EXPERIMENTAL AND CLINICAL TRANSPLANTATION

Cancer is the third most common cause of death (after cardiovascular disease and infection) for patients who have a functioning kidney allograft. Kidney and liver transplant recipients have similar cancer risks because of immunosuppression but different risks because of differences in primary diseases that cause renal and hepatic failure and the inherent behavior of cancers in the liver. There are 4 types of cancer that may develop in liver allograft recipients: (1) recurrent cancer, (2) donor-transmitted cancer, (3) donor-derived cancer, and (4) de novo cancer. Identification of potential donor cancer transmission may occur at postmortem examination of a deceased donor or when a probable donor-transmitted cancer is identified in another recipient. Donor-transmitted cancer after liver transplant is rare in Australia, the United Kingdom, and the United States. Aging of the donor pool may increase the risk of subclinical cancer in donors. Liver transplant recipients have a greater risk of de novo cancer than the general population, and risk factors for de novo cancer in liver transplant recipients include primary sclerosing cholangitis, alcoholic liver disease, smoking, and increased age. Liver transplant recipients may benefit from cancer screening because they have a high risk, are clearly identifiable, and are under continuous medical supervision.

Key words : Hepatic failure, Immunosuppression, Malignancy, Screening, Tumor

Cancer mortality after all forms of transplant provides both physicians and recipients with a significant concern, ranking only behind cardiovascular disease and infection as the cause of death for patients with a functioning renal allograft. While cancer risk became evident as early as the 1960s in renal transplant, de novo cancer after liver transplant has received less attention until the last 10 years. The similarity between cancer risk with renal and liver allografts is driven by the immunosuppression, while differences are driven by the diseases that lead to hepatic as opposed to renal failure, and the inherent behavior of cancers in the liver. This paper provides a review of the current state of knowledge of de novo cancer after liver allografts.

Origins of cancer after liver transplant
In distinguishing the types of cancer relevant to liver allograft recipients, it is important to understand the different situations that define the types of cancer risk. There are 4 different circumstances:

Recurrent cancer
These include cholangiocarcinoma, hepatocellular carcinoma, and extrahepatic cancers, which pre-existed in the recipient and may have been an indication for liver transplant. These are the subject of intense debate and protocols and will not be further discussed here.

Donor transmitted cancers
These are present in the organ at transplant and which, despite great care, may escape donor screening. These are generally avoided quite effectively in most programs by thorough attention to previous medical history, examination, and investigation such that the incidence is general low,¹ for example, in the United States, the risk may be up to 1.0%² of transplants from deceased donors. In Australia there have been only 2 known donor transmitted cancers.³

Donor-derived cancers
In these patients the cancer develops in the transplanted liver after transplant, as may occur in a cirrhotic liver after hepatitis reinfection, or occasionally when a lymphoma is derived from donor rather than recipient lymphocytes.

De novo cancers
De novo cancers account for the majority of cancer deaths in all forms of transplant and which develop under the influence of immunosuppression and are often driven by viral causes. The data collected and reported by the Australia and New Zealand Liver Transplant Registry are shown in Table 1 and demonstrate the relative risks of the 4 different cancer situations.

Donor-derived and donor-transmitted cancers
The 3 concepts defined in the global transplant vigilance and safety program—Project Notify—run by the Italian organ donor agency and the World Health Organization are outlined in Table 2.⁴ The certainty with which the donor origin of a tumor can be stated varies from patient to patient. The strongest evidence comes when the tumor is genotyped, for example, using HLA typing showing donor origin, combined with detection of the same tumor in more than 1 recipient or in the donor after death. There are more occasions when the origin is less certain: ie, when the cancer was not detected in the donor, did not occur in any other recipients, but may have originated in the recipient.

The most difficult clinical decision making comes when a potential donor cancer transmission is identified, usually through postmortem examination of a deceased donor, or when a probable donor transmitted cancer is identified in another recipient. For example, colon cancer is found at donor after death, with no lymph node involvement. Should one remove the liver with retransplant or not?

