SEARCH

SEARCH BY CITATION

Keywords:

  • Cancer;
  • epidemiology;
  • heart transplantation;
  • liver transplantation;
  • lung transplantation;
  • pediatric recipients

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

Population-based evidence on the relative risk of de novo cancer in liver and cardiothoracic transplant recipients is limited. A cohort study was conducted in Australia using population-based liver (n = 1926) and cardiothoracic (n = 2718) registries (1984–2006). Standardized incidence ratios (SIRs) were computed by cancer type, transplanted organ and recipient age. Cox regression models were used to compare cancer incidence by transplanted organ. During a median 5-year follow-up, the risk of any cancer in liver and cardiothoracic recipients was significantly elevated compared to the general population (n = 499; SIR = 2.62, 95%CI 2.40–2.86). An excess risk was observed for 16 cancer types, predominantly cancers with a viral etiology. The pattern of risk by cancer type was broadly similar for heart, lung and liver recipients, except for Merkel cell carcinoma (cardiothoracic only). Seventeen cancers (10 non-Hodgkin lymphomas), were observed in 415 pediatric recipients (SIR = 23.8, 95%CI 13.8–38.0). The adjusted hazard ratio for any cancer in all recipients was higher in heart compared to liver (1.29, 95%CI 1.03–1.63) and lung compared to liver (1.65, 95%CI 1.26–2.16). Understanding the factors responsible for the higher cancer incidence in cardiothoracic compared to liver recipients has the potential to lead to targeted cancer prevention strategies in this high-risk population.


Abbreviations
aHR,

adjusted hazard ratio

ANZLTR,

Australian and New Zealand Liver Transplant Registry

ANZCOTR,

Australian and New Zealand Cardiothoracic Organ Transplant Registry

ACD,

Australian Cancer Database

BCC,

basal cell carcinoma

EBV,

Epstein–Barr virus

HPV,

human papilloma virus

HR,

hazard ratio

IQR,

interquartile range

MCC,

Merkel cell carcinoma

NDI,

National Death Index

NHL,

non-Hodgkin lymphoma

NMSC,

nonmelanoma skin cancer

PSC,

primary sclerosing cholangitis

SCC,

squamous cell carcinoma

SIR,

standardized incidence ratio

UC,

ulcerative colitis.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

Liver and cardiothoracic transplantation are established procedures for patients with end-stage organ failure. In Australia, excellent outcomes are achieved, with 1-year graft survival rates of 80–88% [1, 2]. Cancer, however, has become the leading cause of death in liver and heart transplant recipients surviving for more than 5 years with a functioning graft [1-4]. Immunosuppression is the primary risk factor for cancer in the transplanted population, as reinforced by the remarkably similar cancer profile in those with HIV/AIDS [5].

Population-based cohort studies show a 2.5- to 3-fold excess risk of cancer after solid organ transplantation compared to the general population [5, 6]. As it is the most common form of transplantation, kidney recipients dominate this estimate, and the spectrum of cancer risk for recipients of other organs is less clear [7-9]. Non-population-based estimates suggest that the risk of non-Hodgkin lymphoma (NHL) is highest in lung transplant recipients, followed by heart and liver [10], but the age- and sex-adjusted relative risk of lymphoma and solid cancer between these transplanted organs has not been assessed. Furthermore, few studies have reported cancer incidence by indication for transplantation, and population-level data on cancer in pediatric recipients are scarce [11]. A comparison of cancer risk in different subsets of the transplanted population will provide insights into cancer etiology, and thus cancer prevention, in the context of immune suppression. Quantifying and understanding differences in cancer incidence between recipients of different transplanted organs will generate evidence-based clinical strategies for minimizing cancer risk, and for identifying high-risk patient subsets that would benefit most from these interventions.

We estimated the incidence of de novo cancer in Australian liver, heart and lung transplant recipients over a 23-year period. We computed cancer risk relative to the general population, and compared cancer risk for recipients of different transplanted organs.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

Study population

Our study population comprised Australian residents who received a liver or cardiothoracic transplant in Australia between 1984 and 2006. Transplant recipients were registered on the Australian and New Zealand Liver Transplant Registry (ANZLTR) or the Australia and New Zealand Cardiothoracic Organ Transplant Registry (ANZCOTR). These population-based registries recorded all liver and cardiothoracic transplantations since 1985 and 1984, respectively. They systematically recorded the name, date of birth, sex and primary indication for transplantation.

