Structured Abstract
Background
To maximize the benefit of lung transplantation, the effect of size mismatch on survival in lung transplant recipients with restrictive lung disease (RLD) was examined.
Methods
All single and bilateral RLD lung transplants from 1987–2011 in the UNOS Database were identified. Donor pTLC:Recipient pTLC ratio (pTLCr), quantified mismatch. pTLCr was segregated into 5 strata. A Cox proportional hazards model evaluated the association of pTLCr with mortality hazard. To identify a critical pTLCr, a Cox model using a restricted cubic spline for pTLCr was used.
Results
6,656 transplants for RLD were identified. Median pTLCr for SOLT and BOLT was 1.0 (0.69–1.47) and 0.98 (0.66–1.45). Examination of pTLCr as a categorical variable revealed that undersizing (pTLCr<0.8) for SOLT and moderate oversizing (pTLCr=1.1–1.2) for SOLT and BOLT had a harmful survival effect (for SOLT pTLC < 0.8: HR 1.711 [95% CI 1.146–2.557], P = 0.01 and for BOLT pTLC 1.1–1.2: HR 1.717 [95% CI 1.112–2.651], P = 0.02). Spline analysis revealed significant changes in SOLT mortality by variation of pTLCr between 0.8–0.9 and 1.1–1.2.
Conclusions
RLD patients undergoing SOLT are susceptible to detriments of an undersized lung. RLD patients undergoing BOLT have higher risk of mortality when pTLCr falls between 1.1–1.2.
Introduction
Lung transplantation is the most durable and efficacious treatment of end-stage lung disease and the choice of donor organs has long been a subject of debate. Many donor characteristics including age, comorbidities, and smoking status, have been examined with regards to their impact on outcomes following lung transplantation. More recently, some groups have examined the impact of donor and recipient lung size on survival following lung transplantation.1–4
Due to limitations on ability to physically measure lung dimensions, predictors of lung size have been used instead. One example is predicted total lung capacity (pTLC), which estimates lung volumes by using height as an independent variable.5 This method has been validated and utilized to compare the relationship between donor and recipient lung size in previous studies.1–3, 6–8
However, previous efforts have not specifically focused on restrictive lung disease (RLD), instead including a heterogenous collection of patients. Lung transplant recipients with RLD represent a unique population as the pathophysiology of their disease makes choice of donor lungs important. In particular patients with RLD often have chest wall contraction that affects pulmonary mechanics differently than other transplantation indications.9 Moreover, it is common to have single orthotopic lung transplantation (SOLT) in restrictive disease, and there is currently minimal data regarding size matching in SOLT.9 As such the primary purpose of this study was to examine how the relationship of donor to recipient predicted lung capacity impacts long-term survival following both SOLT and bilateral orthotopic lung transplantation (BOLT) in patients with RLD.
Methods
Data Source
The protocol for this retrospective analysis was approved by the Institutional Review Board at our institution. The United Network for Organ Sharing (UNOS) administers the Organ Procurement and Transplantation Network (OPTN) under contract with the U.S. Department of Health and Human Services (HHS). The UNOS Standard Transplant Analysis and Research file was queried to identify all lung transplants from 1987–2011. All patients with RLD as an indication for transplant were identified as defined by the primary diagnosis at the time of transplantation in the UNOS STAR file. Pediatric, multi-organ, and redo lung transplants were excluded.
Patient Classification/Outcome
The pTLC was calculated for all patients in the final cohort using the equation described by Stocks et al.5; pTLC=7.99*Height−7.08 or 6.60*Height−5.79 for males and females respectively. For purposes of comparison donor and recipients pTLC were compared as a ratio of the donor pTLC: recipient pTLC (pTLCr). Patients were then grouped based on the pTLCr with a pTLCr < 0.9 being defined as undersized, pTLCr between, and including, 0.9 and 1.1 being similary sized, and pTLCr > 1.1 termed oversized. Given the differences in expected outcomes between single (SOLT) and bilateral (BOLT) orthotopic lung transplantation patients were further stratified based on procedure. The primary outcome of the study was long-term survival. Secondary outcomes included postoperative need for dialysis, post-operative bronchial stricture, and length of stay.
Statistical Methods
Following stratification as detailed above, comparisons amongst the three previously described groups were conducted using the analysis of variance (ANOVA) or the Chi-square test for continuous and categorical variables respectively. A Cochran-Armitage trend test was utilized to examine trends in sizing over time, by year. Given the concern that the initial stratification was too simple with regards to sizing the under and oversized, each category was broken into two groups. For undersized they were pTLCr < 0.8 and pTLCr between 0.8 and 0.9, while for oversized it was pTLCr between 1.1 and 1.2 and pTLCr > 1.2. Thus, a total of five groups were established for both SOLT and BOLT. As the primary endpoint was long-term survival, a Cox proportional hazard model for both SOLT and BOLT, was employed to examine the independent effect of pTLCr on long-term survival. Variables included in the model were patient, donor, and procedural characteristics known to influence post-transplant survival. Additionally, Kaplan-Meier analysis was utilized to provide a visual representation of survival amongst the five groups.
