Excellent quality of life after liver transplantation for patients with perihilar cholangiocarcinoma who have undergone neoadjuvant chemoradiation


  • Sarwa Darwish Murad was a recipient of the 2010/2011 Clinical and Translational Research Fellowship in Liver Diseases Award from the American Association for the Study of Liver Diseases/the Liver Institute and Foundation for Education and Research (AASLD/LIFER).

Address reprint requests to W. Ray Kim, M.D., Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First Street Southwest, Rochester, MN 55905. Telephone: 507-538-0254; E-mail: kim.woong@mayo.edu


Patients with perihilar cholangiocarcinoma (CCA) undergoing neoadjuvant chemoradiation followed by liver transplantation (LT) have excellent survival. However, little is known about their quality of life (QOL). We assessed the QOL of these patients and compared it to the QOL of patients who underwent transplantation for other liver diseases. From 1993 to 2010, 129 CCA patients underwent LT, and 93 (72%) were alive as of November 2010. All recipients were sent a previously validated QOL questionnaire composed of disease-specific QOL metrics (liver disease symptoms, Karnofsky score, health perception, and index of well-being) and generic QOL metrics [Short Form 36 (SF-36) and European Quality of Life (EuroQol)]. These recipients were compared to 110 transplant recipients with other liver diseases (excluding hepatitis C). Among the recipients with CCA, the response rate was 85% (n = 79). Patients with CCA did significantly better on liver disease symptoms (3.3 versus 3.2, P = 0.05), the Karnofsky score (90.8 versus 86.6, P = 0.03), the SF-36 Physical Functioning domain (52.0 versus 46.3, P < 0.001), and the EuroQol Mobility category (10% versus 33%, P = 0.001), and they rated their overall health better in comparison with non-CCA patients (85.9 versus 80.7, P = 0.02). CCA patients scored consistently higher on all other domains, albeit without significant differences. The observed differences in QOL remained unchanged when adjustments were made for demographic factors, including the level of education. In conclusion, patients who underwent neoadjuvant chemoradiation followed by LT for perihilar CCA reported excellent QOL that was equal to or better than that of recipients with other liver diseases. These results are important in light of the continued debate about the feasibility of this aggressive treatment in patients with perihilar CCA. Liver Transpl 19:521–528, 2013. © 2013 AASLD.




European Quality of Life


liver transplantation


National Institute of Diabetes and Digestive and Kidney Diseases


quality of life


standard error


Short Form 36.

Early-stage perihilar cholangiocarcinoma (CCA) has undergone a therapeutic paradigm shift over the last 2 decades. Although surgical resection remains the treatment of choice for resectable disease, liver transplantation (LT), when it is combined with neoadjuvant chemoradiation, has emerged as a curative option for select patients with unresectable disease. Indeed, previous studies from our institution[1-4] as well as others[5] have shown that a 5-year recurrence-free survival rate of approximately 70% is achievable in these highly selected LT recipients. The posttransplant recurrence of CCA has been reported in less than 20% of patients successfully completing neoadjuvant chemoradiotherapy.[2, 6]

The neoadjuvant treatment used to achieve these results is not free of toxicities. Patients who undergo this protocol suffer from a wide range of side effects, including common toxicities from chemoradiotherapy such as nausea, vomiting, diarrhea, and fatigue and more serious events such as recurrent bacterial cholangitis, gastroduodenal ulcerations, and gastrointestinal dysmotility.[7] Moreover, radiotherapy can cause vascular damage and increase the risk of perioperative complications. For example, a stricture or thrombosis of the hepatic artery or portal vein after LT has been reported in approximately 20% of patients.[8] Most of these side effects are manageable and are considered reasonable when they are weighed against the substantial gain in long-term survival for patients with this otherwise lethal disease. Thus, it is important to understand the impact of this treatment on the quality of life (QOL) of these patients after transplantation.