Non-Australian residents were excluded from the study population as they are not eligible for cancer registration in Australia. Patients with a history of cancer prior to transplantation (n = 367), including those whose indication for transplantation was a hepatobiliary tumor (n = 93), were not excluded [6, 7, 12, 13]. However, these patients did not contribute person-years at risk for that cancer type. A sensitivity analysis was performed to assess the impact of excluding patients with a history of cancer. In addition, all cancers and person-years follow-up time within 30 days of transplantation [8, 14-17] were not included in our analyses. This exclusion was necessary to avoid counting as incident those cancers that were prevalent at transplantation but not registered until they were histopathologically diagnosed in the explanted organ [13]. Patients who received a combined liver and kidney transplant (n = 23), and those who received a combined liver, heart and lung transplant (n = 3), were classified as liver transplant recipients; those who received a combined heart and lung transplant (n = 137) were classified as lung transplant recipients.

Data collection

Deaths and incident cancers were ascertained by record linkage with population-based administrative health datasets. Deaths were obtained from the National Death Index (NDI; 1980–2006). Cancers were identified from the Australian Cancer Database (ACD), a register of incident primary invasive neoplasms, other than basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) of the skin which are not ascertained by the Australian population cancer registries. The date of diagnosis, topography and morphology was ascertained for each primary neoplasm diagnosed in the cohort between 1982 and 2006. The Australian cancer registries apply international rules when registering multiple primary cancers. Solid cancers were classified according to the International Classification of Diseases (ICD), 10th revision, while hematopoietic neoplasms and Kaposi sarcomas were classified according to the ICD for Oncology, 3rd edition.

Registrant's name, sex, date of birth, date of death and state of residence were used during record linkage that was performed utilizing an established probabilistic algorithm. A linkage probability or weight was given to each potential matching record pair, and a subset of paired records underwent clerical review.

Cancer incidence rates for the Australian population were obtained from the Australian Cancer Database by 5-year age group, sex, calendar year and state, for 1982–2006.

Ethical approval was obtained and the requirement for informed participant consent was waived because the researchers received only de-identified data.

Data analysis

Cancer incidence

Person-years of follow-up accrued from 30 days posttransplantation until the date of cancer diagnosis, death, age 80, or December 31, 2006, whichever occurred first. Crude and age- and sex-standardized cancer incidence rates (ASR), standardized to the 1996 Australian population, and 95% confidence intervals (CIs; based on the normal approximation to the binomial distribution) were calculated using annual Australian population estimates obtained from the Australian Bureau of Statistics. These rates were computed for any cancer, solid cancers and lymphomas for pediatric recipients (0–15 years) and adult (16+ years) recipients, by transplanted organ.

Risk of cancer relative to the general population

Cancer incidence rates in transplant recipients were compared with the Australian general population using the standardized incidence ratio (SIR), defined as the ratio of the observed and the expected numbers of cancers. The likelihood ratio method was used to calculate 95% CIs for cancers with ≥10 expected cases, while exact CIs were used for cancers with <10 expected cases [18]. The expected numbers of incident cancers were calculated by multiplying cohort person-years at risk by the corresponding 5-year age-, sex-, state- and calendar year-specific cancer incidence rates for the Australian population. The exception was Kaposi sarcoma, where 1982 population rates were applied to avoid the impact of AIDS on the incidence of this cancer.

SIRs were computed for the entire cohort and by transplanted organ, recipient age at transplantation (pediatric: ≤15 years), current age (time dependent, 0–39, 40–55, ≥55 years) and primary indication for transplantation. We compared patterns of cancer risk for the different patient subgroups but could not compare SIRs statistically because of the heterogeneity in subgroup age and sex distributions [19].

Comparison of cancer risk by transplanted organ

Within the cohort of all transplant recipients, the risk of any cancer, solid cancers and lymphomas was estimated for pediatric recipients and all recipients. A marginal Cox model, the Wei–Lin–Weissfeld (WLW) model, was applied to enable inclusion of multiple cancers for a recipient [20]. Hazard ratios (HR) with 95% confidence intervals (CI) were calculated comparing the incidence of cancer by transplanted organ, adjusted for age at transplant, sex, multiple transplants (one, more than one; time-dependent variable) and calendar period (1984–1989, 1990–1997, 1998–2006). All variables other than the time-dependent covariate satisfied the proportional hazards assumption.