While the initial Cox proportional hazard model identified the impact of pTLCr on survival it did not indicate where changes in pTLCr had the greatest impact on survival. Thus, to address this limitation we utilized a separate Cox proportional hazard model with a restricted cubic spline to determine where changes in pTLCr were greatest for both SOLT and BOLT. Relative mortality hazard was assessed in relation to a pTLC of 1.0 for both SOLT and BOLT.
In all cases, type I error was controlled at the level of comparison and a p-value of 0.05 was determined a priori to be the level of significance. All statistics were performed using R version 3.3.1, R Foundation for Statistical Computing, Vienna, Austria. Restricted cubic splines were developed using the package rms: Regression Modeling Strategies, R package version 4.5-0.
Results
From 1987 to 2011 6,656 lung transplants met inclusion criteria with n=3,532 (53.1%) undergoing SOLT and n=3,124 (46.9%) undergoing BOLT. Overall, in patients undergoing SOLT 66.9% (n=2,362) were similarly sized, 19.7% (n=696) were undersized and 13.4% (n=474) were considered oversized. For those undergoing BOLT, 64.0% (n=1,998) were similarly sized, 24.1% (n=753) were undersized, and 11.9% (n=373) were oversized. In both cases there was a trend towards the use of similarly sized and undersized lungs during the study time period (Figure 1; p< 0.01 for both SOLT and BOLT).
Figure 1.
Sizing of transplant by era; A) Single orthotopic lung transplant B) Bilateral orthotopic lung transplant (*Note: For Cochran-Armitage trend test sizing was analyzed by individual year, not era)
Regarding donor characteristics, older donors were associated with under-sizing, while younger donors were associated with over-sizing in both SOLT and BOLT. With regards to recipients, under-sizing was associated with older recipients, while over-sizing was associated with younger patients for both SOLT and BOLT. In general, undersized SOLT and BOLT recipients were also associated with an increased incidence of diabetes, were more likely to be white race, and were more likely to be hospitalized or in the intensive care unit prior to transplant (p<0.01 in all cases). In patients with available data, undersized patients had a statistically significant higher lung allocation score (LAS) (44.0 and 49.2 for SOLT and BOLT respectively) than similarly sized (42.6 and 46.3 for SOLT and BOLT respectively) and oversized recipients (41.7 and 46.3 for SOLT and BOLT respectively), although this is unlikely to represent a clinically meaningful difference. Interestingly, in both SOLT and BOLT patients, being female was heavily associated with oversized pTLCr (92% and 96% female for SOLT and BOLT respectively). Also notable was the fact that patients who were undersized had a significantly shorter waitlist length (59 and 56 days for SOLT and BOLT respectively) while patients who were oversized had a significantly longer waitlist time (155 and 97 days respectively) (p<0.01 for all comparisons) (Tables 1 and 2). Unadjusted analysis revealed no significant differences in post-operative dialysis requirement (for SOLT patients: 4.0%, 4.2%, and 3.9%; p=0.96; for appropriately-, under-, and over-sized respectively; BOLT patients: 8.3%, 8.4%, and 6.6% for appropriately-, under-, and over-sized respectively; p=0.52) and median length of stay (for SOLT patients: 12, 12, and 13 days; for appropriately, under, and over-sized respectively; p=0.24; BOLT patients: 17, 17, and 17 days; p=0.24 for appropriately, under, and over-sized respectively; p=0.37). However undersized patients had a statistically significant higher incidence of post-operative airway dehiscence (2.1% and 3.0% for SOLT and BOLT respectively) compared to appropriately (0.8% and 1.4% for SOLT and BOLT respectively) and over-sized recipients (1.1% and 1.7% for SOLT and BOLT respectively) (p=0.03).
Table 1.
