Summary Of An Article

Original Article

⁎ Corresponding au Bethesda, MD 20814

E-mail address: a 1569-1993/© 2016 E Journal of Cystic Fibrosis 16 (2017) 91–97

Cancer risk among lung transplant recipients with cystic fibrosis

Aliza K. Fink a,⁎, Elizabeth L. Yanik b, Bruce C. Marshall a, Michael Wilschanski c, Charles F. Lynch d, April A. Austin e, Glenn Copeland f, Mahboobeh Safaeian b, Eric A. Engels b

a Cystic Fibrosis Foundation, Bethesda, MD, USA b Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA

c Department of Pediatric Gastroenterology, Hadassah Hebrew University Medical Center, Jerusalem, Israel d Department of Epidemiology, The University of Iowa, Iowa City, IA, USA

e New York State Cancer Registry, New York State Department of Health, Albany, NY, USA f Michigan Cancer Surveillance Program, Michigan Department of Health and Human Services, Lansing, MI, USA

Received 26 April 2016; revised 17 July 2016; accepted 24 July 2016 Available online 15 August 2016


Background: Previous studies demonstrated increased digestive tract cancers among individuals with cystic fibrosis (CF), particularly among lung transplant recipients. We describe cancer incidence among CF and non-CF lung recipients. Methods: We used data from the US transplant registry and 16 cancer registries. Standardized incidence ratios (SIRs) compared cancer incidence to the general population, and competing risk methods were used for the cumulative incidence of colorectal cancer. Results: We evaluated 10,179 lung recipients (1681 with CF). Risk was more strongly increased in CF recipients than non-CF recipients for overall cancer (SIR 9.9 vs. 2.7) and multiple cancers including colorectal cancer (24.2 vs. 1.7), esophageal cancer (56.3 vs. 1.3), and non-Hodgkin lymphoma (61.8 vs. 9.4). At five years post-transplant, colorectal cancer was diagnosed in 0.3% of CF recipients aged b50 at transplant and 6.4% aged ≥50. Conclusions: CF recipients have increased risk for colorectal cancer, suggesting a need for enhanced screening. © 2016 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.

Keywords: Cancer; Transplant

1. Introduction

Cystic fibrosis (CF) is one of the most common life- shortening, genetic disorders and occurs in approximately one in 3500 Caucasian births in the US [1]. Mutations in the cystic fibrosis transmembrane regulator (CFTR) gene lead to malfunctioning or absent CFTR protein, which impairs mucosal clearance mechanisms causing recurring lung infec- tions, inflammation, and airflow obstruction. Over time there have been tremendous improvements in CF treatment. In 2014,

thor at: Cystic Fibrosis Foundation, 6931 Arlington Road , USA. (A.K. Fink).

016/j.jcf.2016.07.011 uropean Cystic Fibrosis Society. Published by Elsevier B

Downloaded for Anonymous User (n/a) at University of To For personal use only. No other uses without permis


.V. A

ledo fr sion. C

the median predicted survival in the US was 39.3 years, reflecting steady increases in survival since the early 1960s [2]. Bilateral lung transplantation is a treatment option for some individuals with CF with severe lung disease. Currently, approximately 200–250 individuals with CF receive a lung transplant each year in the US [2].

Cancer is emerging as a long-term complication of CF. During the 1980s, numerous case reports described individuals with CF developing malignancies of the digestive tract, pancreas, and gallbladder [3–7]. These observations led to single center and registry-based studies. While these studies consistently showed that people with CF did not have elevated overall cancer risk, they demonstrated 3–6 fold increased risks for cancers of the digestive tract [8–11].

ll rights reserved.

om by Elsevier on September 16, 2018. opyright ©2018. Elsevier Inc. All rights reserved.

