Abstract
Rationale:
Low body mass index (BMI) may influence lung transplant decisions for patients with advanced cystic fibrosis (CF) lung disease.
Objective:
Determine whether patients with advanced CF lung disease and BMI ≤17 kg/m2 are less likely to be listed for lung transplant or have a higher risk of death without listing compared to those with higher BMI.
Methods:
Using merged United Network for Organ Sharing and CF Foundation Patient Registries, we identified adults with onset of advanced lung disease (FEV1 ≤ 40% predicted) between May-2005 and December-2016. We analyzed survival using competing risks regression with cause-specific risks of listing for lung transplant and death without listing. BMI ≤ 17 kg/m2 was our predictor.
Measurements and main results:
Among 5,121 CF patients with advanced lung disease, 23% were listed for lung transplant (n=1,201), 23% died without listing (n=1,190), and 44% were alive without listing (n=2,730) as of December-2016. Patients with BMI ≤ 17 kg/m2 were less likely to be listed for transplant (HR 0.69; 95% CI 0.57, 0.83) and more likely to die without listing (HR 1.63; 95% CI 1.41, 1.88). We identified important regional variations in the likelihood of referral and listing, based on BMI.
Conclusions:
Patients with advanced CF lung disease and BMI ≤ 17 kg/m2 are less likely to be listed for lung transplant and have a higher risk of dying without listing, compared to those with higher BMI. Regional differences suggest access to transplant for malnourished CF patients may be limited by location.
Keywords: lung diseases, lung transplantation, nutritional status, survival analysis
Introduction
Malnutrition and low body mass index (BMI) are common among patients with advanced cystic fibrosis (CF) (1–5). The etiology of malnutrition in advanced CF is multifactorial, including anorexia due to dyspnea and air trapping, and high basal energy requirements due to chronic respiratory infection (1, 6–13). Among patients with advanced CF lung disease, low BMI is considered a marker for severity of illness and, per recent CF Foundation consensus guidelines, indicates the need for urgent consideration of lung transplantation (14). However, this need for urgent transplant consideration is often coupled with concern that nutritional status will have a negative effect on post-transplant outcomes (3, 15–20), and some lung transplant programs consider low BMI an absolute contraindication to listing (9). The International Society for Heart and Lung Transplantation (ISHLT) recommendations for lung transplant candidate selection cite BMI < 16 kg/m2 as a high-risk factor that impacts transplant candidacy (21) but also acknowledge recently published data that show acceptable survival for recipients with CF and BMI < 17 kg/m2 (4).
We hypothesized that low BMI is a barrier to listing for lung transplant, but that it is simultaneously a marker of disease severity and associated with increased risk of death without transplant. The aim of this study was to determine whether patients with advanced CF lung disease and BMI ≤17 kg/m2 were listed for lung transplant less often than those with higher BMI, and whether they faced a higher risk of death without listing. In addition, we examined the relationship between BMI and referral for lung transplant and death without referral. Our outcomes of listing and death without listing were felt to reflect the practices of both lung transplant centers and CF centers, whereas the outcomes of referral for lung transplant and death without referral reflect the practices of CF centers. By examining outcomes related to listing and referral, we hoped to support our understanding of the barriers to transplantation for patients with low BMI and whether they exist at the level of CF centers, transplant centers, or both. Finally, we evaluated the association between BMI ≤ 17 kg/m2 and referral and listing for lung transplant by geographic region, hypothesizing that the association between BMI and both referral and listing would differ based on regional CF Center and lung transplant program practices. A portion of our analysis was previously reported in the form of an abstract (22).
Methods
Study design and participants
This retrospective cohort study used data from a merged United Network for Organ Sharing (UNOS)/Organ Procurement and Transplantation Network (OPTN) Registry and Cystic Fibrosis Foundation Patient Registry (CFFPR) (23, 24). Adult patients (≥18 years) with CF in the United States (US) were included if they had no previous history of lung transplant and their first occurrence of advanced lung disease (FEV1 percent predicted ≤ 40%; not marked exacerbation or hospitalization, “stable”) started between 5-01-2005 and 12-31-2016.
Measures
Patient characteristics were obtained from the combined UNOS CFFPR database at the time of onset of advanced lung disease. For our primary outcomes of listing for transplant and death without listing, we obtained listing date from UNOS and death date from CFFPR if the patient was not listed. Additional outcomes included lung transplant referral and death without referral. Only annual lung transplant referral status was available in the CFFPR (referral date is not captured by UNOS); referral status in the CFFPR can include: “not pertinent”; “accepted, on waiting list”; “evaluated, final decision pending”; “evaluated, rejected”; “had transplantation”. Thus, to represent referral dates in our analyses we selected the last day of the calendar year when the patient was first reported as referred, listed, or underwent lung transplant. Our predictor of interest was BMI, and potential confounders, selected a priori, included: age, sex, race/ethnicity, education level, insurance (Medicaid at any time in the 3 years prior to first occurrence of advanced lung disease), diabetes requiring insulin, primary CF center, FEV1 percent predicted, use of supplemental oxygen (yes/no), use of supplemental feeding, number of pulmonary exacerbations requiring antibiotics, and presence of Burkholderia cepacia or Mycobacterium abscessus. For patient education level, diabetes status, and primary CF center we used their measurement at the first occurrence of advanced lung disease.