Donor, Recipient, and Transplant Characteristics for Single Orthotopic Lung Transplant Patients (N=3532)
| Variable | Overall | Appropriately Sized |
Under Sized | Over Sized | P-Value | |
|---|---|---|---|---|---|---|
| Donor Characteristics | ||||||
| Age in Years | 31 (21, 45) | 29 (20, 43) | 41 (29, 51) | 24.5 (18, 37) | < 0.001 | |
| BMI (kg/m^2) | 24.3 (21.7, 27.5) | 24.4 (21.8, 27.4) | 24.9 (21.9, 29.1) | 23.7 (21.2, 26.4) | < 0.001 | |
| Tobacco Abuse | 643 (19%) | 419 (18.5%) | 143 (21%) | 81 (18.4%) | 0.322 | |
| Drug Use | 1,023 (29%) | 751 (31.8%) | 146 (21%) | 126 (26.6%) | < 0.001 | |
| Diabetes | 158 (4.6%) | 90 (3.9%) | 45 (6.6%) | 23 (5.2%) | 0.013 | |
| Recipient Characteristics | ||||||
| Age in Years | 60 (54, 65) | 60 (54, 65) | 63 (58, 66) | 57 (50, 63) | < 0.001 | |
| BMI kg/m^2 | 27.2 (24.2, 29.8) | 27.3 (24.2, 29.9) | 27.1 (24.6, 29.5) | 26.8 (23.1, 29.5) | 0.087 | |
| GFR | 88.6 (72.3, 97.8) | 88.7 (73, 98.3) | 89.3 (74.7, 97.1) | 83.9 (68, 97.7) | 0.063 | |
| Diabetes | 493 (14.7%) | 316 (14%) | 130 (19.4%) | 47 (11%) | < 0.001 | |
| Ethnicity | < 0.001 | |||||
| White | 2,943 (83.3%) | 1,988 (84.2%) | 586 (84.2%) | 369 (77.8%) | ||
| Black | 254 (7.2%) | 166 (7%) | 30 (4.3%) | 58 (12.2%) | ||
| Other | 335 (9.5%) | 208 (8.8%) | 80 (11.5%) | 47 (9.9%) | ||
| Functional Status | 0.008 | |||||
| ADL No Assistance | 1,073 (33.3%) | 749 (35%) | 208 (31.7%) | 116 (27.6%) | ||
| ADL With Assistance | 2,147 (66.7%) | 1,393 (65%) | 449 (68.3%) | 305 (72.4%) | ||
| Female Gender | 1,152 (32.6%) | 707 (29.9%) | 9 (1.3%) | 436 (92%) | < 0.001 | |
| Lung Allocation Score | 42.9 (38.2, 51.7) | 42.6 (38, 50.2) | 44 (38.7, 55.8) | 41.7 (37.9, 52.4) | 0.003 | |
| Operative Characteristics | ||||||
| Type of Lung Transplant (Right or Left) |
0.002 | |||||
| Single Right | 1,515 (42.9%) | 990 (41.9%) | 338 (48.6%) | 187 (39.5%) | ||
| Single Left | 2,017 (57.1%) | 1,372 (58.1%) | 358 (51.4%) | 287 (60.5%) | ||
| Medical Condition Before Transplant |
0.562 | |||||
| Not Hospitalized | 3,185 (90.6%) | 2,136 (91%) | 618 (89.2%) | 431 (91.1%) | ||
| Hospitalized Not in ICU | 188 (5.4%) | 119 (5.1%) | 46 (6.6%) | 23 (4.9%) | ||
| In ICU | 141 (4%) | 93 (4%) | 29 (4.2%) | 19 (4%) | ||
| Waitlist Days | 92 (27, 256) | 93 (28, 253.8) | 59 (18, 191.5) | 155 (42.2, 353.5) | < 0.001 | |
| Center Volume | 524 (312, 686) | 520 (312, 661) | 525 (312, 750) | 423 (284, 617) | < 0.001 | |
All statistics displayed as median (interquartile range) or number (percent).
ADL, Activities of Daily Living; BMI, Body Mass Index; GFR, Glomerular Filtration Rate; ICU, Intensive Care Unit, IQR, Interquartile Range.
Center Volume defined as number of transplants during entire study period
Table 2.