Self-reported health-related QOL has been studied for a variety of chronic liver diseases, both before[9-11] and after transplantation.[12, 13] Studies comparing QOL before and after LT have uniformly shown highly statistically significant and clinically meaningful improvements after the procedure. Those studies have commonly used generic instruments such as Short Form 36 (SF-36) and European Quality of Life (EuroQol) as well as disease-specific ones. For example, in the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Liver Transplantation Database study, a QOL instrument was designed for LT recipients to capture the overall QOL, subjective health status, and functional status as well as liver-specific symptoms.[14]

To date, QOL in CCA patients who have been treated with LT has not been studied. The posttransplant QOL of patients with perihilar CCA not only is determined by pretransplant events such as irreversible consequences of the neoadjuvant therapy but also is influenced by the fear of the cancer recurring as well as common posttransplant complications such as adverse effects from immunosuppressive therapy, the risk of infections, and the need for frequent health care contacts. Therefore, the aim of our study was to assess QOL in patients who had undergone neoadjuvant chemoradiation and LT for perihilar CCA in comparison with recipients of LT for indications other than CCA.


Patient Population

Since January 1993, all patients presenting to our institution with unresectable perihilar CCA have been considered for a multimodality treatment protocol consisting of neoadjuvant chemoradiation, surgical staging, and eventually LT as previously described.[4] All patients are evaluated by an experienced hepatobiliary surgeon to determine resectability, and the neoadjuvant protocol is considered only for those deemed unresectable because of bilobar involvement, involvement of major hilar structures, or underlying primary sclerosing cholangitis. Patients who are eligible receive external-beam radiotherapy (target dose = 4500 cGy) with concomitant chemotherapy (usually a continuous infusion of 5-fluorouracil) followed by intraluminal brachytherapy (target dose = 2000-3000 cGy), after which maintenance chemotherapy is continued until LT. Further details on the selection criteria, neoadjuvant regimen, surgical techniques of the staging operation and LT, and results of post-LT follow-up have been described previously.[2-4, 15] This treatment has not changed substantially over time.

From January 1993 to October 2010, 199 patients were enrolled in this protocol, and 129 underwent successful LT (65%). Thirty-six patients (28%) died because of cancer recurrence (n = 24), multiorgan failure (n = 3), invasive intracranial aspergillosis (n = 1), graft-versus-host disease (n = 1), a massive pulmonary embolism (n = 1), sepsis (n = 1), a subdural hematoma (n = 1), a pulmonary infection (n = 1), massive intraoperative intra-abdominal hemorrhaging (n = 1), or posttransplant lymphoproliferative disorder (n = 1). One patient died in an outside hospital of unknown cause. Thus, as of November 2010, 93 patients were alive (72%), and these patients constituted our patient population. These patients were contacted by mail for a QOL survey with the questionnaire described later. For initial nonresponders, reminder mailings were sent up to 2 times (for a total of 2 or 3 mailings). The study was approved by the Mayo Foundation institutional review board, and all participants gave informed consent.

Comparison Population

The comparison group in this study consisted of patients who underwent LT for an indication other than CCA. These patients were taken from a prospective study conducted to investigate the impact of recurrent hepatitis C after LT.[16] In that study, the same QOL instrument was administered to ambulatory patients seen at our posttransplant clinic. There were 110 patients without hepatitis C who belonged in the comparison group; these patients were used in this study as the comparison group for the CCA patients. Patients with hepatitis C were not included in the present analyses because those patients in general and patients with recurrent hepatitis C in particular reported significantly lower QOL than patients without hepatitis C. Hence, the inclusion of these patients would have created a bias in favor of the CCA patients. Our comparison patients underwent transplantation for hepatocellular carcinoma (n = 20), primary sclerosing cholangitis (n = 20), primary biliary cirrhosis (n = 9), alcoholic liver disease (n = 19), cryptogenic cirrhosis (n = 12), nonalcoholic steatohepatitis (n = 11), autoimmune hepatitis (n = 5), polycystic liver disease (n = 4), amyloidosis (n = 4), hepatitis B (n = 3), alpha-1-antitrypsin deficiency (n = 2), or Budd-Chiari syndrome (n = 1). Henceforth, they are called non-CCA patients.

QOL Instrument

This study used instruments for both generic and disease-specific QOL. The disease-specific instrument was the modified NIDDK Liver Transplantation Database QOL form, which captures the overall QOL, subjective health status, and functional status as well as liver-specific symptoms.[14] The instrument version used in this study was previously validated for LT and non-LT patients.[17, 18] It consists of 87 items and addresses the following domains: liver disease symptoms, Karnofsky performance status,[19] health perception, and index of well-being.

