Amount of tobacco consumption is associated with superficial bladder cancer progression

objective: To evaluate the association between smoking habits and outcome of patients with superficial bladder cancer. Methods: A retrospective study was performed evaluating 99 patients (67.0 ± 13.2 years, ranging from 31.4-93.4 years, 72.7% males and 27.3% females) treated at our institution with non muscle-invasive bladder cancer, between 1994 and 2000, with a mean follow-up of 49.3 months (range 4.0-177.9 months). Patients were divided according to smoking status, and the main measured outcome was progression to invasive disease. Additional cohort analysis was performed dividing patients according to previous tobacco exposure: smokers and non-smokers. Smokers were stratified into former smokers, early-quitters, late quitters and continued smokers. results: Smoking habit was significantly more common in males (p = 0.03). Cancer also occurred at an earlier age among smokers (70.8 versus 64.8 years, p = 0.030). Tobacco consumption was present in 62.7% of the patients with bladder cancer. There was a significant higher progression rate to muscle-invasive disease in patients that had more than 60 pack-years of exposure (52.9 versus 26.2%, p = 0.037). These patients had a mean progression time of 59.3 months, whereas patients who had smoked less than 60 packyears progressed after a mean time of 131.8 months. Conclusions: A direct association between the amount of tobacco consumed and disease progression is observed in patients with bladder cancer, as suggested by the present study. Tobacco consumption has a direct association with progression of superficial bladder cancer to invasive disease and also shortens the period of time for muscle invasion. Larger and prospective studies are still necessary to bring further definitive conclusions about reproducibility of our data and to better understand how smoking cessation affects progression of superficial bladder cancer.


OBJectiVe
To evaluate the association between smoking habits and the outcome of patients with superficial bladder cancer.

MetHODS
A retrospective case-control study was performed evaluating 99 patients (67.0 ± 13.2 years, range of 31.4-93.4years) treated at our institution with non muscle-invasive bladder cancer between 1994 and 2000, with a mean follow-up of 49.3 months (range 4.0-177.9months).All patients with incomplete registers were excluded from the study.Other reported bladder conditions included recurrent urinary tract infection (20%) and chronic indwelling bladder catheter (6%).
Oncologic follow-up was performed according to an institutional protocol for superficial bladder cancer: after transurethral resection of the tumor, reresection was performed one month later for highgrade tumors; cistoscopy, cytology, chest X-ray and abdominal ultrasound were performed every three months for the first year, every six months for the first three years, and yearly thereafter in case of no recurrence.
Additional adjuvant treatments were performed when indicated: intravesical BCG (80 mcg/week for six weeks for induction and after this period of time 80 mcg/ week for three weeks every six months as maintenance treatment).Intravesical mitomycin C was applied after the first recurrence following a BCG course, or when BCG was contraindicated.
Patients were divided according to disease progression (progression versus non-progression to muscle invasive disease), and smoking status.Additional cohort analysis was performed dividing patients according to previous tobacco exposure: smokers and non-smokers.Smokers were stratified into former smokers, early-quitters, late quitters and continued smokers.Former smokers were considered all patients who had quit smoking at least one year prior to cancer diagnosis.Early quitters were patients who quit smoking in the first year after diagnosis of cancer, and late quitters were those who had quit more than a year after the initial diagnosis.Measured outcomes included recurrence free survival, progression to muscle-invasive disease and cancer specific-mortality.
Statistical analyses were performed using the Statistical Package for Social Sciences software (SPSS for Mac OS X, SPSS, Inc., Chicago, Illinois).Groups were compared by Pearson's χ 2 test and Student's t test.Disease-specific survival and progression to muscle-invasive disease plots were made using the Kaplan-Meyer method.Survival rates were analyzed for significance using the log rank test.Statistical significance was determined at p < 0.05.

reSUltS
Of 99 patients, 72.7% were males and 27.3% were females.Tobacco use was reported by 62.7% of patients with bladder cancer.Smoking was significantly more common among males than females (p = 0.03, Table 1).Cancer also occurred at an earlier age for smokers (70.8 versus 64.8 years, p = 0.03, Table 1).Mean smoking history was of 37.5 ± 13.3 years (Table 2).Of the 62 smokers, 16 were classified as former smokers, 19 as early-quitters, 7 as late-quitters and 20 as continued-smokers.When patients were divided according to muscle progression, there was no statistical difference between smokers and non-smokers or between subgroups of smokers.Of the 20 continued-smoking patients, 40.0% had progressed to muscle-invasive disease, as compared to 28.6% of the former smokers and early quitters (p = 0.160, Figure 1).If distributed according to tobacco exposure, there was a significant higher rate of progression to muscle-invasive disease in patients that had more than 60 pack-years of exposure (52.9 versus 26.2%, p = 0.037, Table 3).These patients had a mean progression time of 59.3 ± 16.4 months, whereas patients who smoked less than 60 packyears progressed after a mean time of 131.8 ± 14.2 months (p = 0.041, Figure 2).