Analyses were performed using SAS® software v9.2 (SAS Institute Inc., Cary, NC, USA). Person-years were calculated using the%stratify macro [21] and the WLW marginal model was computed using the PHREG program.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

The eligible cohort comprised 4644 transplant recipients; 1926 (41%) liver, 1518 (33%) heart and 1200 (26%) lung (Table 1). Second or higher order transplants were received by 162 (4%) patients. Recipients were followed up for a total of 29 713 person-years. The median duration of follow-up was 5.2 years (interquartile range, IQR 2.0–9.9) and it was highest for heart and lowest for lung recipients. Approximately 9% of patients (n = 415) received a liver or heart transplant before the age of 15 years; no pediatric patients received a lung transplant. The median age at first transplantation was 47 years (IQR 33–55) and it was highest for lung and lowest for liver recipients. Overall, 66% of recipients were male, and the proportion of males by transplanted organ ranged from 53% to 80%. The most common primary indication for transplantation was viral hepatitis (26% of liver), nonischemic cardiomyopathy (46% of heart) and obstructive lung disease (33% of lung).

Table 1. Characteristics of recipients of liver and cardiothoracic transplants in Australia between 1982 and 2006
 LiveraHeartLungb
  Person-years Person-years Person-years
CharacteristicsNo. (%)TotalMediancNo. (%)TotalMediancNo. (%)TotalMedianc
  1. a

    Liver transplant (n = 1900), combined liver and kidney transplant (n = 23), or combined liver, heart and lung transplant (n = 3).

  2. b

    Lung transplant (n = 1063) or combined lung and heart transplant (n = 137).

  3. c

    Median and interquartile range;

  4. d

    Chronic obstructive pulmonary disease.

Total1926 (100)127036.0 (2.2–10)1518 (100)11 7197.1 (2.9–12)1200 (100)52923.3 (1.5–6.5)
Sex         
 Male1191 (61.8)70784.9 (1.8–8.9)1218 (80.2)9 4947.2 (3.1–12)641 (53.4)28273.3 (1.5–6.3)
 Female735 (38.2)56246.9 (2.7–12)300 (19.8)2 2256.4 (2.6–11)559 (46.6)24643.4 (1.4–6.5)
Age at first transplantation (years)
 0–9248 (13.0)19066.7 (2.7–13)39 (2.6)3248.5 (3.6–13)1 (0.1)0.5
 10–19104 (5.4)9509.5 (3.8–14)88 (5.8)6395.8 (2.0–12)72 (6.0)3303.1 (1.3–6.5)
 20–2993 (4.8)7357.1 (3.1–12)111 (7.3)7935.7 (2.8–10)212 (17.7)10033.6 (1.5–7.5)
 30–39174 (9.0)14508.2 (3.7–13)168 (11.0)1 4578.2 (2.8–14)189 (15.7)8523.1 (1.4–7.1)
 40–49540 (28.0)33715.3 (2.1–9.4)337 (22.2)2 9208.2 (3.9–13)245 (20.4)13064.4 (1.8–7.7)
 50–59591 (30.7)34674.8 (1.7–9.0)587 (38.7)4 5177.5 (3.1–12)385 (32.1)15123.1 (1.2–6.0)
 ≥60176 (9.1)8243.6 (1.5–6.9)188 (12)1 0674.8 (1.7–8.7)96 (8.0)2882.2 (1.2–4.3)
Primary indication for transplantation
 Nonischemic cardiomyopathy  692 (45.6)5 5217.5 (2.9–12)  
 Ischemic heart disease  582 (38.3)4 4426.9 (3.2–12)  
 COPDd    349 (29.0)14463.5 (1.6–6.0)
 Obstructive lung disease    398 (33.2)17193.2 (1.5–6.4)
 Congenital heart disease  76 (5)6057.9 (2.8–13)54 (4.5)3595.3 (2.1–12)
 Viral hepatitis494 (25.6)25294.1 (1.7–7.9)    
 Hepatobiliary tumor93 (4.8)3502.3 (0.8–5.0)    
 Autoimmune-related liver disease442 (23)33747.1 (2.6–12)    
 Alcoholic liver disease223 (11.6)14065.8 (2.6–9.2)    
 Congenital biliary disease189 (9.8)15717.7 (2.9–13)    
 Miscellaneous485 (25.5)34726.3 (2.5–11)168 (11.0)1 1505.7 (2.5–11)399 (33.2)17673.0 (1.2–6.8)

After transplantation we observed 499 (10.7%) incident primary cancers in 463 patients; 33 patients had multiple cancers. The median age at diagnosis of first cancer was 57 years (IQR 49–63). The most frequently occurring cancers were NHL (n = 100), lip cancer (n = 58) and cutaneous melanoma (n = 50). The crude overall cancer incidence rate was 1679 per 100 000 person-years and the ASR was 1693 per 100 000 (95% CI 832–2553). Rates for all cancers, solid cancers and lymphomas for pediatric and adult recipients by transplanted organ are shown in Table 2.

