Orthotopic liver transplantation (OLT) is an accepted therapy for end-stage liver disease. In the last decades, short-term survival after transplantation has improved with nowadays a 1-year survival rate of 85% and a 5-year survival rate of 75% (http://www.optn.org). Besides recurrence of the primary disease, the main causes of death in liver recipients over the long term are cardiovascular events and malignancies.1–6 Therefore, it becomes more and more important to prevent diseases and health problems that are not associated with the original liver disease. Smoking is an important risk factor for cardiovascular events as well as several malignancies. Smoking is considered one of the leading causes of preventable death in the general population and could also be an important one in liver transplant recipients. However, studies about the smoking behavior of liver transplant recipients are scarce. DiMartini et al.7 prospectively studied the smoking behavior in patients transplanted for alcoholic liver disease (ALD), and they found that recipients resume smoking early after OLT and increase consumption over time. Ehlers et al.8 reported an active smoker rate of 15% after OLT and a relapse rate of 20% in former smokers. Both studies were performed in the United States.
Studies on the effects of smoking on the long-term course after liver transplantation are also scarce. Some studies have reported a possible relation between smoking and malignancies after OLT,9, 10 especially in recipients with ALD.11–13 In a previous study from our group, a relation between smoking and cardiovascular events was suggested in a group of 331 OLT recipients.6 In recipients of other solid organ transplants, many more studies on the effects of smoking have been performed. These have shown that tobacco use is associated with graft loss and mortality in renal transplant patients,14, 15 graft loss in pancreas transplant patients,16 cardiovascular disease in renal transplant patients,14, 17–19 and malignancies in renal,20 lung,21 and heart transplant patients.22
The primary aim of the present study was to examine smoking behavior before and after OLT and to look for patient groups at risk for resuming smoking after OLT. In addition, we looked for a relation between smoking and survival and morbidity after OLT.
ALD, alcoholic liver disease; FU, follow-up; HBV, hepatitis B virus; HCV, hepatitis C virus; OLT, orthotopic liver transplantation; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; SD, standard deviation.
PATIENTS AND METHODS
All adult patients who had undergone liver transplantation in our hospital between 1979 and May 2005 and had a follow-up of at least 2 years after transplantation were included. Patient characteristics, pretransplant malignancies (including skin) and cardiovascular status, and outcome variables were obtained through a retrospective analysis of the medical charts. The immunosuppression protocol has been described before.23, 24 Briefly, patients transplanted between 1979 and 1986 received azathioprine and steroids. After 1986, cyclosporine was added, and after 1997, patients received a calcineurin inhibitor (cyclosporine or tacrolimus) and steroids, with or without azathioprine. Some patients used sirolimus because of renal dysfunction.
Data on smoking behavior were collected in 2 ways: first, retrospectively from the medical charts and second, by a questionnaire. From the charts, we collected data at 4 time points: during evaluation for OLT (before listing), shortly before OLT, 2 years after OLT, and at the end of follow-up. At each time point, it was determined if a patient was an active smoker, a former smoker, or a never smoker. A questionnaire was sent to all patients presently alive. In this questionnaire, the patients were asked about their smoking habits at the 4 time points mentioned previously. It was asked if the patient was an active smoker, a former smoker, or a never smoker at each time point. In addition, questions were included on the number of years of tobacco use and, in the case of a former smoker, on when tobacco use was stopped, on the type of tobacco product, on the average number of cigarettes or otherwise per day, and on plans for smoking cessation and willingness to participate in a free smoking cessation course in our hospital. Smokers were defined as smoking 7 or more cigarettes per week. For converting cigars to cigarettes, we equated 1 cigar with 4 cigarettes, considering that an average cigar contains 4 g of tobacco and an average cigarette contains 1 g of tobacco.25
Pretransplant Cardiovascular Status and Outcome Variables
For pretransplant cardiovascular status, we recorded treatment for diabetes mellitus and hypertension and vascular events (angina pectoris, myocardial infarction, coronary interventions, transient ischemic attack, cerebral vascular accident, intermittent claudication, and intervention for large vessel disease).