Donor, Recipient, and Transplant Characteristics for Bilateral Orthotopic Lung Transplant Patients (N=3124)
| Variable | Overall | Appropriately Sized |
Under Sized | Over Sized | P-Value | |
|---|---|---|---|---|---|---|
| Donor Characteristics | ||||||
| Age in Years (IQR) | 32 (21, 46) | 30 (21, 45) | 40 (26, 50) | 22 (18, 35) | < 0.001 | |
| BMI (IQR) kg/m^2 | 24.5 (21.9, 27.8) | 24.5 (22.1, 27.6) | 25.1 (22.1, 29.3) | 23.9 (21, 26.6) | < 0.001 | |
| Tobacco Abuse | 430 (13.9%) | 273 (13.8%) | 114 (15.3%) | 43 (11.6%) | 0.24 | |
| Drug Use | 969 (31%) | 644 (32.2%) | 191 (25.4%) | 134 (35.9%) | < 0.001 | |
| Diabetes | 170 (5.5%) | 111 (5.6%) | 44 (5.9%) | 15 (4.1%) | 0.428 | |
| Recipient Characteristics | ||||||
| Age in Years (IQR) | 56 (48, 62) | 56 (48, 61) | 58 (50, 63) | 53 (43, 59) | < 0.001 | |
| BMI (IQR) kg/m^2 | 26.5 (23.4, 29.4) | 26.6 (23.5, 29.5) | 26.6 (24, 29.5) | 25.5 (21.7, 28.6) | < 0.001 | |
| GFR | 93.8 (77.4, 104.5) | 94.1 (77.9, 104.2) | 92.9 (78.6, 103.5) | 94.2 (75, 106.9) | 0.855 | |
| Diabetes | 466 (15.2%) | 290 (14.7%) | 136 (18.3%) | 40 (11.1%) | 0.005 | |
| Ethnicity | < 0.001 | |||||
| White | 2,314 (74.1%) | 1,472 (73.7%) | 588 (78.1%) | 254 (68.1%) | ||
| Black | 441 (14.1%) | 303 (15.2%) | 64 (8.5%) | 74 (19.8%) | ||
| Other | 369 (11.8%) | 223 (11.2%) | 101 (13.4%) | 45 (12.1%) | ||
| Functional Status | 0.819 | |||||
| ADL No Assistance | 856 (28.9%) | 552 (29.2%) | 209 (28.7%) | 95 (27.5%) | ||
| ADL With Assistance | 2,110 (71.1%) | 1,340 (70.8%) | 520 (71.3%) | 250 (72.5%) | ||
| Female Gender | 1,125 (36%) | 758 (37.9%) | 9 (1.2%) | 358 (96%) | < 0.001 | |
| Lung Allocation Score | 46.9 (40.2, 62.2) | 46.3 (39.8, 60.9) | 49.2 (41, 67.5) | 46.3 (39.3, 59.1) | 0.001 | |
| Operative Characteristics | ||||||
| Life Support at Transplantation |
335 (10.8%) | 206 (10.4%) | 89 (11.9%) | 40 (10.8%) | 0.541 | |
| Medical Condition Before Transplant |
0.237 | |||||
| Not Hospitalized | 2,534 (81.8%) | 1,632 (82.5%) | 598 (79.9%) | 304 (82.2%) | ||
| Hospitalized Not in ICU | 241 (7.8%) | 144 (7.3%) | 62 (8.3%) | 35 (9.5%) | ||
| In ICU | 321 (10.4%) | 202 (10.2%) | 88 (11.8%) | 31 (8.4%) | ||
| Waitlist Days | 68 (19, 216) | 71 (19, 222) | 56 (15, 170) | 97 (26, 310) | < 0.001 | |
| Center Volume | 525 (312, 1,042) | 525 (312, 1,042) | 570 (334, 1,042) | 511 (275, 695) | 0.001 | |
All statistics displayed as median (interquartile range) or number (percent).
ADL, Activities of Daily Living; BMI, Body Mass Index; GFR, Glomerular Filtration Rate; ICU, Intensive Care Unit, IQR, Interquartile Range.
Center Volume defined as number of transplants during entire study period
Following stratification into the five aforementioned groups, initial unadjusted survival analysis demonstrated that for both SOLT and BOLT patients there was a significant difference amongst the groups with regards to long-term survival. For SOLT patients, a pTLCr of <0.8 or 1.1–1.2 portended the worst survival, while for patients undergoing a BOLT, a pTLCr of 1.1–1.2 was associated with the worst survival (Figure 2). Following adjustment with a Cox proportional hazards model, the same pTLCr for both SOLT and BOLT was found to be independently associated with an increased risk of long-term mortality (Table 3). The following covariates were examined in the Cox model: Donor age, BMI, history of cigarette use, and history of diabetes. Recipient factors examined included: gender (and gender mismatch), age, BMI, GFR, history of diabetes, functional status, medical condition, waitlist days, transplant era, center transplant volume, and CMV status.
Figure 2.
Kaplan-Meier Analysis of Survival Stratified by pTLCr A) SOLT B) BOLT
Table 3.
