Yayoi Nishida

Utilization of health care databases for pharmacoepidemiology

Pharmacoepidemiology is the study of the utilization and effects of drugs in clinical and population settings, and the outcomes of drug therapy. The growing trend of recording computerized data that will increasingly be automated into health care delivery is making the use of large datasets more and more common in pharmacoepidemiologic research. Most retrospective databases offer large populations and longer observation periods with real-world practice and can answer a variety of research questions quickly and cost-effectively. Observational studies, specifically using large databases, can complement findings from randomized clinical trials (RCTs) by assessing treatment effectiveness in patients encountered in daily clinical practice, although they are more exposed to bias and certainly are lower on the hierarchy of evidence than RCTs. Furthermore, careful defining of the research question with appropriate design and application of advanced statistical techniques, e.g., propensity-score analysis or marginal structural models, can yield findings with validity and improve causal inference of treatment effects. Some existing guidelines for comparative effectiveness help decision makers to evaluate the quality of observational studies comparing the effectiveness of various medical products and services. Thus, the trend for utilization of databases for pharmacoepidemiology will continue to grow in coming years.

Transcriptional profiling in hepatoblastomas using high-density oligonucleotide DNA array.

Hepatoblastoma is a common hepatic tumor in children. Although evidence regarding cytogenetic and molecular genetic alterations in hepatoblastomas has been reported, the molecular events affecting the biologic characteristics of this tumor, including alterations of the gene expression profile, are largely unknown. To identify genes differentially expressed between nondiseased liver (NDL) and hepatoblastoma tumor (HBT), we analyzed the gene expression profile in 14 NDL and 16 HBT samples using a high-density oligonucleotide DNA array. Using Mann-Whitney U test followed by the k-nearest neighbor algorithm, we identified 26 genes (predictor genes) that were able to assign unknown samples derived from NDL and HBT to either the NDL group or HBT group with 100% accuracy. Using a cross-validation approach, we confirmed that the k-nearest neighbor algorithm assigned the particular samples derived from NDL and HBT to either the NDL or HBT group with 93.3% (28/30 samples) accuracy. In the 26 predictor genes, we found alteration of the expression of genes regulating cell division (NAP1L1, STMN1, CCNG2, and CDC7L1) and tumor cell growth (IGF2 and IGFBP4) in HBT. Four predictor genes (ETV3, TPR, CD34, and NR1I3) were also found to be mapped to the chromosomal region 1q21 approximately q32, which has been reported to be frequently involved in the development of hepatoblastoma. The findings obtained in this study suggest that alteration of the expression of some genes regulating cell division and tumor cell growth may be characteristics of the gene expression profile in HBT, and that alteration of the expression of the four predictor genes mapped to chromosomal region 1q21 approximately q32 may also contribute to the differences in gene expression profile between NDL and HBT.

Effect of candesartan monotherapy on lipid metabolism in patients with hypertension: a retrospective longitudinal survey using data from electronic medical records

BackgroundStudies focusing on the add-on effects of angiotensin II type 1 receptor blockers (ARBs) other than their antihypertensive effect are receiving attention. However, the effects of prolonged administration of ARBs on lipid metabolism in clinical cases are unclear. Our aims were to survey the changes in plasma lipid profile in patients with hypertension over a one-year period, and to examine the correlations between these values and the time after the start of ARB monotherapy with candesartan.MethodsWe carried out candesartan monotherapy in patients with mild to moderate hypertension and examined the longitudinal changes in plasma lipid profile. Data from 405 patients for triglyceride (TG), 440 for total cholesterol (TC), 313 for high density lipoprotein cholesterol (HDL-C) and 304 for low density lipoprotein cholesterol (LDL-C) were obtained from the electronic medical records (EMRs) in the Clinical Data Warehouse (CDW) of Nihon University School of Medicine (NUSM). The inverse probability of treatment weighting (IPTW) method (calculated from the inverse of the propensity score) was used to balance the covariates and reduce bias in each treatment duration. Linear mixed effects models were used to analyse the relationship between these longitudinal data of blood examinations and covariates of patient sex, age, diagnosis of diabetes mellitus (DM) and duration of candesartan monotherapy.ResultsPlasma HDL-C level was associated with sex, duration of treatment, and interaction of sex and treatment duration, but not with age or diagnosis of DM. HDL-C level was significantly decreased during the 6~9 months period (p = 0.0218) compared with baseline. TG and TC levels were associated with sex, but not with age, diagnosis of DM or treatment duration. LDL-C level was not associated with any covariate. Analysis of the subjects divided by sex revealed a decrease in HDL-C in female subjects (during the 6~9 months period: p = 0.0054), but not in male subjects.ConclusionsOur study revealed that administration of candesartan slightly decreased HDL-C in female subjects. However, TG, TC and LDL-C levels were not influenced by candesartan monotherapy. Candesartan may be safely used for patients with hypertension with respect to lipid metabolism, because the effect of candesartan on lipids may be small.

