Cigarette smoking has been shown to have an adverse impact on a number of health-related conditions. Smoking causes or is a contributory factor to five of the ten leading causes of death in the United States, including coronary artery disease, malignancy, and emphysema (1).
Although the long-term risks of cigarette smoking can be diminished with smoking cessation, the highly addictive properties of nicotine make cessation extremely difficult for cigarette smokers, even in the face of established smoking-related illnesses.
Given the breadth of disease entities that are smoking-related, it appears likely that cigarette use would have an adverse impact on the general health of transplant recipients. Transplant recipients are already at risk for disease entities such as accelerated atherosclerosis and neoplasia as a result of immunosuppression and its side effects. Smoking in this population has been associated with an increased risk of cardiovascular disease and malignant and premalignant oral lesions (2–5). However, relatively few studies have addressed the effects of smoking on kidney graft outcome (6–7) and on death-censored graft survival in particular. In addition to the effects on cardiovascular risk, which may diminish patient survival, effects on accelerated atherosclerosis in the allograft and on immune function may potentially influence graft loss not caused by patient death.
The aims of this single-center, cohort study of 645 renal transplant recipients were to establish the impact of cigarette smoking on long-term allograft survival and overall patient survival.
The printed and electronic records of 1034 patients who received a renal transplant at the University of Michigan Health System between September 1985 and August 1995 were reviewed. Criteria for inclusion in the study population were (1) recipient age ≥18 years; (2) first renal transplant at our institution during the study period; (3) no history of multiorgan transplant; and (4) adequate documentation of at least 2 years of follow-up. By these criteria, 61 kidney-pancreas, 14 kidney-liver, and 141 pediatric kidney recipients were excluded. There were 47 kidney-alone retransplant recipients who were excluded because their first transplant was done before September 1985 or at another institution. The study group consisted of the remaining 645 patients for whom complete records with respect to graft survival, patient survival, and smoking status at the time of evaluation for transplant were available. Data concerning date of transplantation, donor source, date of graft failure, date of patient birth, and date of patient death were obtained from a computerized database. All other information was obtained from hospital records. Follow-up ranged from 2 to 12 years.
The evaluated risk factors included patient gender; age at time of first kidney transplant; donor source (cadaver, living related, living unrelated); rejection episodes; duration of pretransplant dialysis; presence of diabetes; pretransplant and posttransplant hypertension; pretransplant and posttransplant hypercholesterolemia; and history of smoking before and after transplantation. Smoking history was defined as the number of years of tobacco use multiplied by the average number of packs per day. A positive smoking history was defined as any tobacco use before the pretransplant evaluation. Pretransplant smoking was defined as active use of tobacco at the time of pretransplant evaluation. Posttransplant smoking was defined as the use of tobacco at any time during follow-up before kidney graft loss. Smokeless tobacco use was not included in the analysis. Pretransplant smoking status was the primary risk factor evaluated.
Standard triple-drug maintenance therapy with cyclosporine, azathioprine, and corticosteroids was used for the vast majority of recipients. Induction therapy with polyclonal antithymocyte globulin (Pharmacia and Upjohn, Kalamazoo, MI) was used in all patients. Rejection episodes were treated primarily with pulse steroids and/or anti-CD3 monoclonal antibody (Ortho Biotech, Raritan, NJ).
Cumulative graft survival, death-censored graft survival, and patient survival curves were estimated with the method of Kaplan and Meier (8). Survival rates between groups defined by smoking history were compared using the Breslow-Gehan-Wilcoxon test. Comparisons of risk factors between groups of patients were made by contingency tables and chi-square analysis for categorical variables and by the unpaired two-tailed Student’s t test for continuous variables. A backwards stepwise Cox proportional hazards model was used to determine independent risk factors for graft survival and death-censored graft survival. The covariates were pretransplant dialysis duration, age at transplant, gender, donor type, diabetes, pretransplant smoking, any smoking history, acute rejection, pretransplant hypertension, and pretransplant hypercholesterolemia. Risk factors’ data were >95% complete for each; cases with missing values were omitted from the Cox regression analysis. P <0.05 was considered significant. All data were analyzed by Statview software (Abacus Concepts, Inc., Berkeley, CA).
