SEARCH

SEARCH BY CITATION

Keywords:

  • de novo malignancy;
  • liver transplantation

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Orthotopic liver transplant (OLT) is an established life saving procedure for both acute and chronic liver failure, but incidences and risk factors for development of these malignancies are yet to be established. To determine the incidences and risk factors associated with de novo malignancy after OLT. We performed a systematic review of relevant epidemiological studies available on MEDLINE, which provided information on the incidences and risk factors for the development malignancies in adult OLT recipients published from 1983 to 2009. All data was compiled from retrospective studies. Independent risk factors for the development of de novo malignancy in adult OLT recipients were identified to be statistically significant including immunosuppression, hepatitis C virus infection, smoking, alcoholic cirrhosis and sun exposure. OLT recipients with smoking and alcohol history are of particular risk for head and neck and lung cancers. Primary sclerosing cholangitis and inflammatory bowel disease were found to be independent risk factors for colon cancer. Adult OLT recipients are at increased risk for the development of post-transplant malignancies and obviates the need for surveillance protocols that are safe and cost-effective. OLT recipients should be advised on taking proper precautions in the sun, smoking cessation, and eliminating alcohol consumption. Given the emergence of alcoholic cirrhosis as a leading indication for liver transplantation, the early detection of lung and head and neck cancers is of particular importance.

Orthotopic liver transplantation (OLT) was first recognized as a treatment modality for acute and chronic liver failure in 1983 (1). While recent advances in immunosuppressive regiments have decreased the incidence of graft rejection, cases of de novo malignancy following transplant are being increasingly reported in recent years and the incidence of de novo malignancy ranges from 3 to 26% (2). Although de novo malignancy is currently the second leading cause of death after cardiovascular complications, recent studies suggest the incidence of deaths from malignancy than increasing faster than that from cardiovascular complications (3–5).

The reported incidence of de novo malignancy following OLT is significantly greater than malignancy rates in the general population and this has been attributed, at least in part, to the life-long immunosuppression that OLT recipients must endure following transplant (6, 7). Cytotoxic T cells, macrophages, and natural killer cells are known to inhibit the growth of oncogenic viruses and even destroy malignant cells in vitro (8). More specifically, natural killer cells are thought invade developing tumours and promote tumour lysis through an interferon-γ-mediated process (9). Other mechanisms such as a direct carcinogenic effect of immunosuppressive medications and cyclosporine-induced cancer cell progression have also been suggested (10).

Hence, it is biologically plausible that immunosuppression would predispose patients to de novo malignancy following OLT (11). First, immunosuppression from any cause (genetic, AIDS or drug-induced) is associated with a higher incidence of malignancy than the immunocompetent population (12). Second, some retrospective studies have suggested a dose-dependent relationship to de novo malignancy. A study by Benlloch et al. (13) showed a higher incidence of de novo haematological malignancies because the introduction of tacrolimus in the mid-1990's and that neoplasms are now being found at shorter follow-up times. The authors concluded that higher degrees of immunosuppression caused by tacrolimus may be contributing to more prevalent and aggressive cancers following OLT. They also found that use of azathioprine was an independent risk factor for a higher incidence of de novo malignancies. Furthermore, Dantal et al. (14) showed that renal transplant patients on low-dose cyclosporine (trough range: 75–125 ng/ml) had lower rates of malignancy than high-dose cyclosporine (trough range: 150–250 ng/ml) though they also had higher rates of graft rejection and there was no mortality benefit. Case reports have cited accelerated progression from no detected premalignant disease to malignant disease in OLT recipients (15, 16), which authors have attributed to immunosuppression and possibly suboptimal pre-transplant screening. However, not all studies have demonstrated immunosuppression to be an independent risk factor for de novo malignancy (17). Interpretation of the results is limited by types of malignancies studied and number of patients followed.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Search strategy and identification of studies

We searched database MEDLINE for all epidemiological studies on OLT and de novo malignancy. We used combinations of the keywords: ‘liver transplantation’, ‘malignancy’, ‘de novo tumour’, ‘de novo malignancy’, ‘smoking’, ‘alcohol’, ‘Barrett's oesophagus’, ‘Budd–Chiari syndrome’, ‘myeloproliferative disorder’, ‘skin cancer’, ‘Kaposi's sarcoma’, ‘lymphoma’, ‘post-transplant lymphoproliferative disorder’, ‘Epstein–Barr virus’, ‘cytomegalovirus’, ‘hepatitis C virus’, (HCV) ‘colorectal cancer’, ‘head and neck cancer’, ‘oropharyngeal cancer’, lung cancer', ‘prostate cancer’, ‘bladder cancer’, ‘renal cancer’, ‘breast cancer’, ‘cervical cancer’, ‘uterine cancer’, and ‘ovarian cancer.’ We searched all available data from 1983, the year that liver transplantation was first recognized as a treatment modality for liver disease. We also searched the bibliographies of identified review articles for additional studies.