Global profiling of influence of intra-ischemic brain temperature on gene expression in rat brain.

Mild to moderate differences in brain temperature are known to greatly affect the outcome of cerebral ischemia. The impact of brain temperature on ischemic disorders has been mainly evaluated through pathological analysis. However, no comprehensive analyses have been conducted at the gene expression level. Using a high-density oligonucleotide microarray, we screened 24000 genes in the hippocampus under hypothermic (32 degrees C), normothermic (37 degrees C), and hyperthermic (39 degrees C) conditions in a rat ischemia-reperfusion model. When the ischemic group at each intra-ischemic brain temperature was compared to a sham-operated control group, genes whose expression levels changed more than three-fold with statistical significance could be detected. In our screening condition, thirty-three genes (some of them novel) were obtained after screening, and extensive functional surveys and literature reviews were subsequently performed. In the hypothermic condition, many neuroprotective factor genes were obtained, whereas cell death- and cell damage-associated genes were detected as the brain temperature increased. At all intra-ischemic brain temperatures, multiple molecular chaperone genes were obtained. The finding that intra-ischemic brain temperature affects the expression level of many genes related to neuroprotection or neurotoxicity coincides with the different pathological outcomes at different brain temperatures, demonstrating the utility of the genetic approach.

Screening for control genes in rat global cerebral ischemia using high-density oligonucleotide array

From conventional relative gene expression analyses (Northern blotting, in situ hybridization, and RT‐PCR), it has been reported that the expression of control genes, such as glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) and β‐actin, used as references may be affected by ischemia. Therefore, we extended searching and evaluation at the mRNA level of transcripts whose expression levels were not changed by cerebral ischemia, using a high‐density oligonucleotide array and statistical analysis in a rat global cerebral ischemia and reperfusion model. We added a hyperthermic factor and localization factor to ischemia and identified transcripts with a stable expression level under conditions even more disadvantageous than ischemia only. Screening of more than 8,000 transcripts with the Rat Genome U34A array yielded 28 transcripts, which we listed and classified according to their expression level. Widely used control genes, GAPDH and β‐actin, were not included, although cyclophilin A was included. In addition, we conducted a functional classification based on gene ontology. Under the functional classification of the 28 transcripts, many genes tended to be associated with metabolism. In conclusion, use of several transcripts is recommended, such as those we identified, as references in the analysis of gene expression in pathological models of ischemia.

Comparative effect of angiotensin II type I receptor blockers and calcium channel blockers on laboratory parameters in hypertensive patients with type 2 diabetes

