Abstract
Background.
Accumulated knowledge on the outcomes related to size mismatch in lung transplantation derives from predicted total lung capacity equations rather than individualized measurements of donors and recipients. The increasing availability of computed tomography (CT) makes it possible to measure the lung volumes of donors and recipients before transplantation. We hypothesize that CT-derived lung volumes predict a need for surgical graft reduction and primary graft dysfunction.
Methods.
Donors from the local organ procurement organization and recipients from our hospital from 2012 to 2018 were included if their CT exams were available. The CT lung volumes and plethysmography total lung capacity were measured and compared with predicted total lung capacity using Bland Altman methods. We used logistic regression to predict the need for surgical graft reduction and ordinal logistic regression to stratify the risk for primary graft dysfunction.
Results.
A total of 315 transplant candidates with 575 CT scans and 379 donors with 379 CT scans were included. The CT lung volumes closely approximated plethysmography lung volumes and differed from the predicted total lung capacity in transplant candidates. In donors, CT lung volumes systematically underestimated predicted total lung capacity. Ninety-four donors and recipients were matched and transplanted locally. Larger donor and smaller recipient lung volumes estimated by CT predicted a need for surgical graft reduction and were associated with higher primary graft dysfunction grade.
Conclusion.
The CT lung volumes predicted the need for surgical graft reduction and primary graft dysfunction grade. Adding CT-derived lung volumes to the donor-recipient matching process may improve recipients’ outcomes.
Donors and recipients in lung transplantation are matched based on blood type and organ size. Most knowledge regarding size matching has been generated using donor-to-recipient predicted total lung capacity (TLC) ratios. Generally, a larger predicted TLC ratio is associated with improved outcomes [1−5], and ratios from 0.8 to 1.2 are recommended. Although predicted TLC ratios are practical and easily calculated from patient characteristics, they also have important limitations. They do not avoid size mismatch, as evidenced by 12.5% of recipients requiring surgical graft reduction [5,6]. This occurrence of size mismatch may be partly explained by the wide CIs of predicted TLC equations [7]. Additionally, predicted TLC ratios are calculated using the age, sex, and height of donors and recipients; the observed relationships to outcomes may be related to any of these variables, obscuring the true influence of lung volumes [8]. A possible way to overcome these limitations is to measure lung volumes in donors and recipients.
Pulmonary function tests are performed while evaluating transplant candidates; however, they are not routinely obtained in mechanically ventilated donors. An alternative way to measure lung volumes, which is considered equivalent in clinical practice [9] and clinical research, is computed tomography (CT) volumetry [10,11]. The CT scans of donors are increasingly available in the United States [12], and CT volumetry has already been adopted successfully in living donor lobar transplantation [13−16]. Evaluating CT volumetry in orthotopic lung transplants could help advance size matching toward real personalization. This study aims to understand the relationship between predicted TLC and measured CT lung volumes in donors and recipients. We hypothesize that CT lung volumes predict surgical graft reduction and primary graft dysfunction.
MATERIALS AND METHODS
This retrospective cohort study was conducted at Washington University School of Medicine−Barnes Jewish Hospital and Mid America Transplant in St. Louis, Missouri, following the revised Declaration of Helsinki and the International Society for Heart and Lung Transplantation ethical statements. The Washington University School of Medicine Institutional Review Board reviewed and approved the protocol waiving the need for informed consent (IRB No. 202012069).
All local donors and adult candidates with available CT scans from January 2012 to December 2018 were eligible for inclusion (Fig 1). Demographic data, medical history, plethysmography lung volumes, surgical variables, and outcomes are prospectively collected at our program and were available for the study in both the recipients evaluated at the Washington University Lung Transplant Clinic and donors from Mid-America Transplant, organ procurement organization. Predicted TLC with its corresponding 95% CI was calculated for donors and recipients based on age, sex, and height using the Global Lung Function Initiative online calculator [17]. Primary graft dysfunction (PGD) was evaluated and scored according to the 2016 Consensus Group Statement of the International Society for Heart and Lung Transplantation [18]. The highest score between 24 and 72 hours was selected for each patient.
Fig 1.
Population flow chart, the red box on the left includes candidates (A), the green box on the right includes donors (B), and the bubble over the red and green boxes includes the matched cohort.
The outcomes of interest were the need for surgical graft reduction and PGD grade. Analyses were performed using STATA SE version 15.1 (StataCorp, LLC, College Station, Tex, United States).
Computed Tomography
Transplant candidate CT scans were obtained following standard clinical procedures after asking patients to inhale and perform a breath hold. Donor images were obtained while supported by mechanical ventilation using a tidal volume of 8 mL/kg and a positive end-expiratory pressure (PEEP) of 8 to 15 cmH2O, according to the donor’s needs. The scans were not synchronized with the respiratory cycle.
