 62.3) in the early era group (1992 to 1999) and 56.1% (95% CI, 41.9 to 75.2) in the more recent era group (2000 to 2013; p [ 0.0199). The cumulative incidence among patients who underwent re-LTx because of BOS and developed BOS again after re-LTX was analyzed. The cumulative incidence curves for time periods 1992 to 1999 and 2000 to 2013 are not statistically different for repeat BOS (p [ 0.5087), but they are highly significant for time periods among patients who died (p [ 0.02381). Conclusions. Results for re-LTx have improved over time, especially when BOS is the primary indication. The cumulative incidence among patients who underwent reLTx because of BOS and developed repeat BOS after reLTX showed equal risk between 1992 to 1999 and 2000 to 2013 in the aspect of developing repeat BOS, but in the later era the patients had a significantly higher chance of surviving.
T
regimens in antibiotic prophylaxis and treatment [3, 4]. However, acute and chronic rejection and dysfunction continue to be major problems. The primary cause of death after LTx is chronic rejection or bronchiolitis obliterans syndrome (BOS), and pulmonary retransplantation (re-LTx) is currently the only treatment option for these patients. The development of BOS is rare in the first year after LTx, but the rate increases quickly, with cumulative incidence reported to be as high as 40% to 80% within the first 5 years [5–7]. Other complications that could lead to irreversible graft dysfunction and respiratory failure after LTx, where re-LTx remains the only treatment option, is primary graft dysfunction (PGD), severe airway complication, and refractory acute graft rejection. Given the extent of late graft failure, it is inevitable that some patients, especially those who undergo transplantation at a young age (ie, patients with cystic
he introduction of cyclosporin A made it possible to achieve successful heart and lung transplantations [1]. Clinical isolated lung transplantation (LTx) with longterm survival was initially performed in the early 1980s by the Toronto Lung Transplantation Group [2], and has evolved into an established treatment, and the only lifesaving therapy, for patients with a variety of end-stage pulmonary diseases. Survival in LTx recipients has increased over the years, mainly as a result of careful patient selection; improved lung preservation, surgical techniques, immunosuppressive regimes, and management of ischemia/reperfusion injury; and forceful
Accepted for publication Jan 16, 2015. Address correspondence to Dr Wierup, Department of Cardiothoracic Surgery, Skåne University Hospital, Lund University, 221 85 Lund, Sweden; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;99:1781–7) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.01.052
GENERAL THORACIC
Departments of Cardiothoracic Surgery and Pulmonary Medicine, Skåne University Hospital, Lund University, Sweden; Departments of Pulmonary Medicine and Cardiothoracic Surgery, Sahlgrenska University Hospital, Gothenburg, Sweden; Department of Pulmonary Medicine, Rigshospitalet, Copenhagen, Denmark; Departments of Pulmonary Medicine and Cardiothoracic Surgery, Rikshospitalet, Oslo, Norway; and Department of Cardiac Surgery, Helsinki University Hospital, Helsinki, Finland
GENERAL THORACIC
1782
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
fibrosis), will eventually exhibit recurrent respiratory failure, for which re-LTx will have to be considered. Pulmonary retransplantation is believed to carry a higher risk compared with primary LTx as reflected in the outcome of various studies. Historically, re-LTx showed disappointing results; however, a few recent studies report 1- and 5-year survival rates for re-LTx comparable to current survival rates for primary LTx [8–12]. The timing and indication, the presence of other comorbidities, immunosuppressive management, infection prophylaxis, and the likelihood of success are only a few factors that have to be considered before choosing re-LTx. The value of re-LTx has been heavily debated on ethical grounds as an overall organ shortage means that higher risk re-LTx patients will displace some patients on the waiting list for primary LTx [3, 4, 13, 14]. Owing to an increasing number of potential candidates for re-LTx in the Nordic countries, a review of the current results for re-LTx in this region was made.
Patients and Methods Patients Seventy-five patients underwent re-LTx from 1992 through 2013 in the Nordic countries at five Nordic transplantation centers (Skåne University Hospital, Lund University, Sweden; Rigshospitalet, Copenhagen, Denmark; Sahlgrenska University Hospital, Gothenburg, Sweden; Rikshospitalet, Oslo, Norway; and Helsinki University Hospital, Helsinki, Finland). The median age was 50 years (range, 22 to 64 years). Forty-six of the patients were men and 29 were women. Fifty-three patients received a single re-LTx, and 21 patients had a double re-LTx. Thirty-five patients underwent a contralateral single re-LTx, 18 patients underwent an ipsilateral single re-LTx. Indications for re-LTx were BOS in 62 patients, PGD in 9 patients, and airway complications in 4 patients. Ten patients had ventilator support preoperatively; 2 of those patients underwent re-LTx as a result of BOS, and the rest of the patients had PGD. Two patients had extracorporeal membrane oxygenation support preoperatively, of whom 1 patient underwent reLTx because of PGD, and 1 patient because of BOS. The early era was defined between the years 1992 through 1999, and the more recent era was 2000 until June 2013. Bronchiolitis obliterans syndrome was defined by spirometry as more than 20% decline in forced expiratory volume in 1 second from the highest obtained baseline, according to International Society for Heart and Lung Transplantation (ISHLT) guidelines [15]. Primary graft dysfunction was retrospectively identified in patients with retransplantation owing to initial graft failure in the first months before the ISHLT definition in 2005 [16]. After 2005, the ISHLT definition of PGD had been used prospectively.
