Urinary tract stone raises
subsequent risk for
urinary tract cancer: a
population-based
cohort study
Li-Min Sun, Cheng-Li Lin*?, Yen-Jung Chang*?, Ji-An Liang
†§, Shu-Hui
Liu
††?,
Fung-Chang Sung
?and Chia-Hung Kao
‡§Introduction
A urinary tract stone (UTS) usually forms in the kidney and may migrate down the ureter and pass into the bladder. Diet plays an important role in the pathogenesis of kidney stones [1], but the infection or inflammation that may result, for example, from the use of a urinary catheter can cause bladder stones [2]. Epidemiological surveys have shown that in
economically developed countries, the prevalence of UTS ranges between 4% and 20% [3], but may be as high as 20–25% in the Middle East [4]. The condition occurs more frequently in men than in women [5], but the sex difference in prevalence appears to be diminishing [6].
Several studies have suggested a possible link between UTS and urinary tract cancer (UTC), based on case reports [7–9] case-control studies [10–12], and a cohort study [13]. However, separate case-control studies did not show a significantly higher risk of UTC among patients with UTS compared with control patients [14–18]. To clarify this issue, we investigated the UTC risk for patients with UTS in Taiwan. Bladder cancer is a global health problem, and is ranked the seventh most common cancer type in men and the 17th most common in women worldwide. The incidence of bladder cancer varies considerably among countries, with the highest rates inWestern countries and the lowest in Asian countries
[19]. According to 2008 Taiwanese cancer statistics, bladder cancer was the eighth and 13th most common malignancy among men and women, respectively [20]. However, an unusually high incidence of upper UTS was also found among
Taiwanese patients [21]; therefore, we were also interested in this topic and extended our investigation to cover all UTCs,
except RCCs. The results of the present study were generated using the data of a population-based retrospective cohort obtained from the database of the National Health Insurance (NHI) system of Taiwan.
Patients and Methods
Data analysed in the present study were obtained from the Longitudinal Health Insurance Database (LHID) provided by the Bureau of the NHI (BNHI) in Taiwan. This database is administered by the Department of Health and is managed by the National Health Research Institutes (NHRI). The BNHI grants researchers access to medical claims data after scrambling the identification of patients to protect their confidentiality. The LHID includes all medical records logged between 1996 and 2010, with 1 million people having been randomly selected from all insured beneficiaries. The NHRI reports that no statistically significant differences have been found for sex and age distributions between the LHID
and all NHI insurees. The universal NHI programme was implemented in Taiwan in March 1995, and by 2009 it had achieved a coverage rate of ≈99% of the 23.7 million
Taiwanese citizens [22]. Patient diagnoses are coded according to the International Classification of Diseases (ICD), ninth revision, Clinical Modification (ICD-9-CM). This study was approved by the Ethics Review Board of China Medical University (CMU-REC-101-012).
Participants
Patients who received an initial diagnosis of UTS (ICD-9-CM codes 592.0, 592.1, 594.0, and 594.1) between 1997 and 2010 were selected as the study cohort. The index date for patients with UTS was the date of their first medical visit related to UTS. Patients with a history of cancer diagnosed before the index date, patients with missing data on age or sex, and
patients aged <20 years were excluded from the study. For the comparison cohort, we used a simple random sampling
method to select two insured people with a non-UTS status for every person with UTS during the same period. Patients and controls were matched for age in 5-year categories, and were also matched for sex and the index year.We conducted the matching in a 1:2 ratio to increase the power of statistical analysis.
Outcome Measures
UTC (ICD-9-CM codes 188 and 189) was identified from the Registry for Catastrophic Illness Patient Database in Taiwan. The registration of a case of catastrophic illness requires the diagnosis to be made by a physician, with confirmatory pathological results or other supporting medical information; these documents are formally reviewed by the BNHI. All patient cases were followed to measure the incidence of UTC until the end of 2010 or the date of patient withdrawal from the NHI. Baseline comorbidities, including diabetes (ICD-9-CM code 250), hypertension (ICD-9-CM codes 401 to 405), hyperlipidaemia (ICD-9-CM code 272), and UTI (ICD-9-CM code 590, 595) were also identified among the patients.
Statistical Analysis
We analysed the data to compare the distributions of age, sex, socio-economic status (monthly income and occupation), and baseline comorbidities between the study cohort (patients with UTS) and the comparison cohort (people without UTS). The chi-square test was used for categorical variables, and the
t-test was used for continuous variables. The incidence
densities for UTC were assessed for each cohort by sex, age, socio-economic status, and comorbidities. The rate ratio of incident UTC was determined for the study cohort vs the comparison cohort. Univariate and multiple Cox proportional hazard-regression analyses were used to calculate hazard ratios (HRs) and 95% CIs for the risk of UTC associated with UTS. The multiple models were simultaneously adjusted for age, sex, socio-economic status, and comorbidities. The UTC-free survival rates were estimated using the Kaplan–Meier method,
and the difference between the study and comparison cohorts was compared using the log-rank test.