A) Hazard ratio of effect of pTLCr on long term mortality in SOLT patients B) Hazard ratio of effect of pTLCr on long term mortality in BOLT patients
| A) | ||||
|---|---|---|---|---|
| Variable | Hazard Ratio |
Lower 95% CI |
Upper 95% CI |
P-Value |
| pTLC Ratio (Reference pTLC Ratio 0.9–1.1) | ||||
| pTLC Ratio <0.8 | 1.711 | 1.146 | 2.557 | 0.01 |
| pTLC Ratio 0.8–0.9 | 1.204 | 0.832 | 1.744 | 0.33 |
| pTLC Ratio 1.1–1.2 | 1.398 | 1.019 | 1.917 | 0.04 |
| pTLC Ratio > 1.2 | 1.33 | 0.891 | 1.983 | 0.16 |
| Donor BMI | 0.986 | 0.975 | 0.997 | 0.009 |
| Donor history of significant cigarette use | 1.111 | 0.982 | 1.257 | 0.095 |
| Donor history of Diabetes | 1.166 | 0.912 | 1.493 | 0.221 |
| Gender Mismatch | 0.776 | 0.542 | 1.11 | 0.165 |
| Recipient Age (per 10 years) | 1.202 | 1.118 | 1.292 | <0.001 |
| Recipient BMI | 1.01 | 0.997 | 1.022 | 0.128 |
| Recipient GFR at Transplant | 0.997 | 0.995 | 1 | 0.061 |
| Recipient history of Diabetes | 0.989 | 0.852 | 1.148 | 0.883 |
| Recipient functional status: ADL With Assistance | 1.36 | 1.219 | 1.517 | <0.001 |
| Recipient Gender, female | 0.897 | 0.784 | 1.027 | 0.116 |
| Recipient Medical Condition: Hospitalized, Not in ICU | 1.325 | 1.002 | 1.752 | 0.048 |
| Recipient Medical Condition, In ICU | 2.517 | 1.853 | 3.419 | <0.001 |
| Recipient on Life Support at time of transplant, yes | 1.459 | 1.082 | 1.968 | 0.013 |
| Waitlist Years | 0.95 | 0.889 | 1.015 | 0.129 |
| Transplant Era, 1987–1995 | 1.104 | 0.837 | 1.457 | 0.484 |
| Transplant Era, 1996–2004 | 1.238 | 1.087 | 1.41 | 0.001 |
| Transplant Center Volume (per 100 transplants) | 0.985 | 0.971 | 0.999 | 0.047 |
| CMV Mismatch, Yes | 1.035 | 0.922 | 1.163 | 0.56 |
| B) | ||||
|---|---|---|---|---|
| Variable | Hazard Ratio | Lower 95% CI | Upper 95% CI | P-Value |
| pTLC Ratio (Reference pTLC Ratio 0.9–1.1) | ||||
| pTLC Ratio <0.8 | 1.199 | 0.737 | 1.952 | 0.47 |
| pTLC Ratio 0.8–0.9 | 0.919 | 0.579 | 1.459 | 0.72 |
| pTLC Ratio 1.1–1.2 | 1.717 | 1.112 | 2.651 | 0.02 |
| pTLC Ratio > 1.2 | 0.916 | 0.553 | 1.516 | 0.73 |
| Donor Age (per 10 years) | 1.032 | 0.983 | 1.083 | 0.208 |
| Donor BMI | 0.996 | 0.983 | 1.01 | 0.605 |
| Donor history of significant cigarette use | 1.248 | 1.05 | 1.482 | 0.012 |
| Donor history of diabetes | 1.154 | 0.871 | 1.53 | 0.318 |
| Gender Mismatch | 0.961 | 0.617 | 1.497 | 0.86 |
| Recipient Age (per 10 years) | 1.121 | 1.043 | 1.205 | 0.002 |
| Recipient BMI | 0.992 | 0.977 | 1.008 | 0.337 |
| Recipient GFR at Transplant | 0.996 | 0.993 | 0.999 | 0.007 |
| Recipient history of Diabetes | 1.007 | 0.835 | 1.213 | 0.944 |
| Recipient functional status: ADL with assistance | 1.114 | 0.96 | 1.292 | 0.155 |
| Recipient Gender, female | 0.999 | 0.845 | 1.181 | 0.993 |
| Recipient Medical Condition: Hospitalized, Not in ICU | 0.991 | 0.75 | 1.308 | 0.949 |
| Recipient Medical Condition, In ICU | 1.774 | 1.355 | 2.323 | <0.001 |
| Recipient on Life Support at time of transplant, yes | 1.083 | 0.832 | 1.411 | 0.552 |
| Waitlist Years | 0.978 | 0.914 | 1.047 | 0.523 |
| Transplant Era, 1987–1995 | 2.02 | 1.227 | 3.324 | 0.006 |
| Transplant Era, 1996–2004 | 1.198 | 0.998 | 1.437 | 0.052 |
| Transplant Center Volume (per 100 transplants) | 0.968 | 0.951 | 0.985 | <0.001 |
| CMV Mismatch, Yes | 1.06 | 0.925 | 1.216 | 0.401 |
Restricted cubic splines were then developed to ascertain the role of pTLCr in relative mortality hazard in both SOLT and BOLT (mortality relative to a pTLC of 1.0). For patients undergoing SOLT, decreasing pTLCr below 0.9 increased relative mortality. For patients undergoing BOLT there was no particular degree of over- or under-sizing in which a change in pTLCr was statistically associated with an increase or decrease in mortality (Figure 3).
