Incidence of urological tumors in Down’s syndrome: a systematic review and meta-analysis

Background Some authors have estimated that the incidence of testicular germ cell tumors in individuals with trisomy 21 is more than fivefold higher than that in the general population. Objective This systematic review aimed to estimate the incidence of urological tumors in patients with Down’s syndrome. Study design We conducted a search strategy in MEDLINE (OVID), EMBASE, LILACS, and the Cochrane Central Register of Controlled Trials (CENTRAL) from inception to nowadays. We assessed the risk of bias and performed a meta-analysis. Also, the heterogeneity between trials was evaluated by the I2 test. We completed the subgroup analysis based on the type of urological tumor (testis, bladder, kidney, upper urological tract, penile, retroperitoneum). Results We found 350 studies by the search strategy. After carefully reviewing, full-text studies were included. 16,248 individuals with Down’s syndrome were included, and 42 patients presented with urological tumors. There was a total incidence of 0.1%, 95%CI (0.06–0.19), I2 61%. The most common urological tumor reported was testicular. We found six studies describing 31 events and an overall incidence of 0.19%, 95%CI (0.11–0.33), I2: 51%. Other studies reported kidney, penile, upper urinary tract, bladder, and retroperitoneum tumors with a very low incidence, 0.02%, 0.06%, 0.03%, 0.11%and 0.07%, respectively. Discussion Regarding non-testicular urological tumors, we found incidences as low as 0.02% in kidney cancer or 0.03% in the upper-urothelial tract tumors. It is also lower than the general population. Compared to the age of onset of patients, it is also lower than the general population, perhaps related to a shorter life expectancy. As a limitation, we found a high heterogeneity and a lack of information regarding non-testicular tumors. Conclusion There was a very low incidence of urological tumors in people with Down’s syndrome. Testis tumor was the most frequently described in all cohorts and within a normal distribution range.


Introduction
Down syndrome is a congenital abnormality characterized by the presence of all or part of a third copy of chromosome 21 [1]. It is the most common chromosomal disorder associated with a spectrum of physical and functional disability and a predisposition to developing hematopoietic malignancies, mainly leukemia [1][2][3]. The solids tumors are less frequent than the general population, except for testicular cancers [3][4][5]. Some authors have estimated that the incidence of testicular germ cell tumors in individuals with trisomy 21 is more than fivefold higher than in the general population [6]. In line with that, a decreased risk of dying from other urological cancers, such as bladder or kidney, has been seen in persons with Down's syndrome [7].
In chromosome 21, the Down syndrome critical region (DSCR), an overexpression is responsible for the phenotypic features observed, such as mental retardation and developmental heart defects [8,9]. Many studies indicate a possible role as a tumor suppressor gene; however, others prove its pro-oncogenic activity [10,11]. Some hypotheses have been raised that the extracellular matrix secreted by fibroblasts from individuals with DS suppresses the proliferation of stromal-rich tumors (e.g., breast cancer), but not stromalpoor tumors (e.g., lymphoproliferative disorders, retinoblastoma, and germ cell tumors) [12].
Data on the incidence or frequency of solid tumors (specifically urological tumors) in this population are heterogeneous, in contrast to hematopoietic malignancies, where they are widely known [5]. The need to check this information has caught our attention-this systematic review aimed to estimate the incidence of urological tumors in patients with Down's syndrome.

Methods
We conducted this systematic review following the Cochrane Collaboration recommendations and PRISMA statement.

Eligibility criteria
Study designs: We included clinical trials and observational studies.
Participants: Patients with Down's syndrome and any urological tumor were included. There were no preferences in any other demographic characteristic of participants.
Outcomes: The primary outcome was the incidence of urological tumors in patients with Down's syndrome.
Exclusion criteria: Studies included fewer than five patients with Down's syndrome and any urological tumor.

Information sources and search strategy
We designed a search strategy in MEDLINE (OVID), EMBASE, LILACS, and the Cochrane Central Register of Controlled Trials (CENTRAL) from inception to nowadays (Appendix 1). The search strategy was specific for each database and included a combination of the medical subject headings for incidence and Down's syndrome. To ensure literature saturation, other electronic sources were used to find additional studies' references from relevant articles identified through the search, conferences, thesis databases, Open Grey, Google Scholar, and clinicaltrials.gov, among others. If some information was missing, we contacted the authors by e-mail in case of missing data. No language or setting restrictions were imposed.

Data collection
Two researchers (F.K, MPG) reviewed each reference by title and abstract. Then, reviewers confirmed all data in full texts of relevant studies, applied pre-specified inclusion and exclusion criteria, and extracted the data. Disagreements were resolved by consensus, and where disagreement could not be solved, a third reviewer (H.A.G) dissolved the conflict.
Relevant data were collected using a standardized form to extract the following information from each article: study design, geographic location, authors' names, title, objectives, inclusion and exclusion criteria, number of patients included, losses to follow-up, timing, definitions of outcomes, outcomes, and association measures and funding source.

Risk of bias
We assessed the risk of bias by the STROBE statement.

Data analysis/synthesis of results
We performed the analyses in R. We performed a metaanalysis of proportions with the command metaprop and the method inverse (logit transformed proportions). We performed this statistical approach and the subgroup analysis according to the high clinical heterogeneity expected and a vast proportion of variation among studies. We assessed the heterogeneity through the I 2 test, and the values of < 50% and > 50% in the I 2 test represent low and high levels of heterogeneity, respectively. We reported the results as forest plots with a 95% confidence interval (95% CI).

