Fax: (011) 81-6-6879-3539
Risk factors for malignancy in Japanese renal transplant recipients
Article first published online: 3 APR 2007
Copyright © 2007 American Cancer Society
Volume 109, Issue 10, pages 2109–2115, 15 May 2007
How to Cite
Imao, T., Ichimaru, N., Takahara, S., Kokado, Y., Okumi, M., Imamura, R., Namba, Y., Isaka, Y., Nonomura, N. and Okuyama, A. (2007), Risk factors for malignancy in Japanese renal transplant recipients. Cancer, 109: 2109–2115. doi: 10.1002/cncr.22636
- Issue published online: 25 APR 2007
- Article first published online: 3 APR 2007
- Manuscript Accepted: 29 JAN 2007
- Manuscript Revised: 24 JAN 2007
- Manuscript Received: 11 AUG 2006
- renal transplantation;
- risk factor;
Among recipients of renal transplants, the incidences of renal cancer and gastrointestinal cancer are higher and that of skin cancer is much lower in Japan than in Europe and North America.
The risk factors for the development of malignant tumors were examined in Japanese recipients of renal transplants. A total of 556 patients underwent renal transplantation at the Department of Urology, Osaka University Faculty of Medicine between March 1, 1965, and April 31, 2004. Of these patients, 366 were retrospectively studied in whom risk factors potentially related to the development of malignancy could be evaluated on the basis of medical records. The incidence of malignancy, survival rate, and risk factors for malignancy were examined.
The overall incidence of malignancy was 6.8% (25/366 patients). Six of the 25 patients with malignancy died of cancer, but there was no correlation between the occurrence of malignancy and the survival rate (P = .8058, log-rank test). A Cox proportional-hazards model identified treatment with tacrolimus (hazard ratio [HR] = 4.376; 95% confidence interval [CI]: 1.647–11.627; P = .0031) and age at transplantation (HR = 1.562; 95% CI: 1.089–2.240; P = .0155) as risk factors for malignancy.
The results of multivariate analysis suggested that age at transplantation and the use of tacrolimus were independent risk factors for the development of malignancy in recipients of renal transplants. Cancer 2007. © 2007 American Cancer Society.
Recent advances in immunosuppressive therapy have decreased the rate of acute rejection and markedly improved early graft survival in renal-transplant recipients. Despite these encouraging trends, long-term survival after renal transplantation has remained largely unchanged.1 High mortality among renal-transplant recipients is attributed mainly to increased risks of cardiovascular disease, associated with conditions such as diabetes mellitus and hyperlipidemia,2, 3 as well as infection and malignancy. In Europe and North America the incidence of malignant neoplasms among recipients of renal transplants ranges from 6% to 8.3%.4–7
In Japan, few studies have examined the incidence of malignancy in renal-transplant recipients. Hoshida et al.8 estimated that the incidence of malignancy after renal transplantation was 2.6% in Japan. This figure is higher than the incidence of malignancy in the general population, but lower than that in renal transplant-recipients in Europe and North America. Kasiske et al.7 performed a follow-up study in the US and found that the incidence of malignancy was lower in Asians than in Caucasians. Epidemiologic studies have shown that skin cancer and malignant lymphoma commonly occur in European and North American recipients of renal transplants, whereas gastric cancer and gastrointestinal cancer are more frequent in Japanese recipients.8 However, no study has examined the risk factors for malignancy in Japanese recipients of renal transplants.
In Europe and North America the development of malignancy after renal transplantation has been associated with the age at transplantation,3, 7, 9, 10 number of years after transplantation,4, 7, 9, 10 duration of hemodialysis,7 race,7 and type of immunosuppressive therapy.8, 9, 11 Furthermore, graft recipients have an increased risk of infection during intensive immunosuppressive therapy, potentially leading to the development of hepatocellular carcinoma, associated with posttransplant lymphoproliferative disorders (PTLD) due to Epstein-Barr virus (EBV) or chronic hepatitis C infection. The use of antithymocyte globulin (ATG) and muromonab-CD3 monoclonal antibody (OKT-3) has been suggested to be a risk factor for PTLD.11, 12
Concerning immunosuppressive therapy after organ transplantation, basic immunosuppressants are given for the duration of the patient's life to prevent graft rejection. Additional immunosuppressants capable of strongly suppressing immune responses are transiently used to prevent or treat acute rejection, usually occurring within the first few months after transplantation. In the early era of transplantation medicine, azathioprine and prednisolone were mainly used as basic immunosuppressants. Subsequently, cyclosporine became available in the 1980s and tacrolimus in the 1990s. At present, 3-drug therapy combining cyclosporine or tacrolimus with antimetabolites such as azathioprine, mycophenolate mofetil (MMF), or mizoribine plus steroids is primarily used. As for additional immunosuppressants, in the early era of transplantation ATG was mainly used and then OKT3 appeared. Since 2002, basiliximab has been used. ATG is an antihuman lymphocyte globulin. OKT3 is a mouse monoclonal antibody that binds specifically to CD3 antigen recognition complex on human T lymphocytes, thereby inactivating its function. Basiliximab is a human/mouse chimeric monoclonal antibody with specific affinity for CD25, the α chain of the interleukin-2 receptor selectively expressed on the activated T cells. Binding of basiliximab to CD25 inhibits the binding of interleukin-2 to interleukin-2 receptor.