Figure 3.
Restricted cubic spline of relative mortality hazard by pTLCr in A) SOLT B) BOLT
Discussion
The present study demonstrates that in patients with end-stage lung disease due to RLD requiring lung transplantation, the relationship between donor and recipient pTLC has a significant impact on survival, particularly in patients undergoing SOLT. While the effect on survival is relatively pronounced in patients undergoing SOLT, patients undergoing BOLT do not appear to be as affected by the relationship between donor and recipient lung size. Additionally, in recent years, it appears that transplant teams have been utilizing lungs that are more similarly sized between donor and recipient or undersizing the lungs, while shying away from oversizing transplants.
Given the limited number of available organs for lung transplantation, the importance of identifying the best possible match for a patient is essential. As we have described the size matching of donor lungs to recipients has been discussed in the literature as a potential variable affecting outcomes following transplantation. However, a recent evidence based review from Barnard and colleagues demonstrated that the evidence for size matching is somewhat lacking and guidelines are not well defined.9 Most of the original studies examining this topic were very small case series with less than 100 patients and although these studies provided recommendations on sizing, the number of patients examined limited them.10–14 Mason et al. published institutional data from over 400 patients and found no correlation between lung size matching and patient outcomes.4
Perhaps the largest data for this topic comes from Eberlein and colleagues and they have described their experience with heterogenous patient populations including all lung transplant patients in both institutional data sets and UNOS1–3, 7, 8, as well as in patients with pulmonary hypertension.6 With institutional data they demonstrated that patients undergoing BOLT who had undersized lungs were more likely to have post-operative airway complications and primary graft dysfunction as well as higher hospital costs.1 In a separate study they also demonstrated that oversizing of BOLT with regards to pTLCr was associated with improved survival, increased expiratory airflow capacity as well as a decreased incidence of bronchiolitis obliterans syndrome (BOS).2, 8 Using a similar analysis in the UNOS dataset they confirmed the conclusions from their institutional data and found that oversizing of the pulmonary allograft in BOLT was independently associated with improved survival, but there was no independent association seen in SOLT.3 Similar benefits were found with oversizing in patients with pulmonary hypertension as well.6
While these data are important for current guidelines, there are a few notable limitations to these studies. Perhaps most importantly these studies all examined a heterogeneous patient population, including multiple indications for transplantation such as obstructive lung disease, RLD, cystic fibrosis and others. As noted earlier, the pulmonary mechanics and effects on chest physiology can be dramatically different in a patient with RLD as compared to someone with obstructive lung disease. Because of the heterogeneity of the patient population, the current guidelines for sizing of donor organs are relatively vague with no clear consensus in the most recent recommendations.9 Another limitation of these studies centers on the definitions of over and undersizing. In the institutional studies previously noted, oversized was defined as a pTLCr of > 1 and undersizing was defined as a pTLCr <1. This strict criteria forces one to almost identically match organs for appropriate sizing. It should be noted, however that this manuscript utilizing UNOS data divided patients based on quartiles of pTLCr.3
Our study, as previously emphasized, attempted to capture only the unique population of patients with RLD. Due to the unique pathophysiology of the disease - particularly on chest wall mechanics - it is important that these patients be examined separately from all patients undergoing lung transplantation. The results demonstrate that in patients who undergo SOLT for RLD with considerable undersizing (pTLCr < 0.8) there was a significantly increased risk of mortality as well as incidence of airway dehiscence. Moreover, increasing pTLCr between 0.8 and 0.9 had a significant impact on improving of survival for patients undergoing SOLT. These results are important as the majority of prior research has focused on BOLT. However, our study demonstrates that patients with RLD are slightly more likely to receive a SOLT as compared to a BOLT. As patients receiving a SOLT are subsequently primarily reliant on only one allograft, undersizing the donor lung appears to have a detrimental effect as the transplanted lung may be insufficient for the recipient. Additionally, airway complications in undersized grafts may be the result of the technical difficulty of the bronchial anastomosis and/or increased risk for donor airway devascularization. Furthermore, these data suggest that transplant teams should strive to obtain a pTLCr as close to or greater than 0.9 as possible with small increases in pTLCr above 0.8 potentially have a significant impact on survival.