Inclusion and exclusion criteria

We included studies published in scientific journals that provided information about risk factors for development of de novo skin, lymphoproliferative, and colorectal cancer. We also included case reports and searched all article bibliographies for relevant articles. We excluded data from other transplant populations (for instance: kidney, heart and pediatric) in lieu of adult OLT data unless no data otherwise existed. We only used large clinical trials (number of subjects at least 400) in our calculation of de novo tumour incidences and mean time to diagnosis of de novo tumour so that our results would be more representative of the general OLT population, but still reported data regarding risk factors from smaller trials when not available from larger trials.

Major risk factors and predisposing conditions

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Smoking and alcohol

Smoking's detrimental effect on cardiovascular mortality is universally accepted but its role in malignancy after OLT is emerging. A retrospective cohort study of 326 OLT patients by van der Heide et al. (18) showed that at 10 years, smoking increased an OLT recipient's risk of non-skin malignancies 2.1 vs 12.7% (P=0.019). In a study of 2260 patients, Lee et al. (19) found no difference in all cause mortality at 2 years after OLT, but the Kaplan–Meier survival curves show cumulative survivals of 65 and 80% among active smokers and lifetime non-smokers at transplantation, respectively, at 6 years (P-value not reported). Further, Leithead et al. (20) showed that among OLT recipients, smokers had an increased risk of cardiovascular death and sepsis-related mortality compared with non-smokers. The authors did not find any difference in malignancy-related mortality. Numerous case series have also showed history of smoking as a risk factor for developing malignancy after OLT though mechanisms are yet to be elucidated (21, 22). Hence, new retrospective data has suggested that smoking may increase OLT recipient's odds of non-skin malignancies and all cause mortality after OLT.

Using multivariate analysis, a retrospective study of 722 liver transplant patients, previous alcohol abuse was associated with a three-fold risk of de novo tumour following OLT [P=0.002, 95% confidence interval (1.5–5.8)] (13). Eight of the 10 patients who developed oropharyngeal cancer in this study were transplanted for alcoholic cirrhosis. Similarly, the King's College Hospital study showed that patients who underwent OLT for alcoholic cirrhosis had increased risk of malignancy (P=0.001) (17). Other studies have confirmed this result and suggest that the prognosis is poorer for de novo malignancy in OLT recipients for alcoholic cirrhosis compared with other indications for OLT (23, 24).

Even among the immunocompetent population, alcohol abuse is associated with increased risk of liver and gastrointestinal tract cancers (25). Interestingly, acute alcohol intoxication has been shown to decrease natural killer cell activity and promote tumour metastases in mice (26). Natural killer cells may play an important role in tumour surveillance in vivo and suppression of their activity could promote tumorigenesis.

Tobacco smoking and alcohol appear to increase the risk of lung and oropharyngeal cancers in OLT recipients. According to the most recent Surveillance Epidemiology and End Results (SEER) statistics, the incidences of lung and oropharyngeal cancer in the general population of the US are 0.06 and 0.02% respectively (27). In comparison, the incidence of these cancers in the OLT recipients is and ranges from 0 to 1.2% for lung cancer (17, 28) and 0.1 to 2% oropharyngeal cancer (Table 1).

Table 1.   Incidence of de novo tumour in orthotopic liver transplant recipients published in large case series
Reference# pts#de novo tumourType of de novo tumour # of tumours (% incidence)
Skin (includes Kaposi's)PTLDHead and NeckLungColon/ rectumProstateBladder/ kidneyBreastCervical/uterine/ ovaryOther*
  • *

    Other cancers included: pancreatic, small bowel, liver, seminoma, thyroid, metastatic cancer of unknown primary, angiosarcoma, pituitary, cholangiocarcinoma, gastric, embryonic testicular, glioblastoma, haemangioblastoma, gastric Kaposi's.

  • †The study was limited to non-lymphoid tumours, but did report two leukemias.

  • ‡The study was limited to non-lymphoid tumours, but did report five non-hodgkin's lymphoma.

  • §

    §This study only reported the 45 tumours listed, but did not name the remaining 17 tumours.

  • N/A, not applicable; NR, not reported; OP, oropharygeal; PTLD, post-transplant lymphoproliferative disorder.

Frezza et al. (55)16575635 (2.1)NR2 (0.1)2 (0.1)4 (0.2)1 (0.1)0 (0)3 (0.2)4 (0.2)3 (0.2)
Jain et al. (6)10005724 (2.4)NR7 (0.7)8 (0.8)4 (0.4)3 (0.3)2 (0.2)3 (0.3)0 (0)6 (0.5)
Saigal et al. (17)11403014 (1.2)NR3 (0.3)0 (0)1 (0.1)0 (0)3 (0.3)2 (0.2)1 (0.1)4 (0.4)
Sanchez et al. (41)142112542 (3.0)35 (2.5)4 (0.3)11 (0.8)9 (0.6)0 (0)3 (0.2)7 (0.5)2 (0.1)12 (0.8)
Yao et al. (2)10435317 (1.6)9 (0.9)3 (0.3)5 (0.5)6 (0.6)0 (0)2 (0.2)4 (0.4)2 (0.2)5 (0.5)
Kelly et al. (36)888318 (0.9)NR6 (0.7)1 (0.1)3 (0.3)0 (0)0 (0)2 (0.2)1 (0.1)5 (0.6)
Jonas et al. (21)458338 (1.7)7 (1.5)2 (0.4)3 (0.7)0 (0)0 (0)0 (0)3 (0.7)7 (1.5)3 (0.4)
Jimenez et al. (23)§5056216 (3.2)13 (2.6)10 (2.0)6 (1.2)NRNRNRNRNRNR
Bellnoch et al. (13)77241NR10 (1.3)9 (1.2)8 (1.0)2 (0.3)1 (0.1)3 (0.4)2 (0.3)2 (0.3)4 (0.5)
Incidence of tumours in general population (%) (27)0.020.030.010.060.050.20.030.10.04N/A  