The CT scan lung volumes were measured using the three-dimensional (3D) Slicer extension Chest Imaging Platform [19,20]. The CT lung volumes reflect lung volume to the hilum, excluding the trachea and the mediastinal portion of the main left and right bronchus.
Statistical Analysis
The continuous variables are presented as means with SDs or medians and IQR and compared using t test, analysis of variance, or Kruskal-Wallis test as appropriate. The categorical variables are presented as percentages and compared using Pearson’s χ2 or Fisher’s exact test as appropriate.
We first performed an agreement analysis among the candidate CT lung volumes and the predicted and plethysmography TLC [21] in the entire cohort, as well as for subgroups according to the presence of hyperinflation, restriction, or normal plethysmography TLC (above, below, or within the 95% CI of their predicted TLC) [22]. Second, we explored if age, sex, height, body mass index (BMI), and the diagnostic group were predictors of recipient CT lung volume using linear regression. The same 2-step analysis was repeated in donors. Additionally, for recipients, we implemented the analysis of variance to evaluate the influence of time elapsed from CT scan acquisition to transplantation by partitioning the CTs into 4 bins.
Next, we used logistic regression to assess the association between predicted TLC, plethysmography TLC, CT lung volumes, and the need for surgical graft reduction. Under the assumption that lung volumes are the only variable of importance, these models are unadjusted. Model performance was evaluated using receiver operator characteristics analysis.
Finally, the association between predicted TLC, CT lung volumes, and PGD grade was assessed using ordinal logistic regression. The Brant-Wald test was used to evaluate the proportional odds assumption. The sensitivity analysis included 3 additional models to examine the effect of BMI, donor cause of death, and recipient diagnosis when added to the CT lung volume model. Because age, sex, and height are the basis of the predicted TLC equation and represent CT lung volumes, they were not considered for inclusion in the logistic models.
All analyses avoid assumptions related to the use of ratios in regression models [8] by including independent terms for donor and recipient lung volumes and their ratio as an interaction term in the sensitivity analysis.
RESULTS
Candidate Lung Volumes and Agreement Analysis
A total of 315 candidates with 575 chest CTs were evaluated. The 3D slicer software failed to measure lung volumes in 12 CTs (2%) from 7 (2%) candidates. The remaining 308 candidates and 563 chest CTs were included in the analysis (Fig 1). Overall, CT lung volumes were smaller than predicted TLC by 1.52 ± 2.09 L (Table 1, Fig 2). There was minimal within-subject variance (0.24 ± 0.03), indicating good reproducibility when measuring CT scans from the same individual. The inter-subject variance was high (4.13 ± 0.42), leading to wide 95% limits of agreement (LoA) when comparing CT lung volumes and predicted TLC (Table 1). Agreement improved between CT lung volumes and plethysmography TLC. The CT volume was smaller than plethysmography TLC by 0.47 ± 0.81 L (Table 1, Fig 2). Subgroup analyses revealed that in the normal TLC group, CT volumetry overestimated plethysmography TLC by 0.40 L (95% LoA, −1.60 to 0.76). In the hyperinflated and restricted subgroups, CT volumetry underestimated plethysmography TLC by 0.88 L (95% LoA, −2.38 to 0.62) and 0.38 L (95% LoA, −2.07 to 1.31), suggesting that recipient disease is an important variable affecting lung volume measurement.
Table 1.
Agreement Analysis
| Parameter | CT Lung Volume vs Predicted TLC, Candidates | CT Lung Volumes vs TLC From Pulmonary Function Tests, Candidates | CT Lung Volume vs Predicted TLC, Donors |
|---|---|---|---|
| Bias (SE) | −1.52 (±0.14) | −0.47 (±0.07) | −3.35 (±0.12) |
| SD of the differences (95% CI) | 2.09 (±0.1) | 0.81 (±0.03) | 1.26 |
| Lower LoA (95% CI) | −5.6 (−6.13, −5.16) | −2.06 (−2.22, −1.93) | −5.82 (−6.28, −5.45) |
| Upper LoA (95% CI) | 2.58 (2.13, 3.09) | 1.12 (0.99, 1.28) | −0.88 (−1.25, −0.42) |
| Within-subject variance (SE) | 0.24 (±0.03) | 0.31 (±0.04) | - |
| Between subject variance (SE) | 4.13 (±0.42) | 0.35 (±0.06) | - |
Candidate CT lung volumes are compared to predicted TLC and TLC from pulmonary function tests. Donor CT lung volumes are compared to predicted TLC. Results are presented in liters.
CT, computed tomography; LoA, limit of agreement; TLC, total lung capacity.
Fig 2.