The liver disease symptoms scale is composed of 19 items for females and 20 (including impotence) for males and includes symptoms specific to cholestatic liver disease (eg, itching of skin and joint aches or pains), symptoms found with chronic liver disease (eg, abdominal swelling, fluid retention, and sleeplessness/insomnia), symptoms related to immunosuppression (eg, trembling or shakiness and change in facial appearance), and more general health symptoms (eg, fatigue and difficulty concentrating). All items are scored on a 5-point Likert scale ranging from 0 (extremely) to 4 (not at all), and the calculated mean score ranges from 0 (worst outcome) to 4 (best outcome). The Karnofsky performance status is represented by its single standard question and ranges from 30 (worst) to 100 (best). Health perception is based on a single question regarding the level of health satisfaction and is scored from 1 (completely dissatisfied) to 7 (completely satisfied). The index of well-being is the weighted average of overall life satisfaction, which is represented by 1 item scored from 1 (completely dissatisfied) to 7 (completely satisfied), and the index of general affect, which is composed of 8 bipolar adjective pairs and is scored on a visual 7-point scale to reflect where the respondent feels himself or herself to be between these poles (eg, boring-interesting and enjoyable-miserable). The weighted mean score ranges from 2.1 (worst) to 14.7 (best). The scoring algorithm has been previously published and is available from the corresponding author.[18]

In addition, our study used 2 generic instruments, SF-36[20, 21] and EuroQol,[22] which are both extensively used for chronic liver diseases in pretransplant and posttransplant settings.[10, 12] The SF-36 survey is composed of 36 items divided into 8 individual scales: General Health, Physical Functioning, Role–Physical, Role–Emotional, Social Functioning, Mental Health, Bodily Pain, and Vitality. Each scale score is a simple algebraic sum of responses for all items within that scale under the assumption that all items in the same scale carry equal weight. These are subsequently grouped into 2 composite scores: the Physical Component Score (composed of Physical Functioning, Role–Physical, Bodily Pain, and General Health) and the Mental Component Score (composed of Vitality, Social Functioning, Role–Emotional, and Mental Health). Each scale is directly transformed into a 0 (worst) to 100 (best) scale (ie, raw scores). In addition, these scores are calibrated to those of an age- and sex-matched general population, which is represented by a reference score of 50 ± 10 (ie, a normalized score). EuroQol consists of 5 categorical items (Mobility, Self-Care, Usual Activities, Pain/Discomfort, and Anxiety/Depression) and 1 continuous item evaluating overall health on a visual analogue scale. The former categories are scored as (1) no problem, (2) some problem, or (3) major problem, and the summary scores are presented in a dichotomous fashion as any problem (2 or 3) versus no problem (1). The latter is represented by a horizontal ruler ranging from 0 (worst health) to 100 (perfect health).

Statistical Analysis

The results are expressed as numbers and percentages for categorical variables and as means and standard deviations for continuous variables. Comparisons between groups are based on the chi-square test for categorical variables and on the Mann-Whitney U test for continuous variables. Univariate and multivariate linear regression analyses were used to identify specific determinants of QOL domains. Subgroup analyses for patients completing the survey early ≤ 1 year and late (> 1 year) after LT were performed for each group to examine whether there was any effect of the time elapsing since LT. All analyses except the SF-36 analysis were performed with SPSS version 16.0 (SPSS, Chicago, IL). The SF-36 scoring was analyzed with SAS version 9.2 (SAS Institute, Cary, NC).


Respondent Characteristics

Seventy-nine of the 93 known surviving LT recipients who had undergone neoadjuvant chemoradiation for perihilar CCA (all without tumor recurrence) completed the questionnaire (ie, the CCA group; response rate = 85%). These were compared to the 110 non-CCA patients who underwent transplantation for other liver diseases. Table 1 summarizes the demographics for the 2 groups. The median ages were similar, but as expected, the CCA group was disproportionately male (77% versus 56%). Patients with CCA completed the survey at a median of 4.3 years (range = 0.2-17.8 years) after transplantation, whereas non-CCA patients completed the survey at a median of 2.1 years (range = 0.3-7.1 years, P < 0.001); however, the proportions of patients completing the survey within the first year and patients completing the survey after the first year were similar (P = 0.87). There was no difference in the marital status between the 2 groups, whereas the level of education was significantly higher in the CCA group versus the non-CCA group. At the time of completion of the QOL form, 60% (n = 46) were employed in the CCA group, and 43% (n = 46) were employed in the non-CCA group (P = 0.03).