Data Acquisition and Follow-Up The study was performed for the Nordic Thoracic Transplantation Study Group, which is a scientific and
Ann Thorac Surg 2015;99:1781–7
clinical working group comprising members from all thoracic transplant centers in the Nordic countries (Oslo, Gothenburg, Lund, Copenhagen, Aarhus, Helsinki) and which includes all transplant centers in Scandia transplant. Data were collected retrospectively from lung transplantation databases in the centers and were reviewed for demographic, operative, perioperative, and outcome data. Follow-up data were obtained through patient files, and follow-up was 100% complete. A retrospective study like this does not require approval from a scientific ethical committee in the Nordic countries.
Statistical Analysis Primary stratification of the material was made into two sets of cohorts. The first cohort was based on the main indication for re-LTx, with the following indicator cohorts: BOS, PGD, or airway complications. The second set divides the material based on time period for reLTx: 1992 to 1999 and 2000 to 2013. Time to re-LTx and transplant re-LTx type—single lung (SL) or double lung (DL)—were compared between the sets of cohorts. Primary outcomes include post-re-LTx survival. Overall survival was measured at 1 year and 5 years after the re-LTx by the Kaplan-Meier method, and compared between the two sets of cohorts; indications for re-LTx and time period were accounted for with the log-rank test. Kaplan-Meier survival was also measured for combinations of the cohort sets and transplant re-LTx type, sample sizes allowing. Patients were censored if they reached the end of the study period, June 1, 2014, or were lost to follow-up. All causes of death were incorporated into the survival analyses. Continuous variables are expressed as mean ( standard deviation) and categorical variables as number (percent). Statistical comparisons of continuous variables were made by Student’s t test or analysis of variance when necessary. Statistical comparisons of categorical variables were made by c2 analysis or Fisher’s exact test. Kaplan-Meier survival estimates for 1 and 5 years are expressed as years (95% confidence interval [CI], point estimate). For all statistical analyses, a probability value of less than 0.05 was considered significant. All calculations above were performed using SPSS version 19.0 (IBM Corp, Armonk, NY). A subanalysis of this study was the recurrence of bronchiolitis obliterans syndrome (re-BOS) after retransplantation among patients with BOS as a primary indicator for re-LTx. In this analysis, death will act as a competing risk event to the event of re-BOS. In a competing-risks model, we analyzed incidence of re-BOS and death as two separate outcomes. Specifically, we estimated and compared the cumulative incidence functions for re-BOS and death using Gray’s test (1988). All calculations regarding competing risks were performed using R with the CMPRSK package (available at http:// www.r-project.org). The statistics were performed by Roy Ollila at Sidsoft AB, Malmo, Sweden.
Ann Thorac Surg 2015;99:1781–7
Results
Fig 1. Survival after redo lung transplantation during the years 1992 to 1999 (early era patient group; solid line), and during the years 2000 to 2013 (more recent era patient group; dashed line) in the Nordic countries. Patients in the more recent era had a significantly higher chance of surviving compared with the early era patient group.
1783
The cohort was divided into groups regarding type of transplant, SL versus DL. During the period 1992 to 2013, a total of 53 patients received SL and 21 patients received DL. Overall 1-year survival for the whole cohort (1992 to 2013) was 75.0% (95% CI, 64.1 to 87.8) for SL and 65.9% (95% CI, 48.2 to 90.1) for DL. Corresponding 5-year survival was 51.2% (95% CI, 38.3 to 68.3) for SL and 50.2% (95% CI, 30.7 to 82.2; p ¼ 0.514) for DL. The cohort was divided into the two periods, 1992 to 1999 and 2000 to 2013. Between 1992 and 1999 a total of 13 patients underwent SL re-LTx and 2 patients underwent DL re-LTx. Kaplan-Meyer survival was not applicable because of the small patient group. Regarding the time period 2000 to 2013, 1-year survival was 84.8% (95% CI, 74.3 to 96.8) for SL (n ¼ 40) and 72.9% (95% CI, 55.1 to 96.4) for DL (n ¼ 19). Corresponding 5-year survival was 59.1% (95% CI, 44.4 to 78.7) for SL and 55.5% (95% CI, 34.6 to 89.1; p ¼ 0.546; Fig 2). The patients in the cohort who underwent re-LTx because of BOS were analyzed. From 1992 to 2013, a total of 62 patients underwent re-LTx because of BOS. The 1-year survival was 85.4% (95% CI, 77.0 to 94.7), and the 5-year survival was 50.9% (95% CI, 38.6 to 67.1). The cohort was divided into the two time periods 1992 to 1999 and 2000 to 2013. Patients with BOS who underwent re-LTx in the period 1992 through 1999 (n ¼ 13) had a 1-year survival of 38.5% (95% CI, 19.3 to 76.5), whereas patients with BOS who underwent re-LTx in the period 2000 through 2013 (n ¼ 49) had a 1-year survival of 85.4% (95% CI, 75.9 to 96.0). Corresponding 5-year survival was 23.1% (95% CI, 8.6 to 62.3) in the early era group (1992 to 1999) and 56.1% (95% CI, 41.9 to 75.2) in the more recent era group (2000 to 2013; p ¼ 0.0199; Fig 3). The same cohort (re-LTx because of BOS) was divided regarding type of transplant. A total of 62 patients
Fig 2. Survival after redo lung transplantation during the years 2000 to 2013 (more recent era patient group) in the Nordic countries, comparing single redo lung transplantation (solid line) and double redo lung transplantation (dashed line). No significant difference was seen when comparing the two patient groups.