Sensitivity analysis
We analyzed the sensitivity by extracting weighted studies and running the estimated effect to find differences.

Subgroup analysis
We performed the subgroup analysis based on the type of urological tumor (testis, bladder, kidney, upper urological tract, penile, retroperitoneum).

Characteristics of included studies
We found a total number of 16,248 individuals with Down syndrome. The proportion of males and females was not described in all studies. Primary studies came from Sweden, Denmark, Japan, Australia, Finland, and Britain, and they collected information between 1963 and 2012. All studies were retrospective and multi-center, of which five of six were cohorts, and one was case-control. Furthermore, Hill's (2003) study was based on cohorts and databases, while all the others considered databases and autopsy data.
The age of presentation of those tumors ranged from 15 weeks to 45 years (Table 1).
Urological tumors were located in the testicles, penis, kidney, pelvis, ureter, bladder, and retroperitoneum. Only histological subtypes of testicular tumors were described: seminoma, teratocarcinoma, embryonal carcinoma, yolk sac tumor, and mixed germ cell.

Characteristics of excluded studies
The exclusion criteria were the minimum number of reported individuals, study design, full text unavailable, or no mention of the primary outcome. Hasle 2016, Goldcare 2004, and Hill 2003 showed low risk in most items [4,13,15]. In addition, Sullivan 2006 did not describe how they performed their statistical analysis or calculation of the sample size or inclusion criteria, even though all the characteristics are represented in Table 1. The selection bias was reduced in Patja 2006, because patients from all levels of care were included in the Danish database, and its coverage was 99%. Nevertheless, there is a risk of information bias due to patients who migrated and being lost to follow-up. Ehara's 2011 study was based on an autopsy database, and the data are incomplete in some cases, which can induce information bias, as mentioned by the authors. Though they describe in tables the distribution by age, sex, and tumor type, these have not been adequately described [2,14,16] (Table 2).

Subgroup analysis
The most common urological tumor was testicular. We found six studies describing 31 events and an overall incidence of 0.19%, 95%CI (0.11-0.33), I 2 : 51%. Other studies reported a very low incidence of kidney, penile, upper urinary tract, bladder, and retroperitoneal tumors (Fig. 2).

Sensitivity analysis
We did not find any difference when performing the sensitivity analysis.

Discussion
In contrast to the general population, an estimated rate of new cases of testicular cancer of 6.0 per 100,000 men per year was observed in the USA (SEER). Approximately, 0.4% of men will be diagnosed with testicular cancer during their lifetime, and almost 75% of these cases will occur between 20 and 44 years of age. Furthermore, less than 20% of testicular cancer cases will appear after age 45 [17]. Likewise, in our findings, testicular cancer was present between 20 and 45 years, and none after this age.
Our study has found 31 cases of testicular cancer among 8720 males (16,248 individuals), representing a lifetime incidence of 355 new cases for 100,000 patients with Down syndrome. These numbers mean that 1 in every 281 males   [2][3][4].
According to the SEER database, 0.4% is the estimated lifetime risk of developing testicular cancer for the general population, and this is very similar for patients with Down syndrome in our results. In the same way, regarding the histological subtypes in testicular tumors, seminoma is the most common (58.38%), followed by non-seminoma (40.7%) [17]. Accordingly, we found a similar distribution of testicular tumor subtypes in patients with Down syndrome.
In our study, non-testicular urological tumors were present much less frequently in patients with Down syndrome. However, there is no previous meta-analysis in this regard. The standardized incidence ratio for kidney cancer of 0.6 is mentioned in some studies, such as those by Hill et al. or Patja et al. [2,13]. In addition, Hesle et al. showed SIR of 2.36 for pelvis and ureter tumors, while SIR was 1.0 for bladder tumors [4]. Based on SEER, kidney cancer incidence was 16.4 per 100,000, and approximately 1.7% of the population will be diagnosed with this tumor sometime during their lifetime.
Regarding non-testicular urological tumors, we found incidences as low as 0.02% in kidney cancer or 0.03% in upper-urothelial tract tumors. It was also lower than that in the general population. Comparing the age of onset of patients, it was also lower than that in the general population, perhaps related to a shorter life expectancy.

Strengths and limitations
We carried out a very well-designed systematic review of a specific population. We included studies worldwide, pooling general and subtype information on urological tumors.
As a limitation, we found a high heterogeneity and a lack of information regarding non-testicular tumors. However, these findings were consistent with the literature. Other limitations were that follow-up time was not possible to be determined. Likewise, information was generally based on national databases, which might under-report cancer in this population and can not only lead to information bias, but also conduce a selection bias because this population has perhaps stricter follow-up than the general population. There may be stricter follow-up regarding pathologies, especially at younger ages [17]. However, the incidence of urological

Goldcare 2004
Fully fulfilled or low risk is represented by green, and not met or high risk is represented by red. Items 1-22 are based on the STROBE checklist available for consultation online at https: www. strobe-state ment. org tumors is lower for the population with Down syndrome than for the general population. However, a finding like that described for a similar population was found for testicular cancer. Therefore it cannot be concluded that the selection risk impacted our data, since both incidences would have had an upward behavior.

Conclusions
The Down syndrome population had a very low incidence of urological tumors. Testis tumor was the most frequently described in all cohorts, but similar to the general population. Urologists must carefully examine patients with this condition to prevent missing this tumor.