Objectives: In the Asian population, patterns and risk factors for de novo malignancies after solid-organ transplant are not well understood.
Materials and Methods: Insurance claims from Taiwan’s National Health Institute Research Database from 1997 to 2011 revealed 687 deceased-donor heart transplant recipients, 5038 kidney transplant recipients (50% living related-donor, 50% deceased-donor transplants), and 2127 liver transplant recipients (mainly living related-donor transplants, 30% deceased-donor transplants). During the follow-up period, rates of malignancy incidence were calculated with standardization based on national age, sex, and year-specific incidence. We used multivariate regression analyses to determine risk factors of posttransplant de novo malignancies.
Results: Compared with the general population, several de novo cancers were more common posttransplant (P < .05): lung cancer (2.6-fold), non-melanoma skin cancer (5.8-fold), and non-Hodgkin lymphoma (5.4-fold) in heart recipients; transitional cell carcinoma (31.4-fold), renal cell carcinoma (37.3-fold), and non-Hodgkin lymphoma (3.6-fold) in kidney recipients; and gastric cancer (3.0-fold) and lymphatic-hematopoietic malignancy (4.5-fold) in liver recipients. Independent risk factors for posttransplant malignancy in kidney transplant recipients were increased age, female, hepatitis B virus, and mycophenolate use (adjusted hazard ratio 1.5; 95% confidence interval, 1.2-1.8; P < .001). In liver transplant recipients, old age was an independent risk factor. Kidney transplant recipients without diabetes or hypertension had higher risk of transitional cell carcinoma (adjusted hazard ratio 3.0; 95% confidence interval, 2.1-4.4; P < .001) and renal cell carcinoma (adjusted hazard ratio 1.9; 95% confidence interval, 1.1-3.3; P < .05).
Conclusions: Regional endemic epidemiologic factors play significant roles in the development of de novo cancers, particularly in kidney transplant recipients due to causes of renal failure other than diabetes and hypertension. Each regional organ transplant program should tailor and establish its surveillance protocol based on epidemiologic data. However, the type and intensity of surveillance require further and long-term investigations in this patient cohort.
Key words : Asian population, Cancer, Posttransplant, Solid organ, Survival
Although solid-organ transplant has been an effective salvage therapy for end-stage solid-organ diseases in past decades,1 the development of posttransplant de novo malignancies has become a significant cause of late mortality among various types of transplant recipients.2 Posttransplant de novo malignancies are caused by complex interactions among genetic factors, immune status, environmental factors, and oncologic viruses.3 Therefore, to improve posttransplant outcomes, specific surveillance protocols for posttransplant de novo malignancies should be individually tailored based on the endemic epidemiologic data.
Previous population-based epidemiologic studies from registry data, obtained from Western populations, have revealed that skin cancer, basal cell carcinoma, Kaposi sarcoma, and posttransplant lymphoproliferative disorder are the most common posttransplant de novo malignancies among all types of solid-organ transplant recipients.4-7 However, in Asian populations, studies on posttransplant de novo malignancies among different types of solid-organ transplant recipients are limited, and these limited studies have only focused on specific transplant cohorts, such as heart,8 kidney,9-12 and liver transplant recipients.13,14 More importantly, most of the previous studies were mainly designed to investigate posttransplant malignancies but not de novo malignancies, which theoretically should occur after a longer incubation period, so that interactions among immunosuppressive therapy, biological factors, and environmental carcinogens could be better observed. In addition, influences of immunosuppressive therapy and other risk factors on posttransplant de novo malignancies have not been previously well investigated in Asian populations.
Therefore, using a Taiwan nationwide register database, we conducted this study to investigate specific patterns of posttransplant de novo malignancies among heart, kidney, and liver transplant recipients in a Taiwanese population. Risk factors of posttransplant de novo malignancies, including immunosuppressive therapy, were also investigated.
Materials and Methods
Source of data
Research data were obtained from reimbursement claims of the Taiwan National Health Insurance Program, which was implemented in March 1995 and which mandates health insurance and covers more than 99% of the 22.6 million Taiwanese residents. The National Health Insurance Research Database was established to collect all national insurance beneficiary inpatient and outpatient demographics, registered diagnoses, procedures, prescriptions, and medical expenditures for public research interest. Research articles that have utilized this database as a data resource have been published in prominent scientific journals worldwide.15-17 To protect patient privacy, the electronic database was deidentified, with patient identifications scrambled for further public access for research. This study was evaluated and approved by the committee of Taiwan’s National Health Research Institutes (NHIRD-103-121) and conducted in accordance with the Helsinki Declaration.