Table 1. Demographics at the Time of Completion of the QOL Survey of Patients Who Underwent Transplantation for CCA and Patients Who Underwent Transplantation for Other Reasons
DemographicsCCA Patients (n = 79)Non-CCA Patients (n = 110)P Value
  1. NOTE: The results are expressed as numbers and percentages for categorical variables and as medians and ranges for continuous variables. Percentages are based on the number of patients with complete data.

Age (years)56 (29-71)59 (24-73)0.12
Sex: male [n (%)]61/79 (77)62/110 (56)0.04
QOL survey completed [n (%)]   
Within the first posttransplant year7/79 (9)12/110 (11)0.87
After the first posttransplant year72/79 (91)98/110 (89) 
Marital status [n (%)]   
Married/cohabitating61/78 (78)90/110 (82)0.56
Single/divorced17/78 (22)20/110 (18) 
Highest education degree [n (%)]   
High school or less21/77 (27)48/110 (44) 
Trade school8/77 (10)17/110 (15)0.04
College/university29/77 (38)24/110 (22) 
Advanced19/77 (25)21/110 (19) 
Employment [n (%)]   
Employed46/77 (60)46/107 (43)0.03
Unemployed31/77 (40)61/107 (57) 

Generic Health-Related QOL (SF-36 and EuroQol)

Table 2 summarizes the SF-36 scale and summary scores. Without adjustments for age or sex, patients who underwent transplantation for CCA scored significantly better on Physical Functioning (P < 0.001), Role–Emotional (P = 0.006), and Bodily Pain (P = 0.03). CCA patients had higher scores on the other 5 scales, but the difference did not reach statistical significance. Likewise, the Physical Component Score was significantly higher for those who underwent transplantation for CCA (P = 0.01), whereas the Mental Component Score was not statistically different. When the scores were normalized to those of an age- and sex-matched general population, essentially all of the scores in both groups aggregated around the population norm of 50, and this indicated the recovery of health-related QOL after LT. Patients who underwent transplantation for CCA scored significantly higher on Physical Functioning (P < 0.001). The differences seen in the raw scores persisted in the normalized Role–Emotional and Bodily Pain domains, although they were statistically nonsignificant.

Table 2. SF-36 Questionnaire Scores
 SF-36 Raw Scores
CCA Patients (n = 79)Non-CCA Patients (n = 110)P Value
Scale scores   
General Health68.3 ± 22.266.0 ± 21.20.35
Physical Functioning86.2 ± 13.369.4 ± 25.4<0.001
Role–Physical65.4 ± 41.559.6 ± 40.80.38
Role–Emotional88.7 ± 25.776.1 ± 35.10.006
Social Functioning85.4 ± 21.280.7 ± 22.40.11
Mental Health82.3 ± 13.880.0 ± 15.30.31
Bodily Pain76.4 ± 20.368.3 ± 23.80.03
Vitality60.6 ± 23.158.2 ± 21.90.36
Summary scores   
Physical Component Score47.2 ± 9.643.0 ± 10.90.01
Mental Component Score53.6 ± 7.652.7 ± 9.50.92
 SF-36 Normalized Scores
CCA Patients (n = 79)Non-CCA Patients (n = 110)P Value
  1. NOTE: The results are presented as raw scores (ie, scale scores measured in the current groups) and normalized scores (ie, scale scores adjusted for an age- and sex-matched norm population with a default reference value of 50 with a standard deviation of 10).