GENERAL THORACIC
Levine’s test for equality of variances tests the hypothesis that the two populations’ variances are equal. For the two time periods, 1992 to 1999 and 2000 to 2013, we found Levine’s statistic F ¼ 8.303 to be significant (p ¼ 0.005), and thus the assumption of equal variances are violated. Hence we used the equal variance not assumed t test statistic to evaluate H0: equality of means, here (H0: average time to re-LTx for 1992 to 1999 ¼ average time to re-LTx for 2000 to 2013). The results indicate that there is a significant difference between the time periods in time to re-LTx, t (df ¼ 61) ¼ 3.036, p ¼ 0.004 (p ¼ 0.00352). The mean values indicate that patients during the time period 1992 to 1999 had a significantly shorter time to re-LTx (1992 to 1999 ¼ 2.29 years to re-LTX) than patients treated during 2000 to 2013 (2000 to 2013 ¼ 4.17 years to re-LTx). A total of 75 patients underwent re-LTx between 1992 and 2013. Overall 1-year survival for the whole cohort was 71.4% (95% CI, 61.8 to 82.6). Corresponding 5-year survival was 49.2% (95% CI, 37.9 to 63.9). The cohort was divided into two time periods, 1992 to 1999 and 2000 to 2013. Patients who underwent re-LTx in the period 1992 through 1999 (n ¼ 16) had a 1-year survival of 37.5% (95% CI, 19.9 to 70.6), whereas patients who underwent re-LTx in the period 2000 through 2013 (n ¼ 64) had a 1-year survival of 81.0% (95% CI, 71.5 to 91.8). Corresponding 5-year survival was 25.0% (95% CI, 10.7 to 58.4) in the early era group (1992 to 1999) and 57.2% (95% CI, 44.3 to 73.7) in the more recent era group (2000 to 2013; p ¼ 0.0151; Fig 1).
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
GENERAL THORACIC
1784
Fig 3. Survival after redo lung transplantation because of bronchiolitis obliterans syndrome during the years 1992 to 1999 (early era patient group; solid line), and during the years 2000 to 2013 (more recent era patient group; dashed line) in the Nordic countries. Patients in the more recent era had a significantly higher chance of surviving compared with the early era patient group.
underwent re-LTx because of BOS between 1992 and 2013 (SL n ¼ 46, DL n ¼ 16). One-year survival for SL (n ¼ 46) was 75.5% (95% CI, 63.9 to 89.2) and 74.5% (95% CI, 55.7 to 99.6) for DL (n ¼ 16). Corresponding 5-year survival was 47.2% (95% CI, 33.5 to 66.5) for SL and 54.3% (95% CI, 32.0 to 92.3; p ¼ 0.921; Fig 4). The cumulative incidence amongst patients who underwent re-LTx because of BOS and developed re-BOS
Fig 4. Survival after redo lung transplantation for bronchiolitis obliterans syndrome during the years 1992 to 2013 in the Nordic countries, comparing single redo lung transplantation (solid line) and double redo lung transplantation (dashed line). No significant difference was seen when comparing the two patient groups.
Ann Thorac Surg 2015;99:1781–7
after re-LTX was analyzed. Figure 5 shows the results of Gray’s test for equality of cumulative incidence functions across groups. The cumulative incidence curves for time periods 1992 to 1999 and 2000 to 2013 are not statistically different for re-BOS (p ¼ 0.5087), but they are highly significant for time periods among patients who died (p ¼ 0.02381). Nine patients underwent re-LTx because of PGD between the years 1992 and 2013. Between the time period 1992 to 1999, a total of 2 patients underwent re-LTX. Corresponding figures for the time period 2000 to 2013 was 7 patients. A descriptive analysis is shown in Table 1. Kaplan-Meyer survival was not applicable because of small group size. Four patients underwent re-LTx because of airway complications between the years 1992 and 2013. Between the time period 1992 to 1999, only 1 patient underwent reLTx. Corresponding figures for the time period 2000 to 2013 was 3 patients. A descriptive analysis is shown in Table 1. Kaplan-Meyer survival was not applicable because of small group size. Ten patients received preoperative ventilator support between the years 1992 and 2013. Between the time period 1992 to 1999, a total of 2 patients underwent reLTX. Corresponding figure for the time period 2000 to 2013 was 8 patients. Kaplan-Meyer survival was not applicable because of small group size.
Comment Despite improving results in LTx, a significant number of grafts fail early or late postoperatively, and some patients will require re-LTx. Redo LTx carries a higher risk than primary LTx, and the outcome of re-LTx has traditionally not been good [3, 4]. Worldwide, the demand for re-LTx has increased. From an ethical point of view, it is important to outline the actual benefit of re-LTx as the donor supply does not meet the needs of the listed candidates waiting for primary LTx. The ISHLT Pulmonary Retransplant Registry was founded in 1991 to determine the predictors of outcome after re-LTx and facilitate decisions concerning the appropriateness of re-LTx in individual patients. In a 1996 multicenter study (35 centers worldwide), Novick and colleagues [17] reported that the most important indicator for early survival after re-LTx (n ¼ 160) was the ambulatory status of the recipients. The follow-up analysis made by the same group in 1998, which included 230 patients from 47 centers across the United States, Canada, Europe, and Australia, identified factors favoring survival and freedom from BOS after re-LTx: surgery performed after 1991, no preoperative ventilator support, an interval of more than 2 years passed from primary LTx, and transplant center experience [18]. One important underlying cause of improvements in re-LTx after 1991 (the so called “era effect”) seems to be the ability to distinguish between high- and low-risk candidates for re-LTx. The observation that survival is highest in patients with BOS has lately been supported by data from the Hannover Lung Transplantation group [12]. In 2006,
Ann Thorac Surg 2015;99:1781–7
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
1785
GENERAL THORACIC
Fig 5. The cumulative incidence among patients who underwent redo lung transplantation because of bronchiolitis obliterans syndrome and exhibited repeat bronchiolitis obliterans syndrome after redo lung transplantation was analyzed. The cumulative incidence curves for time periods (A, C) 1992–1999 and (B, D) 2000–2013 are not statistically different for (C, D) repeat bronchiolitis obliterans syndrome (ReBOS) (p ¼ 0.5087), but they are highly significant for time periods among (A, B) patients who died (p ¼ 0.02381). The patients in the two different eras had the same risk of exhibiting repeat bronchiolitis obliterans syndrome, but in the later era, the patients had a significantly higher chance of surviving.