Donor transmitted cancers are rare after liver transplant according to US, UK, and Australian data. The cancers that have been reported include neuroendocrine tumor, adenocarcinoma of the colon, and melanoma. The relative rarity of transmission of donor cancer after liver cancer is notable, and implies a high index of suspicion is exercised at organ retrieval. Table 3 shows the 3 top sites of metastasis of the common cancers—all of which metastasize to the liver. Avoidance of any donor with a prior history of cancer, except superficial skin cancers, in situ cancers, or low-grade brain tumors is the backbone of safe decision making. An urgent recipient with life-threatening acute liver failure might justify breaking this rule but only if there was no other option available. It is worth noting the effect of aging of the donor pool with respect to increasing risk of subclinical cancer. Figure 1 shows the overall incidence of cancer by age in the Australian population, clarifying that donors under 50 years of age will have less than 500 cancers per 100 000 population and are less likely to act as a source of cancer than the elderly donors with incidences up to 3 times greater but who are now being accepted by many transplant programs around the world. It is possible that the incidence of donor transmitted cancer will rise in the future years.

In Australia, to date, the 2 liver transplant patients with cancer proven to be transmitted from the donor, through demonstration of the same tumor in other recipients from the same donor, were melanomas and both patients died of their tumor.

De novo cancers
De novo cancers are significantly increased after all forms of organ transplant compared to the general population.⁶ The data from patients after renal transplant has now been widely examined in different populations using registry data. Cohort studies are the only way of examining such data, and there have been a variety of reports. The most reliable methodology is to use transplant registries as the source of data on transplant and general population cancer registries as the source of data on cancer in both the general and the transplant populations. This avoids several biases introduced by using transplant registry collected cancer data such as the different classifications of tumors, the data source for the cancer diagnosis, and reporting biases. The cancer registries are also not without biases, especially because most rely on a pathological diagnosis from an anatomical sample which is not achieved in all definite cancers—myeloma is a classic example missed by some cancer registries. This methodology has been used in the Australian,⁷ Canadian,⁸ US,⁹ and UK¹⁰ populations. Single-center cohort studies have also proved useful, but the number of life years of follow-up after transplant are usually insufficient for detailed analyses, though some specific inferences can be made.¹¹

Analyses of de novo cancer in liver transplant recipients have now followed the renal transplant methodologies, and there are publications from the European,¹² Australian,^13,14 and US registry experiences.⁹

In their analysis of 1926 Australian Liver transplant recipients followed between 2.2 and 10 years after transplant (median, 6 y), Na and associates¹³ compared a sex- and age-matched general Australian population and noted an increased Standardized Incidence Ratio (SIR) of 2.20 (confidence interval (CI) 1.87-2.57). The cancer types that were significantly elevated were Kaposi sarcoma, vulva, lip, non-Hodgkin lymphoma, unknown primary site, melanoma, colorectal and anal, and salivary gland. The factors known to increase cancer risk were age (relative risk [RR] 1.33), smoking (RR 1.72), alcoholic liver disease (RR 2.14), and primary sclerosing cholangitis (RR 2.62). The increased incidence was highest in the first 2 years where lymphoma and Kaposi sarcoma were the highest frequency. The excess risk of colorectal cancers was confined to the recipients with primary sclerosing cholangitis with ulcerative colitis, thus easily defining the high at risk cohort for intensive bowel screening. In a separate analysis of the same cohort, Na and associates demonstrated that de novo cancer also conferred a 1.96-fold increased risk of death after liver transplant.¹⁴

A Europe-wide analysis undertaken by Euro-transplant concluded that the risk of de novo cancer for liver transplant recipients was very similar to the Australian data with an SIR of 1.94 (CI:1.53-2.44).¹² Lung cancer, colorectal cancer, lymphoma, renal cell carcinoma, thyroid cancer, and vulvar cancers were significantly increased in this cohort.