Scale scores   
General Health49.3 ± 11.849.8 ± 9.80.80
Physical Functioning52.0 ± 6.346.3 ± 11.7<0.001
Role–Physical45.8 ± 13.445.9 ± 11.60.94
Role–Emotional52.1 ± 7.848.7 ± 10.30.68
Social Functioning50.7 ± 10.249.4 ± 9.60.52
Mental Health53.5 ± 7.552.4 ± 8.30.35
Bodily Pain51.9 ± 8.749.4 ± 9.40.09
Vitality49.3 ± 11.849.8 ± 9.80.51
Summary scores   
Physical Component Score48.7 ± 10.946.4 ± 10.60.08
Mental Component Score52.6 ± 7.751.6 ± 9.9>0.99

On the EuroQol instrument, patients who underwent transplantation for CCA reported significantly fewer problems with mobility (P < 0.001; Table 3). Problems with self-care, usual activities, and anxiety or depression were also less prevalent, but these failed to reach statistical significance. Figure 1 summarizes the overall evaluation of health on a visual analogue scale. Patients who underwent transplantation for CCA scored their current level of health as significantly better (mean = 85.9 ± 11.2) than patients who underwent transplantation for other diseases (mean = 80.7 ± 16.1, P = 0.02).

Figure 1.

Graphic representation of the EuroQol visual analogue scale in response to the following question: “How do you feel overall?” *A comparison of CCA and non-CCA patients revealed P = 0.02.

Table 3. Summary Scores for the EuroQol Questions
Any ProblemCCA PatientsNon-CCA PatientsP Value
  1. NOTE: The nominator refers to the number of patients scoring either some problem or a major problem with the specific item, and the denominator refers to the total number of patients who completed the question.

  2. a

    Work, study, housework, leisure, and family activities.

Mobility [n/N (%)]7/68 (10)31/93 (33)0.001
Self-Care [n/N (%)]0/68 (0)4/94 (4)0.14
Usual Activities [n/N (%)]a15/68 (22)34/94 (36)0.06
Pain/Discomfort [n/N (%)]37/68 (54)51/94 (54)>0.99
Anxiety/Depression [n/N (%)]14/68 (21)31/92 (34)0.08

Disease-Specific QOL (NIDDK)

Figure 2 illustrates the responses for the liver disease symptoms scale. According to the composite summary score, patients with CCA reported significantly less severe symptoms than non-CCA patients, although the absolute difference was small (3.3 ± 0.5 versus 3.2 ± 0.6, P = 0.05). In general, patients with CCA scored equally as well as or better than non-CCA patients on all individual symptoms except bowel problems (2.7 versus 3.0, P = 0.05) and warts (3.8 versus 3.9, P = 0.35).

Figure 2.

Liver disease symptoms scale scores for patients who underwent transplantation for CCA (black bars) and patients who underwent transplantation for other liver diseases (striped bars). The top bar represents the Liver Symptoms Summary Score, which is the mean score of all the individual symptoms that follow. Significant differences (ie, P < 0.05) are indicated by asterisks.

The performance status as measured by the Karnofsky scale was significantly higher for patients who underwent transplantation for CCA [mean = 90.8 ± 9.4 (corresponding to the category “able to carry out normal activity, minor symptoms”] versus non-CCA patients [mean = 86.6 ± 12.4 (corresponding to the category “able to carry out normal activity with effort, some symptoms”), P = 0.03]. Similarly, patients who underwent transplantation for CCA tended to score higher on health perception (5.8 ± 0.9) than non-CCA patients (5.5 ± 1.0), but this did not reach statistical significance (P = 0.10). Both groups' scores correspond to a response between satisfied (5) and very satisfied (6). The weighted mean score of the index of well-being was marginally higher for those who underwent transplantation for CCA (12.6 ± 1.4 versus 11.8 ± 2.2, P = 0.08).

The mean numbers of sick days in the preceding month were equal for the 2 groups (1.4 ± 4.2 for CCA patients versus 1.3 ± 4.2 for non-CCA patients, P = 0.77). There were no significant differences in the extent to which patients reported that their health affected paid employment (20% versus 30%, P = 0.14), looking after their home (11% versus 19%, P = 0.14), social life (9% versus 11%, P = 0.57), home life (9% versus 5%, P = 0.27), sex life (24% versus 22%, P = 0.72), interests or hobbies (19% versus 23%, P = 0.53), or vacations (17% versus 18%, P = 0.80) after transplantation for CCA versus transplantation for other liver diseases.