For outcome variables, we looked at patient and graft survival, occurrence of hepatic artery thrombosis, de novo malignancies, and de novo vascular events. For graft survival, we recorded the need and reason for retransplantation. Dates were noted for all events, and events were recorded until the end of follow-up (May 2007).
The methods that we used were discussed with the medical ethics committee. According to Dutch legislation, there were no objections against the methods used. A returned questionnaire was considered informed consent.
Categorical variables are presented as frequencies with percentages, and continuous variables are presented as medians (range). The data were analyzed with the Statistical Package for the Social Sciences, version 14.0 (SPSS, Inc., Chicago, IL). Categorical variables were compared with the chi-square test, and continuous variables were compared with the Mann-Whitney test. Outcome variables were tested for significance with the Kaplan-Meier method with the log rank test. Differences are considered significant when P ≤ 0.05.
In our hospital, 867 liver transplants were performed between 1979 and May 2005. Of these, 526 were for adult patients undergoing at least 1 liver transplant. Eleven patients were lost to follow-up (6 living outside the Netherlands, 3 with follow-up elsewhere, and 2 unknown); of the remaining 515 recipients, 401 (78%) survived at least 2 years after transplantation, and these were included in the present study. Clinical characteristics of the 401 recipients are shown in Table 1. At the time of this study, 326 patients were alive, and 75 were deceased. The most frequent indications for transplantation were primary sclerosing cholangitis (PSC) and primary biliary cirrhosis.
Table 1. Patient Characteristics
Abbreviations: OLT, orthotopic liver transplantation; SD, standard deviation.
Number of patients
Age at OLT [years, median (range)]
Date of OLT [median (range)]
August 1997 (April 1979 to May 2005)
August 1998 (April 1979 to May 2005)
September 1993 (October 1979 to November 2003)
Follow-up after OLT (years)
Diagnosis [number of patients (%)]
Primary sclerosing cholangitis
Primary biliary cirrhosis
Viral (hepatitis B and/or C)
Acute liver failure
Malignancy in explant [number of patients (%)]
Immunosuppression 1 year after OLT
Immunosuppression end follow-up
Smoking Behavior According to the Questionnaire
Three hundred one of the 326 patients returned the questionnaire (a response rate of 92%). Next and in Figs. 1 to 4, the results at the following time points are presented: at the evaluation for OLT, at the time of OLT, 2 years after OLT, and at the end of follow-up [a median of 8.6 years (range, 2-26) after OLT].
Both before and after OLT, a majority of the patients had never smoked (53%-54%). About a third of the patients were former smokers (28%-32%), and a substantial minority of the patients (14%-18%) were active smokers. In Fig. 1, it is shown that at the different time points, the percentages are rather stable, with only a small decline in active smokers shortly before OLT.
Smoking Behavior of Never Smokers.
In Fig. 2, it is shown that of the 163 patients who had never smoked at the time of evaluation, most refrained from smoking, with only 2.5% of the patients smoking at 2 years after OLT and 1.8% at the end of follow-up.
Smoking Behavior of Former Smokers.
In Fig. 2, it is shown that of the 85 patients who were former smokers at the time of evaluation, most refrained from smoking until OLT, but up to 12% restarted smoking after OLT. The 10 patients that relapsed after OLT had stopped smoking at a median of 10 months (1-36) before OLT. The 75 former smokers that did not relapse had stopped smoking 168 months (1-540) before OLT. The difference is statistically significant (P < 0.001). Figure 3 shows in more detail that a minority of patients restarted smoking and later stopped again.
Smoking Behavior of Active Smokers.
In Fig. 2, it is shown that of the 53 active smokers at the time of evaluation, 26% succeeded in quitting smoking before OLT, and this remained and even increased a little to 32% after OLT. However, these are overall percentages. Figure 4 shows in more detail that a minority of these 53 patients first stopped but later restarted smoking at some time after OLT and sometimes stopped again later. At 2 years after OLT, there were 50 active smokers, including (re)starters, and 20% succeeded in quitting smoking at the end of follow-up.