BackgroundBoth angiotensin II type I receptor blockers (ARBs) and calcium channel blockers (CCBs) are widely used antihypertensive drugs. Many clinical studies have demonstrated and compared the organ-protection effects and adverse events of these drugs. However, few large-scale studies have focused on the effect of these drugs as monotherapy on laboratory parameters. We evaluated and compared the effects of ARB and CCB monotherapy on clinical laboratory parameters in patients with concomitant hypertension and type 2 diabetes mellitus.MethodsWe used data from the Clinical Data Warehouse of Nihon University School of Medicine obtained between Nov 1, 2004 and July 31, 2011, to identify cohorts of new ARB users (n = 601) and propensity-score matched new CCB users (n = 601), with concomitant mild to moderate hypertension and type 2 diabetes mellitus. We used a multivariate-adjusted regression model to adjust for differences between ARB and CCB users, and compared laboratory parameters including serum levels of triglyceride (TG), total cholesterol (TC), non-fasting blood glucose, hemoglobin A1c (HbA1c), sodium, potassium, creatinine, alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyltransferase (GGT), hemoglobin and hematocrit, and white blood cell (WBC), red blood cell (RBC) and platelet (PLT) counts up to 12 months after the start of ARB or CCB monotherapy.ResultsWe found a significant reduction of serum TC, HbA1c, hemoglobin and hematocrit and RBC count and a significant increase of serum potassium in ARB users, and a reduction of serum TC and hemoglobin in CCB users, from the baseline period to the exposure period. The reductions of RBC count, hemoglobin and hematocrit in ARB users were significantly greater than those in CCB users. The increase of serum potassium in ARB users was significantly greater than that in CCB users.ConclusionsOur study suggested that hematological adverse effects and electrolyte imbalance are greater with ARB monotherapy than with CCB monotherapy.

Comparative effect of olmesartan and candesartan on lipid metabolism and renal function in patients with hypertension: a retrospective observational study

BackgroundAngiotensin II receptor blockers (ARBs), including olmesartan and candesartan, are widely used antihypertensive agents. Many clinical studies have demonstrated that ARBs have organ-protecting effects, e.g., cardioprotection, vasculoprotection and renoprotection. However, the effect of prolonged olmesartan monotherapy on lipid metabolism in patients with hypertension is less well studied. We performed a retrospective observational study to compare the effects of olmesartan with those of candesartan, focusing on lipid metabolism and renal function.MethodsWe used data from the Clinical Data Warehouse of Nihon University School of Medicine obtained between Nov 1, 2004 and Feb 28, 2011, to identify cohorts of new olmesartan users (n = 168) and candesartan users (n = 266). We used propensity-score weighting to adjust for differences in all covariates (age, sex, comorbid diseases, previous drugs) between olmesartan and candesartan users, and compared serum chemical data including serum triglyceride (TG), LDL-cholesterol (LDL-C), total cholesterol (TC), potassium, creatinine and urea nitrogen. The mean exposure of olmesartan and candesartan users was 126.1 and 122.8 days, respectively.ResultsAfter adjustment, there were no statistically significant differences in all covariates between olmesartan and candesartan users. The mean age was 60.7 and 61.0 years, and 33.4% and 33.7% of olmesartan and candesartan users were women, respectively. There were no statistically significant differences in mean values for all laboratory tests between baseline and during the exposure period in both olmesartan and candesartan users. In olmesartan users, the reduction of serum TG level was significant in comparison with that in candesartan users. Other parameters of lipid profile and renal function showed no statistically significant difference in the change from baseline to during the exposure period between olmesartan and candesartan users.ConclusionsIn this study, we observed a more beneficial effect on lipid metabolism, a reduction of serum TG, with olmesartan monotherapy than with candesartan monotherapy. However, there were no clinically significant changes in the levels of all test parameters between baseline and during the exposure period with both drugs. These results suggest that the influence of olmesartan or candesartan monotherapy on lipid metabolism and renal function is small, and that they can be safely used in patients with hypertension.

Adverse effect profile of trichlormethiazide: a retrospective observational study