Strueber and colleagues [12] presented the largest ever single-center series of re-LTx (n ¼ 54). Their most striking finding was the identification of a group of patients for whom re-LTx had a comparable outcome equal to that for primary LTx. This was true for younger patients with BOS with an otherwise good prognosis, such as cystic fibrosis. In this group, 5-year survival as well as freedom from BOS revealed a similar course to that of primary LTx [12]. On the other hand, patients receiving mechanical ventilation showed a discouraging 1-year survival of 50% compared with ambulatory patients (78%). However, these figures were equivalent to the results for patients undergoing primary LTx who were also receiving mechanical ventilation at the same center. They also found that airway complications were
more prevalent in elderly patients with emphysema, and that re-LTx for acute graft failure was more common in patients with fibrosis and idiopathic pulmonary hypertension. They concluded that re-LTx for airway complications and acute graft failure had been rare during the last 5 years in their center, mainly because of better patient selection for re-LTx [12]. In 2008, Aigner and colleagues [8] from Vienna reported their experience for re-LTx (n ¼ 46), stratified for different indications: PGD (n ¼ 23), BOS (n ¼ 19), and airway problems (n ¼ 4). Survival rates at 1 and 5 years were 35% and 29% in the PGD group, and 73% and 61% in the BOS group, respectively. All 4 patients with airway problems were alive at the follow-up time. They concluded that re-LTx for BOS and airway problems
1786
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
Ann Thorac Surg 2015;99:1781–7
Table 1. Comparison of Transplant Characteristics Among Time Period and Indication for Redo Lung Transplantationa Time Period
GENERAL THORACIC
Variable Time period 1992–1999 2000–2013 Indication for re-LTx BOS PGD Airway Time to re-LTx (y) Transplant re-LTx type SL DL HLT Status Alive Dead
Total re-LTx (n ¼ 75)
1992–1999 (n ¼ 16)
2000–2013 (n ¼ 59)
. .
. .
Indication for re-LTX p Valueb
BOS (n ¼ 62)
PGD (n ¼ 9)
Airway (n ¼ 4)
. 16 (21) 59 (79)
0.98 13 (21) 49 (79)
2 (22) 7 (78)
1 (25) 3 (75) .
0.98 62 9 4 3.77
(83) (12) (5) ( 4.36)
13 2 1 2.29
(81) (13) (6) ( 4.17)
49 7 3 4.17
(83) (12) (5) ( 3.73)
53 (71) 21 (28) 1 (1)
13 (81) 2 (13) 1 (6)
40 (68) 19 (32) 0 (0)
38 (51) 37 (49)
3 (19) 13 (81)
35 (59) 24 (41)
0.002 0.067c
0.005
a Continuous data are mean ( standard deviation), categorical data are n (%). c variables) or c2 test (categorical variables). Fisher’s exact test.
BOS ¼ bronchiolitis obliterans syndrome; DL ¼ double lung; PGD ¼ primary graft dysfunction; SL ¼ single lung.
b
. . . 4.36 ( 3.37)
. . . 0.26 ( 0.46)
. . . 2.57 ( 4.18)
46 (74) 16 (26) 0 (0)
4 (44) 4 (44) 1 (11)
3 (75) 1 (25) 0 (0)
31 (50) 31 (50)
4 (44) 5 (56)
3 (75) 1 (25)
0.05 0.097c
0.799c
The p value is based on one-way analysis of variance (continuous
HLT ¼ heart-lung transplantation;
should be performed in a group of carefully selected patients; however, re-LTx for PGD should be avoided [8]. Osaki and colleagues [11] presented 17 patients from their center who underwent re-LTx between 1998 and 2007, of which 12 re-LTx were caused by BOS. They showed a 1-year survival of 67% and a 5-year survival of 44% for BOS. In 2008, Kawut and colleagues [19] presented a retrospective cohort study of patients who underwent re-LTx between January 2001 and May 2006 in the United States (n ¼ 205; modern re-LTx cohort). The characteristics and survival of this cohort were compared with patients who underwent first re-LTx between January 1990 and December 2000 (historical re-LTx cohort), and patients who underwent primary LTx between January 2001 and May 2006 (modern initial transplant cohort). They showed a 1-year survival of 62% and a 5-year survival of 45%. The authors did not specifically address the outcome for patients undergoing re-LTx for BOS [19]. They also showed that early (
Background. The increasing demand for pulmonary retransplantation (re-LTx) raises ethical issues on the correct allocation of the scarce donor pool. Thus, we performed a thorough review of the current results for reLTx in the Nordic countries. Methods. Seventy-five patients with a median age of 50 years (range, 22 to 64 years) underwent re-LTx from 1992 until June 2013, of which 53 had single re-LTx, 21 had double re-LTx, and 1 patient underwent a heart-lung retransplantation. Primary graft dysfunction (PGD) was the primary indication in 9 cases, bronchiolitis obliterans syndrome (BOS) in 62 cases, and airway complications in 4 cases. Results. Patients who underwent re-LTx in the period 1992 to 1999 (n [ 16) had a 1-year survival of 37.5% (95% confidence interval [CI], 19.9 to 70.6), whereas patients who underwent re-LTx in the period 2000 to 2013 (n [ 64) had a 1-year survival of 81.0% (95% CI, 71.5 to 91.8). Corresponding 5-year survival was 25.0% (95% CI, 10.7 to 58.4) in the early era group (1992 to 1999) and 57.2% (95% CI, 44.3 to 73.7) in the more recent era group (2000 to 2013; p [ 0.0151). Patients with BOS who underwent re-LTx in the period 1992 to 1999 (n [ 13) had a 1-year survival of 38.5% (95% CI, 19.3 to 76.5), whereas patients with BOS
who underwent re-LTx in the period 2000 to 2013 (n [ 49) had a 1-year survival of 85.4% (95% CI, 75.9 to 96.0). Corresponding 5-year survival was 23.1% (95% CI, 8.6 to 62.3) in the early era group (1992 to 1999) and 56.1% (95% CI, 41.9 to 75.2) in the more recent era group (2000 to 2013; p [ 0.0199). The cumulative incidence among patients who underwent re-LTx because of BOS and developed BOS again after re-LTX was analyzed. The cumulative incidence curves for time periods 1992 to 1999 and 2000 to 2013 are not statistically different for repeat BOS (p [ 0.5087), but they are highly significant for time periods among patients who died (p [ 0.02381). Conclusions. Results for re-LTx have improved over time, especially when BOS is the primary indication. The cumulative incidence among patients who underwent reLTx because of BOS and developed repeat BOS after reLTX showed equal risk between 1992 to 1999 and 2000 to 2013 in the aspect of developing repeat BOS, but in the later era the patients had a significantly higher chance of surviving.