Patient smoking habits are listed in Table 1. Fifty-one percent of patients reported a positive smoking history at the time of pretransplant evaluation, and 156 (24%) were active smokers at the time of evaluation. Patients with a positive smoking history had a mean of 24 pack-years of smoking exposure. Of 125 evaluable patients who were smokers at the time of transplant evaluation, 112 (90%) continued to smoke after transplantation. Only six patients (1%) who were not smokers at the time of evaluation became smokers after transplantation.
Pretransplant smokers were younger (average age at transplant, 39±12 vs. 42±13 years, P =0.03, Table 2), more frequently male (65% vs. 55%, P =0.03), and more likely to be hypertensive before transplant (93% vs. 85%, P =0.02). There were no significant differences in the proportion of patients with diabetes, in donor source, in posttransplant hypertension, or in pretransplant and posttransplant hyperlipidemia between patients who were smokers at the time of pretransplant evaluation and those who were not.
Patients who were smokers at the time of pretransplant evaluation had kidney graft survival of 84%, 65%, and 48% at 1, 5, and 10 years, respectively, compared with graft survival in nonsmokers of 88%, 78%, and 62% (P =0.007, Fig. 1). Because lower graft survival among smokers could have been caused by a greater incidence of death with a functioning graft, patient survival rates for the two groups were compared. There was an insignificant trend toward lower patient survival among smokers (84% vs. 88% at 5 years, P =0.14, Fig. 1). Death-censored graft survival was significantly lower in smokers (73% vs. 86% at 5 years, P =0.002, Fig. 1), which also indicated that diminished graft survival in smokers was not caused by patient attrition.
The diminished death-censored graft survival rates seen in pretransplant smokers was evident in recipients of cadaveric as well as living donor kidneys (P =0.02 for each group, Fig. 2). The effects of smoking were also seen equally in male and female recipients (not shown). The incidence of patients who experienced at least one acute rejection episode was 64% among pretransplant smokers and 61% in nonsmokers (P =0.35), indicating that a predisposition to acute rejection was unlikely to account for the lower graft survival in smokers.
Patients with a history of smoking at any time before transplantation also had significantly lower graft survival (5-year graft survival, 68% vs. 81%, respectively, P =0.006) and death-censored graft survival (5-year death-censored graft survival, 79% vs. 87%, respectively, P =0.03) compared with patients with no smoking history. These patients also had significantly lower patient survival (5-year patient survival, 82% vs. 91%, respectively, P =0.007).
We questioned whether smoking cessation before transplantation would have a measurable impact on graft outcome. Indeed, among those patients with a smoking history before transplantation, death-censored graft survival was significantly higher for the patients who were no longer smoking at the time of transplant evaluation (P =0.03, Fig. 3). In fact, graft survival for those who had stopped smoking before evaluation was similar to those patients who had never smoked. However, among patients who were former smokers at transplant evaluation, those with a more than 30 pack-year history of smoking had a significantly worse outcome than those former smokers with a less extensive history (P =0.03, Fig. 4A). Indeed, these patients accrued no significant death-censored survival benefit when compared to those with a comparably heavy smoking history who were still smoking at the time of pretransplant evaluation (P =ns, Fig. 4B).
A number of the factors identified in the univariate analyses could have been confounded by the presence of covariates associated with poor graft outcome. Therefore, a backwards stepwise Cox proportional hazards regression model was used to identify independent risk factors for graft and death-censored graft survival (Table 3). Analysis of logarithmic plots of cumulative hazard confirmed our hypothesis of proportionality during early posttransplant years of follow-up, with excess hazard appearing later in the group of patients who were pretransplant smokers.