Using the National Health Insurance Research Database, we identified patients who underwent heart, kidney, or liver transplant procedures between January 1, 1997, and December 31, 2011. The population included 687 deceased-donor heart transplant recipients, 5038 kidney transplant recipients, and 2127 liver transplant recipients. Most liver transplants were from living related donors, with 30% having transplants from deceased donors. In the kidney transplant group, 50% were from deceased donors and 50% were from living related donors. Unrelated living transplants are not permitted in Taiwan. All recipients were followed until diagnosis of de novo malignancy, patient death, or end of the study period. Because the primary interest of our study was posttransplant de novo malignancies, patients who were lost during follow-up or were diagnosed with posttransplant malignancies within 1 year after transplant were not included to avoid pretransplant undiagnosed occult malignancies. In addition, to exclude recurrent hepatocellular carcinoma in liver transplant recipients, all recipients with pretransplant diagnosed hepatocellular carcinoma were not counted toward a posttransplant diagnoses of de novo hepatocellular carcinoma.
Comorbidities of transplant recipients were identified by registered ICD-9-CM diagnosis codes from 2 years before transplant to end of follow-up. Chronic hepatitis B or C virus infections, hypertension, diabetes mellitus, chronic obstructive pulmonary disease, liver cirrhosis, myocardial infarction, stroke, polycystic kidney disease, glomerulonephritis, and obesity were included. In addition, records regarding posttransplant immunosuppressive therapy were also retrieved. Use of thymoglobulin was introduced in Taiwan in 1995 and was identified when recipients received it at least once. Maintenance immunosuppression included calcineurin inhibitors (tacrolimus: introduced in 1997; cyclosporine: introduced in 1995), antimetabolites (mycophenolate: introduced in 2003; azathioprine: introduced in 1995), mammalian target of rapamycin (mTOR) inhibitors (everolimus: introduced in 2008; sirolimus: introduced in 2001), and steroids. Patients who received tacrolimus and cyclosporine as maintenance immunosuppressants were included in the study when they took prescriptions more than 90 days within the first year after transplant. The similar categorizing rule was also utilized to identify recipients who took antimetabolites and mTOR inhibitors.
Cumulative incidence and overall survival
The date of the event was defined as the date of de novo malignancy being diagnosed, and cumulative incidence was calculated among heart, kidney, and liver transplant recipients. The observation period started on the date of solid-organ transplant and continued until the end of 2011, the patient’s death, or withdrawal from National Health Insurance coverage. Overall survival rates were compared among heart, kidney, and liver transplant patients with and without posttransplant de novo malignancies.
Standardized incidence ratio
Posttransplant de novo malignancies were identified based on ICD-9-CM diagnosis codes that were newly registered more than 1 year after transplant with the exception of hepatocellular carcinoma, which was not counted in liver transplant recipients who were diagnosed with hepatocellular carcinoma before transplant. A standardized incidence ratio (SIR) was calculated to determine standardized risks of organ-specific posttransplant de novo malignancies among heart, kidney, and liver recipients. De novo cancers in bladder and kidney were further subcategorized into transitional cell carcinoma (ICD-9-CM code 188 or 189, except 189.0) and renal cell carcinoma (ICD-9-CM code 189.0).
Multivariate analyses were performed to determine risk factors of posttransplant de novo malignancies in heart, kidney, and liver transplant recipients. Furthermore, a subpopulation analysis was performed to determine risk factors in association with specific posttransplant de novo malignancies whose SIR results were extremely high in a previous SIR analyses.
Kaplan-Meier survival analyses and log-rank tests were performed to evaluate and compare survival outcomes. The SIR was calculated as the number of observed cases of organ-specific malignancy in the transplant population divided by the expected number of corresponding malignancies based on age-specific, sex-specific, and year-specific incidences of various malignancies in the Taiwan general population multiplied by the number of person-years at risk. The incidences of malignancies in the Taiwan general population were obtained from the Catastrophic Illness Patient Dataset, a subset of the National Health Insurance Research Dataset.17 We estimated 95% confidence intervals (95% CIs) of the SIR by Poisson regression.