T
regimens in antibiotic prophylaxis and treatment [3, 4]. However, acute and chronic rejection and dysfunction continue to be major problems. The primary cause of death after LTx is chronic rejection or bronchiolitis obliterans syndrome (BOS), and pulmonary retransplantation (re-LTx) is currently the only treatment option for these patients. The development of BOS is rare in the first year after LTx, but the rate increases quickly, with cumulative incidence reported to be as high as 40% to 80% within the first 5 years [5–7]. Other complications that could lead to irreversible graft dysfunction and respiratory failure after LTx, where re-LTx remains the only treatment option, is primary graft dysfunction (PGD), severe airway complication, and refractory acute graft rejection. Given the extent of late graft failure, it is inevitable that some patients, especially those who undergo transplantation at a young age (ie, patients with cystic
he introduction of cyclosporin A made it possible to achieve successful heart and lung transplantations [1]. Clinical isolated lung transplantation (LTx) with longterm survival was initially performed in the early 1980s by the Toronto Lung Transplantation Group [2], and has evolved into an established treatment, and the only lifesaving therapy, for patients with a variety of end-stage pulmonary diseases. Survival in LTx recipients has increased over the years, mainly as a result of careful patient selection; improved lung preservation, surgical techniques, immunosuppressive regimes, and management of ischemia/reperfusion injury; and forceful
Accepted for publication Jan 16, 2015. Address correspondence to Dr Wierup, Department of Cardiothoracic Surgery, Skåne University Hospital, Lund University, 221 85 Lund, Sweden; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;99:1781–7) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.01.052
GENERAL THORACIC
Departments of Cardiothoracic Surgery and Pulmonary Medicine, Skåne University Hospital, Lund University, Sweden; Departments of Pulmonary Medicine and Cardiothoracic Surgery, Sahlgrenska University Hospital, Gothenburg, Sweden; Department of Pulmonary Medicine, Rigshospitalet, Copenhagen, Denmark; Departments of Pulmonary Medicine and Cardiothoracic Surgery, Rikshospitalet, Oslo, Norway; and Department of Cardiac Surgery, Helsinki University Hospital, Helsinki, Finland
GENERAL THORACIC
1782
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
fibrosis), will eventually exhibit recurrent respiratory failure, for which re-LTx will have to be considered. Pulmonary retransplantation is believed to carry a higher risk compared with primary LTx as reflected in the outcome of various studies. Historically, re-LTx showed disappointing results; however, a few recent studies report 1- and 5-year survival rates for re-LTx comparable to current survival rates for primary LTx [8–12]. The timing and indication, the presence of other comorbidities, immunosuppressive management, infection prophylaxis, and the likelihood of success are only a few factors that have to be considered before choosing re-LTx. The value of re-LTx has been heavily debated on ethical grounds as an overall organ shortage means that higher risk re-LTx patients will displace some patients on the waiting list for primary LTx [3, 4, 13, 14]. Owing to an increasing number of potential candidates for re-LTx in the Nordic countries, a review of the current results for re-LTx in this region was made.
Patients and Methods Patients Seventy-five patients underwent re-LTx from 1992 through 2013 in the Nordic countries at five Nordic transplantation centers (Skåne University Hospital, Lund University, Sweden; Rigshospitalet, Copenhagen, Denmark; Sahlgrenska University Hospital, Gothenburg, Sweden; Rikshospitalet, Oslo, Norway; and Helsinki University Hospital, Helsinki, Finland). The median age was 50 years (range, 22 to 64 years). Forty-six of the patients were men and 29 were women. Fifty-three patients received a single re-LTx, and 21 patients had a double re-LTx. Thirty-five patients underwent a contralateral single re-LTx, 18 patients underwent an ipsilateral single re-LTx. Indications for re-LTx were BOS in 62 patients, PGD in 9 patients, and airway complications in 4 patients. Ten patients had ventilator support preoperatively; 2 of those patients underwent re-LTx as a result of BOS, and the rest of the patients had PGD. Two patients had extracorporeal membrane oxygenation support preoperatively, of whom 1 patient underwent reLTx because of PGD, and 1 patient because of BOS. The early era was defined between the years 1992 through 1999, and the more recent era was 2000 until June 2013. Bronchiolitis obliterans syndrome was defined by spirometry as more than 20% decline in forced expiratory volume in 1 second from the highest obtained baseline, according to International Society for Heart and Lung Transplantation (ISHLT) guidelines [15]. Primary graft dysfunction was retrospectively identified in patients with retransplantation owing to initial graft failure in the first months before the ISHLT definition in 2005 [16]. After 2005, the ISHLT definition of PGD had been used prospectively.