Relation Between Smoking and Patient Characteristics
No relation was found between smoking and gender or between smoking and age in smokers and never smokers. Former smokers at evaluation were older than never smokers (median, 51.2 versus 42.9 years; P = 0.001) and active smokers (median, 51.2 versus 45.9 years; P = 0.05). At the time of evaluation for OLT, more patients with ALD and with acute liver failure were active smokers (52% and 46%, respectively) in comparison with the other diagnoses (see Fig. 5; P < 0.001). In this respect, the percentage of active smokers in the group with PSC was the lowest (1.4%). In the alcoholic group, only 8% were never smokers, whereas 77% of PSC patients had never smoked at the time of evaluation (Fig. 5). At the end of follow-up, the same pattern was seen, with 44% in the alcoholic group being active smokers and only 4.3% in the PSC group being active smokers (P < 0.001). The number of active smokers (31%) in the acute liver failure group was not different anymore in comparison with the other diagnoses.
Smoking was also studied in ALD patients with respect to relapse into drinking. Of the 25 patients with alcoholic cirrhosis, only 6 had 1 or more relapses into drinking after OLT, with 1 patient persistently drinking. Of the 6 relapsers, 4 were smokers at the end of follow-up (67%), 1 was a former smoker (17%), and 1 had never smoked (17%). Of the 19 nonrelapsers, 7 patients were smokers at the end of follow-up (37%), 11 were former smokers (58%), and 1 had never smoked (5%). There was a tendency for a lower number of former smokers among the relapsers versus the nonrelapsers (P = 0.078).
Both before and after OLT and among both active and former smokers, the large majority of patients smoked only cigarettes and/or hand-rolled tobacco (92% and 86%, respectively). A minority preferred pipes and/or cigars (about 5%). No statistical differences between the different groups and time points were found in this respect.
At the end of follow-up, the then active smokers had smoked a median of 34 years (range, 2-50), and they reported smoking a median of 11 cigarettes per day (range, 2-60). At the end of follow-up, the then former smokers had smoked a median of 15 years (range, 3-45), and they reported smoking a median of 10 cigarettes per day (range, 1-45). The former smokers had stopped a median of 16.5 years (range, 0.1-46) before the end of follow-up.
Smoking Cessation Plans
In the questionnaire, the still smoking patients were asked about plans to quit smoking. Five of the 49 patients did not answer this item. Of the other 44 patients, 20 patients (45%) wanted to stop within 6 months, 7 patients (16%) wanted to stop within 5 years, 10 patients (23%) wanted to stop at some point, and 7 patients (16%) wanted to stop never. When offered the opportunity to join a cost-free smoking cessation course, 4 of the 49 patients did not answer this item. Of the other 45 patients, 6 patients accepted the offer (13%), 28 patients refused (62%), and 11 patients (24%) were in doubt. For those intending to stop smoking within 6 months, 30% accepted the offer.
Relation Between Smoking and Outcome After Liver Transplantation
Relation to Patient and Graft Survival
For the analysis of the effect of smoking on patient and graft survival, we first had to make a comparison between smoking behavior data acquired from the charts and those data acquired from the questionnaires. The charts turned out to be incomplete in detail. Comparisons could be made, however, with respect to active smoking at the different time points before and after transplantation and with respect to ever smoking before or after liver transplantation. As shown in Table 2, active smoking was underestimated in the charts, with up to one-third fewer active smokers in the charts versus the questionnaires. The same was true for ever smoking data, except for the time after OLT. In both the charts and the questionnaires, 22% of patients ever smoked after OLT.