This study was designed to investigate risk factors for the development of malignancy and the impact of immunosuppressants in Japanese recipients of renal grafts, who have different clinical characteristics from European and North American recipients. We retrospectively studied patients who underwent renal transplantation at the Department of Urology, Osaka University Faculty of Medicine, including a number of patients followed up for at least 5 years after transplantation. We investigated the incidence of malignancy and risk factors for malignancy, especially the impact of each immunosuppressant, because immunosuppressants are indispensable for treatment after organ transplantation and are taken for a long period.
MATERIALS AND METHODS
We followed up patients who underwent renal transplantation at the Department of Urology, Osaka University Faculty of Medicine, between March 1, 1965, and April 31, 2004. The patients were divided into 3 groups according to the basic immunosuppressant used: a conventional therapy group treated with azathioprine and predonine, a cyclosporine group that received cyclosporine-based immunosuppressive therapy, and a tacrolimus group that was given tacrolimus-based therapy. These 3 groups were compared with respect to the development of malignant neoplasms, the relation between the presence/absence of malignancy and survival time, and correlations of malignancy with gender, age at the time of transplantation, the basic immunosuppressants used in the induction phase, and concomitant immunosuppressants (eg, ATG, OKT-3, and basiliximab).
All of the patients were monitored annually by physical examination, laboratory studies, ultrasonography, and occult-blood testing. If malignancy was suspected, diagnostic studies were performed, including endoscopic and histopathologic examinations whenever possible. Since 2000, computed tomographic (CT) scanning was performed every 6 months. If a malignant neoplasm was detected the patient received appropriate treatment and was followed up to assess tumor response and symptoms. If a patient with malignancy died, cancer was considered the cause of death unless tumor progression or metastasis was ruled out as a possible cause.
The Kaplan-Meier method and the log-rank test were used to compare survival rates and cancer-specific survival rates between patients with malignancy (malignancy group) and those without malignancy (nonmalignancy group), as well as among the conventional therapy group, the cyclosporine group, and the tacrolimus group. Statistical significance was set at P < .05.
To determine risk factors for malignancy, univariate analysis of all demographic characteristics of the patients was first performed with a Cox proportional-hazards model. After examining the relations of individual demographic factors to malignancy, a Cox proportional-hazards model taking into account interactions among demographic factors was again used to identify risk factors for malignancy. The reference group had the following characteristics: male sex, age <10 years, and kidneys from living donors. Age was categorized according to decade. A stepwise method was used to select the factors for this analysis. A Fisher exact test and analysis of variance (ANOVA) were used to analyze characteristics.
A total of 556 patients underwent renal transplantation during the study period. We analyzed data from 366 patients, comprising 228 men and 138 women (301 recipients of living donor renal grafts and 65 recipients of cadaveric renal grafts), in whom risk factors potentially related to the development of malignancy could be evaluated on the basis of medical records. As for the induction of immunosuppression, there were 62 patients in the conventional therapy group, 224 in the cyclosporine group, and 80 in the tacrolimus group. OKT-3 was used in 38 patients, MMF in 60, ATG in 121, mizoribine in 168, and basiliximab in 36.
