De novo malignant disease after liver transplantation? Risk and surveillance strategies

Abstract

There is little doubt that the immune system plays a role in cancer prevention.1 It does this through the direct elimination of oncogenic pathogens, through the elimination of pathogens that are associated with chronic inflammation, and through the direct immune surveillance of cancer cells via the recognition of tumor-specific antigens. Thus, immunodeficient mice experience increases in spontaneous, carcinogen-induced, and genetically associated tumors.1 Correspondingly, humans with a congenital immunodeficiency or immunosuppression due to a human immunodeficiency virus infection are at increased risk for cancer.2-4 Furthermore, suppression of the immune system via immunosuppressive drugs leads to a heightened risk of de novo malignancies,1 and the cancer profile is strikingly similar to that observed during the long-term follow-up of people with human immunodeficiency virus infections.5 In particular, long-term studies of large-scale renal transplant populations have generated this evidence. Recent studies comparing age-, sex-, calendar year–, and region-specific cancer rates in transplant and nontransplant populations have shown a standardized incidence ratio (SIR) of 3 to 4 for renal transplant recipients.6, 7 The incident cancers occur at a wide variety of sites, but many are associated with oncogenic viral infections, including human papillomavirus, hepatitis B virus and hepatitis C virus (HCV), Epstein-Barr virus (EBV), human T cell lymphotropic virus, and human herpesvirus 8. These infections are associated with cancers of the lips and oral cavity, anus, cervix, vulva, liver, connective tissue (Kaposi's sarcoma), and lymphatic systems. Evidence that has accumulated over the past decade shows that similarly increased risks are seen for patients undergoing liver transplantation. Furthermore, malignancy is a leading cause of late death for liver transplant recipients.8-10 It is estimated that liver transplant patients have an incidence of cancer between 3.1% and 14.4% and a cancer-related mortality rate between 0.6% and 8.0%.11, 12 These figures are very general, and an analysis of individual cancers is required to understand and develop targeted surveillance and prevention strategies. In many parts of the world, skin cancer is the most common de novo malignancy after liver transplantation, and it occurs in up to 40% of patients at 20 years with an SIR as high as 70.13, 14 Basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) occur in similar numbers, with the peak time to occurrence 5 to 10 years after transplantation.15, 16 Although the absolute numbers of BCCs and SCCs are similar, the risk with respect to the general population is an order of magnitude higher for SCC versus BCC.15 Risk factors include older age, male sex, a fair-skin phenotype, prior sun exposure, prior actinic keratosis, and smoking.16 Patients with longstanding autoimmune hepatitis before transplantation who are on prednisone and azathioprine for many years are at particular risk. In comparison with other solid organ transplant recipients, liver transplant recipients are at a lower risk for skin cancer16; this finding has been attributed to less intense immunosuppression. Deaths associated with skin cancer after transplantation certainly do occur, but the rate may be relatively low in comparison with that observed for other de novo solid cancers.17 There are few reliable estimates, but the standardized mortality ratio (SMR) versus the general population was nonsignificantly elevated (6.9) in a population-based Australian cohort.10 It is well recognized that posttransplant lymphoproliferative disease (PTLD) is a major complication of organ transplantation, with the risk increased up to 12-fold in comparison with nontransplant populations.18 For liver transplant patients, SIRs of 6.15 to 20.8 have been reported. The risk is highest for pediatric recipients and appears to be maximal for heart/lung recipients in comparison with recipients of other organ types.19 Recipient EBV seronegativity is the most consistent risk factor for PTLD in liver and other organ recipients, particularly early after transplantation.20 There are some data suggesting that in non–kidney transplant recipients, the incidence of PTLD may be lower in the most recent decade versus the previous period.21 Induction immunosuppression with anti-thymocyte globulin (ATG) or Muromonab-CD3 (OKT3), but not the selective interleukin-2 receptor antagonists with their shorter duration of effect, is associated with an increased risk. HCV infection after liver transplantation may also be a risk factor.11, 22 Non-Hodgkin lymphoma (NHL) was the most common cause of cancer-related death in an Australian liver transplant population, and the risk of death due to NHL was significantly elevated in comparison with the risk for a matched general population (SMR = 11). Once PTLD is diagnosed, a mortality rate of approximately 40% has been described, and this appears similar to the rate for the nonimmunosuppressed population.23, 24 Several recent population-based registry studies have