Tosiya Sato

Marginal structural models as a tool for standardization

In this article, we show the general relation between standardization methods and marginal structural models. Standardization has been recognized as a method to control confounding and to estimate causal parameters of interest. Because standardization requires stratification by confounders, the sparse-data problem will occur when stratified by many confounders and one then might have an unstable estimator. A new class of causal models called marginal structural models has recently been proposed. In marginal structural models, the parameters are consistently estimated by the inverse-probability-of-treatment weighting method. Marginal structural models give a nonparametric standardization using the total group (exposed and unexposed) as the standard. In epidemiologic analysis, it is also important to know the change in the average risk of the exposed (or the unexposed) subgroup produced by exposure, which corresponds to the exposed (or the unexposed) group as the standard. We propose modifications of the weights in the marginal structural models, which give the nonparametric estimation of standardized parameters. With the proposed weights, we can use the marginal structural models as a useful tool for the nonparametric multivariate standardization.

Health-Care-Associated Pneumonia Among Hospitalized Patients in a Japanese Community Hospital

BACKGROUND Health-care-associated pneumonia (HCAP) is a relatively new concept. Epidemiologic studies are limited, and initial empirical antibiotic treatment is still under discussion. This study aimed to reveal the differences in mortality and pathogens between HCAP and community-acquired pneumonia (CAP) in each severity class, and to clarify the strategy for the treatment of HCAP. METHODS We conducted a retrospective observational study of patients with HCAP and CAP who were hospitalized between November 2005 and January 2007, and compared baseline characteristics, severity, pathogen distribution, antibiotic regimens, and outcomes. In each severity class (mild, moderate, and severe) assessed using the A-DROP scoring system (ie, age, dehydration, respiratory failure, orientation disturbance, and low BP), we investigated the in-hospital mortality and occurrence of potentially drug-resistant (PDR) pathogens. RESULTS A total of 371 patients (141 HCAP patients, 230 CAP patients) were evaluated. The proportion of patients in the severe class was higher in the HCAP patients than in CAP patients. In the moderate class, the in-hospital mortality proportion of HCAP patients was significantly higher than that of CAP patients (11.1% vs 1.9%, respectively; p = 0.008). In moderate-class patients in whom pathogens were identified, PDR pathogens were isolated more frequently from HCAP patients than from CAP patients (22.2% vs 1.9%, respectively; p = 0.002). The occurrence of PDR pathogens was associated with initial treatment failure and inappropriate initial antibiotic treatment. CONCLUSIONS The present study provides additional evidence that HCAP should be distinguished from CAP, and suggests that the therapeutic strategy for HCAP in the moderate class holds the key to improving mortality. Physicians may need to consider PDR pathogens in selecting the initial empirical antibiotic treatment of HCAP.

Effects of Candesartan Compared With Amlodipine in Hypertensive Patients With High Cardiovascular Risks: Candesartan Antihypertensive Survival Evaluation in Japan Trial

The Candesartan Antihypertensive Survival Evaluation in Japan Trial was designed to compare the long-term effects of the angiotensin II receptor blocker candesartan and the calcium channel blocker amlodipine on the incidence of cardiovascular events, represented as a composite of sudden death and cerebrovascular, cardiac, renal, and vascular events in high-risk Japanese hypertensive patients. We conducted a prospective, randomized, open-label study with blinded assessment of the end point in 4728 Japanese hypertensive patients (mean age: 63.8 years; mean body mass index: 24.6 kg/m2). Patients were followed for an average of 3.2 years. Blood pressure was well controlled with both treatment-based regimens (systolic blood pressure/diastolic blood pressure: 136.1/77.3 mm Hg for candesartan-based regimens and 134.4/76.7 mm Hg for amlodipine-based regimens after 3 years). Primary cardiovascular events occurred in 134 patients with both the candesartan- and amlodipine-based regimens. The 2 treatment-based regimens produced no significant differences in cardiovascular morbidity or mortality in the high-risk Japanese hypertensive patients (hazard ratio: 1.01; 95% CI: 0.79 to 1.28; P=0.969). In each primary end point category, there was no significant difference between the 2 treatment-based regimens. New-onset diabetes occurred in fewer patients taking candesartan (8.7/1000 person-years) than in those taking amlodipine (13.6/1000 person-years), which resulted in a 36% relative risk reduction (hazard ratio: 0.64; 95% CI: 0.43 to 0.97; P=0.033). We disclosed that candesartan-based and amlodipine-based regimens produced no statistical differences in terms of the primary cardiovascular end point, whereas candesartan prevented new-onset diabetes more effectively than amlodipine.