DiScUSSiOn
Tobacco use is the main risk factor for bladder cancer, and different aspects of smoking habit have been analytically characterized (9,11) .After bladder cancer is diagnosed, it seems logical that smoking interruption might bring benefits.However, to the present date, there are no conclusive studies evaluating this issue (6) .This study has some important findings.
Prognostic factors upon diagnosis: a recent study observed worse prognostic factors upon initial diagnosis of bladder cancer among smokers, though similar recurrence patterns were found when compared to non-smokers (12) .This finding is different from our observations.In the present study, there were similar prognostic factors after initial transurethral resection when comparing smokers to non-smokers, as observed by other authors (7) .Maybe the relatively small sample can explain our results, as we could not reproduce these findings.However, after a mean time of 37.  Smoking status: the role of smoking status over prognosis after the initial diagnosis of bladder cancer is not well understood yet.A recent study suggested that smoking interruption seems to be an independent risk factor for disease progression (7) .Other studies failed to prove such relation (12) .In a systematic review, Aveyard et al. found only one high quality study evaluating this issue, and concluded that the evidence is weak since most studies were not statistically significant (6) .Although we observed a trend towards higher progression rates among continued-smokers, our study also failed to demonstrate a statistically significant difference in recurrence patterns when comparing former smokers, quitters and continued-smokers.What we observed was an earlier age of cancer onset when comparing smokers to non-smokers, a data similar to what was previously reported (10) .It is worth mentioning that our relatively small sample becomes even smaller when these subgroups of patients are stratified.But the amount of tobacco consumption was associated with higher and earlier progression rates.Previous studies already reinforced the association between the amount of tobacco use and cancer diagnosis (13) .Although these studies have demonstrated a direct association between the amount of tobacco smoking and higher risk of developing bladder cancer (9) , the risk of disease progression has not been demonstrated.
A significant higher progression rate was observed in patients who were exposed to more than 60 pack-years.This group of patients had also a faster progression to muscle-invasive disease.
We believe that our study can bring very useful information for both patients and clinicians who treat bladder cancer patients.It might be stimulating for patients with diagnosis of a highly recurrent cancer to know that there is something they can do to reduce the risk of disease progression.Patients should be encouraged to reduce and/or quit smoking.Additional measures that could reduce exposure of bladder urothelium to carcinogens might also be encouraged.In a study involving almost 8,000 men, researchers found that men who drank higher amounts of water had a significantly reduced incidence of bladder cancer.Men who drank more than five cups of coffee daily had also an increased risk of bladder cancer (14) .As suggested by the authors, bladder cancer might partly be caused by contact with carcinogens excreted in urine.Therefore, patients with a prior diagnosis of bladder cancer should be stimulated not only to quit smoking but also to quit drinking coffee and to drink large amounts of water.A close follow-up of these patients is also essential to allow early diagnosis and to avoid cancer metastasis.The present study has several limitations.Other factors seem to be associated with tobacco-related bladder cancer, such as the type of tobacco consumed (11) or concomitant coffee ingestion (15) , and these factors were not evaluated in the present study.Our sample size is also relatively small, limiting the power of statistical analysis.However, the present study brings several relevant insights for both clinicians and patients.

cOnclUSiOnS
This study suggested a direct association between the amount of tobacco consumed and disease progression is observed in patients with bladder cancer.Tobacco consumption has direct association with progression of superficial bladder cancer to invasive disease and also shortens the period of time for muscle invasion.This information might help clinicians to convince their patients with bladder cancer that by quitting smoking they will improve their prognosis.Larger and prospective studies remain necessary to bring further definitive conclusions about the reproduction of our data and to better understand how smoking cessation affects the progression of superficial bladder cancer.

Figure 1 .Figure 2 .
Figure 1.Kaplan-Meyer plot according to smoking amount related to progression to muscle invasion

table 1 .
Patients' demographic data and comparison between non-smokers and smokers at initial diagnosis of cancer 5 ± 13.3 years smoking, our patients had cancer diagnosed at an earlier age (70.8 versus 64.8 years, p = 0.03).As early diagnosis of bladder cancer has *p < 0.05.

table 2 .
Smoking history

table 3 .
Distribution of patients according to their past tobacco exposure *p < 0.05.