Data Acquisition and Follow-Up The study was performed for the Nordic Thoracic Transplantation Study Group, which is a scientific and
Ann Thorac Surg 2015;99:1781–7
clinical working group comprising members from all thoracic transplant centers in the Nordic countries (Oslo, Gothenburg, Lund, Copenhagen, Aarhus, Helsinki) and which includes all transplant centers in Scandia transplant. Data were collected retrospectively from lung transplantation databases in the centers and were reviewed for demographic, operative, perioperative, and outcome data. Follow-up data were obtained through patient files, and follow-up was 100% complete. A retrospective study like this does not require approval from a scientific ethical committee in the Nordic countries.
Statistical Analysis Primary stratification of the material was made into two sets of cohorts. The first cohort was based on the main indication for re-LTx, with the following indicator cohorts: BOS, PGD, or airway complications. The second set divides the material based on time period for reLTx: 1992 to 1999 and 2000 to 2013. Time to re-LTx and transplant re-LTx type—single lung (SL) or double lung (DL)—were compared between the sets of cohorts. Primary outcomes include post-re-LTx survival. Overall survival was measured at 1 year and 5 years after the re-LTx by the Kaplan-Meier method, and compared between the two sets of cohorts; indications for re-LTx and time period were accounted for with the log-rank test. Kaplan-Meier survival was also measured for combinations of the cohort sets and transplant re-LTx type, sample sizes allowing. Patients were censored if they reached the end of the study period, June 1, 2014, or were lost to follow-up. All causes of death were incorporated into the survival analyses. Continuous variables are expressed as mean ( standard deviation) and categorical variables as number (percent). Statistical comparisons of continuous variables were made by Student’s t test or analysis of variance when necessary. Statistical comparisons of categorical variables were made by c2 analysis or Fisher’s exact test. Kaplan-Meier survival estimates for 1 and 5 years are expressed as years (95% confidence interval [CI], point estimate). For all statistical analyses, a probability value of less than 0.05 was considered significant. All calculations above were performed using SPSS version 19.0 (IBM Corp, Armonk, NY). A subanalysis of this study was the recurrence of bronchiolitis obliterans syndrome (re-BOS) after retransplantation among patients with BOS as a primary indicator for re-LTx. In this analysis, death will act as a competing risk event to the event of re-BOS. In a competing-risks model, we analyzed incidence of re-BOS and death as two separate outcomes. Specifically, we estimated and compared the cumulative incidence functions for re-BOS and death using Gray’s test (1988). All calculations regarding competing risks were performed using R with the CMPRSK package (available at http:// www.r-project.org). The statistics were performed by Roy Ollila at Sidsoft AB, Malmo, Sweden.
Ann Thorac Surg 2015;99:1781–7
Results
Fig 1. Survival after redo lung transplantation during the years 1992 to 1999 (early era patient group; solid line), and during the years 2000 to 2013 (more recent era patient group; dashed line) in the Nordic countries. Patients in the more recent era had a significantly higher chance of surviving compared with the early era patient group.
1783
The cohort was divided into groups regarding type of transplant, SL versus DL. During the period 1992 to 2013, a total of 53 patients received SL and 21 patients received DL. Overall 1-year survival for the whole cohort (1992 to 2013) was 75.0% (95% CI, 64.1 to 87.8) for SL and 65.9% (95% CI, 48.2 to 90.1) for DL. Corresponding 5-year survival was 51.2% (95% CI, 38.3 to 68.3) for SL and 50.2% (95% CI, 30.7 to 82.2; p ¼ 0.514) for DL. The cohort was divided into the two periods, 1992 to 1999 and 2000 to 2013. Between 1992 and 1999 a total of 13 patients underwent SL re-LTx and 2 patients underwent DL re-LTx. Kaplan-Meyer survival was not applicable because of the small patient group. Regarding the time period 2000 to 2013, 1-year survival was 84.8% (95% CI, 74.3 to 96.8) for SL (n ¼ 40) and 72.9% (95% CI, 55.1 to 96.4) for DL (n ¼ 19). Corresponding 5-year survival was 59.1% (95% CI, 44.4 to 78.7) for SL and 55.5% (95% CI, 34.6 to 89.1; p ¼ 0.546; Fig 2). The patients in the cohort who underwent re-LTx because of BOS were analyzed. From 1992 to 2013, a total of 62 patients underwent re-LTx because of BOS. The 1-year survival was 85.4% (95% CI, 77.0 to 94.7), and the 5-year survival was 50.9% (95% CI, 38.6 to 67.1). The cohort was divided into the two time periods 1992 to 1999 and 2000 to 2013. Patients with BOS who underwent re-LTx in the period 1992 through 1999 (n ¼ 13) had a 1-year survival of 38.5% (95% CI, 19.3 to 76.5), whereas patients with BOS who underwent re-LTx in the period 2000 through 2013 (n ¼ 49) had a 1-year survival of 85.4% (95% CI, 75.9 to 96.0). Corresponding 5-year survival was 23.1% (95% CI, 8.6 to 62.3) in the early era group (1992 to 1999) and 56.1% (95% CI, 41.9 to 75.2) in the more recent era group (2000 to 2013; p ¼ 0.0199; Fig 3). The same cohort (re-LTx because of BOS) was divided regarding type of transplant. A total of 62 patients
Fig 2. Survival after redo lung transplantation during the years 2000 to 2013 (more recent era patient group) in the Nordic countries, comparing single redo lung transplantation (solid line) and double redo lung transplantation (dashed line). No significant difference was seen when comparing the two patient groups.