Multivariate analyses were performed using Cox proportional hazard regression models and adjusted for statistically significant covariates obtained from univariate analyses. P < .05 was considered to indicate statistical significance; all tests were 2-tailed. All analyses were carried out with the SAS statistical package (version 9.4, SAS Institute, Inc., Cary, NC, USA).
Characteristics of heart, kidney, and liver transplant recipients are shown in Table 1. Between January 1997 and December 2011, there were 7852 patients who received a heart (n = 687), kidney (n = 5038), or liver transplant (n = 2127). The transplant recipient population comprised 60.3% male patients, and 62.1% of patients were less than 50 years old. Among comorbidities, liver-specific comorbidities, including chronic viral hepatitis B (57.0%), viral hepatitis C (23.4%), and liver cirrhosis (94.9%), were highest in liver recipients. In kidney transplant recipients, 42.3% were affected by glomerulonephritis, with 85.3% of these recipients diagnosed with hypertension. Myocardial infarction (25.8%), diabetes (44.0%), and stroke (22.1%) were highest in heart transplant recipients.
Approximately 79% of heart transplant recipients used thymoglobulin; in contrast, less than 5% of kidney or liver transplant recipients used this agent. With regard to calcineurin inhibitors, most heart transplant recipients (64.8%) used cyclosporine but most kidney (53.4%) and liver (77.3%) transplant recipients used tacrolimus. With regard to use of antimetabolites, 65% of all transplant recipients took long-term mycophenolate, with 16% of heart, 4% of kidney, and 1% of liver transplant recipients using azathioprine. Liver transplant recipients (48.9%) had the lowest proportion of long-term steroid use versus that shown in heart (92.9%) and kidney (93.7%) transplant recipients. During the study period, mTOR inhibitors were not commonly used in heart, kidney, and liver transplant recipients. The overall follow-up time from solid-organ transplant to occurrence of de novo malignancy or censoring was 4131 person-years in heart transplant recipients, 34 754 person-years in kidney transplant recipients, and 9197 person-years in liver transplant recipients.
Survival and cumulative incidence
Heart transplant recipients had 2-, 4-, 6-, and 8-year overall survival rates of 94.1%, 85.5%, 77.5%, and 70.9%. These rates were 98.5%, 95.3%, 91.6%, and 87.3%, respectively, in kidney transplant recipients and 95.6%, 89.1%, 85.1%, and 82.1%, respectively, in liver transplant recipients (Figure 1). Transplant recipients with posttransplant de novo malignancies had significantly poorer overall survival than those without malignancies regardless of type of transplant (Figure 2). Heart transplant recipients had the lowest cumulative incidence of posttransplant de novo malignancy, with 2-, 4-, 6-, and 8-year cumulative incidence rates of 0.8%, 2.5%, 3.6% and 5.1%, respectively. These rates were 1.4%, 3.7%, 4.8%, and 6.4%, respectively, in liver transplant recipients and 1.6%, 5.0%, 8.3%, and 11.2%, respectively, in kidney transplant recipients (log-rank test; P < .001) (Figure 3).
Standardized incidence ratio
Standardized incidence ratios and associated confidence intervals for various types of posttransplant de novo malignancies, categorized by ICD-9-CM classification, are shown in Table 2.
In heart transplant recipients, 31 (4.5%) of 687 heart recipients developed de novo malignancies (SIR = 1.2; 95% CI, 0.8-1.7; P = not significant) on follow-up. Significantly higher incidences of nonmelanoma skin cancer (SIR = 5.8; 95% CI, 1.5-23.3; P < .05), lung cancer (SIR = 2.6; 95% CI, 1.3-5.2; P < .01), and non-Hodgkin lymphoma (SIR = 5.4; 95% CI, 1.7-16.6; P < .01) were seen in heart transplant recipients.