Table 2. Cancer incidence in Australian pediatric and adult liver and cardiothoracic transplant recipients
 Pediatric recipients (0–15 years)aAdult recipients (≥16 years)
  Liver, heart, or lungLiverHeartLungb
  1. a

    Liver and heart transplantations only.

  2. b

    Lung or combined heart and lung transplantation.

  3. c

    Excluding BCC and SCC of the skin.

  4. d

    Includes non-Hodgkin lymphoma, Hodgkin lymphoma and lymphoma not otherwise specified.

Crude incidence rate per 100 000
 Any cancerc5161825140421951873
 Solid cancersc1521431114017491332
 Lymphomad364397273446541
Age- and sex-standardized incidence rate (95% CI)
 Any cancerc468 (233–702)2060 (944–3176)1149 (889–1409)3038 (1534–4542)2268 (1371–3165)
 Solid cancersc156 (14–299)1701 (588–2814)931 (696–1166)2508 (1022–3995)1714 (860–2569)
 Lymphomad311 (124–498)361 (272–450)222 (111–334)530 (299–761)553 (271–836)
Standardized incidence ratio (95% CI)
 Any cancerc23.8 (13.8–38.0)2.54 (2.32–2.78)2.29 (1.94–2.69)2.69 (2.36–3.04)4.28 (3.49–5.19)
 Solid cancersc51.3 (16.6–120)2.26 (2.03–2.49)2.04 (1.69–2.43)2.35 (2.04–2.70)3.36 (2.63–4.22)
 Lymphomad88.5 (45.7–154)7.82 (6.33–9.53)5.60 (3.59–8.33)6.99 (5.02–9.48)16.8 (11.1–24.4)

Cancer risk relative to the general population: all transplant recipients

Risk of any cancer was significantly higher in transplant recipients than the matched Australian population (SIR = 2.62, 95% CI 2.40–2.86). SIRs were greater than unity for 16 different cancers (Figure 1), and relative risk was highest for Kaposi sarcoma, then cancer of the vulva, lip, nonmelanoma skin cancer (NMSC; excluding BCC and SCC) and salivary gland, followed by NHL. Most (n = 6, 75%) salivary gland cancers were squamous cell carcinomas. Risk of some epithelial cancers common in the general population, including prostate, breast and pancreas, was not significantly elevated. Risk of all HPV-related anogenital cancers (cervix, vulva, vagina, anus and penis) was increased (SIR = 6.03, 95% CI 3.12–10.5), as was the risk of all cancers causally related to alcohol consumption (oral cavity, pharynx, esophagus, colorectum, liver, larynx and breast) [22] (SIR = 1.49, 95% CI 1.18–1.86).

image

Figure 1. Site-specific cancer risk for Australian liver and cardiothoracic transplant recipients relative to the general population. 1NMSC, excluding BCC and SCC. 2Myeloid neoplasms including lymphoid/myeloid not otherwise specified.

Download figure to PowerPoint

Cancer risk relative to the general population: by transplanted organ

Relative to the general population, the risk of any cancer was more than twofold for liver (SIR = 2.20, 95% CI 1.87–2.57) and heart (SIR = 2.64, 95% CI 2.32–2.98) recipients, and more than threefold for lung recipients (SIR = 3.70, 95% CI 3.01–4.48). The risk of lip cancer, NHL and cancer of unknown primary site were significantly elevated for recipients of all three transplanted organs (Supporting Table 1). When recipients with a history of cancer were excluded from the cohort the key findings were unchanged.

The pattern of cancer risk by transplanted organ was similar, with some noteworthy exceptions (Figure 2). Risk of NMSC (other than BCC and SCC) was significantly elevated in both heart (SIR = 32.8, 95% CI 17.9–55.0) and lung (SIR = 61.4, 95% CI 24.7–126) recipients, but there were no incident cases in liver recipients. Seventeen of the 21 skin cancers in cardiothoracic patients were Merkel cell carcinoma (MCC; SIR = 103, 95% CI 60.4–166). In contrast, melanoma risk was significantly increased in liver (SIR = 2.13, 95% CI 1.22–3.46) and heart (SIR = 3.04, 95% CI 2.03–4.36) but not lung (SIR = 1.64, 95% CI 0.53–3.83) recipients. Colorectal cancer risk was significantly raised in liver (SIR = 2.40, 95% CI 1.49–3.68) and lung (SIR = 2.58, 95% CI 1.12–5.09) recipients, but not in heart (SIR = 0.99, 95% CI 0.54–1.63). Only heart and lung transplant recipients exhibited significantly elevated risk of lung cancer (SIR = 2.18, 95% CI 1.39–3.22 and SIR = 3.82, 95% CI 1.65–3.53, respectively).

image

Figure 2. Site-specific cancer risk by transplanted organ relative to the general population. 1NMSC, excluding BCC and SCC of skin. 2Myeloid neoplasms including lymphoid/myeloid not otherwise specified.