GENERAL THORACIC
Levine’s test for equality of variances tests the hypothesis that the two populations’ variances are equal. For the two time periods, 1992 to 1999 and 2000 to 2013, we found Levine’s statistic F ¼ 8.303 to be significant (p ¼ 0.005), and thus the assumption of equal variances are violated. Hence we used the equal variance not assumed t test statistic to evaluate H0: equality of means, here (H0: average time to re-LTx for 1992 to 1999 ¼ average time to re-LTx for 2000 to 2013). The results indicate that there is a significant difference between the time periods in time to re-LTx, t (df ¼ 61) ¼ 3.036, p ¼ 0.004 (p ¼ 0.00352). The mean values indicate that patients during the time period 1992 to 1999 had a significantly shorter time to re-LTx (1992 to 1999 ¼ 2.29 years to re-LTX) than patients treated during 2000 to 2013 (2000 to 2013 ¼ 4.17 years to re-LTx). A total of 75 patients underwent re-LTx between 1992 and 2013. Overall 1-year survival for the whole cohort was 71.4% (95% CI, 61.8 to 82.6). Corresponding 5-year survival was 49.2% (95% CI, 37.9 to 63.9). The cohort was divided into two time periods, 1992 to 1999 and 2000 to 2013. Patients who underwent re-LTx in the period 1992 through 1999 (n ¼ 16) had a 1-year survival of 37.5% (95% CI, 19.9 to 70.6), whereas patients who underwent re-LTx in the period 2000 through 2013 (n ¼ 64) had a 1-year survival of 81.0% (95% CI, 71.5 to 91.8). Corresponding 5-year survival was 25.0% (95% CI, 10.7 to 58.4) in the early era group (1992 to 1999) and 57.2% (95% CI, 44.3 to 73.7) in the more recent era group (2000 to 2013; p ¼ 0.0151; Fig 1).
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
GENERAL THORACIC
1784
Fig 3. Survival after redo lung transplantation because of bronchiolitis obliterans syndrome during the years 1992 to 1999 (early era patient group; solid line), and during the years 2000 to 2013 (more recent era patient group; dashed line) in the Nordic countries. Patients in the more recent era had a significantly higher chance of surviving compared with the early era patient group.
underwent re-LTx because of BOS between 1992 and 2013 (SL n ¼ 46, DL n ¼ 16). One-year survival for SL (n ¼ 46) was 75.5% (95% CI, 63.9 to 89.2) and 74.5% (95% CI, 55.7 to 99.6) for DL (n ¼ 16). Corresponding 5-year survival was 47.2% (95% CI, 33.5 to 66.5) for SL and 54.3% (95% CI, 32.0 to 92.3; p ¼ 0.921; Fig 4). The cumulative incidence amongst patients who underwent re-LTx because of BOS and developed re-BOS
Fig 4. Survival after redo lung transplantation for bronchiolitis obliterans syndrome during the years 1992 to 2013 in the Nordic countries, comparing single redo lung transplantation (solid line) and double redo lung transplantation (dashed line). No significant difference was seen when comparing the two patient groups.
Ann Thorac Surg 2015;99:1781–7
after re-LTX was analyzed. Figure 5 shows the results of Gray’s test for equality of cumulative incidence functions across groups. The cumulative incidence curves for time periods 1992 to 1999 and 2000 to 2013 are not statistically different for re-BOS (p ¼ 0.5087), but they are highly significant for time periods among patients who died (p ¼ 0.02381). Nine patients underwent re-LTx because of PGD between the years 1992 and 2013. Between the time period 1992 to 1999, a total of 2 patients underwent re-LTX. Corresponding figures for the time period 2000 to 2013 was 7 patients. A descriptive analysis is shown in Table 1. Kaplan-Meyer survival was not applicable because of small group size. Four patients underwent re-LTx because of airway complications between the years 1992 and 2013. Between the time period 1992 to 1999, only 1 patient underwent reLTx. Corresponding figures for the time period 2000 to 2013 was 3 patients. A descriptive analysis is shown in Table 1. Kaplan-Meyer survival was not applicable because of small group size. Ten patients received preoperative ventilator support between the years 1992 and 2013. Between the time period 1992 to 1999, a total of 2 patients underwent reLTX. Corresponding figure for the time period 2000 to 2013 was 8 patients. Kaplan-Meyer survival was not applicable because of small group size.