Results
Table 1 shows the distribution of demographic characteristics and baseline comorbidity status for the two cohorts.More than half of the patients with UTS were aged ≤50 years, and 79.5% of this group were men. The mean (SD) age for patients with UTS was 46.9 (14.0) years, whereas for patients without UTS it was 46.4 (14.5) years. Patients with UTS tended to have lessWhite collar occupations (53.5% vs 55.0%). The study cohort showed higher levels of comorbidity than the comparison cohort, including diabetes (11.1% vs 8.48%), hypertension (18.3% vs 11.7%), hyperlipidaemia (27.4% vs 19.0%), and UTI (7.65% vs 3.03%).
The incidence rate ratio (IRR) of UTC for patients with UTS was 5.13-fold that of patients without UTS (4.67 vs 0.91 per 10 000 person-years; Table 2). The Cox proportional hazard-regression model showed that patients with UTS were 4.66-fold more likely to develop UTC (95% CI 2.97–7.30) after controlling for age, sex, socio-economic status, and comorbidities. The incidence of UTC was higher for men than for women in the comparison cohort, but this pattern was reversed in the UTS cohort. Overall, the IRR of UTS was considerably higher for women than for men (11.8 vs 4.3); although the adjusted HR showed that the risk of UTC was 38% greater for men than for women. Patients with UTS who were aged <50 years had a much higher IRR, at 8.28. However,
patients aged ≥65 years showed a 9.84-fold (95% CI 5.40–17.9) hazard of UTC compared with those aged ≤50 years. Patients
with UTS with the highest monthly income had a much higher IRR (15.3), but monthly income did not affect the UTC risk when we pooled the subjects. Patients with UTS with aWhite collar occupation had the highest IRR, at 8.40. However, patients with Blue collar occupations had a significant 4.47-fold (95% CI 1.76–11.4) risk of UTC compared with those with other occupations. The IRR for UTC associated with comorbidities were as follows: 2.58 for diabetes, 3.04 for hypertension, 6.24 for hyperlipidaemia, and
7.51 for UTI. The risk of developing UTC was slightly greater for patients with hyperlipidaemia (adjusted HR 1.67, 95% CI 1.03–2.71) compared with hypertension (adjusted HR 1.21, 95% CI 0.76–1.92), diabetes (adjusted HR 0.86, 95% CI 0.49–1.52) and UTI (adjusted HR 1.35, 95% CI 0.70–2.61) Table 3 shows the incidence of relevant cancers associated with various anatomical levels of UTS. The Table shows the adjusted HRs and 95% CIs of UTC, renal pelvis and ureter cancer (ICD-9-CM codes 189.1 and 189.2), and bladder cancer (ICD-9-CM code 188). In general, patients with lower
UTS(ICD-9-CM code 594) were 4.64-times more likely to develop UTC (95% CI 1.08–20.0) compared with the comparison group, whereas patients with upper UTS
(ICD-9-CM code 592) were 4.63-times more likely to develop UTC than the comparison group (95% CI 2.95–7.68). After controlling for age, sex, socio-economic status and
comorbidities, patients with upper UTS were 7.26-times more likely to develop renal pelvis and ureter cancer compared with the comparison cohort (95% CI 3.16–16.7). By contrast, patients with lower UTS were 7.82-times more likely to
develop bladder cancer compared with the comparison group (95% CI 1.74–35.1) after adjustment. The Kaplan–Meier analysis plot shows that the UTC-free probability was 0.73% lower in the study cohort than in the comparison cohort (log-rank test: P < 0.001; Fig. 1).
Discussion
The results of the present population-based retrospective cohort study showed a significant association between UTS and the risk of UTC. Further analyses showed that the risk was more predominant for women, and that the location of UTC tended to be congruent with the anatomical levels of UTS. In addition, the risk became more pronounced over time.
Cancer has been the leading cause of death in the general population of Taiwan since 1982. The age-adjusted incidence rate has increased steadily and reached 276 new cases per 100 000 patients in 2008 [20]. However, the USA has reported the opposite trend, with a decrease in the rate of
cancer-related deaths in recent years. Data from Surveillance Epidemiology and End Results showed that overall cancer incidence rates for all racial and ethnic groups combined in the USA dropped by 0.8% annually between 2003 and 2007 [23]. In Taiwan, the continuous rise of cancer incidence constitutes a major public health challenge, and has gained the attention of the government. A growing number of population-based investigations on cancer-prevention epidemiology have been conducted since 2007. The NHI database provides adequate data for such studies because of its comprehensive records of health coverage for Taiwanese citizens. The database allows investigators to select specific study groups and matched comparison groups, with ensured representation of the underlying population groups. Our research group recently conducted a cohort study using the LHID database to evaluate the relationship between sleep disorder and subsequent liver cancer risk, and the results showed a positive association [24]. For the present study, we used a similar design to determine whether UTS was associated with the risk of UTC.