Determinants of QOL

Table 4 presents the results of a linear regression analysis modeling the Physical and Mental Component Scores of SF-36 as well as the overall health evaluation (the EuroQol visual analogue scale) according to the demographic characteristics, diagnosis, and time since transplantation. A college education versus other levels of education was associated with significantly better scores on all 3 QOL indicators in both univariate and multivariate linear regression analyses. Patients with CCA continued to have a better Physical Component Score and health evaluation than non-CCA patients when adjustments were made for demographic differences, including education. The time since LT was not a determinant of QOL. There were no significant differences in any of the generic or disease-specific QOL domains between patients with CCA who filled out the QOL form in their first posttransplant year (n = 7) and patients with CCA who completed the form later (n = 72; data not shown). Likewise, the results did not change when we compared patients with CCA to the subset of controls undergoing transplantation for primary sclerosing cholangitis (n = 20; data not shown) or hepatocellular carcinoma (n = 20; data not shown).

Table 4. Linear Regression Analysis of Determinants for the Raw Physical Component Score and Mental Component Score and the Overall Health Evaluation (the EuroQol Visual Analogue Scale) for Patients Who Underwent Transplantation for CCA and Patients Who Underwent Transplantation for Other Liver Diseases
FactorPhysical Component ScoreMental Component ScoreEuroQol Visual Analogue Scale
B Coefficient ± SEP ValueB Coefficient ± SEP ValueB Coefficient ± SEP Value
Univariate analysis      
Age (years)−0.18 ± ± 0.070.82−0.18 ± 0.120.13
Males versus females3.15 ± 1.630.060.41 ± 1.360.770.18 ± 2.360.94
CCA versus non-CCA4.23 ± 1.560.0070.89 ± 1.320.505.21 ± 2.300.03
College education versus other5.93 ± 1.680.0012.02 ± 1.440.168.83 ± 2.42<0.001
Married/cohabitating versus other1.66 ± 1.990.410.75 ± 1.660.65−1.67 ± 2.800.55
Time since LT (years)0.48 ± ± ± 0.390.41
Multivariate analysis      
Age (years)−0.14 ± ± 0.070.74−0.11 ± 0.120.33
Males versus females1.84 ± 1.590.25−0.38 ± 1.470.80−1.33 ± 2.360.57
CCA versus non-CCA3.87 ± 1.700.030.36 ± 1.580.824.42 ± 2.560.09
College education versus other6.24 ± 1.69<0.0012.84 ± 1.560.077.65 ± 2.490.003
Married/cohabitating versus other1.79 ± 1.930.360.90 ± 1.780.61−0.90 ± 2.790.75
Time since LT (years)0.11 ± 0.290.690.21 ± 0.260.43−0.80 ± 0.420.85


A rigorous protocol of neoadjuvant chemoradiotherapy followed by LT has previously been shown to have excellent outcomes in terms of the quantity of life gained by select patients with perihilar CCA.[2-4, 15] However, this is the first study to date to investigate the QOL of these patients. Despite issues specific to their disease and treatment, we expected to find that their QOL would be comparable or noninferior to that of other LT recipients. The results of this study show that patients with perihilar CCA in general have excellent QOL after transplantation that is not only equal to but in some instances even better than that observed in patients who have undergone transplantation for other malignant and nonmalignant liver diseases. Indeed, it was surprising to see that CCA patients mostly did better on domains capturing physical functioning and performance, which are areas expected to be affected the most by pretransplant chemoradiotherapy. Our data are important because they provide an answer to common concerns about the impact of this aggressive protocol on life after LT. Moreover, they provide reassurance that the impact of such treatment on QOL is, for the most part, reversible.