Scatter plot comparing computed tomography and plethysmography lung volumes to predicted total lung capacity according to the global lung function initiative equations in candidates for lung transplantation. The unit of measurement is liters. CT, computed tomography; TLC, total lung capacity.
Transplant candidates had 1 to 4 (median = 1, IQR 1–2) CT scans before transplantation. We evaluated the influence of time elapsed between the date of chest CT image acquisition and transplantation by partitioning CT images into 4 bins. Each bin included the first, second, third, or fourth chest CT, with the first bin being the closest and the fourth farthest from transplantation. The average time elapsed from the CT scan to transplantation was 122, 255, 492, and 664 days across bins 1 to 4. Overall, there was a decline in CT lung volumes from the fourth to first bin from 7.3 ± 1.7 L to 4.5 ± 2.2 L (P = .007). Transplant candidates from the restricted lung disease group were responsible for the overall difference, with stable lung volumes in the hyperinflated and normal lung categories. Finally, plethysmography volume category (hyperinflated β = 1.4 ± 0.4; P < .001; restricted β = −2.9 ± 0.3; P < .001) and height (β = .08 ± 0.01; P < .001; R2= .78) significantly influenced CT lung volume in the multivariable linear regression model. The BMI was also a predictor when substituted for height in the model.
Donor Lung Volume and Agreement Analysis
We repeated the same analysis steps for 379 donors with 379 available CT scans; the 3D Slicer analysis tool failed to measure lung volumes in 22 scans (5.8%) (Fig 1). The CT lung volumes were smaller than the predicted TLC by 3.35 ± 1.26 L (Table 1, Fig 3). A linear regression model including age in years (β = .01, 95% CI 0.001–0.017; P = .044), sex (β = .35, 95% CI= −0.01 to 0.72; P = .055), donor height in centimeters (β = .048, 95% CI 0.026–0.07), and self-reported White race (β = .34, 95% CI 0.03–0.64) predicted donor CT lung volumes. Patients’ BMI was also a predictor when substituted for height.
Fig 3.
Scatter plot comparing computed tomography-derived lung volumes in donors to predicted total lung capacity according to the global lung function initiative equations. The unit of measurement is liters. CT, computed tomography; LLN, lower limit of normal; pTLC, predicted total lung capacity; TLC, total lung capacity; ULN, upper limit of normal.
Outcomes of Matched Donors and Recipients
Of the available donors and recipients, 94 pairs were matched and transplanted at Washington University School of Medicine−Barnes Jewish Hospital (Fig 1). Donors were younger than recipients (37 [IQR: 26–52] vs 59.4 [IQR: 49–66]) and more likely to be of a racial and/or ethnic minority (26 [28%] vs 18 [18%]). Male sex predominated in recipients 61 (65%) and donors 54 (57%). Plethysmography TLC classified recipients into hyperinflated (16 patients [17%]), restricted (57 patients [60.6%]), and normal (21 patients [22.3%]).
The average elapsed time for the most recent CT scan before transplantation was 0.3 ± 6 days for donors and 206 ± 191 days for recipients. The CT scans measured lung volumes ± at 2.8 ± 0.99 L in donors and 4.3 ± 2.2 L in recipients leading to a mean donor-recipient CT lung volume difference of −1.7 ± 2.13 L and a ratio of 0.78 ± 0.5. In contrast, the mean donor-recipient predicted TLC difference and ratio were 0.002 ± 0.9 L and 1.02 ± 0.16.
Graft Reduction Surgery
Ten recipients (10.6%) required surgical graft reduction (Table 2). The smallest donor-to-recipient CT ratio requiring graft reduction was 0.6, with a median and IQR of 1.2 (0.74–1.54). Comparing patients who did and did not require surgical graft reduction, the CT lung volumes were significantly smaller in recipients requiring surgical graft reduction (4.7 ± 2.2 vs 3.0 ± 1.5; P = .027). Recipients requiring surgical resizing were characterized by significantly higher donor-to-recipient CT lung volume ratio (1.3 ± 0.8 vs 0.7 ± 0.4; P < .001) and also predicted TLC ratio (1.1 ± 0.1 vs 1 ± 0.2; P = .03). Four patients (40%) who needed surgical graft reduction had donor CT lung volumes smaller than the recipients, with lobar atelectasis in 2 of the donors. Larger (odds ratio [OR] = 2.34 ± 0.9; P = .026) and smaller recipient CT lung volumes (OR = 0.55 ± 0.13; P = .013) were associated with an increased risk of surgical graft reduction. In our second model, only the recipient’s predicted TLC was significantly associated with the need for surgical graft reduction, whereas the donor’s predicted TLC was not (OR = 0.36 ± 0.2; P = .03; OR = 2.0 ± 0.98; P = .155).
Table 2.