computed cancer-specific SIRs in comparison with matched general populations. Results from the United Kingdom, Australia, the Netherlands, Finland, Sweden, Canada, the United States, and Italy were quite consistent.7, 13, 14, 19, 25-28 The overall SIRs for cancer, including hematopoietic malignancies, ranged from 2.2 (United Kingdom and Australia) to 2.3 (Sweden) and 2.5 (Canada). Higher SIRs were obtained when skin cancers were included in the all-cancer risk estimate: 2.59 (Finland), 3.4 (Sweden), 4.3 (Netherlands), and 6.6 (United Kingdom). Patterns for individual cancers also emerged, but larger cohorts are required to fully understand the breadth and extent of the risk for liver transplant recipients. There is a general pattern of an increased risk for rare cancers with a viral etiology, such as cancers of the lips and oral cavity (SIR = 5-20), vulva (SIR = 26), anus and anal canal (SIR = 10), and Kaposi's sarcoma (SIR = 290).19, 27 Although there are no population-based SIR estimates showing an increased risk of esophageal cancer after liver transplantation, an Italian study reported an SIR of 23.4 on the basis of cases ascertained by medical record reviews.29 This association may be related to prior alcohol exposure; 2 of 3 patients diagnosed with esophageal cancer in a US cohort underwent transplantation for alcoholic liver disease (ALD).23 An excess risk of colorectal cancer (SIR = 2.40-12.5) has been consistently observed in liver transplant recipients,13, 14, 19, 26 but only in those with underlying primary sclerosing cholangitis (PSC) and ulcerative colitis (UC).19, 30 The colorectal cancer risk in PSC patients with UC appears similar to the risk in the nontransplant situation. Furthermore, there is no evidence that PSC patients without UC have an increased risk of colorectal cancer. The risks of lung cancer (SIR = 1.6-2.30) and kidney cancer (SIR = 1.80-30) were increased in some studies7, 13, 27, 28 but not in others. In several studies, patients who underwent transplantation for ALD had a significantly increased malignancy risk.14, 31 Head and neck cancer, combined with lung cancer, represented almost 50% of incident cancers in ALD patients in one study.31 Of interest is the finding that rates of uterine, ovarian, breast, and prostate cancer are not increased beyond population rates. This has implications for cancer surveillance approaches and strategies. Although this discussion indicates a growing literature on the risk of malignancy after liver transplantation, a limited number of studies have addressed mortality associated with de novo malignancies. According to an analysis of European Liver Transplant Registry data, de novo malignancies were responsible for 14% of posttransplant deaths in ALD patients versus 5% to 6% in patients with other indications.31 In an Australian cohort, the risk of cancer-related death after liver transplantation was significantly elevated (SMR = 1.96); the corresponding SMRs for heart and lung transplant recipients were 3.05 and 4.41, respectively.10 In addition to an excess risk of death from any cancer, Australian liver transplant recipients exhibited an increase in melanoma-related death in comparison with a matched general Australian population (SMR = 5.3).10 Apart from hepatic causes, malignancy was the most common cause of death in a US cohort surviving at least a year after transplantation.9 In another analysis of the US cohort, malignancy was associated with a 40% mortality rate 1 year after the diagnosis of cancer.23 A recent study examined colorectal cancer and its outcomes in an Australian and New Zealand liver transplant database.32 The colorectal cancer risk was significantly increased (SIR = 2.8). The average age at diagnosis was 59 years, and the average time since transplantation was 7.3 years, with younger patients presenting with more advanced malignancies. The mortality rate was high, with 19 of 48 patients dying from cancer progression during follow-up. Thus, there is increasing evidence that de novo solid organ malignancies, when they occur, are associated with a high mortality rate; this makes malignancy a major cause of death during the long-term follow-up of liver transplant patients. This finding may be related to heightened tumor aggressiveness in the context of immunosuppression. There are now 4 population-based studies reporting the cancer risk for different populations of organ recipients.7, 19, 21, 28 In a UK cohort, the risk was similar for kidney, liver, and heart recipients (SIR = 2.2-2.5) and higher (SIR = 3.6) for lung transplant recipients.28 A similar trend was seen in an Australian cohort.19 The general pattern of malignancy across Australian organ recipients was similar, although cardiothoracic recipients had a significantly excess incidence of Merkel cell carcinoma that was not seen in liver transplant recipients. Furthermore, an increased risk of lung cancer was observed only in heart and lung transplant recipients. However, in a very large US cohort, the risk of lung cancer was increased for all transplant recipients, regardless of the organ type, and was greatest for lung transplant recipients. On account of demographic