The clinical profiles of the conventional therapy group, cyclosporine group, and tacrolimus group are shown in Table 1. Mean age at the time of renal transplantation (mean ± SD) was 29.0 ± 8.0 years in the conventional therapy group, 34.6 ± 11.4 years in the cyclosporine group, and 36.3 ± 12.4 years in the tacrolimus group. The mean duration of hemodialysis before transplantation was 0.96 ± 1.69 years in the conventional therapy group, 2.72 ± 3.76 years in the cyclosporine group, and 3.27 ± 5.19 years in the tacrolimus group. The mean duration of follow-up was 12.9 ± 9.9 years in the conventional therapy group, 10.5 ± 5.8 years in the cyclosporine group, and 5.0 ± 4.0 years in the tacrolimus group.
|Variable||Azathioprine (AZA) (n = 62)||Cyclosporine (CyA) (n = 224)||Tacrolimus (TRL) (n = 80)||Fisher P||ANOVA|
|Age at Tx (y old: mean ± SD)||29.0 ± 8.0||34.6 ± 11.4||36.3 ± 12.4||—||.0003|
|HD duration (y: mean ± SD)||0.96 ± 1.69||2.72 ± 3.76||3.275 ± 5.19||—||.3652|
|Follow-up period (y: mean ± SD)||12.9 ± 9.9||10.5 ± 5.8||5.0 ± 4.0||—||.0000|
The overall incidence of malignancy during follow-up was 6.8% (25/366 patients): 11.3% (7/62 patients) in the conventional therapy group, 4.9% (11/224 patients) in the cyclosporine group, and 8.8% (7/80 patients) in the tacrolimus group. One patient developed pharyngeal cancer and leukemia. Other tumors were as follows: gastric cancer, carcinoma of the host kidney, breast cancer, malignant lymphoma, uterine cancer, colorectal cancer, rectal cancer, carcinoma of the grafted kidney, liver cancer, carcinoma of the grafted ureter, ovarian cancer, lung cancer, and bladder cancer. No patient had skin cancer. The 25 patients with malignancy consisted of 11 men and 14 women, 20 of whom received living donor renal grafts and 5 cadaveric renal grafts. The mean age at the time of transplantation was 35.1 years (range, 11–56 years), and mean age at the initial detection of malignancy was 44.0 years (range, 24–72 years). The mean interval from transplantation to the diagnosis of malignancy was 8.9 years (range, 12 days to 22 years).
The overall incidence of death was 24.0% (88/366 patients). The incidence of death was 24.0% (6/25 patients) in the malignancy group and 24.0% (82/341 patients) in the nonmalignancy group.
The overall cumulative mortality rates in the malignancy group and nonmalignancy group are shown as Kaplan-Meier curves in Figure 1. Estimated cumulative mortality rates in the 2 groups were 8.0% in the malignancy group and 12.5% in the nonmalignancy group at 5 years and 26.4% and 19.0%, respectively, at 10 years. The overall mortality rate and cancer-specific mortality rate did not significantly different between the 2 groups (P = .8058, log-rank test).
Mortality rates according to basic immunosuppressants were 58.1% (36/62) in the conventional therapy group, 21.0% (47/224) in the cyclosporine group, and 6.3% (5/80) in the tacrolimus group. Cumulative mortality rates estimated by the Kaplan-Meier method were, respectively, 29.0%, 10.4%, and 2.0% at 5 years and 40.8%, 14.5%, and 20.8% at 10 years (Fig. 2). Mortality rates significantly differed among the 3 groups (P < .0001, log-rank test).
Risk Factors for Malignancy
Univariate analysis was performed to assess correlations between the development of malignancy and various patient factors. Age at the time of transplantation was shown to be a significant risk factor, with a hazard ratio (HR) of 1.553 (P = .0255). As for immunosuppressants, the HR was 3.966 (P = .0979) for MMF, 3.573 (P = .1563) for tacrolimus, 2.488 (P = .5093) for basiliximab, and 1.863 (P = .2218) for mizoribine. Thus, the risk of developing malignancy decreased in this order, but there was no significant difference among these immunosuppressants (Table 2). In the case of OKT-3, cyclosporine, and ATG, the HRs were 1.022 (P = .9774), 0.5950 (P = .5665), and 0.498 (P = .1732), respectively.
|Variable||95% CI||Hazard ratio||P|
|Sex (women vs men)||0.958–5.288||2.251||.0626|
|Age (per 10 y older)||1.055–2.286||1.553||.0255|
|Dialysis before transplantation (<2 y vs ≥2 y)||0.175–1.155||0.450||.0967|
|Kidney donor type||0.423–3.879||1.281||.6615|
Multivariate analysis was also performed to detect risk factors for malignancy. Gender, age at transplantation, duration of hemodialysis, and treatment with tacrolimus or MMF were selected as factors. Administration of tacrolimus (HR = 4.376; 95% confidence interval [CI]: 1.647–11.627; P = .0031) and the age at transplantation (HR = 1.562; 95% CI: 1.089–2.240; P = .0155) were found to be risk factors for the development of malignant tumors (Table 3).