There are several potential explanations for why patients with CCA report comparable or even better QOL than other transplant recipients. First, unlike patients with other chronic liver diseases such as alcoholic liver disease and viral hepatitis, these patients were living normal lives before their devastating diagnosis of CCA. Even though two-thirds of the patients have underlying primary sclerosing cholangitis, in one-third of these, CCA represents the first manifestation of their disease.[2, 6] Moreover, common behavioral risk factors such as alcohol and drug use are generally absent in this population. This is further reflected by the higher level of education and employment status for CCA patients versus non-CCA patients observed in our study. As a result, a better socioeconomic status and stronger social support provide a good buffer against adverse events after transplantation. Second, the acute deterioration caused by their illness and the aggressive treatment before transplantation provide a reference against which posttransplant events may be comparatively easy to handle. Moreover, this deterioration, however devastating it may be, is generally of short duration in comparison with more insidious patterns in chronic liver disease. Indeed, others have shown that prolonged disease duration is associated with worse QOL after LT.[12] Finally, without this treatment modality, the prognosis of patients with unresectable perihilar CCA is extremely grim, and life expectancy is generally measured in a matter of months. This aggressive protocol, therefore, provides the only hope for a cure. Once they have undergone transplantation, these patients are generally extremely grateful to be alive and are motivated to pick up their pretransplant lives as quickly as possible.

Although these characteristics of CCA patients were likely to have affected mental health rather than physical health, our data consistently showed that the CCA patients were better off physically than emotionally. The reason for this is not entirely clear. In our multivariate linear regression model explaining differences in the Physical Component Score, we found that having a college education was independently associated with improved scores. Patients with a higher level of education generally hold office jobs instead of jobs that demand more physical exertion. Adams et al.[23] found that transplant recipients with office jobs before transplantation are more likely to be employed in similar jobs after transplantation than those with manual jobs. It is, therefore, likely that for these patients, the level of physical activity achieved after transplantation is not far from their pretransplant baseline. Nevertheless, this does not fully explain why patients with CCA have better physical function because this effect was found to be independent of education. It is, however, possible that other factors not measured in this questionnaire, such as preexisting physical fitness, exercise routines, and dietary habits, are differently distributed in CCA patients and non-CCA patients.

The only symptom significantly more prevalent in patients who underwent transplantation for CCA versus non-CCA patients was a bowel problem. This can be explained by the fact that inflammatory bowel disease is prevalent in patients with CCA, who have underlying primary sclerosing cholangitis in two-thirds of cases. Indeed, 49 of our 79 patients (62%) had concomitant inflammatory bowel disease. Although their liver disease was treated by transplantation, the inflammatory bowel disease was not. Overall, the composite score for liver symptoms in the CCA group was comparable to that measured 1 year after transplantation for patients with nonmalignant cholestatic liver disease, as described previously.[18]

Our study has strengths and weaknesses. Its strengths include a high response rate, a unique patient population that can be studied with a significant sample size only at our institution, and the use of both generic and disease-specific standardized QOL instruments. Its weaknesses include first an asymmetric study design with a confined cohort of CCA patients, whereas the non-CCA patients were collected at random over a 2-year window. However, there is little reason to suspect that this introduced a systematic bias, and the case mix for the non-CCA patients was heterogeneous and fairly representative of our LT practice. Second, the timing of the survey was closer to transplantation for patients without CCA. However, QOL is expected to be most affected in the first year after transplantation,[13] and the proportions of patients taking the survey within and after that first year were equal in the 2 groups. Moreover, the time since transplantation was not found to be a predictor of QOL in our linear regression models. Third, since the start of the protocol, 26 patients experienced tumor recurrence; however, 24 of these patients died before this study. The 2 remaining patients did not respond to our questionnaire, and this theoretically could have led to an overestimation of the QOL of the CCA survivors. However, even if their QOL would have been poor in all aspects, the overall outcome would most likely not have been affected because it would have concerned only 2.5% of the total study population (2/79). Finally, the CCA and non-CCA populations differed in terms of sex and, more importantly, education, which are important factors in QOL. However, in multivariate analyses, the observed difference in QOL between CCA and non-CCA patients persisted even after corrections for differences in education.

In conclusion, patients with perihilar CCA who have undergone a rigorous protocol including neoadjuvant chemoradiotherapy followed by LT in general have excellent QOL after transplantation. Their QOL is equal to or, in the case of physical functioning and performance, better than that of transplant recipients with other liver diseases. The findings of this study are important in light of the continued debate about the feasibility of this aggressive treatment modality in this patient population with limited therapeutic options.