In this model, acute rejection, pretransplant dialysis duration, and pretransplant smoking were independent risk factors for death-censored graft survival. Age at transplant, gender, donor type, diabetes, any smoking history, pretransplant hypertension, and pretransplant hypercholesterolemia were not independently associated with death-censored graft survival. Pretransplant smoking was a strong independent risk factor, with a relative risk of 2.3 (P <0.0005).
Only 13 patients stopped smoking after transplantation (Table 1). Ten of these patients (77%) lost their grafts. Graft losses in this group included one death with a functioning graft, four grafts with primary nonfunction or early thrombosis, three grafts lost to rejection, and two grafts lost in the setting of severe pulmonary edema within 3 months of transplant. Therefore, it is likely that graft loss preceded smoking cessation or both were linked to medical illness in a majority these patients, and poor graft survival in this subgroup was probably not directly caused by smoking cessation.
Transplant recipients are at risk for a number of disease states that are known to be aggravated by cigarette smoking in nonimmunosuppressed individuals. Smoking is associated with an increased risk of a wide variety of cancers, and transplant recipients are likewise at an increased risk for malignancy. The combination of immunosuppression and smoking may also contribute significantly to both premalignant and malignant lesions of the oral cavity and oropharynx (2–3).
Cigarette smoking also aggravates atherosclerosis in transplant recipients who have an increased prevalence of several atherosclerotic risk factors such as hypertension, hyperlipidemia, and hyperglycemia due to adverse effects of immunosuppression (9–10). Smoking is a known risk factor for coronary artery disease and peripheral vascular occlusive disease in kidney and heart transplant recipients (4–5,11). The adverse effects of smoking on atherosclerosis could lead to diminished kidney allograft survival through effects on coronary artery disease, renovascular disease, or small vessel atherosclerosis in the allograft. In fact, chronic allograft rejection, which is the leading cause of late graft loss, is pathophysiologically indistinguishable from atherosclerosis in nonrenal vessels (12).
Smoking is also associated with adverse effects on renal function in nontransplant patients, including acceleration of renal insufficiency, reductions in renal plasma flow, elevations of plasma endothelin-1, renal arteriolar thickening, and increased platelet aggregation (13–17).
The effect of smoking on immunologic graft survival could be due to an increased risk of chronic rejection. Although cigarette smoking has suppressive effects on immunity, including impairments in cytotoxic T-lymphocyte responses, polymorphonuclear leukocyte phagocytosis, and B-cell activity (18–21), patients with thromboarteritis obliterans, a smoking-related disease, have an increased frequency of HLA-DR4 and a lower frequency of HLA-DRW6. This suggests a locus in which cigarette smoking and immunity may interact to produce immunostimulation (22).
Our study found diminished graft and patient survival associated with all classes of smoking, including smoking during the pretransplant and posttransplant interval. Smoking at the time of pretransplant evaluation had the greatest effect on graft survival in both the univariate and multivariate analyses. The adverse effects of pretransplant smoking on graft survival were primarily due to graft loss per se rather than death with a functioning graft. The incidence of acute rejection was similar when pretransplant smokers and nonsmokers were compared. These findings are consistent with a mechanism by which smoking may accelerate graft failure through small vessel arteriopathy or chronic rejection, as previously mentioned.
Smoking at the time of transplant evaluation and a history of heavy smoking were particularly deleterious to graft outcome. Patients who had stopped smoking before transplant evaluation had significantly higher death-censored graft survival rates when compared to those who were still active smokers. Indeed, graft survival in the former group was comparable to those patients who never smoked. However, a history of heavy smoking (>30 pack-years) appeared to preclude the otherwise salutary benefits of smoking cessation; among these heavy smokers, graft survival was not significantly improved by smoking cessation.