Antibody to hepatitis B core antigen and risk for hepatitis C-related hepatocellular carcinoma : A prospective study

Context Retrospective studies suggest that exposure to hepatitis B virus (HBV) may contribute to the development of hepatocellular carcinoma (HCC) in hepatitis C virus (HCV)positive patients with cirrhosis. Contribution These investigators prospectively studied patients with chronic HCV infection and evidence of occult HBV infection (negative results for hepatitis B surface antigen and HBV DNA but positive results for antibody to hepatitis B core antigen [anti-HBc] on serologic testing). Patients with HCV-related cirrhosis and positive results for anti-HBc on serologic testing were at high risk for HCC. Anti-HBc positivity was associated with increased risk for HCC, even in patients with a virologic response to interferon therapy. Caution The effect of alcohol cannot be fully assessed because of the small number of study patients who drank moderately. Implication Anti-HBc serologic testing may be a valuable indicator of special risk for HCC in patients with HCV-related cirrhosis. The Editors Hepatocellular carcinoma (HCC) is one of the most common types of cancer worldwide, and its incidence has been increasing (1, 2). In Japan, an endemic area for hepatitis B virus (HBV) and hepatitis C virus (HCV), it is well known that more than 75% of cases of HCC are attributable to HCV-related chronic liver disease, and nearly 15% are attributable to HBV-related liver disease (3). Several reports have focused on the clinical role of HBV as a unique infection, in which HBV DNA is detectable in the liver despite the absence of serum hepatitis B surface antigen (HBsAg) (46). It is increasingly recognized that after a person is exposed to HBV, infection persists in the liver for a prolonged period (710). This unique persistent infection, known as occult (or latent) HBV infection, is characterized by HBV DNA in the liver but no HBsAg in the serum (1113). In most cases, antibody to hepatitis B core antigen (anti-HBc) is detectable, and thus anti-HBc is believed to be a surrogate marker for latent carriers (14). In fact, we recently showed that HBV infection invariably occurred through grafts from anti-HBcpositive donors in HBV-naive recipients through living-donor liver transplantation (15). In addition to our data, other reports showing frequent HBV transmission from anti-HBcpositive cadaveric donors to recipients indicate that most healthy persons who are positive for anti-HBc, even at low titers, have latent HBV infection in liver tissue (11, 1618). Indeed, we have shown that most anti-HBcpositive healthy persons have a latent episomal form of HBV infection accompanied by ongoing viral replication (19, 20). In contrast, the prevalence of latent HBV infection in anti-HBcpositive patients with HCV-related chronic liver disease, including chronic hepatitis, cirrhosis, and HCC, remains controversial (21). However, several reports have revealed that the HBV genome is frequently detectable in liver tumors in anti-HBcpositive, HBsAg-negative patients with HCV-related liver disease, which suggests that occult HBV infection may contribute to the progression of liver damage and the development of HCC in HCV-positive patients (14, 2227). In a large-scale retrospective study of the prevalence of anti-HBc among 2014 patients with chronic HCV infection, we found that nearly 50% of patients with HCV-related liver disease had anti-HBc (28). Moreover, we found a strong correlation between the prevalence of anti-HBc and the clinical progression of liver disease. The prevalence of anti-HBc was approximately 60% in patients with HCV-related HCC (28). This high prevalence of anti-HBc in HCV-positive patients, particularly those with HCC, strongly suggests that previous exposure to HBV plays an important role in the development of HCC in patients with HCV-related chronic liver disease. Therefore, we performed a prospective study to determine whether previous exposure to HBV affects the clinical course, especially in development of HCC, in patients with chronic HCV infection. Methods Patients Patients with chronic HCV infection who presented to Kyoto University, Kyoto, Japan, and 14 affiliated core hospitals from May 1995 to June 1995 were enrolled. To be eligible, patients had to have serologically confirmed HCV infection without HBsAg and HBV DNA in sera. All patients had been followed with biochemical tests, including -fetoprotein, and ultrasonography or computed tomography (CT) every 3 to 6 months before and after enrollment. We excluded patients who had elevated -fetoprotein levels or those in whom HCC had been diagnosed before enrollment. As a result, 872 patients with chronic HCV infection were enrolled. We discontinued follow-up in patients who moved from the study districts but included their clinical data until they moved. The end of follow-up was defined as the date of diagnosis of HCC, date of death, date of move from the study district, or the closing date of the study (15 