In the United States, an analysis of 175 732 organ transplant recipients of which 37 888 received a liver allograft, were compared with the general US population using 14 US regional cancer registries. Engels and associates found the same order of increased risk as in Australia and Europe with an SIR of 2.1 (CI:2.06-2.14).⁹ They also identified the risk of cancer of the transplanted organ was always greatest amongst heart, liver, kidney, and lung allograft recipients. The SIR of cancer of the liver was 43.83 (CI: 40.90-46.91) in liver transplant recipients, but was not elevated in heart, lung, or kidney recipients, although kidney and lung cancer were both significantly elevated in liver recipients, this was to a lesser extent than it was in kidney and lung recipients.

Role of Specific Immunosuppressants

Analyses of the data from most studies support the view that the majority of the excess cancer risk after liver transplant arises from the effect of immuno-suppression on cancers driven by viral oncogenesis. A recent analysis of immune suppression restricted to immunoglobulin deficiency does, however, suggest that different cancers may be driven by different aspects of immune deficiency,15 and the known excess of colorectal cancers is probably driven by the inflammatory environment of ulcerative colitis. The broad view of differences between the cancer risks of the different organ allograft types is that the higher overall level of immunosuppression drives the higher risk in heart, lung, and kidney transplants than is seen in liver recipients. But are different immunosuppressive agents responsible for different levels of risk?

There are several approaches to the question of which agent is worse and which is better with respect to cancer risk: registry cohorts, single-center analyses, a few posthoc analyses of randomized controlled trials, experimental models, and effect of the agents in clinical cancer treatment. None is conclusive, and as yet there are too few controlled data to drive an evidence-based decision.

In addition to Epstein-Barr virus status of the recipient, induction therapy has been shown to be a risk factor for early posttransplant lymphoproliferative disease using OKT3, antithymocyte globulins,¹⁶ and alemtuzumab,¹⁷ although it is not clear if there is a difference between these agents.

The question of whether there is any difference between the calcineurin inhibitors after liver transplant remains unanswered because of conflicting data. Some studies suggest, in multivariate analysis, a 2-fold increase in cancer with cyclosporine, but there are weaknesses in the study which urge caution.¹¹ Whether or not mycophenolate mofetil adds to the cancer risk of an immunosuppressed liver transplant patient has been examined in the United States.¹⁶ Perhaps the short follow-up times of this study should urge caution, but the differences between 10 099 patients treated with tacrolimus and steroids compared with 9180 treated also with mycophenolate were insignificant.

The one drug group known to affect cancer risk is the mammalian target of rapamycin inhibitors sirolimus and everolimus.¹⁸ The Federal Drug Administration’s black box warning for use of these agents after liver transplant has perhaps inhibited formal study of their use in this circumstance.

The logical approach to prevention of cancers after all forms of transplant would appear to be to screen the at-risk population. The dominant arguments in favor are that the risk is high, and patients are both clearly identifiable and under continuous medical supervision.

The rational economic arguments are not so convincing unfortunately, because the economic benefit of screening in the general population is derived from the additional life years saved by screening.¹⁹ The technologies used to screen may be deficient in the transplant population. For example, fecal occult blood testing followed by colonoscopy relies on a test specificity for fecal occult blood that is highly likely to be different comparing the general population and transplant recipients who have multiple causes for gastrointestinal bleeding. The studies in transplant recipients are hard to undertake with sufficient power to demonstrate efficacy because of the numbers required to be recruited and thus, the evidence may never be forthcoming.

The second problem is that screening in the general population relies on population survival data. The transplant recipient has a 10-fold worse prognosis than an age- and sex-matched individual in the general population,^13,14 so the life years saved by screening a transplant recipient will be 10-fold less, and the cost per transplant life year saved consequently 10-fold greater. The rational arguments do not support screening but most clinicians advocate for the application of national screening guidelines to the transplant population.

Strategies for avoidance, detection, and treatment of cancer are important to the long-term survival and care of liver transplant recipients. Counselling about cancer risks and about individualized screening programs are needed.

Volume: 12 Issue: 1 March 2014 - Supplement - 1

FULL TEXT