The role of genetic variations with low penetrance has earned special concern in urolithiasis research, which, in association with other risk factors, may define the critical threshold necessary for development of clinically significant renal stones. A number of genetic markers in different urolithiasis genes including SPP1, VDR, CaSR, urokinase, prothrombin, interleukins and others have been investigated in this regard (6). Osteopontin has earned a particular prominence among these genetic risk factors as an importance determinant and regulator of renal calcification and stone formation (11, 13, 35). However, the results of genetic association studies in urolithiasis are still to achieve diagnostic and translational significance. The gap in existing knowledge and inconsistent results for potential genetic associations in urolithiasis indicate a need of genetic epidemiology studies performed in diverse populations.
For the first time, we present a genetic association study investigating the role of common genetic variants in SPP1 gene using a sample set of Pakistani patients manifesting clinically significant urolithiasis where we demonstrate significant association of three SPP1 promoter polymorphisms (rs2853744:G>T, rs11730582:T>C and rs11439060:delG>G) with urolithiasis. We also demonstrated that subjects simultaneously harboring G-C-dG alleles of SPP1 rs2853744-rs11730582-rs11439060 polymorphisms, respectively, have 1.68 times increased risk of urolithiasis that is statistically significant as determined by haplotype association analysis.
Moreover, estimates of post-hoc study power showed that the levels of power associated with SPP1 rs2853744:G>T, rs11730582:T>C and rs11439060:delG>G polymorphisms were 78.4%, 79.8% and 99.7%. These results reflect that the sample size of 235 provided fairly adequate power (almost 80%) in determining genetic associations between these polymorphic variants of SPP1 gene and urolithiasis in the indigenous population.
We also analyzed whether additional risk factors, including gender, early age at presentation, severe disease (multiple renal stones, recurrences), presence of familial history of urolithiasis and parental consanguinity modulated the SPP1 polymorphisms based potential genetic risk of urolithiasis. However, sub-group analyses considering these additional risk factors reflected no moderation of genetic risk for SPP1 polymorphisms analyzed by any of the additional risk factors, at least in the context of present sample set.
Considerable heterogeneity in correlation of these SPP1 polymorphisms and risk of urolithiasis have been reported by studies conducted in different ethnic groups. An Iranian study (19) reported a positive association of G allele/GG genotype of rs2853744:G>T with risk of developing urolithiasis. In contrast to our study results, no significant association of rs11730582:T>C with urolithiasis risk was reported in 3 independent studies from Taiwan, Iran and China (19, 20, 36). For rs11439060:delG>G, significant association with urolithiasis phenotype was described in two studies, however, in consistence with our study, first study of Taiwanese origin reported dG/dG genotype as the risk genotype associated with increased susceptibilty of urolithiasis (36), while the second study found insertion allele or genotype (G allele or G/G genotype) to be more prevalent in Chinese urolithiasis patients as compared to controls (20).
The observed overall heterogeneity in the results of different studies regarding association of SPP1 polymorphisms with urolithiasis may be attributable to many factors of genetic and non-genetic (including demographic, environmental and analytical variants) in origin. With respect to genetic modulators, variations in allele/genotype distributions and LD pattern of SPP1 polymorphisms in different populations, reflecting a population specific genetic architecture, may determine substantial heterogeneity observed in genetic risk of urolithiasis. Similarly, non-genetic confounding factors may also explain a part of heterogeneity observed in the results of different genetic studies of SPP1 polymorphisms and urolithiasis risk, including; a) differential prevalence of urolithiasis in different regions home to different ethnicities [e.g. high (12-15%) vs moderate (9.6%) vs low (5.7%) prevalence of urolithiasis in Pakistan, Taiwan and Iran, respectively] (37-39); b) environmental risk factors (most importantly a lithogenic diet and lifestyle, and chronic dehydration) owing to differences in their context and relative contribution (4); and c) lack of standardization and consistency in study methodologies (reflecting selection bias, control source, genotyping method used, statistical analysis approach especially when it comes to HWE conformance and applying correction for multiple testing).
Meta-analysis is a powerful tool that provides evidence based comprehensive and reliable results compared to a single study when investigating association of potential risk factors and disease phenotype. Therefore, in addition to presenting results of indigenous study, we also conducted a meta-analysis to clarify the possible association between SPP1 polymorphisms and risk of urolithiasis. To the best of our knowledge, no meta-analysis has been carried out previously regarding association of SPP1 promoter polymorphisms with urolithiasis risk. The results of present meta-analysis reveal that GG genotype of SPP1 rs2853744:G>T significantly increased the risk of urolithiasis by 1.37 fold in a recessive model. However, no significant association between other SPP1 polymorphisms analyzed (rs11730582:T>C and rs11439060:delG>G) was observed after correction for multiple testing. Several indicators of the robustness of analyses done and results generated in the current meta-analysis can be identified including; a) all the studies included in the meta-analysis were in HWE, b) no publication bias or heterogeneity was observed except for rs11439060:delG>G, c) a single study did not influenced the cumulative results as suggested by sensitivity analysis, and d) correction for multiple testing was applied. Inclusion of Pakistani dataset in the overall meta-analysis also reinforced the results obtained in this study. However, the results should still be interpreted with caution because of the limited number of primary studies available for present meta-analysis.
Currently, there is only one meta-analysis available on the subject which revealed positive association of SPP1 coding region genetic variant (rs1126616:C>T) and lower serum and urine osteopontin levels with increased risk of developing urolithiasis (40), however, they did not include any other SPP1 polymorphism (including SPP1 promoter polymorphisms investigated in this study) in the analysis, which limits the usefulness and broader applicability of that study.
Despite the efforts made to generate evidence based and robust statistical results through current case control and meta-analysis study, a number of limitations should be acknowledged. First, a comprehensive investigation and correlation of biochemical parameters of renal stone disease (including stone analysis and serum/urine osteopontin levels) could not be done due to limited resources available. Second, all the eligible studies, including our own data, could not address all the known risk factors involved. Keeping in view the multifactorial nature of the urolithiasis phenotype, a more comprehensive and precise analysis should be based on adjusted estimates considering covariates such as gender, age, lifestyle, dietary habits including fluid intake, and other genetic factors, thus also investigating gene-gene and gene-environment interactions. Third, sub-group analysis based on ethnicity, source of control samples and other factors, could not be performed due to limited number of published studies available for current meta-analysis. Further, we did not include other SPP1 polymorphisms because we could find only a couple of studies with limited sample size.