92 A.K. Fink et al. / Journal of Cystic Fibrosis 16 (2017) 91–97

Specifically, cancer risk was found to be elevated among CF patients who received a lung transplant, with almost 3-fold increased risk for cancer overall and 17-fold increased risk for digestive tract cancers [11]. Maisonneuve et al. reported a 6-fold increased risk of colon cancer among CF patients without a transplant and a 30-fold increase among CF patients with a lung transplant [11]. Transplant recipients have increased risk for a number of cancers, especially those caused by viruses, as a result of immunosuppression administered to prevent rejection of the transplanted organ [12,13]. However, the strong increase in digestive tract cancers, especially colon cancer, observed in transplant recipients with CF is not typical of other transplant recipients [12–14].

The primary objective of this study is to describe cancer incidence among lung transplant recipients with CF and compare cancer incidence to that in lung recipients without CF. We focused on colorectal cancer and other gastrointestinal tract cancers. Strengths of our study include the evaluation of a large representative population of lung transplant recipients in the US and systematic ascertainment of incident cancer diagnoses.

2. Methods

We used data from the Transplant Cancer Match (TCM) Study [13]. The TCM Study links the US solid organ transplant registry (i.e., Scientific Registry of Transplant Recipients, SRTR) to data from 16 state and metropolitan region cancer registries [13]. The SRTR includes data from the Organ Procurement and Transplant Network on all US transplant recipients, including demographic, clinical, and transplant characteristics, and patient and graft survival. Cancer registries collect data on all malignancies other than basal and squamous cell skin cancers in their defined geographic area. Individuals are eligible for the TCM study based on whether their residence when listed for transplant or when they were transplanted was covered by a participating cancer registry. The linked data in the TCM Study thus capture incident cancers for 45% of all US transplants during 1987–2011. The TCM Study was approved by institutional review boards at the National Cancer Institute and, as required, participating cancer registries.

We included all lung recipients in the TCM Study with post-transplant follow-up transplant during a period covered by a participating cancer registry. We excluded patients who received a combined lung and other organ transplant (n = 41 among the CF subgroup). The cohort was restricted to non- Hispanic whites, non-Hispanic blacks, Hispanics, and Asian/ Pacific Islanders, since data on expected cancer rates were available for these groups. The analysis excluded American Indians/Native Alaskans, multiracial individuals, and those with unknown race.

Analyses were conducted using the following SRTR variables: reason for transplant (CF vs. other), age at transplant, sex, race/ethnicity, type of lung transplant (bilateral sequential, en-bloc double, single right lung, single left lung), calendar year of transplant, baseline Epstein–Barr virus serostatus (negative, positive, missing), and cigarette use (no, yes, not collected during transplant year, collected but missing).

Downloaded for Anonymous User (n/a) at University of Toled For personal use only. No other uses without permissio

Cancers were identified using cancer registry data and categorized according to a modified version of the Surveillance, Epidemiology and End Results (SEER) site recode [13].

Follow-up was calculated separately for each transplant (3.6% of individuals with more than one transplant were evaluated in successive intervals). Individuals were followed for cancer from lung transplantation or start of cancer registry coverage, whichever came later, until the first of the following events: death (50% of recipients exited for this reason) graft failure or a subsequent transplant (5%), loss to follow-up by the SRTR (4%), or end of cancer registry coverage (40%).

We calculated standardized incidence ratios (SIRs) to evaluate cancer risk of lung recipients compared to the general population. Expected cancer rates were calculated in strata defined by age, sex, race/ethnicity, calendar year and registry region. SIRs were calculated for all cancers combined and for specific cancers, separately for CF and non-CF lung recipients. For non-Hodgkin lymphoma (NHL), the most common cancer among CF recipients, we used Poisson regression to calculate SIR ratios comparing risk in CF and non-CF recipients, adjusting for age and EBV serostatus at time of transplant.

For colorectal cancer, we sought to understand how absolute risk among lung recipients compared to risk in the general population at age 50 (the recommended age for routine colorectal cancer screening) [15]. We therefore calculated the cumulative incidence of colorectal cancer, stratified by CF status and age at transplant (b50 vs. 50+ years) [16]. Follow-up was measured from transplant until the first of: colorectal cancer diagnosis, graft failure, re-transplantation, death, or loss to follow-up. Cumulative incidence accounts for the competing risk of death [17]. This method could not accommodate delayed entry, so we excluded a small number of recipients whose transplant occurred before cancer registry coverage (23 CF recipients and 171 non-CF recipients).