Premalignant disease

Given the role of the immune system in tumour surveillance, it is likely that OLT recipients with premalignant conditions before OLT could be at higher risk for developing malignant disease in the setting of immunosuppression. Patients with Barrett's oesophagus are at increased risk of developing oesophageal cancer (29). Interestingly, one patient developed high-grade dysplasia of the oesophagus that previously only had metaplasia consistent with Barrett's oesophagus on pretransplant endoscopy 9 months prior. The patient was on cyclosporine A, azathioprine, and prednisone for immunosuppression. Other case series have also demonstrated the relationship between Barrett's oesophagus and oesophageal cancer (30, 31).

Myeloproliferative disorder (MPD) is a risk factor for leukemia which commonly presents as Budd–Chiari syndrome. Occlusion of the hepatic veins in Budd–Chiari syndrome can lead to liver failure necessitating the need for liver transplant in some cases. Saigal and colleagues conducted a study of 1097 OLT recipients, six of whom had Budd–Chiari syndrome. One of these patients developed acute leukemia after 72 months of follow-up suggesting that this was likely normal progression of MPD to acute leukemia, not an accelerated course caused by OLT (32). Dousett et al. (33) also reported two cases of relatively late progression of MPD to leukemia (29 and 31 months). More recently, a retrospective study of 606 OLT recipients (seven of which had Budd–Chiari), none developed leukemic transformation (34).

A recent retrospective study of 175 patients analysed de novo tumours that developed from other premalignant conditions (31). In all, five cancers were determined to have developed from pre-malignant conditions: cervical carcinoma in situ (cervical atypia), oesophageal carcinoma (Barrett's esophagus), colon cancer (colonic polyp and ulcerative colitis). This study also found one patient who was transplanted for Caroli's disease that developed cholangiocarcinoma 10 months after OLT. Caroli's disease is a rare congenital disorder causing dilatation of intra-hepatic bile ducts. While the more typical presentation of the disease is recurrent cholangitis and hepatomegaly, cholangiocarcinoma can develop in 7% of patients and thus would be considered a pre-malignant condition (35).

Age and gender

In a single center trial in the UK, Kelly et al. (36) reported the average age of diagnosis de novo tumour was 54 years old compared to 56 years old in the recurrent tumour group suggesting no difference in age. Another single center trial from Germany also found similar ages for transplant recipients who did not develop de novo malignancy and those that did (45±18 years and 46±14 years, respectively) (21). However, more recent studies were able to find that OLT patients greater than 40 years old and 51 years old at time of transplant was an independent risk factor for de novo malignancy (7, 37). The aforementioned single center studies did not find a statistically significant difference between males and females in the development of de novo tumour (21, 36). But one study found male sex to be an independent risk factor for the development of skin cancer after OLT (38).

Incidences of major de novo malignancies following orthotopic liver transplant

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Skin cancers

Overall, skin cancers are the most common de novo malignancies following organ transplantation (39, 40). The overall incidence of skin cancer following OLT has been reported to be 0.9–3.2% in large case series (23, 36) (Table 1) compared with 0.02% in the general population (27). In a smaller study of 151 OLT recipients, Mithoefer et al. (38) found a skin cancer incidence of 22.5%, a value they claim may have been underestimated due to having lesions removed my primary care physicians, though only 83% of OLT recipients responded to their questionnaire which may have resulted in reporting bias. In large retrospective trials, the mean time of diagnosis of skin cancer after OLT has been reported to range from 36.4 and 50.2 months (6, 17). Sanchez et al. (41) reported the mean time of diagnosis of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) to be 19.4 and 47.2 months respectively.