|Variable||95% CI||Hazard ratio||P|
|Sex (women vs men)||0.983–5.196||2.260||.0549|
|Age (per 10 y older)||1.089–2.240||1.562||.0155|
|Dialysis before transplantation (<2 y vs ≥2 y)||0.170–1.007||0.413||.0519|
During the present follow-up study (conducted until April 31, 2004), the incidence of malignancy was 6.8% in renal-transplant recipients. This figure is higher than the incidence of 2.6% reported in 1999 by Ichimaru et al.13 of our department. It is also higher than the incidence reported by Hoshida et al.8 in a 1997 study involving several Japanese centers, including Osaka University Faculty of Medicine, but is similar to the range of 6% to 8.3% reported in Europe and North America.4–7 In the present study the follow-up period was longer than 10 years in most patients. This long period after transplantation might be related to the higher incidence of malignancy in the present study as compared with the results of Ichimaru et al. However, the interval after transplantation was not extracted as an independent risk factor. In contrast to our findings, a posttransplantation period of over 10 years was reported by Agraharkar et al.10 to be a risk factor for the development of malignancy. Physicians should therefore remain alert to the risk of malignancy in long-term survivors after renal transplantation.
The most common cancers occurring in our patients differed from those in European and North American patients, consistent with the findings of previous studies.8 In our series, gastric cancer, carcinoma of the native kidney, breast cancer, malignant lymphoma, uterine cancer, and colorectal cancer were the most common. Thus, renal and gastrointestinal cancers were common, whereas no patient had skin cancer. Tang et al.14 reported that the incidence of malignant tumors was 9% (4/45) in 45 Chinese recipients of renal grafts who survived for at least 10 years after transplantation. Gastrointestinal carcinoma developed in 2 patients, cervical cancer in 1, and hepatocellular carcinoma in 1. No patient had skin cancer, supporting the results of previous studies in Japanese recipients and suggesting that the low incidence of skin cancer is a characteristic of Asian patients. In the present study renal cancer was the most common malignancy (4 patients had cancer of a native kidney and 1 had cancer of the grafted kidney) and gastric cancer and breast cancer were the second most common, in agreement with the findings of Hoshida et al.8 They found that renal-transplant recipients have a particularly high risk of renal cancer, about 80-fold higher than the risk in the general population. The risk of gastric cancer was about 1.40-fold higher and the risk of breast cancer was about 1.53-fold higher in renal transplant recipients. In 1994 a survey by the Japanese Society for Transplantation15 reported that the incidence of renal cancer was 8.6-fold higher in renal-transplant recipients and 5.6-fold higher in patients receiving dialysis compared with the risk in the general population. The incidence of gastric cancer was 0.5-fold that of the general population in renal-transplant recipients and 0.8-fold in patients receiving dialysis. Agraharkar et al.10 reported that the risk of renal cancer is about 7-fold higher in European and North American recipients of renal transplants compared with the general populations. It has also been reported that the risk of renal cancer is 3.3- to 9.9-fold higher in patients with end-stage renal failure managed by hemodialysis than in the general populations in Europe and North America, although the reasons for this higher incidence remain unclear.16 Therefore, screening for renal cancer should be done periodically before and after renal transplantation to detect malignancy at an early stage. Screening examinations should include abdominal ultrasound and other imaging methods, such as abdominal CT scanning.
The survival rate of our renal transplantation patients with malignancy was not significantly lower than that of patients without malignancy, but showed a decreasing trend after a long follow-up period (Fig. 1). After renal transplantation at our hospital, all patients annually undergo general physical examinations, laboratory studies, ultrasonographic examinations, and occult-blood testing. If necessary, endoscopic and histopathologic examinations are also performed. Since 2000, CT scanning has been performed every 6 months to screen for malignancy. We are intent on the early detection and diagnosis of any malignancy.