3. Results

We evaluated 10,179 lung recipients comprising 1681 CF recipients and 8498 non-CF recipients. Patient characteristics are displayed in Table 1. Age at transplant differed substantially as 77% of CF recipients were younger than age 35, compared with only 7% of non-CF recipients. CF recipients primarily received a bilateral transplant versus fewer than half of non-CF recipients (94% versus 43%). CF recipients were more likely than non-CF recipients to be EBV seronegative at the time of transplant (although there was substantial missing data). In addition, although data were not collected for the entire time period and there were also missing data, only 1% of CF recipients were reported to have smoked N10 pack years, compared to 30% of non-CF recipients. Use of induction and maintenance immunosuppressive medications did not differ between CF and non-CF recipients (data not shown).

During 35,514 person-years of follow-up (mean 3.7 and 3.4 person-years for CF and non-CF recipients, respectively), 845 cancers were diagnosed (85 among CF recipients and 760 in non-CF recipients). This reflects an unadjusted rate of 1.3 per 100 person years among CF recipients and 2.6 per 100 person

o from by Elsevier on September 16, 2018. n. Copyright ©2018. Elsevier Inc. All rights reserved.

Table 1 Characteristics of U.S. lung transplant recipients.

Characteristic CF recipient N (% of total) Non-CF recipient N (% of total) p-Value

Total 1681 (100) 8498 (100) Sex Male 846 (50.3) 4522 (53.2) 0.03

Female 835 (49.7) 3976 (46.8) Age at Transplant, years 0–19 381 (22.7) (22.717.07) 234 (2.8) b0.001

20–34 915 (54.4) 386 (4.5) 35–49 335 (19.9) 1778 (20.9) 50–64 48 (2.9) 5150 (60.6) 65+ 2 (0.1) 950 (11.2)

Race/Ethnicity White, Non-Hispanic 1580 (94.0) 7121 (83.8) b0.001 Black, Non-Hispanic 17 (1.0) 719 (8.5) Hispanic 81 (4.8) 526 (6.2) Asian/Pacific Islander 3 (0.2) 132 (1.6)

Transplant Procedure Type Bilateral sequential Lung 1575 (93.7) 3660 (43.1) b0.001 En-Bloc Double Lung 57 (3.4) 259 (3.0) Single left lung 22 (1.3) 2305 (27.1) Single right lung 27 (1.6) 2274 (26.8)

Year of Transplant 1987–1994 174 (10.4) 980 (11.5) 0.02 1995–1999 405 (24.1) 1808 (21.3) 2000–2004 472 (28.1) 2257 (26.6) 2005–2009 562 (33.4) 3062 (36.0) 2010–2011 68 (4.0) 391 (4.6)

EBV Serostatus Negative 268 (15.9) 658 (7.7) b0.001 Positive 761 (45.3) 4383 (51.6) Missing 652 (38.8) 3457 (40.7)

Cigarette use N10 pack-years years No 456 (27.1) 734 (8.6) b0.001 Yes 20 (1.2) 2585 (30.4) Missing 455 (27.1) 919 (10.8) Not collected 1 750 (44.6) 4260 (50.1)

1 Data on smoking status were only collected for transplants during 1995–2004.

93A.K. Fink et al. / Journal of Cystic Fibrosis 16 (2017) 91–97

years among non-CF recipients. Compared with the general population, CF recipients had an almost 10-fold elevation in overall cancer risk (SIR = 9.9, 95% confidence interval (CI) = 7.9, 12.3) and non-CF recipients had a 2.7-fold elevation (SIR = 2.7, 95%CI = 2.5, 2.9). Among the cancer types evaluated, there were 24 with at least one case diagnosed among a CF lung transplant recipient. SIRs for these are displayed in Fig. 1, and observed counts and SIRs for all cancers are presented in Supplemental Table 1. The most common cancers among CF recipients were NHL (n = 34, 40% of the total) and colorectal cancer (n = 15, 18%). For non-CF recipients, the most common cancers were lung cancer (n = 215, 25%) and NHL (n = 128, 17%).