Among heart and renal transplant recipients, SCC seems to have higher incidence than BCC in a ratio of 4:1 (42). In OLT recipients, however, the incidence of SCC and BCC seems more evenly distributed (17, 21, 43). However, there is some evidence that the SCC:BCC ratio would increase at longer periods of follow-up (41, 44). Melanoma was recently reported to have an increased risk of 10.1-fold according to a recent single-center Italian study among OLT recipients and has a lower incidence in transplant recipients than BCC and SCC (45). Aside from the more general risk factors for de novo malignancy mentioned above, risk factors specific to the development of de novo SCC and BCC after OLT have included, red hair, number of second degree sunburns and total cumulative lifetime sun exposure (Table 2) (38). This is consistent with what was found among heart transplant recipients (46–48). Polymorphisms in p53 tumour surveillance have also been associated with SCC and may be relevant to the OLT population (49). In a side by side comparison, patients treated with cyclosporine developed skin cancer at shorter times of follow-up than patients treated with tacrolimus suggesting that cyclosporine is an independent risk factor for the development of skin cancer specifically (38).

Table 2.   Site-specific risk factors to de novo malignancy in adults after liver transplantation
SiteRisk factor*
  • *

    Statistically significant.

  • †Statistically significant association with patients transplanted for alcoholic cirrhosis.

  • ‡There is no data in the liver transplantation literature that as identified statistically significant risk factors to this site specific cancer.

SkinSmoking (23)
Alcoholic cirrhosis (17, 39)
Age >40, male gender, red hair, brown eyes, PSC, cyclosporine (38)
Skin type and sun exposure (43)
Post-transplant lymphoproliferative disorder (PTLD)Age>50, HCV, alcoholic cirrhosis, use of anti-lymphocyte Ab (60)
Transplantation for acute fulminant liver failure within 18 months after OLT, rejection therapy with high-dose steroids (92)
HCV (13,61)
ColorectalPSC, IBD
OropharygealAlcohol use (23, 74, 75)
Smoking (73)
LungAlcohol use (79)
Smoking (79)
GynaecologicInsufficient data
GenitourinaryInsufficient data

Organ transplantation confers up to a 500-fold risk of Kaposi's sarcoma (KS) compared with the general population (50). Among OLT recipients, the incidence varies from 0.14 to 2.8% (51). KS is caused by human herpesvirus 8 (HHV-8) and its incidence appears to vary by regional differences in seroprevalence. The highest incidence is found in Saudi Arabia while it has lowest incidence in Western nations like the US (52). Among the renal transplant population, HHV-8 seropositivity before transplant is a risk factor for development of de novo KS and some evidence has suggested that donor HHV-8 seropositivity may increase the recipients risk for KS as well (53). Interestingly, KS seems to be directly related to the degree of immunosuppression and lesions have been known to regress when immunosuppression is discontinued (54). All cases of KS have been of the skin, but there is one reported case of gastric KS after OLT (55).

Post-transplant lymphoproliferative disorder

The second most common cancer in OLT recipients are post-transplant lymphoproliferative disorders (PTLD). In the adult OLT population, the incidence has varied from 0.9 to 2.6% in large case series (2, 23) (Table 1) compared with 0.03% in the general population (27). The mean time to diagnosis of PTLD has been reported to be between 26 and 32 months in these large case series (2, 16, 21, 41).

Epstein–Barr virus (EBV) is considered a risk factor for development of PTLD in patients with congenital and acquired immunodeficiency (56). EBV infects the B cell in two ways. Firstly, it causes a lytic infection where the virus replicates and lyses host cells releasing its progeny to infect other B cells. Secondly, and more important to the pathophysiology of PTLD, is a latent infection where EBV incorporates itself into B cell DNA allowing the virus to replicate unchecked (56). Naturally, immunosuppression designed to suppress T-cell activity would increase an OLT recipient's risk of developing PTLD. Immunosuppressants like the calcineurin inhibitors, OKT3, and antibodies to thymocyte, for this reason, have been shown to increase one's risk for PTLD. Conversely, drugs that do not suppress T-cell activity like mycophenolate mofetil and anti-IL2 receptor monoclonal antibodies have not been shown to increase OLT recipient's risk of PTLD (57).

Most of the data supporting the role of EBV in the development of PTLD is found in the pediatric transplant literature. Children have a higher likelihood of being EBV seroconversion following transplant (most donor livers are EBV positive) along with the required immunosuppression seem to place younger people at considerably higher risk for PTLD (58). Similarly, cytomegalovirus-(CMV) negative patients who received a CMV positive liver have a four- to six-fold higher risk of PTLD indicating that CMV seroconversion may be a risk factor for PTLD in adults as well (59).

Hepatitis C virus may also be a risk factor for the development of PTLD. Bellnoch and colleagues found that HCV was a independent risk factor for the development of PTLD and this has been supported by other reports as well (13, 60, 61). The mechanism by which HCV can cause PTLD is intriguing to consider. Unlike EBV, HCV is an RNA virus and cannot incorporate into the human genome. However, the HCV envelope protein E2 binds to CD81on the surface of hepatocytes and B cells allowing it to translocate into host cells and cause monoclonal expansion of B cells via activation of the anti-apoptotic gene Bcl-2 (62).

