Liver transplant recipients are at greater risk of developing de novo malignancies than the age- and sex-matched normal population. Those tumors that are increased in incidence are primarily skin cancers and the posttransplant lymphoproliferative diseases (PTLD) (1, 5, 7, 25, 26). This increased incidence occurs primarily as a consequence of immunosuppression which may affect tumor development and progression by a variety of mechanisms: thus those tumors associated with oncogenic viruses, such as lymphoma or Kaposi sarcoma, occur as a direct consequence of reduced lymphocyte regulation, whereas some immunosuppressive drugs may affect the behavior of the malignant cells. Thus, cyclosporin may enhance the malignant potential of some cells whereas sirolimus may inhibit tumor progression (2, 3, 4).
The largest international registry was set up by the late Israel Penn and the registry continues to record the incidence of posttransplant cancers (5) However, although such national and international registries provide invaluable information on the prevalence and incidence of de novo cancers in the transplant population, they do have some disadvantages: data are not always validated, reporting of both the number of people with cancer and the length of follow-up of the cohort is not always complete, and the age- and sex-associated risk factors cannot readily be assessed.
In this study, we have compared the incidence of de novo posttransplant tumors in a single center with national data to determine the extent of increased cancer risk. We also estimated the cancer risk in relation to sex, age, and indication for liver transplantation.
PATIENTS AND METHODS
In all, 1,778 adults (849 male) underwent transplantation (OLT) between January 1982 and March 2004. The median duration of follow-up was 65.6 months (range 32.5–76.9 months). Patients were followed until death or until the end of the study (March 2004). The diagnosis of cancer was established histologically. Patients were treated with triple therapy for the first 3 months with a calcineurin inhibitor (CNI), azathioprine (1–2 mg/kg/day), and prednisone (20 mg/day). With the exception of those transplanted for autoimmune hepatitis, corticosteroids were withdrawn by 3 months. Cyclosporine was used as the primary CNI between 1982 and 1994 (target whole blood trough levels were 100–150 ng/ml); after this time, patients were maintained primarily on tacrolimus (target whole blood levels 5–10 ng/ml). Induction therapy with monoclonal or polyclonal antibodies was not used. Mycophenolate (2 g/day), introduced into our routine clinical practice in 2002, was used as a CNI sparing agent or when the patient had two episodes of acute rejection or developed histological features suggesting ductopenic rejection.
All our patients had a thorough transplant assessment. This includes chest x-ray, ultrasonography of abdomen and computerized tomography of chest and abdomen, upper and lower gastrointestinal endoscopy if indicated. Females also had breast examination and mammography and gynaecological examination and investigation if clinically indicated.
Cancer incidence in the survey patients was expressed as a standardized incidence ratio (SIR); this is the ratio of the observed number of de novo cancer cases to the expected number of cases in the general population expressed as a percentage. A SIR value of greater than 100 suggests excess risk and a value of less than 100 suggests a decreased risk. Expected numbers of cancers were calculated from cancer incidence rates for England and Wales (specified by age, sex, and calendar year) applied to similarly- defined arrays of person-years-at-risk (pyr) generated by the data. Study subjects entered the pyr on the date of transplantation and left the pyr in the date of death, date last known to follow-up or the closing date of the survey, whichever was the earlier. Overall standardized mortality ratios (SMRs) were calculated in an analogous way as the ratio of observed death to expected deaths, expressed as a percentage. The statistical significance of the differences between the observed numbers and their corresponding expectations were assessed by means of the Poisson distribution. All significance tests were two-tailed. National mortality and cancer incidence rates for each site of cancer were obtained from the Department of Public Health and Epidemiology, Birmingham.
A total of 1,778 adults liver transplant recipients were studied. Overall, 250 patients (14.1%) were age <35 years, 339 patients (19.1%) were aged between 35–44 years, 578 patients (32.5%) were aged between 45–54 years, 495 patients (27.8%) were aged between 55–65 years, and 116 patients (6.5%) were aged above 65 years. The indications for transplantation are shown in Table 1.
Of the 1,778 patients, 1,178 (66.3%) were alive at the time of censoring (March 2004) and the median survival was 84 months. Six hundred (33.7%) had died at a median time of 29 months. Comparison of observed mortality rates in the transplant population with an age- and sex-matched population’s expected mortality in relation to years since OLT is shown in Table 2. The SMR for the overall population was 814 (95% CI 750–881); the SMR is greatest in the first 4 years after transplantation but for those who survived for more than 15 years, there was no statistically significant difference in mortality in the transplant population compared with a matched group. Overall, the choice of calcineurin inhibitor did not significantly affect survival. There was no difference in SMR between males and females (Table 3). Those transplanted for autoimmune liver diseases (primarily primary biliary cirrhosis, autoimmune hepatitis, and primary sclerosing cholangitis) had a significantly better survival (as shown by SMR) compared to those transplanted for viral hepatitis (hepatitis B and C) (Table 3).