Having a pTLCr between 1.1 and 1.2 was also independently associated with increased mortality for BOLT; however, this association may have partly been a result of the relatively small number of patients with a pTLCr that fell in this range. However, in a restricted patient, an over-sized allograft placed in to an already restrictive chest cavity can potentially lead to diminished respiratory mechanics or even pulmonary tamponade. It is important to note however, there was no obvious change in pTLCr for patients undergoing BOLT which appeared to have a significant impact on survival. This result is disparate from those derived from Eberlein and colleagues and this may be a result of the heterogeneity of their study population. Notably, survival for patients undergoing BOLT were higher than for those with SOLT, and further comparisons about the benefit of BOLT vs. SOLT with regards to size mismatch may be warranted. Based on our results it appears that for patients undergoing BOLT the fact the recipient receives two good lungs may be enough to provide appropriate survival within current norms of size matching. These data may further encourage the use of deceased donor lobar transplantation if necessary for appropriate sizing in the setting of a BOLT, but should lend some caution to this approach in patients undergoing SOLT. It should also be noted that pneumoreduction techniques are not captured in UNOS, so it is unclear if the over-sized allografts underwent downsizing prior to implantation.
Given the aforementioned results, it ultimately raises the question of how these data should impact current listing and organ allocation practices. The recent guidelines for sizing state that transplant teams should consider lower and upper limits for pTLCr for patients being listed for transplant.9 It does not appear that these size considerations are as important for patients undergoing BOLT for RLD, especially in regards to under-sizing of the allografts. However, for patients undergoing SOLT for RLD it may be of benefit to establish what transplant teams feel is an appropriate organ based on size at the time of listing. There are also data suggesting that patients with RLD undergoing SOLT have somewhat disparate outcomes based on the laterality of the transplant. Furthermore, given our results it may be reasonable to consider organ size as part of the organ allocation process for SOLT in RLD to help improve survival. If sizing were to be considered as part of the listing and/or organ allocation process, it would then be prudent to reach a consensus of how to estimate or measure the total lung capacity. One alternative to estimation of pTLC is actual TLC, which is derived from plethysmography. However in RLD previous data has demonstrated that actual TLC is not well correlated with donor pTLC and may underestimate the lung size that is acceptable for a recipient.15 As such pTLC appears to be the optimal means to estimate lung capacity, particularly as it is easy to use and does not require significant effort, which is important for donor lung capacity calculation. However, as technology continues to evolve, particularly with regards to imaging, other modalities may provide more accurate measurements. For example, CT scanning should be able to provide a volumetric approach to approximating chest cavity size and may provide for better size matching.
Limitations
It should be noted that our study has some notable limitations. Given the retrospective nature of the study and that the data was derived from a large national database, there is potential for bias associated with large retrospective cohort studies. Furthermore, the premise of calculating lung capacity is based on an estimate utilizing height alone. As this may not be the most accurate measurement, the data may be biased, however, as mentioned before pTLC appears to be more accurate for lung capacity measurements as compared to actual TLC in restrictive patients.15 Furthermore, given our use of the UNOS dataset we were unable to quantify the effect of size mismatch on various important outcomes including postoperative ventilator requirements, incidence of primary graft dysfunction, and incidence of BOS. Finally, our dataset includes patients from 1987 to 2011 and thus, there are many patients transplanted in the non-lung allocation score era and this may limit the applicability of the results to current transplant practices.
Conclusions
Given the limitation of available organs for lung transplantation strategies to ensure appropriate allocation so as to maximize post-transplant survival are essential for the most efficient use of these organs. Size matching of the organs appears to be important for patients undergoing SOLT, but does not appear to have a significant effect on patients undergoing BOLT. Based on these data transplant teams should aim to avoid undersizing potential organs for SOLT and continue to seek the best available organ regardless of size for patients listed for BOLT.