Colorectal cancer

The incidence of colorectal cancer in the OLT population has been reported to be 0–0.6% (2, 21, 41) (Table 1) compared with the incidence in the general population of 0.05% according to SEER data (27) and thus has been reported more frequently after OLT than in the general population. This finding of increased risk is confirmed by a recent systematic review (63). A study from the Netherlands showed a 12-fold increased risk of colorectal cancer after OLT (7). But it has been suggested that this database includes adenomas as well which can lead to an exaggerated disease burden (52). The mean time to diagnosis of colon cancer after OLT has been reported to be between 16 and 50.6 months (6, 17, 41). Similarly, a study by Albright et al. (64) found that colonic polyp formation in OLT recipients was comparable to the immunocompetent population.

Colorectal cancer when detected in the transplanted population tends to appear at an earlier age and confer a poorer prognosis compared to the general population. In a study that included 150 solid organ transplants (62% kidney, 19.3% heart, 18% liver and 0.7% lung), the mean age of diagnosis of colorectal cancer was 54 years old compared with 72 years in the SEER database (65). These patients also had decreased 5-year survival compared with the general population. These results are supported by a study of over 5600 solid organ transplants performed at the University of Wisconsin (66). In order to minimize the development of colorectal cancer from pre-cancerous lesions, the authors of this study suggested obtaining baseline colonoscopies on patients 50 years old and older who will be receiving a transplant and considering a baseline colonoscopy on patients younger than 50. They also suggest obtaining a screening colonoscopy 2 years after transplant. A study that investigated ‘intensive’ screening after OLT showed that performing screening colonoscopy 3 years after OLT was showed to decrease mortality, so the actual timing of colonscopy requires additional investigation (67).

Interestingly, patients receiving OLT for primary sclerosing cholangitis (PSC) appear to be at additional increased risk for developing de novo malignancy compared with OLT recipients for other causes with an incidence as high as 9.6% (18, 68, 69). Papatheodoridis et al. (70) was unable to detect any cases of colorectal cancer among OLT recipients with PSC, but reported new cases of ulcerative colitis and worsening ulcerative colitis among 30 OLT patients being transplanted for PSC both of which increases one's risk of developing colon cancer. Lastly, Fabia et al. (71) showed in a retrospective study of 1 085 OLT recipients, the incidence of colorectal cancer was found to be 7.9% in patients with known ulcerative colitis (UC) vs 0.1% in patients without UC, suggesting an increased risk of colorectal cancer in patients with UC following OLT (Table 2).

Head and neck cancers

The incidence of head and neck cancers in OLT recipients has been reported to range from 0.1(55)% to 2% in large retrospective trials (23, 55) (Table 1) compared to 0.01% in the general population (27). The mean time to diagnosis of head and neck cancer after OLT has been reported to range between 34.3 months and 61.2 months (6, 17, 41, 72, 73).

History of alcohol use seems to predispose OLT recipients to oropharyngeal cancer. In a study by Duvoux et al. (73), the incidence of oropharyngeal cancer among patients transplanted for alcohol was 16.7%, vs none in patients transplanted for nonalcoholic indications (P=0.001) (Table 2). Among transplants for alcoholic cirrhosis, Jain et al. (74) showed a risk 25.5 times of oropharyngeal cancer greater than transplants for non-alcoholic cirrhosis. Similarly, Bellamy et al. has shown that upper aerodigestive squamous carcinomas were overrepresented in OLT recipients transplanted for alcoholic cirrhosis (75). Interestingly a single centre study where none of the patient's had a history of alcohol nor cigarette use, there were no cases of oropharyngeal cancer (76).

In a study by Jonas et al. (21), two of 33 cancers detected after OLT were tongue cancers, both of which occurred in patients with a history of smoking. Jain et al. (6) found that seven of 57 OLT recipients developed oropharyngeal cancers, which was found to be 7.6 times higher than the general population according to SEER data. Furthermore, four of the seven patients that developed laryngeal cancer had a significant tobacco history. These results suggest that smoking may also be an individual risk factor for the development of head and neck cancers. But the effects of smoking and alcohol have been difficult to separate as long term alcohol use has been shown to potentiate the carcinogenic effects of cigarette smoke (77) and that heavy drinkers tend to also be heavy smokers (78).

Lung cancer

The incidence of lung cancer among OLT recipients is higher than that of the general population. The incidence of lung cancer in OLT recipients has been reported to be from 0 to 1.2% (17, 23) (Table 1) compared with 0.06% in the general population. The mean time to lung cancer diagnosis in OLT recipients ranged from 42 to 50 months in large case series (6, 21, 41, 55). Smoking as a cause of lung cancer in the general population has been well established and it appears that smoking history would also predispose OLT recipients given their immunosuppressant state (Table 2). In the Jonas et al. (21) series three lung cancers were detected in OLT recipients – all three patients were former smokers. In the Jain et al. (6) series, eight lung cancers were detected with five of the eight patients having a history of smoking, while in the other three patients, history of smoking was never documented in the medical record. Thus, like the general population, it would appear that smoking increases the OLT population's risk of lung cancer.