Of the 1,778 patients, 141 (7.9%) developed new tumors following transplantation (Table 4). Of these, 60 were male and 81 were female; the median age at transplant was 50 years (range 16–73). 45 patients died due to cancer and 15 patients died due to other causes. In all, 81 of the 141 patients who had de novo cancer are still alive. Overall, patients were twice as likely to get cancer as the age- and sex-matched general population (SIR 207, 95% CI 174–244), although males were at greater risk than females (SIR 248 [95% CI 189–319] and 184 [95% CI 146–229] for male and female respectively), this difference was not statistically significant (Table 4).
The greatest increase in cancers was seen in lymphoid and solid organ tumors. The incidence of de novo lymphoid tumors is ten times higher (SIR 1026), skin cancers is 5.8 times higher (SIR 580), large bowel cancer is 4.8 times higher (SIR 489), and lung cancer is 1.9 times (SIR 196) higher than the general population. The most common de novo tumor was skin cancer (basal cell or squamous cell carcinoma [51/141 patients]) and was similar in males and females; the mean duration form OLT to skin tumor diagnosis was 65 months. We observed 18 lymphoid tumors (8 male) at a mean time of 45 months.
Cancer of the large bowel was much more commonly seen in those with ulcerative colitis (SIR 2727) than those without (SIR 347). There were 18 patients who had a de novo colon cancer after transplantation; primary sclerosing cholangitis was the indication for transplantation in 10 patients. Of these, 6 of 10 had ulcerative colitis at the time of transplantation. Three of six recipients (50%) who had UC at the time of OLT have died and the tumor was the cause of death in two. The mean duration from OLT to tumor diagnosis was 60 months. Three patients were on tacrolimus and the remainder were on cyclosporine as their primary immunosuppressive regime. Eight patients died and the cause of death was recurrent PSC in five, septicaemia in one, renal failure in one, and chronic rejection in one. All 10 male recipients and 7 of the 8 females over 40 years at the time of diagnosis suggesting that older age may be a higher risk for colonic cancer.
Females (but not males) had a greater risk of lung cancer and had a significant higher SIR compared to general population (SIR 336, 95% confidence interval: 174–587, P< 0.001). There were 14 patients (12 female) who developed lung cancer postOLT. Eight patients had squamous cell carcinoma, five patients had small cell carcinoma, and one had large cell carcinoma. The mean time from OLT to tumor diagnosis was 57 months. Of these 14, 11 had a documented smoking history. Only 2 of the 14 patients are alive.
There was no increase in cancers of the breast, cervix or rectum compared with a matched population. The occurrence of other de novo cancers were malignant neoplasm of mouth (1), oropharynx (2), esophagus (2), stomach (2), rectum (3), pancreas (2), malignant melanoma (1), uterine cervix (1), uterus (1), ovary (2), prostate (3), bladder (1), kidney (2), and Hodgkin’s disease (2). One patient developed de novo hepatocellular carcinoma in a cirrhotic graft. We did not analyze the SIR for these patients, as the numbers are too small.
Those transplanted for alcoholic liver disease had a three times greater de novo cancer incidence (SIR 316) than those grafted for other indications; of interest, those grafted for drug-induced liver failure had no increase risk of cancer compared to those with liver failure from other causes (SIR 94 and 208, respectively) (Table 5). SIR for all cancers was greater in recipients aged less than 40 years than those who were older at transplantation and this was largely due to significantly higher than expected skin cancers in the <40s age group and females <40 had a SIR of 484 for breast cancer. (Table 6).
As most of the problems associated with the perioperative period have been overcome, attention has focused on factors affecting long-term survival and many units have developed well-patient clinics to focus on maintaining good health. Causes of premature patient and graft loss include recurrent disease and complications of immunosuppression, including cardiovascular disease, renal failure and malignancy. In this study, we have shown that while survival of liver allograft recipients is excellent, it is significantly reduced compared with an age- and sex-matched population. This difference may be greater than is apparent because patients selected for transplantation are screened for cardiac, respiratory, and other diseases, so they might be expected to have a better than average survival if they had not had liver disease. That those who have survived more than 15 years have a life expectancy no different from the matched population may be a consequence of the relatively small numbers at risk.