Baseline Donor, Recipient, and Surgical Variable According to Need for Surgical Size Reduction
| Variables | Surgical Size Reduction = No | Surgical Size Reduction = Yes | P Value | |
|---|---|---|---|---|
| N | 84 | 10 | ||
| Recipient Age, mean (SD) | 55.1 (14.7) | 55.9(12.7) | .87 | |
| Sex | Female | 33 (39%) | 7 (70%) | .063 |
| Recipient height (cm), mean (SD) | 173 (10) | 157 (1) | .15 | |
| Body mass index (kg/m2), mean (SD) | 24 (4.5) | 26.6 (4.6) | .09 | |
| Diagnostic group* | Normal lung volume | 21 (25%) | 0 (0%) | .11 |
| Hyperinflation | 15 (18%) | 1 (10%) | ||
| Restriction | 48 (57%) | 9 (90%) | ||
| Donor Age, mean (SD) | 38.2 (15.2) | 39.8 (11.2) | .74 | |
| Sex | Female | 37 (44%) | 6 (60%) | .34 |
| Height cm, mean (SD) | 171.1 (8.45181) | 171.2 (8.8) | .96 | |
| Body mass index (kg/m2), mean (SD) | 26.8 (6.4) | 26.2 (2.7) | .773 | |
| Smoking history | 34 (56.7%) | 6 (85.7%) | .85 | |
| Ischemic time (min), mean (SD) | 221 (46) | 255 (64) | .24 | |
| MV >48 h | 23 (28%) | 4 (40%) | ||
| Unexpected return to OR | 12 (15%) | 2 (20%) | ||
| Max PGD 24–72 | 0 | 14 (17%) | 0 (0%) | .52 |
| 1 | 47 (56%) | 7 (70%) | ||
| 2 | 11 (13%) | 1 (10%) | ||
| 3 | 12 (14%) | 2 (20%) | ||
| Recipient lung volume, mean (SD) | 4.70 (2.21) | 3.04 (1.5) | .027 | |
| Donor lung volume, mean (SD) | 2.71 (0.96) | 3.19 (1.15) | .14 | |
| CT lung volume ratio, mean (SD) | 0.72 (0.42) | 1.32 (0.81) | < .001 | |
| Median pTLC donor, mean (SD) | 6.24 (1.11) | 5.54 (1.13) | .063 | |
| Median pTLC recipient, mean (SD) | 6.15 (0.99) | 6.15 (1.14) | .98 | |
| Predicted TLC ratio, man (SD) | 1.01 (0.16) | 1.12 (0.13) | .03 |
MV, mechanical ventilation; OR, operating room; PGD, primary graft dysfunction; TLC, total lung capacity.
Diagnostic group is based on the presence of obstructive, restrictive, or normal physiology on preoperative pulmonary function tests.
Primary Graft Dysfunction
The mean recipient CT lung volume progressively decreased with increasing PGD grade. Ischemic time was also longer with higher PGD grades (Table 3). One ordinal logistic regression model was fit to evaluate the association between PGD grade to donor and recipient CT lung volume and another for predicted TLC. Larger donor and smaller recipient CT lung volumes predicted increasing PGD grade, whereas donor and recipient predicted TLC did not (Table 4). Three additional models accounting for donor mechanism of death, recipient diagnosis, and donor and recipient BMI did not substantially alter the results. Ischemic time and CT lung volumes were colinear (variable inflation factor = 11.5) and thus were not included in the same model. For a median lung volume donor, the probability of PGD 0 increased with higher recipient CT lung volumes, and the probability of PGD 2 and 3 increased with decreasing recipient CT lung volumes (Fig 4). Donor-to-recipient CT lung volume ratios of 0.66, 0.50, 0.30, and 0.25 carried a risk of 30%, 25%, 20%, and 15% for PGD grades 2 or 3 (Fig 5). The minimal ratio associated with surgical graft reduction of 0.6 was associated with a PGD risk <30%.
Table 3.