A study by Jimenez et al. (79) showed that patients being transplanted for alcoholic cirrhosis qhad higher rates of lung cancer than those transplanted for other causes (4.3 vs 0.7%, P<0.001). This result, however, could be confounded by the fact that patients who drink alcohol are more likely to smoke cigarettes as 73.3% of de novo lung cancer patients in this study were both heavy drinkers and intermediate or heavy smokers. So just as alcoholic cirrhosis is associated with head and neck cancers, alcoholic cirrhosis is also associated with lung cancer. Smoking is associated with alcoholic cirrhosis and is presumably associated with lung cancer in the OLT population as well, though no study has reported this data to date.

Jimenez et al. (79) also found that de novo lung cancer was usually diagnosed in advanced stages and carried a worse prognosis. Lung cancers were diagnosed from a mean range of 86±standard deviation of 43 months after OLT. Of the lung cancers detected, 66.6% were Stage IV, 20% were Stage IIIB, and 13.3% were Stage IIB. This is similar to the findings other studies (22, 41). Thus in high risk patients (i.e. smokers and patients undergoing transplant for alcoholic cirrhosis), post-OLT surveillance is paramount. Some clinicians have used annual chest X-rays or computed tomography (CT) scans to screen for lung malignancies (67).

Gynaecological cancers

In large retrospective trials of OLT recipients, the most common gynaecological cancer was breast cancer with incidence ranging from 0.2% to 0.7% (21, 55) (Table 1). Non-breast gynaecological cancers (cervical and ovarian) in the OLT population had an incidence of 0–1.5% (6, 21) (Table 1). The incidences of breast, cervical, uterine, and ovarian cancers in the general population according to age-adjusted SEER statistics are 0.1, 0.008, 0.02 and 0.01% respectively (27). Thus it appears that gynaecological cancers are increased in the OLT population compared to the general population. In breast cancer, mean time to diagnosis ranged from 41 to 124 months in large retrospective trials (6, 17, 41, 55). The mean time to diagnosis of non-breast gynaecological cancers ranged from 1 to 59 months (17, 21, 41, 55).

The case series by Jain et al. (6) showed that breast cancer occurred at a rate 1.9 times lower, but this finding was not statistically significant. The authors suggested that this trend to decreased frequency was because of breast cancer screening (i.e. mammograms) that women obtain regularly through their primary care physicians. Hence, many of these cancers were detected early and eliminated before patients even received liver transplantation.

Further, there is evidence to suggest that this early detection leads to decreased mortality in OLT recipients who develop gynaecological cancers. Jonas et al. (21) described 10 de novo gynaecological cancers diagnosed among 458 OLT recipients. Three patients were diagnosed with breast cancer. All forms of cancer had no evidence of metastases and were treated with either lumpectomy or mastectomy followed by adjuvant radiotherapy or anti-hormonal therapy with all patients being disease free at 8, 26 and 45 months of follow-up respectively. Similarly, 7 patients who developed de novo cervical intraepithelial neoplasia (CIN), grades 1–3, had no recurrence of disease after either conization or hysterectomy suggesting that when gynaecological cancers did appear after OLT, they appeared locally and carried a relatively good prognosis. This is supported by an even larger study by Sanchez et al. which showed that of all cancers detect post-OLT, breast cancer had the lowest tumour specific mortality (of seven cases only one patient died from complications of breast cancer) (41). Given this data, yearly screening of OLT patients with mammograms and Papanicolau (pap) smears has been used to detect gynaecological cancers in the OLT population and has been shown to increase survival after OLT (67). This approach is more aggressive than recent screening recommendations by the American College of Obstetrics and Gynecology for the general population (80).

Genitourinary cancer

In large series of OLT recipients, this incidence of non-prostate genitourinary cancer ranged from 0 to 0.4% (13, 55). The incidence of prostate cancer has ranged from 0 to 0.3% trials (2, 6, 17, 21, 36, 41), but five large trials did not detect a single case of prostate cancer (Table 1). The combined incidence of bladder and renal cancer is 0.03% while the incidence of prostate cancer is 0.2% according to SEER statistics (27). Thus it appears that the incidence of non-prostate genitourinary cancer is increased, but prostate cancer may be comparable to the general population. Mean time to diagnosis of prostate cancer ranged from 5.8 to 18.4 months while mean time to diagnosis of non-prostate genitourinary cancer ranged from 20 to 55.3 months (6, 17, 41, 55).

Compared with bladder and renal cancer, it appears that patients diagnosed with prostate cancers were diagnosed sooner, had more localized disease, and decreased tumour-specific mortality (6). Of the five patients that developed genitourinary cancers in the series, three were prostate cancers and occurred locally. They were treated with surgery, radiation and anti-androgen therapy with all three living at range of 41–72 months of follow-up. The other two patients had renal cell carcinoma and bladder cancer, but both died less than 2 years after diagnosis.