Analysis
We examined the relationship between BMI ≤ 17 kg/m2 and the outcomes of listing for lung transplant, death without listing, referral for transplant, and death without referral using four separate competing risks regression models with cause-specific hazards (25, 26). BMI was modeled as a time-varying covariate, adjusted for the following time-varying covariates: age, FEV1 percent predicted, use of supplemental oxygen, use of supplemental feeding, number of pulmonary exacerbations requiring antibiotics, and presence of Burkholderia cepacia or Mycobacterium abscessus. Our analysis relied on longitudinal survival models, thus each patient’s hazard was updated when they recorded a new measurement and did not change until a successive measurement was made. Fixed covariates included sex, race/ethnicity, diabetes requiring insulin, education level, and insurance status. We also performed sensitivity analysis using BMI separated into the following categories: BMI ≤17 kg/m2, BMI 17.01–18.49 kg/m2 or BMI ≥18.5 kg/m2. Nutritional differences were explored by fitting an interaction between BMI and supplemental feeding. Regional effects were explored by looking at the interaction between BMI and patient location, defined by UNOS Region. All analyses were performed using R software, version 3.6.3 with two-sided significance value of 0.05. This project was approved by the institutional review board at the University of Washington (STUDY #7923).
Results
Description of cohort
We identified 5,344 adult patients with CF in the US who were found to have advanced lung disease (FEV1 percent predicted ≤ 40%) for the first time after 5-01-2005. Of these, 223 patients were excluded for either having previously received a lung transplant (n=86), inaccuracies in lung transplant dates (n=14), or because they had no follow up data (n=123) (Figure E1). Among the remaining cohort of 5,121 patients, 1,201 (23%) were listed for transplant and 1,190 (23%) died without being listed prior to 12-31-2016. Patients with a BMI ≤ 17 kg/m2 accounted for only 14% of listings but 23% of deaths without listing. Additional patient characteristics by listing status are outlined in Table 1.
Table 1.
Characteristics of patients by lung transplant listing status by end of study period
| Variable | Alive without listing (n=2,730) | Listed (n=1,201) | Death without listing (n=1,190) |
|---|---|---|---|
| Age, median years (IQR) | 25.6 (21.1–33.5) | 22.5 (19.1–28.3) | 23.1 (19.3–30.5) |
| Female sex, n (%) | 1,292 (47%) | 665 (56%) | 609 (51%) |
| Education, n (%) | |||
| Graduate | 72 (3%) | 18 (2%) | 12 (1%) |
| College | 1,325 (49%) | 646 (34%) | 380 (32%) |
| High school or less | 975 (36%) | 500 (35%) | 620 (52%) |
| Unknown | 358 (13%) | 112 (4%) | 178 (15%) |
| Race, n (%) | |||
| White | 2,577 (94%) | 1,167 (97%) | 1,113 (94%) |
| Black | 122 (4%) | 22 (2%) | 60 (5%) |
| American Indian or Native Alaskan | 15 (1%) | 5 (<1%) | 8 (1%) |
| Asian | 15 (1%) | 4 (<1%) | 3 (<1%) |
| Pacific Islander | 2 (<1%) | 1 (<1%) | 0 (0%) |
| Other | 26 (1%) | 9 (1%) | 11 (1%) |
| Hispanic ethnicity, n (%) | 151 (6%) | 71 (6%) | 64 (5%) |
| Insurance, n (%) | |||
| Private | 1,547 (57%) | 771 (64%) | 481 (40%) |
| Medicare | 422 (15%) | 130 (11%) | 226 (19%) |
| Medicaid | 981 (36%) | 469 (39%) | 638 (54%) |
| State/Special Needs | 237 (9%) | 60 (5%) | 60 (5%) |
| Tricare/Military Service Plan | 57 (2%) | 29 (2%) | 22 (2%) |
| Indian Health Service | 3 (<1%) | 0 (0%) | 2 (<1%) |
| Other | 59 (2%) | 32 (3%) | 35 (3%) |
| BMI kg/m2 at listing or death, n (%) | |||
| ≤ 17 | -- | 163 (14%) | 271 (23%) |
| 17.01 – 18.49 | -- | 290 (24%) | 251 (21%) |
| 18.5 – 25 | -- | 675 (56%) | 571 (48%) |
| 25 – 29.99 | -- | 62 (5%) | 75 (6%) |
| ≥ 30 | -- | 11 (1%) | 22 (2%) |
| FEV1, median (IQR) | 36.3 (32.3 – 38.5) | 34.7 (28.7 – 38.0) | 34.4 (28.7 – 37.6) |
| Oxygen therapy | |||
| None | 1,875 (69%) | 580 (48%) | 491 (41%) |
| Only during exacerbations | 424 (16%) | 230 (19%) | 248 (21%) |
| Continuous or nightly | 431 (16%) | 391 (31%) | 451 (38%) |
| Pulmonary exacerbations requiring antibiotics/year, median (IQR) | 1 (0–2) | 2 (1–3) | 2 (1–3) |
BMI: body mass index; FEV: forced expiratory volume in 1 second
Analysis for listing and referral
Adjusting for potential confounders, we found that patients with BMI ≤17 kg/m2 were less likely to be listed for transplant (HR 0.69; 95% CI 0.57, 0.83) and more likely to die without listing (HR 1.63; 95% CI 1.41, 1.88) than those with a higher BMI (Table 2). In an adjusted analysis with BMI as a three-category variable, the significant relationship between BMI ≤17 kg/m2 and listing and death without listing remained (HR 0.71, 95% CI 0.59, 0.87; and HR 1.74, 95% CI 1.50 – 2.03, respectively) (Table E1). Using referral and death without referral as outcomes in adjusted analysis, we found that patients with BMI ≤17 kg/m2 were less likely to be referred for lung transplant (HR 0.76; 95% CI 0.65, 0.89) and more likely to die without referral (HR 1.38; 95% CI 1.16, 1.64) (Table 3). Hazard ratios for all predictors included in multivariable models are shown in Tables 2 and 3. Competing risks regression fitting BMI as a three-category variable with an interaction on supplemental feeding suggested that patients with BMI ≤ 17 kg/m2 had a lower risk of death without listing (HR 0.75; 95% CI 0.59, 0.95) than patients with BMI ≥ 18.5 kg/m2.