Two previous studies by Cosio et al. (7), and Kasiske et al. (8) have demonstrated an adverse impact of a positive smoking history on patient survival in kidney transplant recipients, which is consistent with our findings. An association between any smoking history and death-censored graft survival was not found in the latter study and was not addressed in the former. Although our study did note a significant association between any smoking history and death-censored graft survival, graft losses due to patient death had a greater impact on overall graft survival. Conversely, the association between smoking in the peritransplant period and death-censored graft survival noted in our study was not specifically addressed by either of the aforementioned studies. Therefore, while those patients with any smoking history are at increased risk for death with a functioning graft relative to those who never smoked, those specifically smoking before transplantation appear to be at risk primarily because of graft failure.
It remains unclear from the results of our study whether the impact of smoking on graft survival results from the effects of cigarette use before transplant or from continued usage after transplantation. Because so few patients who smoked at the pretransplant evaluation quit voluntarily, and even fewer who were not smokers before transplant began afterwards, the specific effect of smoking after transplantation could not be assessed. It is likely that both chronic and acute processes are involved, because smoking cessation before evaluation was beneficial for all but the heaviest smokers.
The adverse effects of cigarette use did not appear to be mediated by effects on other conditions associated with smoking, such as hypertension or hyperlipidemia. The two groups were also similar with respect to pretransplant factors that affect graft outcome, such as the immunosuppressive regimen, frequency of diabetes, cadaveric donor origin, and duration of pretransplant dialysis. Furthermore, in the multivariate analysis, which included all pretransplant risk factors, pretransplant smoking was independently associated with reduced death-censored graft survival. However, potential differences in socioeconomic status and compliance, both of which are risk factors known to affect graft outcome (23) and which are associated with smoking (24), were not addressed in this study and, therefore, cannot be ruled out.
We noted a 24% prevalence of pretransplant smoking in our study population, which parallels the prevalence of smoking in the entire state of Michigan in 1977 (25). A limitation of the study design is that pretransplant smoking was defined at the time of transplant evaluation, which may be as little as a few months or as long as many years before transplant. We chose this definition because the smoking history was prospectively obtained at the time of the pretransplant evaluation in all patients. Smoking status at the time of evaluation probably closely reflects the status at transplant, because more than 90% of patients who were smokers at evaluation smoked after transplant, and many of those who stopped smoking posttransplant did so as a result of prolonged medical illness or operative complications after transplantation. Finally, the adverse effects of smoking were equivalent for living donor kidney recipients (in whom the interval from evaluation to transplantation is short) and cadaver kidney recipients (for whom the interval is longer). Therefore, differences in waiting time did not appear to have influenced our findings.
Although the adverse effects of smoking on graft survival could not be explained by patient death, it is possible that gradual health deterioration in these patients as a result of smoking could lead to both graft loss and then subsequent death. To address this possibility we split the patients by era (1985–1990, 1991–1995) and again performed death-censored survival estimates grouped by pretransplant smoking. In both eras, pretransplant smoking had a significant adverse effect on death-censored graft survival but not patient survival (data not shown). Delayed death after graft loss would have been selectively reflected in diminished patient survival in the earlier cohort because of longer follow-up. Therefore, graft losses among smokers are not likely to be preterminal events related to smoking-related disease states that result in later patient demise.
Intervention to effect smoking cessation should be done during the initial discussion on transplantation. It is difficult to predict whether cessation of smoking at the time of evaluation reduces risk, because so few of our patients actually stopped smoking. However, the demonstrated benefit of smoking cessation before transplant evaluation, which may predate transplantation by many years, justifies an aggressive approach to smoking cessation in patients with end-stage renal disease who are being considered for renal transplantation.
1. Centers for Disease Control and Prevention. National vital statistics report 1998: Vol 47(4), p. 5–7.
2. Duvoux C, Delacroix I, Richardet JP, et al. Increased incidence of oropharyngeal squamous cell carcinomas after liver transplantation for alcoholic cirrhosis. Transplantation 1999; 67: 418.
3. King GN, Healy CM, Glover MT, et al. Increased prevalence of dysplastic and malignant lip lesions in renal-transplant recipients. N Engl J Med 1995; 332: 1052.
4. Radovancevic B, Poindexter S, Birovljev S, et al. Risk factors for development of accelerated coronary artery disease in cardiac transplant recipients. Eur J Cardiothorac Surg 1990; 4: 309.