Baseline Characteristics According to Highest PGD Grade 0 to 3 From 24 to 72 Hours
| Variables | PGD = 0 | PGD = 1 | PGD = 2 | PGD = 3 | P Value | |
|---|---|---|---|---|---|---|
| N | 14 | 54 | 12 | 14 | ||
| Recipient Age, mean (SD) | 57.0 (13.4) | 54.5(14.8) | 52.7 (17.4) | 58.3 (12) | .72 | |
| Sex | Female | 5 (36%) | 25 (46%) | 7 (58%) | 3 (21%) | .23 |
| Height (cm), mean (SD) | 170.4 (11) | 168.8 (9.5) | 165.8 (8.4) | 174.5 (9.2) | .029 | |
| Body mass index (kg/m2), mean (SD) | 24.6 (4.1) | 23.5 (4.8) | 24.6 (4.6) | 26.6 (3.4) | .156 | |
| Diagnostic group* | Normal Lung volume | 3 (21%) | 16 (30%) | 1 (8%) | 1 (7%) | .085 |
| Hyperinflation | 4 (29%) | 10 (19%) | 2 (17%) | 0 (0%) | ||
| Restrictive | 7 (50%) | 28 (52%) | 9 (75%) | 13 (93%) | ||
| Donor Age, mean (SD) | 33.3 (15.3) | 39 (15.5) | 42.8 (13.5) | 36.6 (12.2) | .40 | |
| Sex | Female | 7 (50%) | 26 (48%) | 6 (50%) | 4 (29%) | .58 |
| Height cm, mean (SD) | 172.0 (8.2) | 170.0 (8.6) | 171.5 (9) | 173.8 (7.7) | .49 | |
| Body mass index (kg/m2), mean (SD) | 26.7 (8.2) | 27.4 (6.1) | 25.4 (6.3) | 25.3 (3.5) | .595 | |
| Smoking history | >20 packs year | 5 (100%) | 25 (59.5%) | 3 (30%) | 7 (70%) | .025 |
| Ischemic time (min), mean (SD) | 187.0 (15.6) | 210.1 (39.4) | 220.8 (29.9) | 280.6 (45.3) | .002 | |
| Cause of death | Anoxia | 1 (50%) | 4 (20%) | 0 (0%) | 3 (43%) | .55 |
| CVA | 0 (0%) | 6 (30%) | 1 (25%) | 1 (14%) | ||
| Intracranial Hemorrhage | 0 (0%) | 1 (5%) | 1 (25%) | 0 (0%) | ||
| Trauma | 1 (50%) | 9 (45%) | 2 (50%) | 2 (29%) | ||
| Other | 0 (0%) | 0 (0%) | 0 (0%) | 1 (14%) | ||
| Mechanical ventilation >48 h | 2 (14%) | 10 (19%) | 4 (36%) | 11 (79%) | < .001 | |
| Unexpected return to the OR | 4 (29%) | 3 (6%) | 2 (18%) | 5 (36%) | .016 | |
| CT lung volume recipient, mean (SD) | 5.15 (2.52) | 4.79 (2.13) | 3.69 (1.95) | 3.30 (1.86) | .051 | |
| CT lung volume donor, mean (SD) | 2.68 (1.29) | 2.69 (.95) | 2.85 (0.88) | 3.04 (0.09) | .67 | |
| Predicted TLC recipient, mean (SD) | 6.30 (1.21) | 6.07 (1.11) | 5.70 (0.97) | 6.76 (1.06) | .093 | |
| Predicted TLC donor, mean (SD) | 6.11 (1.10) | 6 (0.97) | 6.21 (1.09) | 6.57 (0.93) | .38 |
CT, computed tomography; OR, operating room; PGD, primary graft dysfunction; TLC, total lung capacity.
Table 4.
Ordinal Logistic Regression Models for PGD 0–3
| Variable | Coefficient | 95% CI | P |
|---|---|---|---|
| Model 1 | |||
| CT volume donor | 0.44 | −0.01 to 0.89 | .058 |
| CT volume recipient | −0.32 | −0.52 to −0.11 | .002 |
| Intercept for PGD 1 | −2.15 | −3.6 to −0.7 | |
| Intercept for PGD 2 | 0.85 | −0.51 to 2.2 | |
| Intercept for PGD 3 | 1.7 | 0.3, 3.16 | |
| Model 2 | |||
| Predicted TLC donor | 0.3 | −0.49 to 0.38 | .804 |
| Predicted TLC recipient | −0.05 | −0.19 to 0.78 | .226 |
| Intercept for PGD 1 | −0.25 | −2.89 to 2.38 | |
| Intercept for PGD 2 | 2.46 | −0.22 to 5.13 | |
| Intercept for PGD 3 | 3.26 | 0.54 to 5.98 |
PGD 0 is the reference variable. Model 1 Brant test χ2 = 2.04; P = .73.
CT, computed tomography; PGD, primary graft dysfunction; TLC, total lung capacity.
Fig 4.
Predicted probabilities for each primary graft dysfunction grade with 95% CIs on the vertical axis according to recipient computed tomography lung volume in liters on the horizontal axis for a donor with the mean computed tomography lung volume of 2.8 liters.
Fig 5.
Donor computed tomography lung volumes on the vertical axis with the corresponding recipient computed tomography lung volumes to achieve a 30%, 25%, 20%, and 15% risk of primary graft dysfunction grade 2 and 3.