Because of the increased risk of malignancy in liver allograft recipients, follow-up must include additional surveillance for those cancers where there is an increased risk. Most published reports have shown increased incidence of de novo cancers such as malignant lymphomas and cutaneous neoplasms (1, 5, 7) but a decreased incidence of breast cancer (6). We have found, as previously shown, increase risk of cancer of the skin, lymphoid system, and bowel, but no increased risk of breast or cervical cancer. However, Kelly and colleagues (8) suggested that with current immunosuppressive regimens, liver transplant patients are not at an increased risk for developing nonlymphoid solid organ tumors.
The most common de novo tumor in our study group, as reported by others (1, 7, 8), was skin cancer, either basal cell or squamous cell carcinoma, accounting for 36% of de novo tumors. However, unlike Frezza (1), we found no difference between the two calcineurin inhibitors. Although Xio (9) and Ignacio Herrero (10) found that older age is one of the risk factors associated with skin cancer, we found the SIR in patients aged below 40 years was higher than older patients. This may reflect that the incidence of skin cancer is greater in the older population (expected cancer 0.43 and 8.36 in those below and above 40 years).
Although rare, PTLD is a significant cause of mortality in allograft recipients. Benlloch and colleagues suggested that hematological neoplasms appeared earlier than solid, were more prevalent in those transplanted after 1995 than before and associated with a lower survival than solid organ tumors (11). We found that the incidence of lymphoid tumor was significantly higher compared to general population (SIR 1026). In contrast, leukemia is rare (12) and we found only one patient with acute leukemia.
With respect to colon cancer, the overall SIR was 489 but was much greater in those with preexisting ulcerative colitis (SIR 2727 and SIR 347 for those with and without colitis respectively). Annual surveillance colonoscopy has been recommended in those with colitis (13, 14). It may be appropriate to restrict this additional surveillance to older patients since the SIR was 476 in those aged over 40 years compared to SIR of 100 in younger patients. As all our patients who have ulcerative colitis were treated with ursodeoxycholic acid as this may reduce the risk of colonic polyps and cancer (15–18). In the nontransplant population, those with inflammatory bowel disease have a greater risk of colon cancer than the matched population but the increased risk (which is affected by a variety of factors including duration and degree of inflammation and use of concomitant medication) is around threefold, much less than that seen in the liver allograft recipients with colitis (19, 20)
The most striking finding was an increase in the incidence of lung cancer in female OLT recipients. As reported by others (21, 22), many of our patients had a documented past history of smoking but we do not have a sufficiently detailed record to determine whether smoking alone accounts for the increased cancer risk in this cohort. These findings contrast with the findings of De Perrot who concluded that the risk of bronchogenic carcinoma was low (21).
The incidence of breast cancer after transplantation does not appear to be increased (6, 8); we found that the incidence was not statistically significant compared to that in the nontransplanted population. Of interest, all these patients were on cyclosporine as their primary immunosuppression. Again, age appears to be an important factor with a higher SIR for breast cancer in younger patients.
The extent to which the choice of immunosuppression affects the risk of de novo cancer is uncertain. Potent induction protocols are associated with an increased risk of PTLD (23, 24) but none of our patients receive antibodies such as OKT3, ATG, or ALG. The choice of calcineurin inhibitor may be a factor: Frezza and colleagues found a higher incidence of nonlymphoid cancer incidence in those given cyclosporine compared with tacrolimus (1). In this study, the overall incidence of all neoplasms was significantly greater in those on cyclosporine (SIR 223) compared to general population but not for those who are on tacrolimus (SIR 147). Those who were on cyclosporine had an increased incidence of colonic, lung and lymphoid tumors.
We have compared the observed and expected number of incident cancer (SIR) in relation to the indication for liver transplantation. We found that recipients who were transplanted for alcoholic-related liver disease (ALD) had SIR 1.5 times higher (SIR 316 vs. SIR 199) compared to those grafted for other indications. Our results are similar to previously published data by others (8, 10, 11, 25). The significance of the lack of increased risk in those transplanted for drug induced liver failure group is unclear.
In conclusion, we have found that there is a significant increase in the risk of cancer in this group of liver allograft recipients. Although the absolute risk of cancer is low, we found that the increase in risk is greater in the younger aged recipients than the older ones. Those transplanted for alcohol-related cirrhosis had a higher risk than those transplanted for other indications. In addition to the recognized increased risk in skin cancers and lymphoproliferative diseases, we found an increase in the risk of cancer of the colon, especially in those with colitis. We found no increased risk for cancer of the breast or cervix. These observations may help in the appropriate surveillance of those who have received a liver graft.