5. Kasiske BL. Risk factors for accelerated atherosclerosis in renal transplant recipients. Am J Med 1988; 84: 985.
6. Cosio FG, Falkenhain ME, Pesavento TE, et al. Patient survival after renal transplantation. II. The impact of smoking. Clin Transplant 1999; 13: 336.
7. Kasiske BL, Klinger D. Cigarette smoking in renal transplant recipients. J Am Soc Nephrol 2000; 11: 753.
8. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457.
9. Kasiske BL, Tortorice KL, Heim-Duthoy KL, Awni WM, Rao KV. The adverse impact of cyclosporine on serum lipids in renal transplant recipients. Am J Kidney Dis 1991; 17: 700.
10. Curtis JJ. Hypertension following kidney transplantation. Am J Kidney Dis 1994; 23: 471.
11. Nagele H, Kalmar P, Rodiger W, Stubbe HM. Smoking after heart transplantation: an underestimated hazard? Eur J Cardiothorac Surg 1997; 12: 70.
12. Vollmer E, Bosse A, Bogeholz J, et al. Apolipoproteins and immunohistological differentiation of cells in the arterial wall of kidneys in transplant arteriopathy: morphological parallels with atherosclerosis. Pathol Res Pract 1991; 187: 957.
13. Remuzzi G. Cigarette smoking and renal function impairment. Am J Kidney Dis. 1999; 33: 807.
14. Orth SR, Stockmann A, Conradt C, et al. Smoking as a risk factor for end-stage renal failure in men with primary renal disease. Kidney Int 1998; 54: 926.
15. Gambaro G, Verlato F, Budakovic A, et al. Renal impairment in chronic cigarette smokers. J Am Soc Nephrol 1998; 9: 562.
16. Davis JW, Arnold J, Wiegmann T. Cigarette-smoking affects platelets and endothelium in chronic hemodialysis patients. Nephron 1993; 64: 359.
17. Oberai B, Adams CWM, High OB. Myocardial and renal arteriolar thickening in cigarette smoking. Atherosclerosis 1984; 52: 185.
18. Savage SM, Donaldson LA, Cherian S, White VA, Sopori ML. Effects of cigarette smoke on the immune response. II. Chronic exposure to cigarette smoke inhibits surface immunoglobulin-mediated responses in B cells. Toxicol Appl Pharmacol 1991; 111: 523.
19. Sasagawa S, Kameda H, Sudo J, Tanabe T. Inhibitory effect of nicotine on chemiluminescence response of human polymorphonuclear leukocytes stimulated by opsonized zymosan in vitro. J Toxicol Sci 1984; 9: 1.
20. Sopori ML, Gairola CC, DeLucia AJ, Bryant LR, Cherian S. Immune responsiveness of monkeys exposed chronically to cigarette smoke. Clin Immunol Immunopathol 1985; 36: 338.
21. King TE Jr, Savici D, Campbell PA. Phagocytosis and killing of Listeria
monocytogenes by alveolar macrophages: smokers versus nonsmokers. J Infect Dis 1988; 158: 1309.
22. Papa M, Bass A, Adar R, et al. Autoimmune mechanisms in thromboangiitis obliterans (Buerger’s disease): the role of tobacco antigen and the major histocompatibility complex. Surgery 1992; 111: 527.
23. Butkus DE, Meydrech EF, Raju SS. Racial differences in the survival of cadaveric renal allografts: overriding effects of HLA matching and socioeconomic factors. N Engl J Med 1992; 327: 840.
24. Choiniere R, Lafontaine P, Edwards AC. Distribution of cardiovascular disease risk factors by socioeconomic status among Canadian adults. CMAJ 2000; 162 (9 suppl): S13.
© 2001 Lippincott Williams & Wilkins, Inc.
25. No authors. Cigarette smoking among adults–United States, 1997. MMWR Morb Mortal Wkly Rep 1999; 48(43): 993.