Data in these series occurred in the era of prostate-specific antigen (PSA) screening for men aged 50 and older. In this era of intensive screening, it is possible that many prostate cancers were detected early and eliminated. However, in 2008, the United States Preventative Services Task Force (USPSTF) released guidelines stating that there was insufficient evidence available to mandate prostate cancer screening and as a result the general population is generally not screened for prostate cancer currently (81). This is generally because the PSA can be elevated in benign prostate conditions and potentially expose patients to unnecessary treatments or surgeries. However, given the OLT population's overall increased risk for malignancy and data supporting decreased prostate specific mortality in patients who were screened in an era of intensive screening, recent data has suggested that it is reasonable to obtain baseline PSAs in OLT candidates and measure PSAs annually after OLT to screen for prostate cancer (67).

Miscellaneous cancers

Rarer still were miscellaneous cancers detected after OLT. There are been solitary reports of papillary thyroid cancer, glioblastoma, SCC of the conjunctiva, small bowel, pituitary, stomach, embryonic testicular, haemangioblastoma, pancreatic cancer, liver cancer, metastatic adenocarcinoma of unknown origin, sarcoma, and cholangiocarcinoma and Kaposi's sarcoma of the stomach in the post OLT population (Table 1) (2, 6, 13, 17, 21, 36, 41, 55). Given their relative infrequency in large case series, these cancers are likely sporadic and not necessarily directly related to OLT.

Malignancy surveillance after orthotopic liver transplant

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

As these data suggest, OLT recipients are at increased risk for de novo malignancy and current research has investigated intensive screening protocols for this population to improve survival after OLT (Table 3). Finkenstedt et al. showed that the median non-skin cancer-related survival increased from 1.2 to 3.3 years (P=0.021) years after intensive screening and overall survival increased from 3.1 to 11.3 years (P=0.001) (67). The intensive protocol implemented included the following annual screening: chest and abdomen computed tomography (CT), measurement of PSA, pap smear, mammography, and dermatological screening. Colonoscopy was performed 3 years after OLT and every 5 years thereafter. In patients with adenoma and prior history of inflammatory bowel disease, colonoscopy was performed 1 year after OLT. The non-intensive screening protocol consisted of annual chest X-rays, and abdominal ultrasound. Patients with history of prior malignancy were screened with chest and abdomen CT. Pap smears and mammograms were performed according to standard of care and dermatological screening was not performed routinely. A similar screening protocol for OLT recipients also showed increased survival in the intensive screening group as well (82). Cost-effectiveness analysis of these intensive screening protocols to our knowledge has not been performed.

Table 3.   Intensive screening protocols for tumour surveillance in orthotopic liver transplant recipients
ReferenceTraditional screening*Intensive screening*Major findings
  • *

    Each test was performed annually unless otherwise noted.

  • †Only in patients with pre-liver transplant history of malignancy.

  • ‡According to standard of care.

  • §

    §Performed 3 years after OLT and every 5 years thereafter. In patients with adenoma or inflammatory bowel disease prior to liver transplant, the first colonoscopy was performed 1 year after OLT.

  • Performed 1 year after OLT in patients with adenoma on pre-OLT colonoscopy and repeated every 2–4 years if more adenomas were found. If no adenomas were found, colonoscopy was repeated every 10 years in patients>50 years old.

  • CT, computed tomography; CXR, chest X-ray; ENT, ear, nose, and throat; Pap, Papanicolaou; US, ultrasound.

Finkenstedt et al. (67)CXR Abdominal US Chest and Abdominal CT Mammography and urologic screeningChest and Abdominal CT Urologic screening (PSA) Gynaecologic screening (Pap smear and mammography) Skin examination Colonoscopy§Improved median non-skin tumour-related mortality in the intensive screening group (1.2 vs 3.3 years, P=0.021) Improved overall survival in the intensive screening group (3.1 vs 11.3 years, P=0.001)
Herrero et al. (82)None, patients presented symptoms or were incidentally diagnosedCXR Abdominal US Mammography (every 2 years) Colonoscopy ENT clinic visit (in patients with >20 pack year smoking) CT scan (site not specified, in patients with >20 pack year smoking) PSA (age>55)Improved survival in the intensive screening group. Median survival of traditional screening group was 13.5 months while all patients in the intensive screening group were alive after median follow up 25 months (P=0.002)

Discussion and recommendations

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Orthotopic liver transplant is now an established life saving procedure for patients with acute and chronic liver failure. One increasingly recognized complication of this procedure is that of post-transplant malignancy. The immune system in the immunocompetent population has a variety of mechanisms involved in tumour surveillance and tumour lysis. Ironically, it is the immunosuppression required to prevent allograft rejection that puts OLT recipients at highest risk for the development of de novo tumour. A number of interesting studies have suggested that immunosuppression is indeed the culprit for the development of de novo tumour following transplant. Bellnoch et al. (13) showed that after the introduction of more potent immunosuppressants (i.e. tacrolimus) at their transplant center in 1995, de novo tumours have increased and appeared at shorter duration of follow-up. Higher degrees of immunosuppression were also shown to increase transplant recipient's risk of de novo tumour in a dose-dependent relationship (14). Saigal et al. was unable to detect an increased risk of tacrolimus administration on de novo tumours. However, their study was limited by number of included patients and types of malignancies assessed (17).