Table 2.
Multivariable competing risks regression for predictors of listing for transplant and death without listing
| Variablea | Listing for transplant (HR 95% CI) | Death without listing (HR 95% CI) |
|---|---|---|
| BMI ≤ 17 kg/m2 | 0.69 (0.57, 0.83) | 1.63 (1.41, 1.88) |
| Age, 5-year intervals | 0.96 (0.92, 0.99) | 1.09 (1.06, 1.12) |
| Education (reference: some college or greater) | ||
| High school | 0.56 (0.49, 0.65) | 1.30 (1.14, 1.48) |
| Unknown | 0.64 (0.51, 0.80) | 1.51 (1.26, 1.81) |
| Medicaid | 0.70 (0.61, 0.80) | 1.42 (1.25, 1.63) |
| White race | 1.31 (0.92, 1.88) | 0.75 (0.60, 0.92) |
| Male sex | 0.74 (0.65, 0.84) | 0.93 (0.83, 1.04) |
| FEV1 % predicted | 0.90 (0.89, 0.91) | 0.95 (0.94, 0.95) |
| Oxygen therapy (reference: no supplemental oxygen) | ||
| Only during exacerbations | 1.37 (1.11, 1.69) | 2.55 (2.10, 3.09) |
| Continuous or nightly | 2.52 (2.12, 3.00) | 3.67 (3.10, 4.35) |
| Pulmonary exacerbations treated with antibiotics/year | 1.14 (1.10, 1.17) | 1.09 (1.06, 1.12) |
| Diabetes requiring insulin | 1.57 (1.39, 1.78) | 1.06 (0.94, 1.19) |
| Burkholderia cepacia | 0.56 (0.38, 0.84) | 1.42 (1.13, 1.79) |
| Mycobacterium abscessus | 0.51 (0.18, 1.43) | 0.76 (0.31, 1.84) |
| Supplemental feeding (reference: no) | ||
| Yes | 1.08 (0.95, 1.23) | 0.99 (0.88, 1.11) |
| Unknown | 0.83 (0.45, 1.55) | 0.71 (0.41, 1.23) |
BMI: body mass index; CI: confidence interval; FEV: forced expiratory volume in 1 second; HR: hazard ratio
For non-binary, categorical variables, the reference group is noted.
Table 3.
Multivariable competing risks regression for predictors of referral for transplant and death without referral
| Variable | Referral for transplant (HR 95% CI) | Death without referral (HR 95% CI) |
|---|---|---|
| BMI ≤ 17 kg/m2 | 0.76 (0.65, 0.89) | 1.38 (1.16, 1.64) |
| Age, 5-year intervals | 0.92 (0.90, 0.95) | 1.08 (1.05, 1.12) |
| Education (reference: some college or greater) | ||
| High school | 0.65 (0.58, 0.72) | 1.55 (1.32, 1.81) |
| Unknown | 0.60 (0.51, 0.70) | 1.48 (1.20, 1.82) |
| Medicaid | 0.83 (0.74, 0.92) | 1.33 (1.14, 1.56) |
| White race | 1.46 (1.13, 1.87) | 0.82 (0.63, 1.06) |
| Male sex | 0.82 (0.75, 0.90) | 0.87 (0.76, 1.00) |
| FEV1 % predicted | 0.96 (0.96, 0.97) | 0.98 (0.97, 0.99) |
| Oxygen therapy (reference: no supplemental oxygen) | ||
| Only during exacerbations | 1.27 (1.09, 1.49) | 2.21 (1.77, 2.75) |
| Continuous or nightly | 1.88 (1.66, 2.13) | 2.30 (1.90, 2.77) |
| Pulmonary exacerbations treated with antibiotics/year | 1.05 (1.03, 1.08) | 1.02 (0.99, 1.05) |
| Diabetes requiring insulin | 1.28 (1.16, 1.40) | 0.84 (0.73, 0.97) |
| Burkholderia cepacia | 1.42 (1.14, 1.76) | 1.38 (0.97, 1.96) |
| Mycobacterium abscessus | 1.59 (1.01, 2.50) | 1.20 (0.51, 2.83) |
| Supplementary Feeding (reference: no) | ||
| Yes | 1.12 (1.02, 1.24) | 0.93 (0.81, 1.07) |
| Unknown | 0.97 (0.60, 1.57) | 0.90 (0.50, 1.63) |
BMI: body mass index; CI: confidence interval; FEV: forced expiratory volume in 1 second; HR: hazard ratio
For non-binary, categorical variables, the reference group is noted.