Among kidney transplant recipients, de novo malignancies were diagnosed in 470/5038 patients (9.3%) (SIR = 3.0; 95% CI, 2.7-3.3, P < .001), which was the highest incidence rate among solid-organ transplant recipients. Rates of renal cell carcinoma (SIR = 37.3; 95% CI, 30.4-45.8; P < .001), transitional cell carcinoma (SIR = 31.4; 95% CI, 26.9-36.6; P < .001), hepatocellular carcinoma (SIR = 3.1; 95% CI, 2.4-3.9; P < .001), nonmelanoma skin cancer (SIR = 2.9; 95% CI, 1.2-6.9; P < .05), thyroid cancer (SIR = 2.7; 95% CI, 1.6-4.8; P < .001), and non-Hodgkin lymphoma (SIR = 3.6; 95% CI, 2.1-6.4; P < .001) were all significantly higher in kidney transplant recipients than in the general population.
Among liver transplant recipients, 111/2127 patients (5.2%) developed de novo malignancies (SIR = 1.5; 95% CI, 1.2-1.9; P < .001). Rates of posttransplant de novo gastric cancer (SIR = 3.0; 95% CI, 1.4-6.7; P < .01), prostate cancer (SIR = 2.8; 95% CI, 1.2-6.1; P < .05), and lymphatic-hematopoietic malignancy (SIR = 4.5; 95% CI, 2.4-8.3; P < .001) were significantly higher than in the general population.
Increased age was significantly associated with increased risk of posttransplant de novo malignancies in heart (crude hazard ratio = 7.8; 95% CI, 1.0-57.6; P < .05), kidney (adjusted hazard ratio [aHR] = 8.4; 95% CI, 5.2-13.7; P < .001), and liver transplant recipients (aHR = 4.2; 95% CI, 1.7-10.6; P < .01) (Table 3). Female sex (aHR = 1.5; 95% CI, 1.2-1.8; P < .001), viral hepatitis B carrier (aHR = 1.3; 95% CI, 1.0-1.7; P < .05), and use of mycophenolate (aHR = 1.5; 95% CI, 1.2-1.8; P < .001) were independent risk factors for posttransplant de novo malignancies in kidney transplant recipients. Interestingly, hypertension in kidney transplant patients was significantly associated with reduced risk of posttransplant de novo malignancies (aHR = 0.4; 95% CI, 0.4-0.5; P < .001).
In a subpopulation analysis to determine risk factors of de novo transitional cell carcinoma and renal cell carcinoma after kidney transplant, patients without history of diabetes or hypertension had significantly higher risk of posttransplant de novo transitional cell carcinoma (aHR = 3.0; 95% CI, 2.1-4.4; P < .001) and renal cell carcinoma (aHR = 1.9; 95% CI, 1.1-3.3; P < .05) compared with patients who had diabetes or hypertension (Figure 4).
In addition, kidney transplant recipients with increased age (30-50 y: aHR = 5.0; 95% CI, 2.0-12.2; P < .001; > 50 y: aHR = 9.0; 95% CI, 3.6-22.4; P < .001), those who were female (aHR = 2.9; 95% CI, 2.1-4.2; P < .001), and those who used mycophenolate (aHR = 1.7; 95% CI, 1.2-2.4; P < .01) also had significantly higher risk of posttransplant de novo transitional cell carcinoma. Similarly, patients with increased age, those who were female patients (aHR = 2.3; 95% CI, 1.5-3.7; P < .001), or those who used mycophenolate (aHR = 1.7; 95% CI, 1.1-2.8; P < .05) also had significantly higher risk of posttransplant de novo renal cell carcinoma.
This is the first nationwide study based on the Taiwanese population to investigate posttransplant de novo malignancies in solid-organ transplant recipients. Because the Taiwanese population is ethnically Chinese in origin, the results of this study may be extrapolated to the population in mainland China. This study indicated that patterns and risk factors of posttransplant de novo malignancies vary between different transplant types. Our results showed lung cancer and nonmelanoma skin cancer in heart transplant recipients, transitional cell carcinoma and renal cell carcinoma in kidney transplant recipients, gastric cancers and prostate cancers in liver transplant recipients, and increased lymphatic-hematopoietic malignancies in all transplant types. Development of such cancers posttransplant requires close surveillance. Increased age, female sex, use of mycophenolate in kidney transplant recipients, and older age in liver transplant recipients were significantly associated with increased risks of posttransplant de novo malignancies. Moreover, kidney transplant recipients without diabetes or hypertension and who were on mycophenolate had higher risks of posttransplant de novo transitional cell carcinoma and renal cell carcinoma. The clinical implication of this study is that posttransplant surveillance protocols for malignancies should be specifically established based on the endemic epidemiologic data.