Download figure to PowerPoint

Cancer risk relative to the general population: by recipient age

Seventeen cases of cancer, including 10 NHL, were observed in 415 pediatric recipients, resulting in SIRs exceeding 20 and 70 for any cancer and lymphoma, respectively (Table 2). Cancer of the vulva, colorectum and breast, as well as NHL and Hodgkin lymphoma, all occurred at significantly increased risk. The median age at diagnosis was 7 years (IQR 6–17) for lymphoma and 30 years (IQR 16–32) for solid cancer.

When considering current or attained age for all recipients combined, a significant excess risk of any cancer, lip cancer, NHL and melanoma was observed for all age groups (Table 3). Colorectal cancer risk was significantly increased in the 0–39 year and 40–55 year age groups and was of borderline significance in the >55 year age group. A significantly elevated risk of skin and lung cancer was observed only for those aged at least 55 years, and risk of cancer of unknown primary site was increased for those more than 40 years of age.

Table 3. Cancer risk by current age for the most frequent incident cancers in Australian recipients of liver and cardiothoracic transplants
 Age 0–39 yearsAge 40–55 yearsAge >55 years
Cancer typeObsExpSIR95% CIObsExpSIR95% CIObsExpSIR95% CI
  1. Obs, observed number of cancers; Exp, expected number of cancers; SIR, standardized incidence ratio.

  2. a

    Excluding BCC and SCC of the skin.

  3. b

    NMSC, excluding BCC and SCC of the skin.

  4. c

    Excluding anal cancer.

All cancersa595.1411.58.74–14.813136.73.572.99–4.223101482.091.87–2.33
 Lip20.1315.91.92–57.3240.6835.422.7–52.7321.5320.914.3–29.6
 Skinb10.0423.70.60–13210.185.490.14–30.6190.6330.218.2–47.2
 Non-Hodgkin lymphoma240.4256.836.4–84.5382.5514.910.5–20.5388.924.263.01–5.84
 Unknown primary site10.0518.40.47–10350.806.222.02–14.5193.755.073.05–7.91
 Melanoma71.295.422.18–11.2145.772.421.33–4.072913.02.231.51–3.14
 Lung0---52.462.030.66–4.742815.81.771.19–2.50
 Colorectalc30.2114.12.91–41.294.112.191.00–4.163020.71.450.99–2.03
 Prostate012.630.380.01–2.123334.70.950.66–1.31

Cancer risk relative to the general population: by indication for transplantation

For liver and cardiothoracic recipients, cancer risk was significantly increased regardless of the indication for transplantation, except for those receiving a liver on account of a hepatobiliary tumor (n = 93; Table 4). After liver transplantation, colorectal cancer risk was increased in those with autoimmune-related liver disease (n = 12; SIR = 4.49, 95% CI 2.32–7.84) but not in those without (SIR = 1.48, 95% CI 0.68–2.82). Similarly, it was increased in those with primary sclerosing cholangitis (PSC, n = 10; SIR = 9.58, 95% CI 4.59–17.6) but not in those without (SIR = 1.43, 95% CI 0.71–2.56), and in those with PSC and ulcerative colitis (UC) (n = 5; SIR = 12.5, 95% CI 4.06–29.1), but not in those without (SIR = 1.32, 95% CI 0.63–2.42).

Table 4. Cancer risk by indication for transplantation in Australian recipients of liver or cardiothoracic transplants
  All cancersa
Transplanted organ(s)Primary indicationObsExpSIR95% CI
  1. Obs, observed number of cancers; Exp, expected number of cancers; SIR, standardized incidence ratio.

  2. a

    Excluding BCC and SCC of the skin.