DISCUSSION
Our study evaluates the use of CT volumetry to advance toward personalized size matching in lung transplantation. The CT lung volumes and plethysmography TLC were equivalent in recipients; both methods were poorly correlated to predicted TLC. In donors, CT lung volumes were smaller than predicted TLC. The CT lung volumes significantly changed over time in patients with restrictive lung disease but remained stable from listing to transplantation in the rest of the cohort. The CT lung volumes in donors and recipients were independent predictors of the need for surgical graft reduction and primary graft dysfunction grade outperforming predicted TLC.
In our pragmatic analysis, CT lung volumes from clinical CT scan images predicted clinically significant outcomes outperforming predicted TLC. In recipients, CT lung volumes were remarkably reproducible with small within-individual variability and closely related to plethysmography TLC. The agreement between CT lung volumes and plethysmography TLC was similar to previous reports in patients with end-stage lung disease [23,24]. Multiple factors likely contributed to the average bias of −0.47 L: 1. the recumbent position used to obtain CT scan images results in underestimation of lung volumes when compared with the seated position used for plethysmography [25] due to increased diaphragmatic inspiratory load by abdominal organs; 2. plethysmography TLC may overestimate lung volumes in the presence of severe obstructive ventilatory physiology through incomplete equalization of mouth and alveolar pressures [26]; and 3. CT volumetry measures air density volume within the chest, excluding the anatomic dead space, whereas plethysmography includes all air-filled structures resulting in larger volumes [27].
In contrast, CT volumetry and predicted TLC were poorly correlated. The predicted TLC equations are derived from healthy individuals in the general population [17], whereas patients awaiting lung transplantation have advanced lung disease [28]. A preponderance of restrictive physiology in our cohort helps explain the underestimation of predicted TLC. The opposite result could be expected in cohorts with most patients exhibiting hyperinflation.
In donors, CT lung volumes were systematically smaller than predicted TLC. All donors in our cohort were brain-dead and on mechanical ventilation. The CT lung volumes were measured without regard for the respiratory cycle phase at the set PEEP, usually 8 to 10 cmH2O, and not during a forced inspiratory maneuver. These factors resulted in systematic underestimation of the predicted TLC. Similar to a previous study [29], CT lung volumes in donors could be estimated using variables predictive of TLC, age, sex, height, and race, confirming measurement at the mechanical ventilation equivalent of functional residual capacity, a volume proportional to TLC. We only had one chest CT available for each donor, and the PEEP level during imaging was not available for analysis. Previous studies have described changes in lung volumes with changes in PEEP [30]. Unexpectedly, 4 donors with smaller CT lung volumes than their recipients required surgical graft reduction. This was explained by the presence of lobar atelectasis in 2 donors and measurements within the margin of error of CT volumetry for the other 2 pairs. Standardization of image acquisition, including tidal volumes, PEEP settings, gating of scans to end-inspiration, and attention to atelectasis in clinical practice could help minimize this source of error.
Poor outcomes for the smaller recipient and larger donor lung volumes are counterintuitive compared to previous findings using the predicted TLC ratio [1−5]. However, our results are not inconsistent. The relationship of donor-recipient lung volume is frequently inverted when using measures of actual lung capacity instead of predicted TLC [31]. For example, taller donors are accepted for recipients with hyperinflation and shorter donors for patients with restrictive physiology [28]. Because predicted TLC does not consider the underlying disease process, donors will be considered larger than the recipient in the first case and smaller in the second. The opposite is true in the case of CT lung volumes.
Larger donor and smaller recipient CT lung volumes predicted higher PGD grade. Large recipient CT lung volumes were observed in chronic obstructive pulmonary disease, including that caused by alpha-1 antitrypsin deficiency. Both diseases are considered at low risk for PGD [32]. The opposite is true for patients with restrictive physiology and normal lung volumes, which include the high PGD risk diagnoses of sarcoidosis, connective tissue disease-related interstitial lung disease, and pulmonary hypertension [32,33]. The sample size was too small to adjust our model according to diagnosis; thus, it is still unclear whether lung volume or underlying diagnosis are responsible for the adverse outcomes.
Larger donor lung volume increased the risk for PGD regardless of recipient lung size. This finding is consistent with the only other study using actual TLC measures to assess the risk of PGD [30]. The authors evaluated donor-to-recipient chest X-ray lung height ratios. They found that a smaller ratio and smaller donor for the recipient’s size were associated with decreased risk of PGD grades 2 to 3 at 48 to 72 hours [31]. The presence of atelectasis, poor accommodation of cardiac output resulting in pulmonary edema, atelectrauma during mechanical ventilation, and prolonged ischemic times due to obscuration of the surgical field may explain the association of larger donor lung volumes with an increased risk of PGD.
Our analysis has several limitations. It is a single-center study of bilateral lung transplantation using local donors. The retrospective nature of the analysis means that patients had already been size matched using an alternative method, and we only address surgical graft reduction and primary graft dysfunction. At the same time, long-term outcomes remain to be tested. Before our results are generalized, they must be replicated in other cohorts.