Consistent with studies on Western populations, our study revealed an increased risk of posttransplant de novo malignancies compared with that shown in the general population; the pattern of prevailing posttransplant de novo malignancies has been shown to vary among different solid-organ transplant types and areas of residence.4-7 In addition, our study also reported a universally increased risk of posttransplant lymphatic-hematopoietic malignancy in all types of solid-organ transplant recipients, similar to that shown in Western studies.4-7 This increased risk could be caused by use of posttransplant immunosuppressants, which has been shown to be proportional to the risk of posttransplant lymphoproliferative disorders.18 In contrast, the risk of posttransplant de novo skin cancer in this study was not as high as that shown in Western populations.7 Similar to other Asian studies,9,11,12,19 there may be a predominant genetic influence on the occurrence of skin cancer that could cause less prevalence of such cancers in Asian populations.
With the use of strict inclusion and exclusion criteria, this study avoided pretransplant undiagnosed occult malignancies. Despite these efforts, posttransplant de novo transitional cell carcinoma and renal cell carcinoma (SIR = 31.4/37.3) still remained extremely high in kidney transplant recipients in Taiwan compared with the reported incidences of bladder and kidney cancers in Korea (SIR = 8.6/10.5),10 Sweden (SIR = 2.0/6.2),7 United Kingdom (SIR = 2.4/7.9),4 and Hong Kong (SIR = 8.2/12.5).9 Because risks of posttransplant bladder and kidney cancers in Korea and Hong Kong have been reported to be significantly lower than those shown in this study, merely genetic background cannot account for the high risk of transitional cell carcinoma and renal cell carcinoma that we found.9,10 Li and associates reported a similar finding of extremely high risk of posttransplant bladder and kidney cancers in kidney transplant recipients.20 However, the study enrolled all posttransplant malignancies, even those who were diagnosed within 1 year after transplant. Theoretically, malignancies occurring shortly after transplant are less likely to be de novo malignancies because, logically, time would be needed for the development of de novo malignancies due to the interplay among immunosuppressive therapy, biological factors, and environment carcinogens.3
Consistent with previous studies,11,21,22 increased age and female sex were also risk factors for posttransplant de novo malignancies in our study. A novel finding reported here was that, in kidney transplant recipients without diabetes or hypertension as the cause of renal failure who were on mycophenolate for immunosuppression, mycophenolate was an independent risk factor for posttransplant de novo transitional cell carcinoma and renal cell carcinoma. Diabetes, hypertension, chronic glomerulonephritis, and polycystic kidney disease have been reported to be the main causes of chronic renal failure that consequently necessitate kidney transplant.23 Patients without these diseases may present with other unusual causes of renal failure, such as herbal supplements containing aristolochic acid, compound analgesics, and exposure to environmental arsenic, which are well-known environmental carcinogens.22,24-27
Mycophenolate was initially introduced to replace carcinogenic azathioprine, and previous studies indeed support the concept that use of mycophenolate is not associated with increased risk of posttransplant malignancies.23,28-30 Surprisingly, this study revealed that kidney transplant recipients with nondiabetic and nonhypertensive causes of renal failure who used mycophenolate were at increased risk of posttransplant de novo malignancies, particularly for transitional cell carcinoma and renal cell carcinoma. This novel finding may imply that interactions between mycophenolate and endemic environmental carcinogens24-27 may partially account for the extremely high incidence of posttransplant de novo transitional cell carcinoma and renal cell carcinoma that we found in kidney transplant recipients. However, the real influences and mechanisms of mycophenolate with environmental carcinogens on the development of transitional cell carcinoma and renal cell carcinoma require elucidation in further studies. Thus, our study may underpin the importance of enhanced cancer screening protocols in patients with end-stage renal diseases that are not due to hypertension or diabetes mellitus. However, this finding may require validation in long-term multicenter studies.