LiverViral hepatitis3016.81.791.22–2.50
 Hepatobiliary tumor22.130.940.11–3.39
 Autoimmune-related liver disease6021.42.802.15–3.56
 Alcoholic liver disease2412.61.911.24–2.77
 Congenital biliary disease40.517.832.13–20.0
 Miscellaneous liver disease3316.02.061.44–2.85
CardiothoracicNon-ischemic cardiomyopathy10536.42.882.36–3.47
 Ischemic heart disease11650.02.321.91–2.78
 Congenital heart disease122.115.673.04–9.51
 Obstructive lung disease233.626.354.03–9.53
 Chronic obstructive pulmonary disease3512.62.781.96–3.81
 Miscellaneous cardiothoracic disease5515.93.462.63–4.46

Comparison of cancer risk by transplanted organ

Pediatric heart recipients were at higher risk of any cancer (aHR 3.10, 95% CI 1.01–9.47) and lymphoma (aHR 6.67, 95% CI 2.37–18.8) compared to liver recipients. For all transplant recipients, the risk of any cancer and of lymphoma was significantly greater for heart and lung recipients compared to liver recipients (Table 5), whereas the risk of solid cancer was only increased in lung compared to liver recipients. Solid cancer risk was also increased in lung compared to heart transplant recipients (aHR 1.41, 95% CI 1.03–1.93); there was no significant difference in the risk of any cancer (aHR 1.28, 95% CI 0.98–1.69) or lymphoma (aHR 1.11, 95% CI 0.65–1.89).

Table 5. Risk factors for cancer in Australian recipients of liver and cardiothoracic transplants
 Adjusted hazard ratio (95% CI)a
 All cancersAll solid cancersAll lymphomas
  1. a

    Adjusted for age at transplant (single years), sex, number of transplants (as a time-dependent covariate) and calendar year at first transplantation.

  2. b

    Time dependent; yes from second transplant.

Age at first transplantation (single years)1.04 (1.03–1.05)1.05 (1.04–1.06)1.00 (0.98–1.01)
Sex
 Male1.00 (ref)1.00 (ref)1.00 (ref)
 Female0.80 (0.63–1.00)0.74 (0.56–0.97)0.99 (0.63–1.56)
Number of transplantsb
 One1.00 (ref)1.00 (ref)1.00 (ref)
 More than one0.78 (0.36–1.69)0.56 (0.21–1.50)1.09 (0.34–3.45)
Calendar period of first transplantation
 1984–19890.96 (0.72–1.29)1.15 (0.83–1.59)0.51 (0.25–1.07)
 1990–19971.00 (ref)1.00 (ref)1.00 (ref)
 1998–20060.73 (0.57–0.94)0.73 (0.55–0.97)0.69 (0.42–1.12)
Transplanted organ
 Liver1.00 (ref)1.00 (ref)1.00 (ref)
 Heart1.29 (1.03–1.63)1.15 (0.88–1.49)1.89 (1.14–3.14)
 Lung1.66 (1.27–2.17)1.61 (1.17–2.21)2.10 (1.25–3.54)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

In this population-based study we found a significantly higher risk of any de novo cancer for both heart and lung transplant recipients compared to liver transplant recipients after adjusting for age, sex, multiple transplantations and calendar period of transplantation. An increased risk was also observed for lymphomas and solid cancers, but the excess risk of solid cancers was restricted to lung recipients compared to both liver and heart recipients. Understanding the factors responsible for the greater cancer incidence in heart and lung as compared to liver transplant recipients has the potential to lead to targeted cancer prevention strategies. Compared to the general population matched for age, sex, calendar year and state of residence, the risk of cancer was also increased, from 2.2-fold (liver), to 2.6-fold (heart) and 3.7-fold (lung). The pattern of site-specific risk by transplanted organ was broadly similar, confirming the critical role of immunosuppression in posttransplantation cancer risk. Exceptions to this pattern also suggest patient subgroups at high risk for specific cancer types.

Population-based evidence on the relative risk of cancer in recipients of different types of solid organ is limited. A Swedish study showed an increased risk of NHL for liver, heart and lung transplant recipients relative to kidney transplant recipients, after adjustment for age, sex, year and follow-up time [23]. Our twofold higher risk of lymphoma in heart and in lung recipients compared to liver recipients is in line with findings from the Collaborative Transplant Study [10]. Our study presents the first population-based evidence of an excess risk of any cancer in heart and lung recipients compared to liver recipients, and a significant increased risk of solid cancer in lung compared to heart and liver transplant recipients.