Regional differences in listing and referral
Figure 1 depicts the patients with BMI ≤ 17 kg/m2 at first occurrence of FEV1 ≤ 40% who went on to die without being listed at each CF center across the continental United States, as well as the percentage of transplants involving patients with BMI ≤ 17 kg/m2 at each lung transplant center. Competing risks regression model fitting BMI ≤ 17 kg/m2 as an interaction with UNOS Region found that BMI ≤ 17 kg/m2 was associated with a lower likelihood of listing in four UNOS regions and a higher risk of death without listing in six UNOS Regions (Table 4).
Figure 1.
The map represents the continental United States, with United Network for Organ Sharing (UNOS) regions outlined in black. Blue circles represent lung transplant centers and the proportion of transplants at each center involving patients with BMI ≤ 17 kg/m2 (darker blue indicates larger proportion). Red circles represent cystic fibrosis centers and the proportion of patients with BMI ≤ 17 kg/m2 at first occurrence of FEV1 ≤ 40% who die without being listed (darker red indicates larger proportion). CF: cystic fibrosis; LTx: lung transplant
Table 4.
BMI ≤ 17 kg/m2 interaction with UNOS Region: Listing and death without listing, referral and death without referral
| UNOS Regiona | Listing for transplant (HR 95% CI) | Death without listing (HR 95% CI) | Referral for transplant (HR 95% CI) | Death without referral (HR 95% CI) |
|---|---|---|---|---|
| Region 1 | 0.61 (0.26,1.45) | 1.23 (0.74, 2.03) | 0.51 (0.29, 0.93) | 1.31 (0.67, 2.56) |
| Region 2 | 0.54 (0.30, 0.97) | 1.94 (1.21, 3.11) | 0.53 (0.26, 1.10) | 1.71 (1.18, 2.49) |
| Region 3 | 0.82 (0.51, 1.33) | 1.37 (0.98, 1.90) | 0.85 (0.52, 1.40) | 1.23 (0.58, 2.59) |
| Region 4 | 1.09 (0.62, 1.89) | 0.96 (0.50, 1.84) | 0.74 (0.44, 1.23) | 1.18 (0.60, 2.33) |
| Region 5 | 0.81 (0.48, 1.35) | 1.24 (0.81, 1.91) | 0.93 (0.64, 1.34) | 0.95 (0.49, 1.85) |
| Region 6 | 0.13 (0.02, 0.96) | 2.67 (1.52, 4.69) | 0.68 (0.44, 1.05) | 0.74 (0.36, 1.49) |
| Region 7 | 0.40 (0.18, 0.90) | 2.18 (1.44, 3.31) | 1.91 (1.10, 3.29) | 0.97 (0.50, 1.89) |
| Region 8 | 0.73 (0.44, 1.21) | 1.91 (1.23, 2.97) | 0.84 (0.53, 1.32) | 0.85 (0.42, 1.73) |
| Region 9 | 0.93 (0.32, 2.72) | 1.79 (0.82, 3.88) | 0.68 (0.44, 1.06) | 1.32 (0.67, 2.60) |
| Region 10 | 0.75 (0.46, 1.23) | 2.25 (1.48, 3.41) | 1.19 (0.64, 2.25) | 0.87 (0.44, 1.74) |
| Region 11 | 0.49 (0.28, 0.84) | 1.42 (1.01, 1.99) | 0.65 (0.46, 0.91) | 1.38 (0.68, 2.81) |
BMI: body mass index; CI: confidence interval; HR: hazard ratio; UNOS: United Network for Organ Sharing
Competing risks regression model fitting BMI ≤ 17 kg/m2 as an interaction with UNOS Region.
States included in each region: Region 1: Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Eastern Vermont; Region 2: Delaware, District of Columbia, Maryland, New Jersey, Pennsylvania, West Virginia, Northern Virginia; Region 3: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, Puerto Rico; Region 4: Oklahoma, Texas; Region 5: Arizona, California, Nevada, New Mexico, Utah; Region 6: Alaska, Hawaii, Idaho, Montana, Oregon, Washington; Region 7: Illinois, Minnesota, North Dakota, South Dakota, Wisconsin; Region 8: Colorado, Iowa, Kansas, Missouri, Nebraska, Wyoming; Region 9: New York, Western Vermont; Region 10: Indiana, Michigan, Ohio; Region 11: Kentucky, North Carolina, South Carolina, Tennessee, Virginia
Examining regional differences in referral, BMI ≤ 17 kg/m2 was associated with a lower likelihood of referral in two UNOS regions and a higher likelihood of referral in one region (Table 4). BMI ≤ 17 kg/m2 was only associated with a higher risk of death without referral in one region and was not significantly associated with a higher or lower risk of death without referral in any of the other 10 UNOS Regions (Table 4). United Network for Organ Sharing (UNOS) region-specific hazard ratios for patients with BMI ≤17 kg/m2 and 95% confidence interval bars for listing for lung transplant against the hazard ratio for death without listing are shown and a similar comparison for referral and death without referral are shown in Figure 2. A comparison of the distribution of patient outcomes for lung transplant listing and death without listing based on BMI at referral for those with BMI ≤17 kg/m2 versus those with BMI > 17 kg/m2 is available in Table E2.