Other promising options for immunosuppression following OLT have led to some safety concerns. In vitro, sirolimus, a macrolide antibiotic which interferes with the interleukin-2 pathway, has been shown to suppress hepatoma growth (83). Guba et al. (84) found that unlike cyclosporine, sirolimus inhibited the growth of lung metastases in the murine model for colon cancer The authors also found that sirolimus inhibited vascular endothelial growth factor (VEGF) in vivo and in vitro leading to decreased neovascularization of the lung metastases. Others have demonstrated that sirolimus should be considered in patients transplanted for hepatocellular carcinoma OLT (85).

Emerging evidence seems to suggest that smoking places OLT recipients at higher risk for non-skin malignancies, but is inconclusive in regards to malignancy-related mortality. However, like the immunocompetent smoking population, OLT patients that smoke have higher rates of cardiovascular-related mortality, sepsis-related mortality, and all cause mortality compared to OLT recipients that do not smoke. The data available also suggests that patients receiving OLT for alcoholic cirrhosis have a poorer prognosis than OLT for other indications which is in line with in vitro studies showing alcohol related suppression of natural killer cell activity. To our knowledge, all clinical data has been collected retrospectively and prospective trials would be required to truly confirm a relationship though de novo tumour following OLT is relatively rare. Older age is likely a risk factor for the development of de novo tumour – OLT recipients over 40 years old should be considered at higher risk for de novo tumour though this data is retrospective as well. While smoking and alcohol use, age and existence of premalignant conditions raise suspicion for de novo tumour development, prevention and screening after OLT is of paramount importance as no formal risk stratification has been developed. In the patients with IBD who have undergone transplant for PSC, specific screening protocols have not been firmly established though some have suggested annual colonoscopy (17).

The EBV and CMV seroconversion following OLT seems to place patients at higher risk for developing PTLD. Recent evidence from the pediatric transplant literature has shown that monitoring EBV viral load levels and reducing immunosuppression accordingly (58) has been effective in preventing the onset of PTLD. Further, a study that treated eight PTLD patients (four children aged and four adults) with disease refractory to reduction of immunosuppression with HLA-matched allogeneic cytotoxic T cells. Three of the four children experienced complete remission from PTLD while one adult experienced only partial remission. The remaining four patients experienced no remission (86). Interestingly, success of adoptive T-cell therapy seemed to favour children over adults, but larger controlled trials are needed to confirm this result. Early on, ganciclovir monotherapy or in conjunction with intravenous CMV immune globulin was proposed for prophylaxis against CMV-related diseases after organ transplantation (87). However, due to the poor bioavailability of oral ganciclovir and the infeasibility of long-term intravenous antiviral therapy, oral valganciclovir has replaced ganciclovir for prophylaxis against CMV disease. Whether to provide universal prophylaxis or preemptive therapy is still up for debate (88). Hence, while EBV and CMV seroconversion pose a theoretical risk to the development of PTLD in OLT recipients, but the role of anti-viral therapy, if any, in the peritransplant period has not yet been established.

Based on recent studies that have investigated intensive screening protocols, we would recommend the following annually to OLT recipients to screen for malignancy regardless of the patient's age: physical examination (including examination of the oropharynx and full body dermatological exam), urinalysis, PSA, Pap smear and mammography. Screening protocols involving annual CT chest, abdomen and pelvis, has recently been shown to decrease cancer related mortality and thus has been validated, but radiological burden must also be taken into account. An estimated one in 270 women and one in 600 men who had CT angiography have been found to develop cancer from that CT scan (89) and these numbers would certainly be greater in the OLT population. To decrease radiation induced cancers, annual chest X-rays may be a reasonable alternative to CT of the chest (90). Abdominal and pelvic ultrasounds have not been validated as a screening modality in OLT recipients, but could represent a future alternative to CT.

The current American College of Gastroenterology recommendations for colorectal cancer screening is colonoscopy every 10 years for patients of usual risk and every 5 years for patients at increased risk (i.e. family history of colon cancer) (91). Thus, we would recommend screening OLT patients like increased risk patients by performing colonoscopy at least every 5 years. Published data has suggested that performing the first colonoscopy 3 years after OLT, but 1 year in patients with inflammatory bowel disease or primary sclerosing cholangitis improves survival (67). If dysplastic polyps are found, these patients will need to be screened at even more frequent intervals.

Conclusions

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References

Orthotopic liver transplant recipients are at increased risk for the development of post-transplant malignancies. While the major risk factor for the development of de novo malignancy after OLT appears to be immunosuppression, the emergence of alcoholic cirrhosis as an indication for liver transplant should raise clinicians' awareness of morbidity and mortality associated with head and neck and lung cancers. Surveillance protocols for OLT recipients are currently being developed and if cost-effective, should become standard of care for this high-risk group of patients.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Major risk factors and predisposing conditions
  5. Incidences of major de novo malignancies following orthotopic liver transplant
  6. Malignancy surveillance after orthotopic liver transplant
  7. Discussion and recommendations
  8. Conclusions
  9. References