May 2005). A total of 384 patients were classified into the hepatitis group or cirrhosis group on the basis of histologic findings on liver biopsy. The differential diagnosis of cirrhosis or hepatitis was made in the remaining 488 patients by using the cirrhosis discriminant score (29, 30). This score is based on 3 laboratory variables: platelet count, alanine aminotransferaseaspartate aminotransferase ratio, and prothrombin time. It has been shown to be highly sensitive in identifying cirrhosis in patients with HCV infection. In accordance with the original definition, patients with a high score (8) were classified into the cirrhosis group. Patients with ascites confirmed by ultrasonography or CT or previous variceal bleeding were given a diagnosis of cirrhosis, regardless of their score, because these findings are strong indicators of portal hypertension and most likely cirrhosis. As a result, 597 (68.5%) patients had a diagnosis of chronic hepatitis and 275 (31.5%) patients had a diagnosis of cirrhosis at the time of enrollment. The patients had regular clinical assessments, biochemical tests, and ultrasonography or CT of the liver every 3 to 6 months during the follow-up period. Patients were stratified into 3 categories according to their smoking habits: nonsmokers, light smokers who smoked fewer than 20 pack-years, and heavy smokers who smoked 20 pack-years or more. Similarly, we stratified patients into 3 categories according to their drinking habits: nondrinkers, moderate drinkers with an average ethanol intake less than 30 g/d, and heavy drinkers with an average ethanol intake greater than 30 g/d. Information on average alcohol intake was based on the patients drinking habits during the 15 years before study entry. All patients provided informed consent to participate in the study, and the study was designed in accordance with the Declaration of Helsinki (31). Serologic Studies At study entry, serum samples from each patient were tested for serologic markers of HCV. Detection of HBsAg, anti-HBc, and antibody to hepatitis B surface antigen (anti-HBs) was performed by using commercial enzyme immunoassay kits (Dainabot, Tokyo, Japan) (28). Results of the anti-HBc assays were expressed as the percentage of inhibition, and the specimen was considered to be anti-HBc positive when the percentage of inhibition was greater than 50% (19). Detection of HBV DNA was done by using DNA probe assay (32). Anti-HCV was assessed by using second-generation assays (Dainabot) (28). Serum HCV RNA levels were determined in 254 patients by using a competitive reverse transcriptionpolymerase chain reaction assay, and positivity of HCV RNA was confirmed in all patients who were examined at study entry (33). History of Interferon Therapy Of the 576 patients with chronic hepatitis, 224 had a history of interferon therapy. One hundred ninety-two patients received 5 to 10 million U of interferon- intramuscularly every day for the first 2 weeks and then 3 times weekly for the following 22 weeks. The remaining 32 patients were treated with 3 to 6 million U of interferon- intravenously every day for 8 weeks. No patient received pegylated interferon or combination therapy with ribavirin. Patients who received interferon were divided into 3 groups based on their virologic response to therapy. Patients with a sustained virologic response were defined as those with no detectable HCV RNA by qualitative assay at least 24 weeks after cessation of therapy. Patients with relapse were defined as those with disappearance of viremia at the end of treatment followed by reappearance of viremia with 24 weeks. Nonresponders included patients whose serum HCV RNA remained positive during therapy. Statistical Analysis The incidence rates for HCC are expressed as the number of HCC cases per 1000 person-years. Incidence rate ratios were calculated by dividing rates, and the exact 95% CIs for the rate ratios were calculated on the basis of a binomial distribution, which is a conditional distribution for 2 independent Poisson distributions. The Cox proportional hazards model was used to calculate the incidence rate ratios for the association between HBc seropositivity and HCC incidence. The multivariate model, with adjustment for potential risk factors, included male sex, age, alcohol intake (none, 0 to 30 g/d, and 30 g/d), smoking (none, 0 to 20 pack-years, and 20 pack-years), and history of interferon therapy (yes or no). The associated 95% Wald CIs were calculated. Analyses were done by using PC-SAS, version 8.2 (SAS Institute, Inc., Cary, North Carolina) and JMP, version 4.0 (SAS Institute, Inc.). Role of the Funding Source The Japan Society for the Promotion of Science provided funding for the study. The funding source had no role in the collection, analysis, or interpretation of the data or in the decision to submit the paper for publication. Results Characteristics of Patients at Enrollment We followed 846 of the 872 enrolled patients. The remaining 26 patients were excluded from the analysis. Twenty-on