Comment Despite improving results in LTx, a significant number of grafts fail early or late postoperatively, and some patients will require re-LTx. Redo LTx carries a higher risk than primary LTx, and the outcome of re-LTx has traditionally not been good [3, 4]. Worldwide, the demand for re-LTx has increased. From an ethical point of view, it is important to outline the actual benefit of re-LTx as the donor supply does not meet the needs of the listed candidates waiting for primary LTx. The ISHLT Pulmonary Retransplant Registry was founded in 1991 to determine the predictors of outcome after re-LTx and facilitate decisions concerning the appropriateness of re-LTx in individual patients. In a 1996 multicenter study (35 centers worldwide), Novick and colleagues [17] reported that the most important indicator for early survival after re-LTx (n ¼ 160) was the ambulatory status of the recipients. The follow-up analysis made by the same group in 1998, which included 230 patients from 47 centers across the United States, Canada, Europe, and Australia, identified factors favoring survival and freedom from BOS after re-LTx: surgery performed after 1991, no preoperative ventilator support, an interval of more than 2 years passed from primary LTx, and transplant center experience [18]. One important underlying cause of improvements in re-LTx after 1991 (the so called “era effect”) seems to be the ability to distinguish between high- and low-risk candidates for re-LTx. The observation that survival is highest in patients with BOS has lately been supported by data from the Hannover Lung Transplantation group [12]. In 2006,
Ann Thorac Surg 2015;99:1781–7
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
1785
GENERAL THORACIC
Fig 5. The cumulative incidence among patients who underwent redo lung transplantation because of bronchiolitis obliterans syndrome and exhibited repeat bronchiolitis obliterans syndrome after redo lung transplantation was analyzed. The cumulative incidence curves for time periods (A, C) 1992–1999 and (B, D) 2000–2013 are not statistically different for (C, D) repeat bronchiolitis obliterans syndrome (ReBOS) (p ¼ 0.5087), but they are highly significant for time periods among (A, B) patients who died (p ¼ 0.02381). The patients in the two different eras had the same risk of exhibiting repeat bronchiolitis obliterans syndrome, but in the later era, the patients had a significantly higher chance of surviving.
Strueber and colleagues [12] presented the largest ever single-center series of re-LTx (n ¼ 54). Their most striking finding was the identification of a group of patients for whom re-LTx had a comparable outcome equal to that for primary LTx. This was true for younger patients with BOS with an otherwise good prognosis, such as cystic fibrosis. In this group, 5-year survival as well as freedom from BOS revealed a similar course to that of primary LTx [12]. On the other hand, patients receiving mechanical ventilation showed a discouraging 1-year survival of 50% compared with ambulatory patients (78%). However, these figures were equivalent to the results for patients undergoing primary LTx who were also receiving mechanical ventilation at the same center. They also found that airway complications were
more prevalent in elderly patients with emphysema, and that re-LTx for acute graft failure was more common in patients with fibrosis and idiopathic pulmonary hypertension. They concluded that re-LTx for airway complications and acute graft failure had been rare during the last 5 years in their center, mainly because of better patient selection for re-LTx [12]. In 2008, Aigner and colleagues [8] from Vienna reported their experience for re-LTx (n ¼ 46), stratified for different indications: PGD (n ¼ 23), BOS (n ¼ 19), and airway problems (n ¼ 4). Survival rates at 1 and 5 years were 35% and 29% in the PGD group, and 73% and 61% in the BOS group, respectively. All 4 patients with airway problems were alive at the follow-up time. They concluded that re-LTx for BOS and airway problems
1786
LINDSTEDT ET AL PULMONARY RETRANSPLANTATION IN NORDIC COUNTRIES
Ann Thorac Surg 2015;99:1781–7
Table 1. Comparison of Transplant Characteristics Among Time Period and Indication for Redo Lung Transplantationa Time Period
GENERAL THORACIC
Variable Time period 1992–1999 2000–2013 Indication for re-LTx BOS PGD Airway Time to re-LTx (y) Transplant re-LTx type SL DL HLT Status Alive Dead
Total re-LTx (n ¼ 75)
1992–1999 (n ¼ 16)
2000–2013 (n ¼ 59)
. .
. .
Indication for re-LTX p Valueb
BOS (n ¼ 62)
PGD (n ¼ 9)
Airway (n ¼ 4)
. 16 (21) 59 (79)
0.98 13 (21) 49 (79)
2 (22) 7 (78)
1 (25) 3 (75) .
0.98 62 9 4 3.77
(83) (12) (5) ( 4.36)
13 2 1 2.29
(81) (13) (6) ( 4.17)
49 7 3 4.17
(83) (12) (5) ( 3.73)
53 (71) 21 (28) 1 (1)
13 (81) 2 (13) 1 (6)
40 (68) 19 (32) 0 (0)
38 (51) 37 (49)
3 (19) 13 (81)
35 (59) 24 (41)
0.002 0.067c
0.005
a Continuous data are mean ( standard deviation), categorical data are n (%). c variables) or c2 test (categorical variables). Fisher’s exact test.
BOS ¼ bronchiolitis obliterans syndrome; DL ¼ double lung; PGD ¼ primary graft dysfunction; SL ¼ single lung.
b
. . . 4.36 ( 3.37)
. . . 0.26 ( 0.46)
. . . 2.57 ( 4.18)
46 (74) 16 (26) 0 (0)
4 (44) 4 (44) 1 (11)
3 (75) 1 (25) 0 (0)
31 (50) 31 (50)
4 (44) 5 (56)
3 (75) 1 (25)
0.05 0.097c
0.799c
The p value is based on one-way analysis of variance (continuous
HLT ¼ heart-lung transplantation;
should be performed in a group of carefully selected patients; however, re-LTx for PGD should be avoided [8]. Osaki and colleagues [11] presented 17 patients from their center who underwent re-LTx between 1998 and 2007, of which 12 re-LTx were caused by BOS. They showed a 1-year survival of 67% and a 5-year survival of 44% for BOS. In 2008, Kawut and colleagues [19] presented a retrospective cohort study of patients who underwent re-LTx between January 2001 and May 2006 in the United States (n ¼ 205; modern re-LTx cohort). The characteristics and survival of this cohort were compared with patients who underwent first re-LTx between January 1990 and December 2000 (historical re-LTx cohort), and patients who underwent primary LTx between January 2001 and May 2006 (modern initial transplant cohort). They showed a 1-year survival of 62% and a 5-year survival of 45%. The authors did not specifically address the outcome for patients undergoing re-LTx for BOS [19]. They also showed that early (