Also compatible with previous studies, the incidence of lung cancer in heart transplant recipients was higher than in the general population.4,6,7 Tobacco consumption and exposure to immunosuppressive therapy may partially account for the increased risk. In addition, risk of de novo hepatocellular carcinoma in liver transplant recipients was comparable to that shown in the general population. This finding may imply that recipients who underwent liver transplant for end-stage liver disease are not at an increased risk of developing de novo hepatocellular carcinoma, although most liver transplant recipients may have histories of viral hepatitis. However, whether immunosuppressive therapy imposed an increased risk of de novo hepatocellular carcinoma in liver transplant recipients with viral hepatitis needs further elucidation.
The results of this study should be interpreted cautiously because of the following limitations. First, this dataset lacked information regarding personal habits, such as the amount of tobacco and alcohol consumption, use of herbal medication, and viral loading of hepatitis virus. It also lacked data regarding exposure to environmental carcinogens such as air pollution or arsenic in underground water. Further studies are needed to elucidate the influence of interactions among posttransplant immunosuppressive therapy, biological factors, and environmental carcinogens on risk of posttransplant de novo malignancies. Second, everolimus was only introduced in 2008 for transplant recipients; therefore, the number of users was too small to achieve adequate evaluation of its efficacy on posttransplant de novo cancers. A longer observation period is needed to obtain adequate estimation. Third, to focus on posttransplant de novo malignancies, all transplant recipients who were lost to follow-up due to death or new diagnosis of cancers within 1 year after transplant were not included in this study. Thus, the results of overall survival may be overestimated, and the cumulative incidence of de novo malignancies could be lower than that of all posttransplant malignancies reported in other series.
Risk of posttransplant de novo malignancies has increased universally, but the patterns of malignancies differ among various types of solid-organ transplants. To improve posttransplant outcomes, each regional organ transplant program, based on the population it serves, should tailor and establish its surveillance protocol for posttransplant de novo malignancies based on epidemiologic data. However, the type and intensity of surveillance mandated in such a cohort of patients require further long-term studies.
DOI : 10.6002/ect.2019.0210
From the 1School of Medicine, China Medical University, Taichung, Taiwan; the
2Department of Surgery, China Medical University Hospital, Taichung, Taiwan; the
3Division of Transplantation, University of Illinois at Chicago, Illinois, USA;
the 4Division of Surgery, University of North Dakota, Grand Forks, North Dakota;
the 5Management Office for Health Data, China Medical University Hospital,
Taichung, Taiwan; the 6College of Medicine, China Medical University, Taichung,
Taiwan; the 7Department of Anesthesiology, Taipei Medical University Hospital,
Taipei, Taiwan; the 8Health Policy Research Center, Taipei Medical University
Hospital, Taipei, Taiwan; the 9School of Medicine, Taipei Medical University,
Taipei, Taiwan; and the 10School of Chinese Medicine, China Medical University,
Acknowledgements: This study is supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of Excellence (MOHW105-TDU-B-212-133019), China Medical University Hospital, Academia Sinica Taiwan Biobank Stroke Biosignature Project (BM10501010037), the NRPB Stroke Clinical Trial Consortium (MOST 104-2325-B-039 -005), the Tseng-Lien Lin Foundation, Taichung, Taiwan, the Taiwan Brain Disease Foundation, Taipei, Taiwan, and the Katsuzo and Kiyo Aoshima Memorial Funds, Japan. The authors have no conflicts of interest to disclose.
Corresponding author: Arshad Khan, Altru Health System and University of North Dakota, 711 Delmore Drive, Roseau, MN 56751, USA
Phone: +1 218 463 1365
Table 1. Demographic and Clinical Characteristics in Heart, Kidney, and Liver Transplant Recipients
Table 2. Invidence of Posttransplant De Novo Malignancy in Heart, Kidney, and Liver Transplant Recipients
Table 3. Risk Factors of Posttransplant De Novo Malignancies in Heart, Kidney, and Liver Transplant Recipients
Figure 1. Overall Survival Among Heart, Kidney, and Liver Transplant Recipients
Figure 2. Comparison of Overall Survival Between Patients With and Without Posttransplant De Novo Malignancies
Figure 3. Cumulative Incidence of Posttransplant De Novo Malignancies Among Heart, Kidney, and Liver Transplant Recipients
Figure 4. Risk of De Novo Transitional Cell Carcinoma and Renal Cell Carcinoma After Kidney Transplant