Figure 2.
Panel A United Network for Organ Sharing (UNOS) region-specific hazard ratios for patients with BMI ≤17 kg/m2 and 95% confidence interval bars for listing for lung transplant (x-axis) against the hazard ratio for death without listing (y-axis). On the x-axis, a higher hazard ratio reflects a greater likelihood of listing, while on the y-axis a higher hazard ratio reflects a greater likelihood of death without listing.
Panel B United Network for Organ Sharing (UNOS) region-specific hazard ratios for patients with BMI ≤17 kg/m2 and 95% confidence interval bars for lung transplant referral (x-axis) against the hazard ratio for death without referral (y-axis). On the x-axis, a higher hazard ratio reflects a greater likelihood of referral, while on the y-axis a higher hazard ratio reflects a greater likelihood of death.
HR: hazard ratio; LTx: lung transplant
Discussion
In this large US cohort of adults with advanced CF lung disease, we demonstrated an association between BMI ≤ 17 kg/m2 and lower likelihood of both listing and referral for lung transplant, as well as a higher risk of death without listing. Furthermore, our analysis identified important regional differences by BMI in listing and referral for lung transplant, suggesting that access to transplant for malnourished individuals with CF may differ by geographic location.
Determining candidacy for lung transplant is a complex decision that incorporates not only predicted risks and benefits, but also the limited availability of donor organs. Guidelines suggest that programs select candidates based on a high risk of death without transplant, and a high predicted likelihood of survival after transplant (21), the latter of which requires assessment of risk factors associated with worse post-transplant outcomes. Low pre-transplant BMI has been associated with worse post-transplant outcomes in recipients with CF and other lung diseases (20). For this reason, many centers implement BMI thresholds as contraindications to transplant among malnourished patients with advanced lung disease (9). Our findings likely reflect the implications of these policies, with lower “risk” of listing for patients with BMI ≤ 17 kg/m2. However, there are several arguments against using low BMI as a strict exclusion criterion for lung transplant in candidates with CF.
In certain situations, the importance of predicted survival benefit may be placed above long-term survival in candidacy decisions and allocation policies. For example, despite studies showing worse outcomes after lung transplant, the US lung allocation score prioritizes certain subgroups with high short-term predicted mortality including those on mechanical support, during acute exacerbations of idiopathic pulmonary fibrosis, or other forms of acute respiratory distress syndrome (27). Malnutrition may be inherent to disease progression in many individuals with advanced CF lung disease, is often considered a marker of disease severity, and recent evidence suggests acceptable post-transplant outcomes for recipients with CF and pre-transplant BMI ≤ 17 kg/m2 (4). In fact, our study demonstrated a higher risk of death without listing among people with advanced CF lung disease, and like other higher risk groups awaiting transplant, an argument can be made that malnutrition in isolation should be considered a marker of urgency rather than a barrier to transplant for patients with CF.
Interestingly, our study also demonstrated differences in referral practices based on BMI in some regions of the US, meaning that rather than being declined by a transplant center for low BMI, many individuals with malnutrition are simply not referred by their CF center for evaluation. It is unclear whether variability in referral practices reflects CF centers’ familiarity with their local lung transplant centers’ BMI policies or other CF center-specific factors, but restrictive referral policies conflict with recent CFF guidelines which recommend that low BMI and other barriers to transplant, both modifiable and non-modifiable, should not preclude referral (14). Rather, guidelines suggest earlier referral in malnourished individuals with advanced CF lung disease, partnership between CF and transplant centers to improve nutrition pre-emptively, and an opinion from a second transplant center when an individual is declined based on a center-specific contraindication (14). A related and concerning finding was variability in listing practices across UNOS regions in the US. The origins of this variability are unclear and may reflect transplant center-specific differences within regions including transplant volumes, donor availability, and risk tolerance. Variability in listing policies may translate directly to CF center referral practices and is also important because, while some patients may be able to travel to a UNOS Region where BMI is not as heavily influential in candidacy decisions, those who are socioeconomically disadvantaged may not have that option. To that point, geographic variability in referral and listing practices raises ethical concerns and conflicts with the UNOS Final Rule for access to transplant (28).
Although we report an association between BMI ≤ 17 kg/m2 and lower likelihood of both listing and referral for lung transplant, as well as a higher risk of death without listing, we acknowledge the complex relationship between nutritional status and other barriers to transplant, including economic stability, food insecurity, social support, and access to high-quality education and healthcare (29, 30). Restrictive policies related to BMI may currently be in place to preserve a scarce resource and potentially prevent poor post-transplant outcomes; however, the fact that BMI is tied to social determinants of health should serve as an important motivating factor to address the underlying factors that contribute to low BMI in an effort to address inequities in care.