Differentiation-defective phenotypes revealed by large-scale analyses of human pluripotent stem cells

Significance In the past few years, findings have been controversial in regard to whether human induced pluripotent stem cells (hiPSCs) are distinct from human embryonic stem cells (hESCs) in their molecular signatures and differentiation properties. In this study, hiPSCs and hESCs have overlapping variations in molecular signatures such as RNA expression and DNA methylation. However, some hiPSC clones retained a significant number of undifferentiated cells even after neural differentiation culture and formed teratoma when transplanted into mouse brains. These differentiation-defective hiPSC clones were marked by higher expression levels of several genes, including those expressed from long terminal repeats of specific human endogenous retroviruses. They need to be identified and eliminated prior to applications in regenerative medicine. We examined the gene expression and DNA methylation of 49 human induced pluripotent stem cells (hiPSCs) and 10 human embryonic stem cells and found overlapped variations in gene expression and DNA methylation in the two types of human pluripotent stem cell lines. Comparisons of the in vitro neural differentiation of 40 hiPSCs and 10 human embryonic stem cells showed that seven hiPSC clones retained a significant number of undifferentiated cells even after neural differentiation culture and formed teratoma when transplanted into mouse brains. These differentiation-defective hiPSC clones were marked by higher expression levels of several genes, including those expressed from long terminal repeats of specific human endogenous retroviruses. These data demonstrated a subset of hiPSC lines that have aberrant gene expression and defective potential in neural differentiation, which need to be identified and eliminated before applications in regenerative medicine.

Transarterial chemotherapy alone versus transarterial chemoembolization for hepatocellular carcinoma: A randomized phase III trial

BACKGROUND/AIMS Transcatheter arterial chemoembolization (TACE) is a combination of transarterial infusion chemotherapy (TAI) and embolization, and has been widely used to treat patients with hepatocellular carcinoma (HCC). However, since the impact of adding embolization on the survival of patients treated with TAI had never been evaluated in a phase III study, we conducted a multi-center, open-label trial comparing TACE and TAI to assess the effect of adding embolization on survival. METHODS Patients with newly diagnosed unresectable HCC were randomly assigned to either a TACE group or a TAI group. Zinostatin stimalamer was injected into the hepatic artery, together with gelatin sponge in the TACE group and without gelatin sponge in the TAI group. Treatment was repeated when follow-up computed tomography showed the appearance of new lesions in the liver or re-growth of previously treated tumors. RESULTS Seventy-nine patients were assigned to the TACE group, and 82 were assigned to the TAI group. The two groups were comparable with respect to their baseline characteristics. At the time of the analysis, 51 patients in the TACE group and 58 in the TAI group had died. The median overall survival time was 646 days in the TACE group and 679days in the TAI group (p=0.383). CONCLUSIONS The results of this study suggest that treatment intensification by adding embolization did not increase survival over TAI with zinostatin stimalamer alone in patients with HCC.

Dose‐response Trial of Lactoferrin in Patients with Chronic Hepatitis C

Hepatitis C virus (HCV) is one of the most common causes of chronic hepatitis. Interferon is presently the only effective treatment for chronic hepatitis C (CH‐C), though its effectiveness is limited. Lactoferrin (LF), which is an 80‐kDa, iron‐binding glycoprotein, has several biological activities including anti‐viral activity, and it was recently reported to inhibit HCV infection in cultured human hepatocytes. The present trial was designed to assess the relationship between the dose of bovine LF (bLF) and the effect of bLF on serum alanine aminotransaminase (ALT) and HCV RNA levels in patients with CH‐C. Forty‐five patients entered at each of the three dose levels (bLF of 1.8, 3.6, and 7.2 g/day) received orally an 8‐week course of bLF. There was no significant relation between the dose of bLF and the effect of bLF on serum ALT or HCV RNA levels. Biochemical (a 50% or greater decrease in the serum ALT level) and virological (a 50% or greater decrease in HCV RNA level) responses were observed in two and four patients, respectively, but all responders relapsed during the follow‐up period after bLF treatment. The bLF treatment was generally well tolerated, and no patient had any serious adverse event. In conclusion, the excellent tolerance and potential anti‐HCV activity of bLF shown in this trial suggest that further trials using a large number of patients are mandatory. We are currently conducting a double‐blind randomized controlled trial comparing bLF with placebo to clarify the anti‐HCV activity of bLF in patients with CH‐C.