To the best of our knowledge, this is the first nationwide population-based retrospective cohort study on UTS and UTC in Taiwan.We analysed the data of 21 862 patients with UTS from 1 million randomly selected insured Taiwanese residents. This data source represents the majority of people in Taiwan and offers undisputed generalizability. The comparison group was randomly frequency-matched with the patient group for age, sex, and index year, with two insured people (without
UTS) from the general population matched with every patient with UTS. Unlike our earlier cohort study, we calculated
the HRs instead of standardised incidence ratios (SIRs) to estimate the cancer risk. HRs provided a realistic assessment
because we did not select the data from whole population to were known not to have been diagnosed with UTS. Thus,
using HRs to estimate the endpoint for the two groups with and without a clear risk factor (in this case UTS) was appropriate.
might be associated with kidney stones [7–9]. However, several epidemiological studies have investigated the possible relationship between UTS and UTC, with inconsistent results [10–18]. Chow et al. [13] conducted a population-based cohort study in Sweden, in which patient cases detailing hospitalisation for kidney or ureter stones were followed for up to 25 years, to examine the subsequent risk of UTC. They reported that patients with kidney or ureter stones were at a significantly higher risk of developing renal pelvis, ureter, or bladder cancer compared with the comparison group, even at the 10-year follow-up point. In addition, the SIR was twice as high for women than for men. They also mentioned that kidney and ureter stones were located on the same side of the body as the tumour for most patients with renal pelvis or ureter cancer. The present findings were largely consistent with those obtained by Chow et al. First, patients with UTS showed a significantly higher risk of developing UTC. Second, the present results showed that women were at higher risk than men (IRR for women was 11.8 vs 4.30 for men). Third, as shown in Fig. 1, the risk became more pronounced over time, implying a possible causal association. Fourth, because of limitations in the present data, we categorised the lesion location (stone and cancer) by anatomical level rather than by laterality. The locations of UTC tended to be congruent with the anatomical levels of UTS.
Several studies have suggested that chronic inflammation resulting from bladder and kidney stones raises the risk of bladder cancer [11–14]. UTS might induce chronic irritation that may lead to proliferative urothelial changes; alternatively, the mechanical effect of stones on the epithelium might increase absorption and/or exposure to carcinogens in urine, which in turn may raise the risk of bladder cancer [17]. Animal studies have provided some evidence of chronic irritation by stones on the bladder epithelium, and have linked the presence of stones to neoplastic changes in the adjacent urothelium [25,26]. The proposed mechanism appears consistent with the present finding that most UTC occurred at the same anatomical level of the patient’s body as the UTS; Chow et al. [13] reported similar
results. The greater risk for women than for men remains unexplained, but it might be related to the higher frequency of UTIs reported among women [27].
The strength of the present study was its population-based design using a large sample that was considered representative of the general population of Taiwan. Therefore, the results provide a realistic indication of the incidence of UTC among patients with UTS.
However, the study was subject to some limitations. First, information on potential confounding factors, e.g. cigarette smoking and occupational exposure, was unavailable; thus, analyses adjusting for these variables could not be performed. Second, the LHID provided no information on cancer
histology type. Several studies have shown that squamous cell carcinoma of the bladder is related to UTS [8,13,28], but the present results did not confirm this.We did not include cancer of the renal parenchyma as an outcome for analysis because a previous study showed that UTS was not related to the development of RCC [13]. Third, for the control group, we are unable to identify patients who were definitely UTS-free, as only those who had symptoms or were diagnosed by screening would have the diagnosis of UTS. In addition, we are unable to recognise the patients as first time or recurrent UTS formers based on the database. Fourth, the possibility that a diagnosis of UTS might increase the chance of future medical episodes/imaging, which might increase incidental UTC
detection, and a surveillance bias might exist. Apart from these limitations, the data on UTS and UTC diagnoses used in the present analysis were sound.
In conclusion, the present study showed a positive association between UTS and UTC. Furthermore, the location (anatomical level) of the stones and cancer tended to be linked.Women with UTS were at higher risk for UTC than theirmale
counterparts. The risk of UTC increased with the length of time a patient continued to have UTS, suggesting a possible causal relationship. Further large-scale prospective studies are necessary to confirm the present findings, and possible
stone formers, (ii) metabolic status of the patients, (iii)
composition of the stones, (iv) stones remaining in the urinary tract for long period of time untreated, and (v) UTIs