Our study has several important limitations. First, data was available through 12-31-2016, and it is possible that listing or referral practice patterns have changed since that time. However, among those patients who have advanced lung disease and are appropriate for referral for lung transplant, evidence from 2005–2015 suggests little change in limitations of transplant among those with low BMI (4). Limited change in over a decade would support the notion that a major shift has also not occurred in the past 5 years. Second, as an observational analysis, we acknowledge the inability to invoke causality in describing the association between BMI and our outcomes. There are many factors which influence decisions to list patients for lung transplant, one of which is nutritional status. Although observational, our findings do generate concern that BMI itself is an important contributor to determinations about transplant risk and the regional variations we identified suggest that a more unified approach to consideration of BMI may be indicated. Third, in using the merged United Network for Organ Sharing (UNOS)/Organ Procurement and Transplantation Network (OPTN) Registry and Cystic Fibrosis Foundation Patient Registry (CFFPR), we are reliant on the accuracy and completeness of reporting for those resources. The CFFPR is estimated to include ~81–84% of persons with CF in the US (31), and annual reports note capture of most clinic visits and hospitalizations reported in the electronic health record for patients cared for within CF care centers, with a high level of accuracy (31). Because only annual lung transplant referral status is available in the CFFPR (referral date is not captured by UNOS), we selected the last day of the calendar year when the patient was first reported as referred, listed, or underwent lung transplant. This could result in inaccurate data on patients whose status changed but was not captured given the annual reporting. Finally, our cohort does not include patients receiving highly effective CF modulator treatments, as those were not readily available to a large portion of the CF population until 2019. However, low BMI still complicates advanced CF and will likely continue do so despite these new therapies.
In summary, despite being a marker for worse outcomes among individuals with advanced CF lung disease and thus urgency for lung transplant, BMI ≤ 17 kg/m2 appears to be a barrier to listing for lung transplant in some UNOS regions. Our findings suggest the need for a unified approach to using BMI in decisions about lung transplant, one that is informed by existing evidence about post-transplant outcomes in patients with low BMI and considers the potential harms of variability in care across geographic regions.
Highlights.
In advanced CF lung disease low BMI is a barrier to lung transplant in some regions
Low BMI is associated with lower likelihood of referral and listing for transplant
Referral and listing practices vary by UNOS region
Funding sources:
KJR is supported by the National Institutes of Health (NIH) K23HL138154, Cystic Fibrosis Foundation (CFF) RAMOS17A0, and CHEST Foundation Cystic Fibrosis Grant; ALJ is supported by the NHLBI of the NIH under award number K23HL128793 as well as the CFF (LONG17A0); MLA receives funding from NIH/NIDDK [DK089507] and the CFF; CHG receives funding from the NIH [UM1 HL119073, P30 DK089507, U01 HL114589, UL1 TR000423], the CFF, and the FDA [R01 FD003704, R01 FD006848].
Table E1.
Multivariable competing risks regression for predictors of listing for transplant and death without listing
| Variable | Listing for transplant (HR 95% CI) | Death without listing (HR 95% CI) |
|---|---|---|
| BMI (reference: BMI ≥ 18.5 kg/m2) | ||
| BMI ≤ 17 kg/m2 | 0.71 (0.59, 0.87) | 1.74 (1.50, 2.03) |
| BMI 17 – 18.49 kg/m2 | 1.11 (0.96, 1.29) | 1.27 (1.09, 1.49) |
| Age, 5-year intervals | 0.96 (0.93, 0.99) | 1.10 (1.06, 1.13) |
| Education (reference: some college or greater) | ||
| High school | 0.56 (0.49, 0.65) | 1.30 (1.14, 1.48) |
| Unknown | 0.64 (0.51, 0.80) | 1.51 (1.26, 1.81) |
| Medicaid | 0.70 (0.61, 0.80) | 1.42 (1.24, 1.62) |
| White race | 1.31 (0.92, 1.88) | 0.75 (0.61, 0.92) |
| Male sex | 0.74 (0.65, 0.85) | 0.93 (0.83, 1.04) |
| FEV1 % predicted | 0.90 (0.89, 0.91) | 0.95 (0.94, 0.96) |
| Oxygen therapy (reference: no supplemental oxygen) | ||
| Only during exacerbations | 1.37 (1.11, 1.69) | 2.53 (2.09, 3.07) |
| Continuous or nightly | 2.52 (2.12, 3.00) | 3.67 (3.10, 4.35) |
| Pulmonary exacerbations treated with antibiotics/year | 1.13 (1.10, 1.17) | 1.09 (1.06, 1.12) |
| Diabetes requiring insulin | 1.57 (1.39, 1.78) | 1.06 (0.94, 1.19) |
| Burkholderia cepacia | 0.56 (0.38, 0.83) | 1.41 (1.12, 1.78) |
| Mycobacterium abscessus | 0.51 (0.18, 1.44) | 0.76 (0.31, 1.84) |
| Supplemental feeding | ||
| Yes | 1.08 (0.95, 1.22) | 0.98 (0.87, 1.10) |
| Unknown | 0.84 (0.45, 1.56) | 0.71 (0.41, 1.25) |
BMI: body mass index; CI: confidence interval; FEV: forced expiratory volume in 1 second; HR: hazard ratio
For non-binary, categorical variables, the reference group is noted.
Table E2.