Assessing the Effects of Non-steroidal Anti-inflammatory Drugs on Antihypertensive Drug Therapy Using Post-Marketing Surveillance Database

Background Antihypertensive and non-steroidal anti-inflammatory drugs (NSAIDs) are used to treat many common diseases. However, it has been suspected that interactions between these drugs exist. Here, we assessed the interactions between non-selective NSAIDs and several classes of antihypertensive drugs. Methods The study design was a cohort study using “The Antihypertensive Drug Database,” which is a collection of data accumulated from Drug Use Investigations. Subjects newly starting antihypertensive drug therapy were identified in the database. We compared the “User” group, who were co-administered NSAIDs, with the “Non-user” group, who were not. The outcome measure was the change in systolic blood pressure from the baseline after 2 months of treatment. We estimated the non-adjusted and adjusted differences in the change in systolic blood pressure between the “User” and “Non-user” groups. Results Data were collected for a total of 1,204 subjects, of whom 364 were prescribed beta blockers, 60 were prescribed diuretics, 628 were prescribed angiotensin-converting enzyme inhibitors, and 152 were prescribed calcium channel blockers. The adjusted difference in the change in systolic blood pressure between the User (n = 301) and Non-user (n = 903) groups was 2.88 mmHg (95% confidence interval: 0.89, 4.87); thus, systolic blood pressure in the Non-User group decreased further from the baseline than that in the User group. In subjects administered beta blockers, diuretics, angiotensin-converting enzyme inhibitors, and calcium channel blockers, the corresponding differences were 0.37 mmHg (-3.24, 3.98), 6.11 mmHg (-3.16, 15.37), 3.85 mmHg (1.16, 6.66), and 3.50 mmHg (-2.03, 9.02). Conclusion The effectiveness of antihypertensive drugs was attenuated by the co-administration of NSAIDs. The differences in the effects of NSAIDs varied with different classes of antihypertensive drugs.

Multicenter phase II Study of gemcitabine and S-1 combination therapy (GS Therapy) in patients with metastatic pancreatic cancer

OBJECTIVE The aim of this multicenter Phase II study was to assess the efficacy and toxicity of gemcitabine and S-1 combination therapy for metastatic pancreatic cancer. METHODS Chemotherapy-naïve patients with histologically or cytologically proven metastatic pancreatic adenocarcinoma were eligible for this study. Gemcitabine was administered at a dose of 1000 mg/m(2) over 30 min on days 1 and 8, and oral S-1 at a dose of 40 mg/m(2) twice daily from days 1 to 14, repeated every 3 weeks. RESULTS A total of 55 patients were included and the efficacy and toxicity were analyzed in 54 patients who received at least one dose of gemcitabine and S-1 combination therapy. Although no complete response was seen, a partial response was achieved in 24 patients, resulting in an overall response rate of 44.4% (95% confidence interval: 30.9-58.6%). The median progression-free survival was 5.9 months (95% confidence interval: 4.1-6.9 months) and the median overall survival was 10.1 months (95% confidence interval: 8.5-10.8 months) with a 1-year survival rate of 33.0%. The major Grade 3-4 toxicities were neutropenia (80%), leucopenia (59%), thrombocytopenia (22%), anorexia (17%) and rash (7%). Hematological toxicity was mostly transient and there was only one episode of febrile neutropenia ≥Grade 3. CONCLUSIONS Gemcitabine and S-1 combination therapy produced a high response rate with good survival in patients with metastatic pancreatic cancer. A randomized Phase III study to confirm the efficacy of gemcitabine and S-1 combination therapy is ongoing.

Postoperative radiotherapy is effective for thymic carcinoma but not for thymoma in stage II and III thymic epithelial tumors: the Japanese Association for Research on the Thymus Database Study.

The efficacy of postoperative radiotherapy (PORT) for thymic epithelial tumors is still controversial. Using the Japanese Association for Research on the Thymus (JART) database, this study was aimed at clarifying the efficacy of PORT for Masaoka stage II and III thymic carcinoma and thymoma.