Acknowledgments
This study was supported with department funding. MSM is supported by the American Society for Transplantation. MGH and BCG are supported by the NIH Cardiothoracic Trials Network. This work was supported in part by Health Resources and Services Administration contract 234-2005-370011C. The content is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
Abbreviations
- BOLT
Bilateral Orthotopic Lung Transplant
- BOS
Bronchiolitis Obliterans Syndrome
- LAS
Lung Allocation Score
- LOS
Length of Stay
- pTLC
Predicted Total Lung Capacity
- pTLCr
Predicted Total Lung Capacity Ratio
- RLD
Restrictive Lung Disease
- SOLT
Single Orthotopic Lung Transplant
- UNOS
United Network for Organ Sharing
Footnotes
Disclosure
The authors have no relevant disclosures
References
- 1.Eberlein M, Arnaoutakis GJ, Yarmus L, Feller-Kopman D, Dezube R, Chahla MF, et al. The effect of lung size mismatch on complications and resource utilization after bilateral lung transplantation. J Heart Lung Transplant. 2012;31(5):492–500. doi: 10.1016/j.healun.2011.12.009. [DOI] [PubMed] [Google Scholar]
- 2.Eberlein M, Permutt S, Chahla MF, Bolukbas S, Nathan SD, Shlobin OA, et al. Lung size mismatch in bilateral lung transplantation is associated with allograft function and bronchiolitis obliterans syndrome. Chest. 2012;141(2):451–460. doi: 10.1378/chest.11-0767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Eberlein M, Reed RM, Bolukbas S, Parekh KR, Arnaoutakis GJ, Orens JB, et al. Lung size mismatch and survival after single and bilateral lung transplantation. Ann Thorac Surg. 2013;96(2):457–463. doi: 10.1016/j.athoracsur.2013.04.064. [DOI] [PubMed] [Google Scholar]
- 4.Mason DP, Batizy LH, Wu J, Nowicki ER, Murthy SC, McNeill AM, et al. Matching donor to recipient in lung transplantation: How much does size matter? J Thorac Cardiovasc Surg. 2009;137(5):1234–1240. e1. doi: 10.1016/j.jtcvs.2008.10.024. [DOI] [PubMed] [Google Scholar]
- 5.Stocks J, Quanjer PH. Reference values for residual volume, functional residual capacity and total lung capacity. ATS Workshop on Lung Volume Measurements. Official Statement of The European Respiratory Society. Eur Respir J. 1995;8(3):492–506. doi: 10.1183/09031936.95.08030492. [DOI] [PubMed] [Google Scholar]
- 6.Eberlein M, Diehl E, Bolukbas S, Merlo CA, Reed RM. An oversized allograft is associated with improved survival after lung transplantation for idiopathic pulmonary arterial hypertension. J Heart Lung Transplant. 2013;32(12):1172–1178. doi: 10.1016/j.healun.2013.06.011. [DOI] [PubMed] [Google Scholar]
- 7.Eberlein M, Permutt S, Brown RH, Brooker A, Chahla MF, Bolukbas S, et al. Supranormal expiratory airflow after bilateral lung transplantation is associated with improved survival. Am J Respir Crit Care Med. 2011;183(1):79–87. doi: 10.1164/rccm.201004-0593OC. [DOI] [PubMed] [Google Scholar]
- 8.Eberlein M, Reed RM, Permutt S, Chahla MF, Bolukbas S, Nathan SD, et al. Parameters of donor-recipient size mismatch and survival after bilateral lung transplantation. J Heart Lung Transplant. 2012;31(11):1207–1213. e7. doi: 10.1016/j.healun.2011.07.015. [DOI] [PubMed] [Google Scholar]
- 9.Barnard JB, Davies O, Curry P, Catarino P, Dunning J, Jenkins D, et al. Size matching in lung transplantation: an evidence-based review. J Heart Lung Transplant. 2013;32(9):849–860. doi: 10.1016/j.healun.2013.07.002. [DOI] [PubMed] [Google Scholar]
- 10.Tamm M, Higenbottam TW, Dennis CM, Sharples LD, Wallwork J. Donor and recipient predicted lung volume and lung size after heart-lung transplantation. Am J Respir Crit Care Med. 1994;150(2):403–407. doi: 10.1164/ajrccm.150.2.8049822. [DOI] [PubMed] [Google Scholar]
- 11.Ouwens JP, van der Mark TW, van der Bij W, Geertsma A, de Boer WJ, Koeter GH. Size matching in lung transplantation using predicted total lung capacity. Eur Respir J. 2002;20(6):1419–1422. doi: 10.1183/09031936.02.00294402. [DOI] [PubMed] [Google Scholar]
- 12.Chacon RA, Corris PA, Dark JH, Gibson GJ. Respiratory mechanics after heart-lung and bilateral lung transplantation. Thorax. 1997;52(8):718–722. doi: 10.1136/thx.52.8.718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Miyoshi S, Demertzis S, Eckstein F, Hohlfeld J, Schaefers HJ. Chest size matching in single and double lung transplantation. Jpn J Thorac Cardiovasc Surg. 1999;47(4):163–170. doi: 10.1007/BF03217963. [DOI] [PubMed] [Google Scholar]
- 14.Massard G, Badier M, Guillot C, Reynaud M, Thomas P, Giudicelli R, et al. Lung size matching for double lung transplantation based on the submammary thoracic perimeter. Accuracy and functional results. The Joint Marseille-Montreal Lung Transplant Program. J Thorac Cardiovasc Surg. 1993;105(1):9–14. [PubMed] [Google Scholar]
- 15.Loizzi D, Aigner C, Jaksch P, Scheed A, Mora B, Sollitto F, et al. A scale for decision making between whole lung transplantation or lobar transplantation. Eur J Cardiothorac Surg. 2010;37(5):1122–1125. doi: 10.1016/j.ejcts.2009.11.032. [DOI] [PubMed] [Google Scholar]