Our data support the role of CT volumetry in matching lung transplant recipients. Whether size mismatch is causally related to any outcome beyond the need for surgical graft reduction or it is simply a proxy for other factors such as sex and underlying diagnoses remains unclear. However, independent of its causal association, donor and recipient lung volume matching remains an actionable factor. Before application to clinical practice, our findings should be corroborated prospectively, including populations not previously matched based on predicted TLC.
Acknowledgments
Research reported in this publication was supported by the Washington University Institute of Clinical and Translational Sciences grant UL1TR002345 from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH). The content is solely the responsibility of the authors and does not necessarily represent the official view of the NIH.
Footnotes
DISCLOSURES
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
DATA AVAILABILITY
Data will be made available on request.
REFERENCES
- [1].Eberlein M, Reed RM, Maidaa M, Bolukbas S, Arnaoutakis GJ, Orens JB, et al. Donor-recipient size matching and survival after lung transplantation. A cohort study. Ann Am Thorac Soc 2013; 10:418–25. [DOI] [PubMed] [Google Scholar]
- [2].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;31:492–500. [DOI] [PubMed] [Google Scholar]
- [3].Eberlein M, Reed RM, Bolukbas S, Parekh KR, Arnaoutakis GL, Orens JB, et al. Lung size mismatch in bilateral lung transplantation is associated with allograft function and bronchiolitis obliterans syndrome. Chest 2012;141:451–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Eberlein M, et al. Lung size mismatch and survival after single and bilateral lung transplantation. Ann Thorac Surg 2013;96:457–63. [DOI] [PubMed] [Google Scholar]
- [5].Riddell P, Ma J, Dunne B, Binnie M, Cypel M, Donahoe L, et al. A simplified strategy for donor-recipient size-matching in lung transplant for interstitial lung disease. J Heart Lung Transplant 2021;40:1422–30. [DOI] [PubMed] [Google Scholar]
- [6].Montoya P, Bello I, Ascanio F, Romero L, Pérez J, Rosado J, et al. Graft reduction surgery is associated with poorer outcome after lung transplantation: a single-centre propensity score-matched analysis. Eur J Cardiothorac Surg 2021;60:1308–15. [DOI] [PubMed] [Google Scholar]
- [7].Vazquez Guillamet R Comments and opinions: predicted total lung capacity equations: a barrier to the definition of safe lung size differences in lung transplantation. J Heart Lung Transplant 2022;41:996–7. [DOI] [PubMed] [Google Scholar]
- [8].Kronmal RA. Spurious correlation and the fallacy of the ratio standard revisited. J R Statist Soc A 1993;156:379–92. [Google Scholar]
- [9].Tantucci C, Bottone D, Borghesi A, Guerini M, Quadri F, Pini L. Methods for measuring lung volumes: is there a better one? Respiration 2016;91:273–80. [DOI] [PubMed] [Google Scholar]
- [10].Sieren JP, Newell JD Jr, Barr RG, Bleecker ER, Burnette N, Carretta EE, et al. SPIROMICS protocol for multicenter quantitative computed tomography to phenotype the lungs. Am J Respir Crit Care Med 2016;194:794–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Han MK, Kazerooni EA, Lynch DA, Lio LX, Murray S, Curtis JL, et al. Chronic obstructive pulmonary disease exacerbations in the COPDGene study: associated radiologic phenotypes. Radiology 2011;261:274–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Gauthier JM, Bierhals AJ, Liu J, Balsara KR, Frederkisen C, Gremminger E, et al. Chest computed tomography imaging improves potential lung donor assessment. J Thorac Cardiovasc Surg 2019;157:1711–1718.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Camargo JJP, Irion KL, Marchiori E, Hochhegger B, Porto NS, Moraes BG, et al. Computed tomography measurement of lung volume in preoperative assessment for living donor lung transplantation: volume calculation using 3D surface rendering in the determination of size compatibility. Pediatr Transplant 2009;13:429–39. [DOI] [PubMed] [Google Scholar]
- [14].Chen F, Kubo T, Shoji T, Fujinaga T, Bando T, Date H. Comparison of pulmonary function test and computed tomography volumetry in living lung donors. J Heart Lung Transplant 2011;30:572–5. [DOI] [PubMed] [Google Scholar]
- [15].Date H, Aoyama A, Hijiya K, Motoyama H, Handa T, Kinoshita H, et al. Outcomes of various transplant procedures (single, sparing, inverted) in living-donor lobar lung transplantation. J Thorac Cardiovasc Surg 2017;153:479–86. [DOI] [PubMed] [Google Scholar]
- [16].Chen-Yoshikawa TF, Date H. Three-dimensional image in lung transplantation. Gen Thorac Cardiovasc Surg 2018;66:19–26. [DOI] [PubMed] [Google Scholar]