The key explanation for the difference in cancer risk by transplanted organ is variation in the intensity or type of immunosuppression [15, 24, 25]. Several studies have found an association between posttransplantation cancer risk and receipt of induction therapy with lymphocyte depleting antibodies or maintenance immunosuppression with specific agents [10, 26-30]. There are no published data directly comparing the dose and type of immunosuppressive agents for liver, heart and lung transplant recipients in Australia. However, Australian clinical transplantation practice has generally followed international trends, with a lower prevalence of induction therapies and lower overall immunosuppressive dose for liver transplant recipients compared to heart and lung recipients [15, 31]. In addition to differences in the extent and type of immunosuppression by transplanted organ, other factors may also play a role, either independent of or interacting with immunosuppression. These include patient factors, such as prevalent or acquired infection by carcinogenic agents, autoimmune disease, carcinogenic behaviors and genetic predisposition to cancer, as well as inherent biological differences in the transplanted tissue. A greater volume of lymphoid tissue in the lung compared to other organs, and thus greater potential for the transmission of donor lymphocytes infected with Epstein–Barr virus (EBV), has been suggested to explain the higher NHL risk in lung compared to liver transplant recipients [32].

When the liver and cardiothoracic transplant recipients were considered together, our data agree with prior evidence showing a wide-ranging excess cancer risk relative to the general population [5, 6], especially cancers with a viral cause. The pattern of site-specific cancer risk by transplanted organ was broadly similar, and also largely consistent with prior population-based evidence for liver [7, 9, 12, 15, 33-36] and heart [8, 15, 37] transplantation. Our study adds to existing evidence showing an increased risk of NHL, Kaposi sarcoma, colorectal cancer, lip cancer and cancer of unknown primary site after liver transplantation. Novel findings were an excess risk of cutaneous melanoma and cancer of the thyroid, vulva, anus and salivary gland. We did not confirm previously published findings of an excess risk of cancer of the liver [6], lung [12, 15, 33], oral cavity [15] or kidney [34] after liver transplantation.

Our estimate for colorectal cancer risk after liver transplantation (SIR = 2.40, 95% CI 1.49–3.68) is similar to a meta-estimate for prior population-based studies (SIR = 2.6, 95% CI 1.7–4.1) [38]. PSC is an indication for liver transplantation, and 60–80% of individuals with PSC also have inflammatory bowel disease, predominantly UC. It is established that patients with UC are at high risk of colorectal cancer [39, 40], and findings from a single retrospective study suggest that transplantation may not alter this inherently high risk [41]. In our study, an excess risk of colorectal cancer was confined to liver transplant recipients with a history of PSC and UC. These data support the use of screening in these patients to enable the early diagnosis of UC and colorectal adenomas.

Consistent with prior population-based studies [6, 8, 15], we identified an excess risk of NHL, lip cancer and lung cancer in heart transplant recipients. However, we did not observe an increased risk of kidney [6, 8, 15], oral cancer [8, 15] or multiple myeloma [8, 15]. We quantified the excess risk of cutaneous melanoma in Australian heart transplant recipients [42], and we also identified novel associations; an elevated risk of MCC, myeloid neoplasms and cancer of the salivary gland, eye and unknown primary site. The increased risk of MCC, a neuroendocrine skin cancer possibly associated with infection by Merkel cell polyomavirus (MCPyV) [43], is in broad agreement with the only prior estimate posttransplantation, an SIR of 66 for Finnish kidney transplant recipients [44]. The recent report of an excess risk of Merkel cell carcinoma in US HIV/AIDS patients (SIR = 11, 95% CI 6.3–17) [45] supports an association between risk of this cancer and immunosuppression. However, the absence of MCCs in our cohort of liver transplant recipients suggests that MCC risk may not be related to immunosuppression per se, and that the intensity or types of immunosuppression or other factors are likely to be important.

In our study, lung transplantation was associated with an increased risk of NHL, MCC and cancer of the vulva, lip, lung, colorectum and unknown primary site. Our findings thus confirm prior population-based evidence of an excess risk of NHL and lung cancer [6, 15], but not anal cancer [15].

There are scarce data on cancer risk in pediatric liver and heart transplant recipients [7, 11]. We found an increased risk of NHL, skin cancer and vulvar cancer, concurring with a Swedish pediatric cohort consisting mostly of kidney transplant recipients [11]. However, the excess risk of colorectal cancer, breast cancer and Hodgkin lymphoma we observed have not previously been reported, and require validation in larger cohorts. Notably, most solid cancers occurred in adulthood, highlighting the need for increased clinical surveillance during this phase of life. The striking increased risk of NHL in childhood has consistently been observed [8, 46] and has been associated with EBV seroconversion [47] and intensity of immunosuppression [48]. As a result, antiviral prophylaxis and targeted monitoring of EBV viral load in peripheral blood is recommended for high-risk pediatric patients [49].