Regional differences: outcomes for lung transplant listing and death without listing based on BMI at referrala
| UNOS Regionb | BMI ≤ 17 kg/m2 at referral | BMI > 17 kg/m2 at referral | p-value | ||
|---|---|---|---|---|---|
| Listing for transplant | Death without listing | Listing for transplant | Death without listing | ||
| Region 1 | 7 (47%) | 5 (33%) | 69 (54%) | 15 (12%) | 0.065 |
| Region 2 | 13 (45%) | 10 (34%) | 110 (61%) | 18 (10%) | 0.002 |
| Region 3 | 31 (57%) | 11 (20%) | 142 (58%) | 42 (17%) | 0.837 |
| Region 4 | 23 (64%) | 5 (14%) | 77 (66%) | 18 (16%) | 0.853 |
| Region 5 | 22 (73%) | 2 (7%) | 136 (65%) | 34 (16%) | 0.387 |
| Region 6 | 2 (13%) | 10 (67%) | 58 (72%) | 8 (10%) | <0.001 |
| Region 7 | 7 (35%) | 10 (50%) | 78 (66%) | 12 (10%) | <0.001 |
| Region 8 | 18 (72%) | 3 (12%) | 106 (66%) | 18 (11%) | 0.735 |
| Region 9 | 10 (59%) | 4 (24%) | 46 (52%) | 9 (10%) | 0.138 |
| Region 10 | 13 (42%) | 12 (39%) | 84 (58%) | 21 (14%) | 0.007 |
| Region 11 | 17 (45%) | 12 (32%) | 132 (61%) | 32 (15%) | 0.035 |
BMI: body mass index; UNOS: United Network for Organ Sharing
Comparing outcomes at time of referral for those with BMI ≤17 kg/m2 to those with BMI >17 kg/m2. P-values from chi-square tests comparing the distribution of patient outcomes between groups defined by BMI at time of referral. The denominators for each of the cells are the total number of patients in each region with BMI ≤ (or >) 17 kg/m2 at the time of referral. Percentages do not add to 100% because some patients neither die nor get listed for transplant in the dataset.
States included in each region: Region 1: Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Eastern Vermont; Region 2: Delaware, District of Columbia, Maryland, New Jersey, Pennsylvania, West Virginia, Northern Virginia; Region 3: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, Puerto Rico; Region 4: Oklahoma, Texas; Region 5: Arizona, California, Nevada, New Mexico, Utah; Region 6: Alaska, Hawaii, Idaho, Montana, Oregon, Washington; Region 7: Illinois, Minnesota, North Dakota, South Dakota, Wisconsin; Region 8: Colorado, Iowa, Kansas, Missouri, Nebraska, Wyoming; Region 9: New York, Western Vermont; Region 10: Indiana, Michigan, Ohio; Region 11: Kentucky, North Carolina, South Carolina, Tennessee, Virginia
Figure E1.
CONSORT diagram.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Culhane S, George C, Pearo B, Spoede E. Malnutrition in cystic fibrosis: a review. Nutr Clin Pract. 2013;28(6):676–83. [DOI] [PubMed] [Google Scholar]
- 2.Dray X, Kanaan R, Bienvenu T, Desmazes-Dufeu N, Dusser D, Marteau P, et al. Malnutrition in adults with cystic fibrosis. Eur J Clin Nutr. 2005;59(1):152–4. [DOI] [PubMed] [Google Scholar]
- 3.Lederer DJ, Wilt JS, D’Ovidio F, Bacchetta MD, Shah L, Ravichandran S, et al. Obesity and underweight are associated with an increased risk of death after lung transplantation. American journal of respiratory and critical care medicine. 2009;180(9):887–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ramos KJ, Kapnadak SG, Bradford MC, Somayaji R, Morrell ED, Pilewski JM, et al. Underweight Patients With Cystic Fibrosis Have Acceptable Survival Following Lung Transplantation: A United Network for Organ Sharing Registry Study. Chest. 2020;157(4):898–906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schönenberger KA, Reber E, Bally L, Geiser T, Lin D, Stanga Z. Nutritional assessment in adults with cystic fibrosis. Nutrition (Burbank, Los Angeles County, Calif). 2019;67–68:110518. [DOI] [PubMed] [Google Scholar]
- 6.Barni GC, Forte GC, Forgiarini LF, Abrahão CLO, Dalcin PTR. Factors associated with malnutrition in adolescent and adult patients with cystic fibrosis. Jornal brasileiro de pneumologia : publicacao oficial da Sociedade Brasileira de Pneumologia e Tisilogia. 2017;43(5):337–43. [DOI] [PubMed] [Google Scholar]
- 7.Hadjiliadis D Special considerations for patients with cystic fibrosis undergoing lung transplantation. Chest. 2007;131(4):1224–31. [DOI] [PubMed] [Google Scholar]
- 8.Matel JL, Milla CE. Nutrition in cystic fibrosis. Seminars in respiratory and critical care medicine. 2009;30(5):579–86. [DOI] [PubMed] [Google Scholar]
- 9.Morrell MR, Pilewski JM. Lung Transplantation for Cystic Fibrosis. Clinics in chest medicine. 2016;37(1):127–38. [DOI] [PubMed] [Google Scholar]
- 10.Panagopoulou P, Fotoulaki M, Nikolaou A, Nousia-Arvanitakis S. Prevalence of malnutrition and obesity among cystic fibrosis patients. Pediatrics international : official journal of the Japan Pediatric Society. 2014;56(1):89–94. [DOI] [PubMed] [Google Scholar]
- 11.Pencharz PB, Durie PR. Pathogenesis of malnutrition in cystic fibrosis, and its treatment. Clinical nutrition (Edinburgh, Scotland). 2000;19(6):387–94. [DOI] [PubMed] [Google Scholar]
- 12.Singer LG, Brazelton TR, Doyle RL, Morris RE, Theodore J. Weight gain after lung transplantation. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2003;22(8):894–902. [DOI] [PubMed] [Google Scholar]