- [17].Hall GL, Filipow N, Ruppel G, Okitika T, Thomposon B, Kirkby J, et al. Official ERS technical standard: Global Lung Function Initiative reference values for static lung volumes in individuals of European ancestry. Eur Respir J 2021;57:2000289. [DOI] [PubMed] [Google Scholar]
- [18].Snell GI, Yusen RD, Weill D, Strueber M, Garrity E, Reed A, et al. Report of the ISHLT Working Group on Primary Lung Graft Dysfunction, part I: definition and grading-A 2016 Consensus Group statement of the International Society for Heart and Lung Transplantation. J Heart Lung Transplant 2017;36:1097–103. [DOI] [PubMed] [Google Scholar]
- [19].Ross JC, Est[notdef]epar RSJ, Díaz A, Westin C-F, Kikinis R, Silverman EK, et al. Lung extraction, lobe segmentation and hierarchical region assessment for quantitative analysis on high resolution computed tomography images. Med Image Comput Comput Assist Interv 2009;12:690–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Fedorov A, Beichel R, Kalpathy-Cramer J, Finet J, Fillion-Robin J-C, Pujol S, et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network. Magn Reson Imaging 2012; 30:1323–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Bland JM, Altman DG. Agreement between methods of measurement with multiple observations per individual. J Biopharm Stat 2007;17:571–82. [DOI] [PubMed] [Google Scholar]
- [22].Wanger J, Clausen JL, Coates A, Pedersen OF, Brusasco V, Burgos F, et al. Standardisation of the measurement of lung volumes. Eur Respir J 2005;26:511–22. [DOI] [PubMed] [Google Scholar]
- [23].Jung WS, Haam S, Shin JM, Han K, Park CH, Byun MK, et al. The feasibility of CT lung volume as a surrogate marker of donor-recipient size matching in lung transplantation. Medicine (Baltimore) 2016;95:e3957. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Hwang SH, Lee JG, Kim TH, Paik HC, Park CH, Haam S. Comparison of predicted total lung capacity and total lung capacity by computed tomography in lung transplantation candidates. Yonsei Med J 2016;57:963–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Yamada Y, Yamada M, Chubachi S, Yokoyama Y, Matsuoka S, Tanabe A, et al. Comparison of inspiratory and expiratory lung and lobe volumes among supine, standing, and sitting positions using conventional and upright CT. Sci Rep 2020;10:16203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Rodenstein DO, Francis C, Stănescu DC. Airway closure in humans does not result in overestimation of plethysmographic lung volume. J Appl Physiol Respir Environ Exerc Physiol 1983;55:1784–9. [DOI] [PubMed] [Google Scholar]
- [27].O’Donnell CR, Bankier AA, Stiebellehner L, Reilly J, Brown R, Loring SH. Comparison of plethysmographic and helium dilution lung volumes: which is best for COPD? Chest 2010;137:1108–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Chambers DC, Cherikh WS, Harhay MO, Hayes D Jr, Hsich E, Khush KK, et al. The International Thoracic Organ Transplant Registry of the International Society for Heart and Lung Transplantation: thirty-sixth adult lung and heart-lung transplantation report-2019; focus theme: donor and recipient size match. J Heart Lung Transplant 2019;38:1042–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Konheim JA, Kon ZN, Pasrija C, Luo Q, Sanchez PG, Garcia JP, et al. Predictive equations for lung volumes from computed tomography for size matching in pulmonary transplantation. J Thorac Cardiovasc Surg 2016;151 1163–9.e1. [DOI] [PubMed] [Google Scholar]
- [30].Bikker I, Miranda RD, Bommel JV, Bakker J, Gommers D. Functional residual capacity measurement during mechanical ventilation in order to find the optimal positive end-expiratory pressure. Crit Care 2007;11:1–2. [Google Scholar]
- [31].Li D, Weinkauf J, Hirji A, Nagendran J, Kapasi A, Lien D, et al. Chest x-ray sizing for lung transplants reflects pulmonary diagnosis and body composition and is associated with primary graft dysfunction risk. Transplantation 2021;105:382–9. [DOI] [PubMed] [Google Scholar]
- [32].Diamond JM, Lee JC, Kawut SM, Shah RJ, Localio AR, Bellamy SL, et al. Clinical risk factors for primary graft dysfunction after lung transplantation. Am J Respir Crit Care Med 2013;187:527–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Natalini JG, Diamond JM, Porteous MK, Lederer DJ, Willie KM, Weinacker AB, et al. Risk of primary graft dysfunction following lung transplantation in selected adults with connective tissue disease-associated interstitial lung disease. J Heart Lung Transplant 2021;40:351–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be made available on request.