1. Frezza EE, Fung JJ, van Thiel DH. Non-lymphoid cancer after liver transplantation. Hepatogastroenterology
1997; 44: 1172.
2. Liu X, Powlas J, Shi Y, et al. Rapamycin inhibits Akt-mediated oncogenic transformation tumor growth. Anticancer Res
2004; 24: 2697.
3. Koehl GE, Andrassy J, Guba M, et al. Rapamycin protects allografts from rejection while simultaneously attacking tumors in immunosupressed mice. Transplantation
2004; 77: 1319.
4. Penn I. Cancers following cyclosporine therapy. Transplantation
1987; 43: 32.
5. Penn I. Neoplastic consequences of transplantation and chemotherapy. Cancer Detect Prev Suppl
6. Jain AB, Yee LD, Nalesnik MA, et al. Comparative incidence of de novo nonlymphoid malignancies after liver transplantation under Tacrolimus using surveillance epidermiology end result data. Transplantation
1998; 66: 1193.
7. Sanchez EQ, Marubashi S, Jung G, et al. De novo tumors after liver transplantation: a single-institution experience. Liver Transpl
2002; 8: 285.
8. Kelly DM, Emre S, Guy SR, et al. Liver transplant recipients are not at increased risk for nonlymphoid solid organ tumors. Cancer
1998; 83: 1237.
9. Xiol X, Guardiola J, Menendez S, et al. Risk factors for development of de novo neoplasia after liver transplantation. Liver Transpl
2001; 7: 971.
10. Herrero JI, Lorenzo M, Quiroga J, et al. De novo neoplasia after liver transplantation: an analysis of risk factors and influence on survival. Liver Transplant
2005; 11: 89.
11. Benlloch S, Berenguer M, Prieto M, et al. De novo internal neoplasms after liver transplantation: increased risk and aggressive behavior in recent years? Am J Transplant
2004; 4: 596.
12. Doti CA, Gondolesi GE, Sheiner PA, et al. Leukemia after liver transplant. Transplantation
2001; 72: 1643.
13. Loftus EV Jr, Aguilar HI, Sandborn WJ, et al. Risk of colorectal neoplasia in patients with PSC and UC following OLT. Hepatology
1998; 27: 685.
14. Bleday R, Lee E, Jessurun J, et al. Increased risk of early colorectal neoplasms after hepatic transplant in patients with inflammatory bowel disease. Dis Colon Rectum
1993; 36: 908.
15. Neuberger J. Liver transplantation for cholestatic liver disease. Curr Treat Options Gastroenterol
2003; 6: 113.
16. Vera A, Gunson BK, Ussatoff V, et al. Colorectal cancer in patients with inflammatory bowel disease after liver transplantation for primary sclerosing cholangitis. Transplantation
2003; 75: 1983.
17. Pardi DS, Loftus EV Jr, Kremers, WK, et al. Ursodeoxycholic acid as a chemopreventive agents in patients with ulcerative colitis and primary sclerosing cholangitis. Gasteroenterology
2003; 124: 889.
18. Khare S, Cerda S, Wali RK, et al. Ursodeoxycholic acid inhibits Ras mutations, wild-type Ras activation, and cyclooxygenase-2 expression in colon cancer. Cancer Res
2003; 63: 3517.
19. Eaden JA, Abrams KR, Mayberry JF. The risk of colorectal cancer in ulcerative colitis: a meta-analysis. Gut
2001; 48: 526.
20. Winther KV, Jess T, Langholz E, et al. Long-term risk of cancer in ulcerative colitis: a population-based cohort study from Copenhagen County. Clin. Gastroenterol Hepatol
2004; 2: 1088.
21. de Perrot M, Wigle DA, Pierre AF, et al. Bronchogenic carcinoma after solid organ transplantation. Bronchogenic carcinoma after solid organ transplantation. Ann Thorac Surg
2003; 75: 367.
22. Delcambre F, Pruvot FR, Ramon P, et al. Primary bronchogenic carcinoma in transplant recipients. Transplant Proc
1996; 28: 2884.
23. Opelz G, Dohler B. Lymphoma after solid organ transplantation: a collaborative transplant study group. Am J Transplant
2004; 4: 222.
24. Norin S, Kimby E, Ericzon BG, et al. Post-transplant lymphoma—a single –center experience of 500 liver transplantation. Med Oncol
2004; 21: 273.
25. Jain A, DiMartini A, Kashyap R, et al. Long-term follow-up after liver transplantation for alcoholic liver disease under tacrolimus. Transplantation
2000; 70: 1335.