- 13.Slae M, Wilschanski M. Prevention of malnutrition in cystic fibrosis. Current opinion in pulmonary medicine. 2019;25(6):674–9. [DOI] [PubMed] [Google Scholar]
- 14.Ramos KJ, Smith PJ, McKone EF, Pilewski JM, Lucy A, Hempstead SE, et al. Lung transplant referral for individuals with cystic fibrosis: Cystic Fibrosis Foundation consensus guidelines. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2019;18(3):321–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Allen JG, Arnaoutakis GJ, Weiss ES, Merlo CA, Conte JV, Shah AS. The impact of recipient body mass index on survival after lung transplantation. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2010;29(9):1026–33. [DOI] [PubMed] [Google Scholar]
- 16.Culver DA, Mazzone PJ, Khandwala F, Blazey HC, Decamp MM, Chapman JT. Discordant utility of ideal body weight and body mass index as predictors of mortality in lung transplant recipients. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2005;24(2):137–44. [DOI] [PubMed] [Google Scholar]
- 17.Madill J, Gutierrez C, Grossman J, Allard J, Chan C, Hutcheon M, et al. Nutritional assessment of the lung transplant patient: body mass index as a predictor of 90-day mortality following transplantation. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2001;20(3):288–96. [DOI] [PubMed] [Google Scholar]
- 18.Plöchl W, Pezawas L, Artemiou O, Grimm M, Klepetko W, Hiesmayr M. Nutritional status, ICU duration and ICU mortality in lung transplant recipients. Intensive care medicine. 1996;22(11):1179–85. [DOI] [PubMed] [Google Scholar]
- 19.Schwebel C, Pin I, Barnoud D, Devouassoux G, Brichon PY, Chaffanjon P, et al. Prevalence and consequences of nutritional depletion in lung transplant candidates. The European respiratory journal. 2000;16(6):1050–5. [DOI] [PubMed] [Google Scholar]
- 20.Upala S, Panichsillapakit T, Wijarnpreecha K, Jaruvongvanich V, Sanguankeo A. Underweight and obesity increase the risk of mortality after lung transplantation: a systematic review and meta-analysis. Transplant international : official journal of the European Society for Organ Transplantation. 2016;29(3):285–96. [DOI] [PubMed] [Google Scholar]
- 21.Leard LE, Holm AM, Valapour M, Glanville AR, Attawar S, Aversa M, et al. Consensus document for the selection of lung transplant candidates: An update from the International Society for Heart and Lung Transplantation. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ramos KJ, Pryor JB, Hee Wai T, Kapnadak SG, Aitken ML, Goss CH, Jennerich AL. Low BMI as a barrier to referral for lung transplant in cystic fibrosis. 2021 CHEST Annual Meeting (Workshop session), October 17–20, 2021, Orlando, FL. [Google Scholar]
- 23.Ramos KJ, Sykes J, Stanojevic S, Ma X, Ostrenga JS, Fink A, et al. Survival and Lung Transplant Outcomes for Individuals With Advanced Cystic Fibrosis Lung Disease Living in the United States and Canada: An Analysis of National Registries. Chest. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Stephenson AL, Ramos KJ, Sykes J, Ma X, Stanojevic S, Quon BS, et al. Bridging the survival gap in cystic fibrosis: An investigation of lung transplant outcomes in Canada and the United States. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2021;40(3):201–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Prentice RL, Kalbfleisch JD, Peterson AV Jr., Flournoy N, Farewell VT, Breslow NE. The analysis of failure times in the presence of competing risks. Biometrics. 1978;34(4):541–54. [PubMed] [Google Scholar]
- 26.Coriati A, Sykes J, Lemonnier L, Ma X, Stanojevic S, Dehillotte C, et al. The impact of the high emergency lung transplantation program in cystic fibrosis in France: insight from a comparison with Canada. The European respiratory journal. 2021. [DOI] [PubMed] [Google Scholar]
- 27.Lyu DM, Goff RR, Chan KM. The Lung Allocation Score and Its Relevance. Seminars in respiratory and critical care medicine. 2021;42(3):346–56. [DOI] [PubMed] [Google Scholar]
- 28.Egan TM. How Should Lungs Be Allocated for Transplant? Seminars in respiratory and critical care medicine. 2018;39(2):126–37. [DOI] [PubMed] [Google Scholar]
- 29.Quon BS, Psoter K, Mayer-Hamblett N, Aitken ML, Li CI, Goss CH. Disparities in access to lung transplantation for patients with cystic fibrosis by socioeconomic status. American journal of respiratory and critical care medicine. 2012;186(10):1008–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ramos KJ, Quon BS, Psoter KJ, Lease ED, Mayer-Hamblett N, Aitken ML, et al. Predictors of non-referral of patients with cystic fibrosisfor lung transplant evaluation in the United States. Journal of Cystic Fibrosis. 2016;15(2):196–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Knapp EA, Fink AK, Goss CH, Sewall A, Ostrenga J, Dowd C, et al. The Cystic Fibrosis Foundation Patient Registry. Design and Methods of a National Observational Disease Registry. Annals of the American Thoracic Society. 2016;13(7):1173–9. [DOI] [PubMed] [Google Scholar]