Multivariate analysis showed that age at the time of transplantation (HR, 1.562) and treatment with tacrolimus (HR, 4.376) were significant risk factors for malignancy. Consistent with our results, ‘age at the time of transplantation’ has been extensively reported to be a risk factor for malignancy in Europe and North America.3, 7, 9, 10 We also found that the use of tacrolimus is clearly a risk factor for malignancy. Although patients given azathioprine were followed up for a longer period, patients given tacrolimus had a significantly higher incidence of malignant tumors than those given azathioprine. This finding suggests that the incidence of malignant tumors (particularly solid tumors) would further increase if patients given tacrolimus were followed up for longer periods, thereby having a substantial effect on the outcomes of renal-transplant recipients.
A correlation between the use of immunosuppressants and the development of malignant tumors has been reported by many authors. The cancers observed in renal-transplant recipients have changed since around the time when cyclosporine became available. The incidences of malignant lymphoma, Kaposi sarcoma, and renal cancer have increased, whereas those of skin cancer, uterine cervical cancer, and vulvar carcinoma have decreased.17 Hoshida and Aozasa18 reported that the median interval to the onset of malignancy was 95.5 months before the introduction of cyclosporine therapy, decreasing to 42.5 months after the introduction of cyclosporine, suggesting that cyclosporine may also play a role in the development of malignancy in Japanese transplant patients. It was recently shown that the incidence of malignancy could be decreased by using cyclosporine in combination with sirolimus,19 but the mechanism involved and the contribution of sirolimus are not known.20 Tacrolimus has only been studied for short periods (3–5 years) in general because it became available for clinical use in the 1990s. It should be noted that the 5-year risk of malignancy has not been properly assessed for tacrolimus. According to a US multicenter study of renal transplantation,21 the incidence of malignant tumors 3 years after transplantation was 9.2% in the cyclosporine group and 8.8% in the tacrolimus group, and there was no significant difference between the 2 drug groups. A meta-analysis of 30 recent studies of cyclosporine and tacrolimus also showed no significant difference between the 2 drugs with respect to the incidence of malignancy in renal-transplant recipients.22 However, a study conducted by Opelz and Dohler11 in 145,104 recipients of cadaver renal grafts showed that the incidence of lymphoma 2 years after transplantation was about 2-fold higher in the tacrolimus group than in the cyclosporine group, irrespective of the combined use of MMF and azathioprine. Caillard et al.23 also reported that the use of tacrolimus is associated with a high risk of PTLD. In contrast, Kasiske et al.7 found that tacrolimus was similar to cyclosporine with respect to the risk of malignancies other than skin cancer; the risk of skin cancer was significantly lower with tacrolimus.
The mechanism underlying the association of tacrolimus with the development of cancer is not clearly understood. Generally, however, intensive immunosuppressive therapy increases the risk of infection, which may consequently heighten the likelihood of malignancy. Patients who are receiving long-term immunosuppressive therapy, especially regimens including tacrolimus, should therefore be carefully monitored to avoid excessive immunosuppression.
In recent years the renal graft survival rate early after transplantation has been further improved by the advent of MMF and basiliximab. Outside of Japan, it has been reported that additional medications in the induction phase, such as OKT-3, ATG, and basiliximab, have no impact on the incidence of malignancy in the native kidney, but increase the occurrence of PTLD.11 The present study found that no drug was associated with an increased risk of PTLD. Nevertheless, PTLD is closely associated with viral infections (including EBV infection) in Japanese patients.18 The relation between the occurrence of PTLD and the use of these immunosuppressants should therefore be investigated further in Japanese recipients of renal transplants.
The era of transplantation may be a risk factor for the development of malignant tumors. Because the conventional therapy group, cyclosporine group, and tacrolimus group reflect distinct phases of this era, the era of transplantation was not included in the analysis of risk factors.
In the future, longer survival after transplantation and increased numbers of elderly recipients have caused concern that the risk of cancer associated with long-term immunosuppressive therapy and viral infections may increase in Japanese recipients. Our study suggests that age at transplantation and the use of tacrolimus were risk factors for the development of malignancy. Few previous studies have examined the risk of malignancy associated with tacrolimus therapy 5 years or more after renal transplantation. Large-scale long-term studies are needed to determine the actual risk of malignancy in patients receiving tacrolimus therapy. The development of new immunosuppressants with fewer side effects, such as malignancy, is required.
- 13Malignant neoplasms after renal transplantation. Renal Transplant Vasc Surg. 1999; 11: 3–6., , , et al.
- 15The Japanese Society for Clinical Renal Transplantation. The Japanese Society for Transplantation Annual progress report from the Japanese Renal Transplant Registry, 2003, Part III. Results from 2000 follow-up survey. Ishoku. 2004; 39: 57–64.