Table 2. Tobacco Use from Medical Charts Versus Questionnaires
On the basis of this analysis, patient survival and graft survival were studied through a comparison of patients who smoked in the posttransplant period according to the charts and those who did not smoke or whose smoking status was unknown according to the charts. In the group of 401 patients, no differences were found for patient and graft survival between the 103 ever smokers after OLT and the other 298 patients. Also, after exclusion of the 59 patients with unknown behavior, there were no differences in patient and graft survival between ever smokers and nonsmokers. Furthermore, no differences were found for the causes of graft failure between smokers and the other patients. Overall patient survival was as follows: at 1 year, 100% (inclusion criterion); at 5 years, 93%; at 10 years, 85%; at 15 years, 72%; and at 20 years, 64%. The overall graft survival was as follows: at 1 year, 93%; at 5 years, 82%; at 10 years, 73%; at 15 years, 60%; and at 20 years, 49%.
Relation to Morbidity
Because of a lack of detail in the charts with respect to smoking behavior, we decided to study the possible relation between smoking and posttransplant morbidity only in the group of questionnaire responders. Fifty-nine responders who reported tobacco use at 2 years after OLT and/or at the end of follow-up were compared with 236 responders who reported abstaining from tobacco use after OLT. The comparison is shown in Table 3. De novo malignancies, excluding skin cancer, developed significantly more often in smokers in comparison with nonsmokers (see Fig. 6; P = 0.019, Kaplan-Meier log rank test). The cumulative risk (standard error in parentheses) for the development of de novo malignancies in smokers was 3.5% (2.4%), 12.7% (4.9%), and 18.2% (7.0%) at 5, 10, and 15 years after OLT, respectively. In nonsmokers, this risk was 0.6% (0.6%), 2.1% (1.2%), and 7.7% (3.8%), respectively. Tumor types in smokers were oropharynx cancer (1 patient), cancer of the vulva or cervix (2 patients), colon cancer (2 patients), and posttransplant lymphoproliferative disorder (2 patients). Tumor types in nonsmokers were gastric, colon, and prostate cancer (1 patient each), posttransplant lymphoproliferative disorder (2 patients), and de novo hepatocellular carcinoma (1 patient). We found a tendency for hepatic artery thrombosis occurring more often in smokers compared to nonsmokers (P = 0.078, Kaplan-Meier log rank test).
Table 3. Outcome for Posttransplant Active Smokers and Nonsmokers
Long-term morbidity and survival after OLT are determined to a large degree by the development of cardiovascular diseases and cancer. Tobacco use is a well-known risk factor for both disease entities. In our country, tobacco use is increasingly forbidden in public areas, but until July 1, 2008, smoking was still allowed in restaurants and bars. In our center, discontinuation of tobacco use is encouraged before the actual transplant takes place, but it is not a contraindication. The present study gives us insight into the smoking patterns and some negative effects of smoking. In addition, the study helps to recognize subgroups of patients who could be offered extra help in the future.
As we were mainly concerned with long-term negative effects of smoking, we studied patients who survived at least 2 years after OLT. Detailed data were not available in the medical charts; therefore, we used a questionnaire that was sent to all patients alive. The response rate was high (92%). Although this method has several drawbacks, including a recall bias of events that took place in the past, the fact that the questionnaires remained anonymous to the treating physicians and were sent several years after so far successful transplants must have helped the patients to report the truth.
Both before and after OLT, a small majority of patients had never used tobacco, and around 17% of the patients were active smokers. This figure compares favorable with the Dutch population as in 2007, 28% of the Dutch population were reported to be active smokers (Dutch Central Agency for Statistics: http://statline.cbs.nl). The percentage of active smokers in a German population of kidney transplants was 13%,26 and in Italian heart transplants, the percentage was 12%.27 A positive finding was that, of the patients who smoked at the time of the evaluation for OLT, almost one-third succeeded in smoking cessation, often during the waiting time for OLT, although sometimes with ups and downs (Figs. 2 and 4). That the overall percentages of active smokers before and after OLT were about the same implies that other patients, mainly former smokers at the time of evaluation (12%), restarted smoking after OLT (Figs. 2 and 3). We found that especially former smokers who had succeeded in stopping tobacco use a relatively short time ago were at risk of relapse.