Several potential limitations must be considered when interpreting our findings. The risk of basal cell and squamous cell skin cancer could not be estimated because these neoplasms are not recorded by all Australian cancer registries; consequently, our estimates for ‘any cancer’ risk exclude these cancers. In addition, it is possible that a small number of cancers were donor derived. Moreover, this study relied upon record linkage between transplant registries and routinely collected administrative data, and some false positive and negative linkages do occur. Nevertheless, the record linkage is highly sensitive and specific [50], and linkage errors are unlikely to occur differentially by transplanted organ or cancer type. On the other hand, a potential bias would arise if there was differential participation in national cancer screening programs (breast, cervical and colorectal cancer) by transplant recipients compared to the general population, or by transplanted organ. While there is some evidence that Australian kidney transplant recipients undergo more screening [51], with the exception of patients with PSC, there is no data on screening rates for liver, heart and lung transplant recipients. Furthermore, surveillance bias is unlikely to explain differences between transplanted organs in the risk of other solid cancers or of lymphomas. While linkage accuracy is known, there has been no formal validation of linkage completeness, and we were unable to censor upon migration from Australia. Nevertheless, given the high quality of Australian cancer registries, we expect to have identified most incident cancers. Finally, we did not adjust the p-value for statistical significance (p < 0.05) to take into account multiple statistical tests, and thus it is possible that some of our findings may be due to chance.

The key strengths of this study are the population basis for inclusion of transplant recipients and for ascertaining deaths and cancers. The use of identical methods for transplant recipients, the general population, and for the different transplanted organs, enabled unbiased comparison of risk and also minimized the influence of selection bias and loss to follow-up. The study included site-specific cancer risk estimates for the largest population-based series of nonkidney pediatric transplant recipients published to date. The relatively large population size also allowed the risk of some rare cancers, such as MCC, to be estimated. Finally, the systematic recording of indication for transplantation by the transplant registries allowed insight into patient subgroups at high risk.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

This study found evidence of a higher risk of cancer in heart and lung compared to liver transplant recipients in Australia. Understanding the factors responsible for these associations is expected to lead to strategies that reduce the cancer burden facing this high-risk patient group. Knowledge of the cancer profile by transplanted organ and patient age will facilitate early detection and improve patient outcomes.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

This study was funded by the National Health and Medical Research Council (ID510254). A.G. is supported by an NHMRC principal research fellowship (ID568819). C.V. is supported by a National Health and Medical Research Council Career Development Fellowship (ID1023159) and a Cancer Institute New South Wales Career Development Fellowship (ID10/CDF/242). R.N. is supported by a Translational Cancer Research Network PhD scholarship top-up award.

We thank the Australian and New Zealand Liver Transplant Registry, the Australian and New Zealand Cardiothoracic Organ Transplant Registry, and also the Australian Institute of Health and Welfare for conducting the data linkage.

We thank: Phyllis Larkins (Royal Prince Alfred Hospital, Sydney); Geraldine Lipka (Princess Alexandra Hospital, Brisbane); Cassandra Kastaneas (St Vincent's Hospital, Sydney); Vicki Jermyn and Brooke Andersen (The Children's Hospital, Westmead); Jo Maddicks-Law, Nicole Ostenfeld, Sara Gray and Muhtashimuddin Ahmed (Prince Charles Hospital, Brisbane); Kerrie Beale (Royal Childrens Hospital, Brisbane); Libby John, Nicole Williams (Flinders Medical Centre, Adelaide); Kathryn Marshall, Ailsa Cowie, Connie Kambanaros, Jasmin Board and Colleen Farrell (Alfred Hospital, Melbourne); Lyn Crellin, Kathe Beyerle, Kate Schurmann, Anne Shipp, Janette McEwan, Danielle Kamolins, Angie Wood and Hollie Gilmore (Royal Childrens Hospital, Melbourne); Julie Pavlovic and Betheia Lele (Austin Hospital, Melbourne); Sharon Lawrence, Clare Wood and Sharlene Beinke (Royal Perth Hospital); Barb Chester, Judith Bull, Joanne Plummer, Nikki Copland and Megan O-Dea (Sir Charles Gairdner Hospital, Perth).

The following lead investigators participated in this study: George Alex, Glenda Balderson, Peter Bergin, John Chen, Weng Chin, Lawrence Dembo, Pamela Dilworth, Looi Ee, Allan Glanville, Winita Hardikar, Peter Hopkins, George Javorsky, Garry Jeffrey, Robert Jones, Bronwyn Levvey, Steven Lynch, Michael Musk, Ross Pettersson, Greg Snell, Michael Stormon and Robert Weintraub.

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. ACKNOWLEDGMENTS
  9. Disclosure
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
ajt4302-sup-0001-TableS1.doc76KTable S1: Site-specific SIRs for Australian transplant recipients by transplant type (Figure 2)

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.