BackgroundTrichlormethiazide, a thiazide diuretic, was introduced in 1960 and remains one of the most frequently used diuretics for treating hypertension in Japan. While numerous clinical trials have indicated important side effects of thiazides, e.g., adverse effects on electrolytes and uric acid, very few data exist on serum electrolyte levels in patients with trichlormethiazide treatment. We performed a retrospective cohort study to assess the adverse effects of trichlormethiazide, focusing on serum electrolyte and uric acid levels.MethodsWe used data from the Clinical Data Warehouse of Nihon University School of Medicine obtained between Nov 1, 2004 and July 31, 2010, to identify cohorts of new trichlormethiazide users (n = 99 for 1 mg, n = 61 for 2 mg daily dosage) and an equal number of non-users (control). We used propensity-score matching to adjust for differences between users and control for each dosage, and compared serum chemical data including serum sodium, potassium, uric acid, creatinine and urea nitrogen. The mean exposure of trichlormethiazide of 1 mg and 2 mg users was 58 days and 64 days, respectively.ResultsThe mean age was 66 years, and 55% of trichlormethiazide users of the 1 mg dose were female. In trichlormethiazide users of the 2 mg dose, the mean age was 68 years, and 43% of users were female. There were no statistically significant differences in all covariates (age, sex, comorbid diseases, past drugs, and current antihypertensive drugs) between trichlormethiazide users and controls for both doses. In trichlormethiazide users of the 2 mg dose, the reduction of serum potassium level and the elevation of serum uric acid level were significant compared with control, whereas changes of mean serum sodium, creatinine and urea nitrogen levels were not significant. In trichlormethiazide users of the 1 mg dose, all tests showed no statistically significant change from baseline to during the exposure period in comparison with control.ConclusionsOur study showed adverse effects of decreased serum potassium and increased serum uric acid with trichlormethiazide treatment, and suggested that a lower dose of trichlormethiazide may minimize these adverse effects. These findings support the current trend in hypertension therapeutics to shift towards lower doses of thiazides.

Comparison of the effect of mesalazine and sulfasalazine on laboratory parameters: a retrospective observational study

PurposeMesalazine and sulfasalazine are commonly used drugs for the treatment of inflammatory bowel disease. However, there have been few reports with a strict statistical analysis comparing the effects of mesalazine and sulfasalazine on laboratory test results. Therefore, we designed a retrospective cohort study to investigate whether or not differences in clinical laboratory parameters exist between mesalazine and sulfasalazine users.MethodsWe used data from the Clinical Data Warehouse of Nihon University School of Medicine to identify cohorts of new mesalazine users (n = 303) and sulfasalazine users (n = 67). We used a multivariate regression model and regression adjustment with the propensity score to adjust for differences in baseline covariates between mesalazine and sulfasalazine users, and compared serum levels of creatinine, urea nitrogen, aspartate aminotransferase, alanine aminotransferase, and hematological parameters including red and white blood cell counts and platelet count.ResultsAfter adjustment, in sulfasalazine users, the mean values for all tests showed no significant change between baseline and during the exposure period. In contrast, in mesalazine users, the mean WBC and platelet counts during the exposure period were significantly lower than those at baseline. Furthermore, mean serum urea nitrogen level during the exposure period was significantly higher than that at baseline. In terms of mean changes in laboratory test values during the exposure period compared with baseline, the reduction of platelet count in mesalazine users was significant in comparison to that in sulfasalazine users.ConclusionOur findings suggested that the hematological adverse effects of mesalazine treatment might be greater than those of sulfasalazine treatment.

Comparative effect of clopidogrel plus aspirin and aspirin monotherapy on hematological parameters using propensity score matching

Background Clopidogrel and aspirin are antiplatelet agents that are recommended to reduce the risk of recurrent stroke and other cardiovascular events. Dual antiplatelet therapy with clopidogrel and aspirin has been shown to increase the risk of hemorrhage, but the effects of the drugs on laboratory parameters have not been well studied in real-world clinical settings. Therefore, we evaluated and compared the effects of combination therapy with clopidogrel plus aspirin and aspirin monotherapy on laboratory parameters. Methods We used data from the Nihon University School of Medicine Clinical Data Warehouse obtained between November 2004 and May 2011 to identify cohorts of new users (n = 130) of clopidogrel (75 mg/day) plus aspirin (100 mg/day) and a propensity score matched sample of new users (n = 130) of aspirin alone (100 mg/day). We used a multivariate regression model to compare serum levels of creatinine, aspartate aminotransferase, and alanine aminotransferase, as well as hematological parameters including hemoglobin level, hematocrit, and white blood cell, red blood cell, and platelet counts up to 2 months after the start of administration of the study drugs. Results There were no significant differences for any characteristics and baseline laboratory parameters between users of clopidogrel plus aspirin and users of aspirin alone. Reductions in white blood cell and red blood cell counts, hemoglobin levels, and hematocrit in users of clopidogrel plus aspirin were significantly greater than those in users of aspirin alone. Conclusion Our findings suggest that adverse hematological effects may be greater with combination clopidogrel plus aspirin therapy than with aspirin monotherapy.