and other differences between transplant populations, SIRs do not convey the full picture of the relative cancer risk between organ transplant types. A Swedish analysis showed that in comparison with kidney transplant recipients, liver transplant recipients had a 2.7-fold greater risk of NHL after adjustments for sex, age, follow-up time, and the calendar period of transplantation.21 A similar analysis of Australian liver and cardiothoracic transplant recipients showed that in comparison with liver transplant recipients, the risk of lymphoma was 1.9-fold higher for heart transplant recipients and 2.1-fold higher for lung transplant recipients.19 When all cancers were considered, the excess risk in comparison with liver transplant recipients was 1.3-fold for heart transplant recipients and 1.6-fold for lung transplant recipients. We have yet to see whether these differences in risk by transplanted organ are fully explained by differences in the type or extent of immunosuppression and preexisting cancer risk factors. Overall, it is generally thought that less immunosuppression over the long term will be associated with less malignancy. Avoidance of ATG induction seems sensible for patients at high risk for PTLD (eg, those who are seronegative for EBV or seropositive for HCV). Whether cyclosporine or tacrolimus is associated with differing risks remains controversial, but there is reliable evidence that a higher dose of immunosuppression increases the risk of de novo malignancy.33 For skin cancer, the combination of a regular dermatological review is effective. Minimization of sun exposure and the use of high-protection sunscreens are well-recognized strategies, but they must be reinforced on a regular basis; this should include the use of written materials.34, 35 Patients with longstanding autoimmune hepatitis before transplantation who have been on azathioprine are at particular risk. In such patients, the continuation of long-term azathioprine after transplantation is best avoided. Emerging data show that the use of mammalian target of rapamycin inhibitors may suppress skin malignancies after renal transplantation, and there are anecdotal reports of similar experiences for liver transplant recipients.36 The association of PTLD with EBV infection, particularly in children, has led to the concept of prospectively monitoring EBV DNA levels with the aim of reducing or minimizing immunosuppression once certain levels are reached.37 In one study,38 11 of 44 pediatric patients had their immunosuppression reduced once the EBV viral load reached 4000 copies/μg of DNA. This led to a decrease in EBV-associated PTLD from 16% to 2% (P < 0.05). This approach has not been universally adopted, although it seems sensible for high-risk groups and especially EBV-seronegative children, who are heavily immunosuppressed.37 For solid organ malignancies, it is important that screening guidelines for malignancies in the general population be employed for adult transplant recipients.34, 35 These include screening for cervical and breast cancer in women and urological examinations in men. It is difficult to recommend additional surveillance beyond that applied in the general population because there is no evidence that breast and prostate cancer and even cervical cancer are more prevalent in liver transplant recipients. General-population screening should be undertaken to increase the likelihood of an early diagnosis. Colonoscopy screening in accordance with local screening guidelines should be in place, whereas patients with UC and PSC should undergo yearly colonoscopy. At one center, annual checkups that included chest and abdominal computed tomography scans in addition to the aforementioned examinations were investigated.39 The authors claimed earlier detection of cancers and improved survival, although it is clear that routine screening along general-population guidelines was not in place before this program. It seems wise to provide counsel on alcohol and smoking risks to liver transplant recipients because both increase the incidence of cancer.9, 31, 40, 41 Additional annual chest X-ray and otolaryngology examinations have been proposed for patients with a diagnosis of ALD and smokers.12 Liver transplant patients are at increased risk for the development of de novo malignancies during long-term follow-up (particularly skin cancer and PTLD). Solid organ malignancies appear to be markedly increased in tumors driven by infections (eg, human papillomavirus and human herpesvirus 8), whereas common malignancies such as breast and prostate cancer are not increased beyond general-population rates. The development of malignancies in patients originally undergoing transplantation for ALD is a concern, but reliable evidence for specific cancers and risk factors is lacking. Prevention and surveillance according to general-population guidelines should be undertaken for all liver transplant recipients, and they should be particularly reinforced and enhanced for patients with ALD. Long-term immunosuppression should be minimized because de novo malignancy remains a very significant cause of mortality, particularly for long-term survivors.23, 42