Compared to the general population, CF recipients had significantly elevated incidence for cancers of the lip (n = 3, SIR = 203, 95%CI = 41.8, 592), esophagus (n = 3, SIR = 56.3, 95%CI = 11.6, 165), colorectum (n = 15, SIR = 24.2, 95%CI = 13.5, 39.8), soft tissue (n = 3, SIR = 26.5, 95%CI = 5.5, 77.4), kidney and renal pelvis (n = 3, SIR = 11.0, 95%CI = 2.3, 32.3), and thyroid (n = 4, SIR = 6.2, 95%CI = 1.7, 16.0), as well as for NHL (n = 34, SIR = 61.8, 95%CI = 42.8, 86.3) and poorly specified tumors (n = 3, SIR = 28.8 , 95% CI = 5.9, 84.3). In addition, elevated risk was observed for the following cancers with only one observed case: intrahepatic bile duct (SIR = 130, 95%CI = 3.3, 727), other biliary tract excluding gall bladder (SIR = 65.5, 95%CI = 1.7, 365), acute monocytic leukemia (SIR = 178, 95%CI = 4.5, 991) and Kaposi sarcoma (SIR = 409, 95%CI = 10.3, 2280).

Downloaded for Anonymous User (n/a) at University of Toledo fr For personal use only. No other uses without permission. C

As indicated in Fig. 1 by lack of overlap in the confidence intervals, the SIRs in CF recipients were significantly higher than in non-CF recipients for esophageal cancer, colorectal cancer, and NHL. We further explored these cancers by combining esophageal cancer with other upper digestive tract cancers (consistent with previous studies) [11], and we subdivided colorectal cancers by subsite and NHLs by histologic subtype (Supplemental Table 1). Risk for upper digestive tract cancer was significantly higher in CF recipients compared to non-CF recipients (SIR = 27.5, 95%CI = 8.9, 64.1 vs. SIR = 2.3, 95%CI = 1.3, 3.6).

Risks for all subsites of the colorectum (proximal colon, distal colon and rectum) were significantly higher among CF recipients than in the general population, and the SIRs were also higher than among non-CF recipients (Fig. 2, Supplemental Table 1). Of the 15 cases of colorectal cancer in CF recipients, only 2 were in patients transplanted after age 50. The median age at colorectal cancer diagnosis was 43 years and the median time from transplant to colorectal cancer was 6.4 years. Among CF recipients, 6 colorectal cancers were diagnosed at local stage, 5 were regional stage, and 4 were distant stage. There was no apparent pattern between the timing of cancer diagnosis and cancer stage. Of eight CF recipients with colorectal cancer who died, 6 died from their cancer, and their survival time ranged from 0.3 to 4.4 years after cancer diagnosis.

Cumulative incidence of colorectal cancer in CF and non-CF recipients is displayed in Fig. 3. Among recipients under age 50 at transplant, those with CF had a five-year absolute risk of 0.3% of

om by Elsevier on September 16, 2018. opyright ©2018. Elsevier Inc. All rights reserved.








S ta

n d

a rd

iz e

d I

n ci

d e

n ce

R a

tio (



Non-CF Lung Recipients CF Lung Recipients

Fig. 1. Standardized incidence ratios for selected cancers among CF and non-CF lung recipients. The figure shows standardized incidence ratios (circles) and 95% confidence intervals (vertical lines) among lung recipients with and without cystic fibrosis (CF). Results are shown for cancer for which there was at least 1 observed case among recipients with CF. Cancers where the SIR for CF recipients is statistically significantly different are marked by a ★.

94 A.K. Fink et al. / Journal of Cystic Fibrosis 16 (2017) 91–97

developing colorectal cancer, compared with a zero risk among non-CF recipients. Among CF recipients transplanted when they were younger than age 50, the cumulative incidence of colorectal cancer continued to rise over time, reaching 1.6% at 10 years after transplant. Among individuals who received a transplant





Colorectum Colon R

S ta

n d

a rd

iz e

d I