To the best of our knowledge, there is only 1 study in liver transplant patients reported in the literature that is comparable to our study. Ehlers et al.8 studied patients transplanted in Florida. Their method was a structured interview by telephone with a 42% response rate. They reported an active smoker rate of 15% after OLT and a relapse rate of 20% in former smokers. Both figures are comparable with our findings. However, the response rate of 42% makes the possibility of underestimation higher than in our study with a response rate of 92%, and so a comparison of the data remains problematic. From both studies, it follows that intervention programs should be aimed not only at active smokers but also at former smokers and that these programs should be continued for many years, if not for the rest of their lives.
We found that most former smokers started smoking early in their life and quit after a median of 15 years of smoking, long before transplantation. The range was wide, however. At the end of follow-up, the cumulative number of smoking years was twice as high in the active smokers compared to the former smokers. It was no surprise that almost all patients used tobacco by smoking cigarettes and/or hand-rolling tobacco. Pipes or cigars were enjoyed by only 5% of the patients.
Tobacco use was highest in patients with ALD and in those with acute liver failure. During the evaluation, 52% of the alcoholic group were active smokers (and only 8% were never smokers). The combination of alcohol and tobacco addiction is well known and has been reported to be as high as 90% in alcohol abusers.28, 29 This combination is also shown in our finding of a tendency for a higher number of former smokers among ALD patients without a relapse into drinking, and this suggests that nonrelapsers are more likely to stop smoking than relapsers. As alcohol abstinence for at least 6 months was a prerequisite for OLT in our center, it is disturbing that so many still continued active smoking. At the end of follow-up, active smoking in this group had decreased from 52% to 44%; this decrease is remarkable in comparison with the whole group (18%-16%). There is probably already more attention for this group of patients, or attention to stop alcohol use also influences smoking behavior, but clearly more is needed to further bring down these figures. DiMartini et al. from Pittsburgh reported tobacco use after OLT for ALD to be an underestimated problem as these patients were reported to have a high incidence of lung and pharyngeal cancer.7, 30–32 They found that on average more than 40% of patients with ALD were smoking after OLT and that smoking was resumed already at 3 months post-OLT. Our data seem comparable, except that we show an overall decrease if it is reckoned from the evaluation for OLT onward.7
The PSC group is remarkable, with only 1.4% being active smokers and 77% being never smokers at the time of evaluation for OLT. A possible protective effect of smoking on the development of PSC33–36 and ulcerative colitis37–39 has been shown before. The exact protective mechanism of smoking is not yet known. Studies with nicotine as therapy for PSC showed no beneficial effects.40, 41 It would be interesting to study the effects of smoking on recurrent PSC after OLT, but because of the low number of tobacco users in this patient group, this will be possible only in a multicenter study group.
We have not found a negative effect of active smoking after OLT on long-term patient and graft survival so far, but this item could be studied only with smoking data from the medical charts, and these data were incomplete. However, in the group of survivors who responded to the questionnaire, we did find that the prevalence of nonskin cancer was significantly increased in active smokers after OLT. At 10 years, the cumulative malignancy rate was 12.7% in smokers versus 2.1% in nonsmokers after OLT. Malignancies seemed to develop much earlier after OLT in tobacco users (Fig. 6). We did not notice a difference in the types of tumors that developed. Also, no effect on skin cancer was found. Tobacco use as a risk factor for malignancy after OLT has been recognized before.9–13 Previous studies, however, relied on data in medical charts and most often not on actual posttransplant smoking data.
We did not find an increased prevalence of cardiovascular disease in smokers. Most likely, tobacco use is only one of many risk factors for cardiovascular disease after OLT, such as overweight, diabetes, hypertension, and the use of immunosuppressive agents.
In summary, both before and after OLT, about 17% of patients are active smokers, but after OLT, this percentage includes restarting former smokers and a decreasing number of actively smoking ALD patients. Proportionally, most tobacco users are found in the ALD group, and the lowest number is found in the PSC group. Pretransplant intervention programs should be aimed not only at pretransplant active smokers but also at former smokers, especially those who stopped smoking quite recently. The higher prevalence of malignancies in active smokers after OLT warrants intervention programs after OLT and regular screening for malignancies.