Quality assessment of observational studies in psychiatry: an example from perinatal psychiatric research

In perinatal psychiatry, randomized controlled trials are often not feasible on ethical grounds. Many studies are observational in nature, while others employ large databases not designed primarily for research purposes. Quality assessment of the resulting research is complicated by a lack of standardized tools specifically for this purpose. The aim of this paper is to describe the Systematic Assessment of Quality in Observational Research (SAQOR), a quality assessment tool our team devised for a series of systematic reviews and meta-analyses of evidence-based literature regarding risks and benefits of antidepressant medication during pregnancy.

Exemestane for primary prevention of breast cancer in postmenopausal women: NCIC CTG MAP.3—A randomized, placebo-controlled clinical trial

LBA504

Background: Limited efficacy and serious toxicities have limited uptake of tamoxifen or raloxifene as preventatives of breast cancer. Aromatase inhibitors (AIs) prevent contralateral breast cancers more than tamoxifen in adjuvant trials and have fewer serious side effects. This is the first report of an AI used in primary prevention. Methods: NCIC CTG MAP.3 is a randomized trial designed to detect a 65% reduction in annual incidence of invasive breast cancer (IBC) on exemestane (E) versus placebo (P). Eligible postmenopausal women had ≥ one of the following risk factors: Gail score >1.66%, prior ADH, ALH, LCIS or DCIS with mastectomy, age over 60. Health-related and menopause-specific quality of life (QOL) were assessed by SF-36 and MENQOL questionnaires. Results: From 2004-2010, 4,560 women were randomized: age 62.5 yrs (37-90); Gail Score 2.3 % (0.6-21); BMI 28.0 kg/m2 (15.9-65.4). Risk factors included: age >60 yrs (49%); Gail score >1.66 (40%); and prior ADH, ALH, LCIS or DCIS with mastectomy (11%). At median follow-up of 35 months there were 11 IBCs on E and 32 on P (annual incidence 0.19% vs 0.55%; HR= 0.35, 95% CI 0.18-0.70, p = 0.002); ductal (10E/27P), lobular (1E/5P). Most tumors were ER positive (7E/27P); Her2/neu negative (10E/26P); TNM stage T1 (8E/28P), N0 (7E/22P), M0 (11E/30P). E was superior in all subgroups: by Gail score, age, BMI, prior LCIS and DCIS. The annual incidence rate of IBC or DCIS was 0.35% E and 0.77% P (HR=0.47;95% CI 0.27-0.79; p = 0.004) based on 64 IBCs or DCISs (20E/44P). Clinical bone fractures, osteoporosis, hypercholesterolemia or cardiovascular events were equal in both arms. No clinically meaningful differences in QOL were detected. Conclusions: Exemestane significantly reduced invasive and pre-invasive breast cancers in postmenopausal women at increased risk for breast cancer with no serious toxicities. Exemestane should be considered a new option for primary prevention of breast cancer. Supported by the Canadian Cancer Society; Pfizer Inc. PEG supported in part by Avon Foundation.

Recurrent triploid and dispermic conceptions in patients with NLRP7 mutations

To understand the mechanisms leading to hydatidiform mole formation in patients with NLRP7 mutations, we used a combination of various approaches to characterize five products of conception, from two patients, shown by flow cytometry to contain non-diploid cells. We demonstrate that four of these conceptions are triploid and two of them originated from fertilization with more than one sperm. We show that three of these triploid conceptions fulfill the histopathological criteria of partial hydatidiform mole and one fulfills the histopathological criteria of spontaneous abortion. Our data demonstrate that some oocytes from one patient with NLRP7 mutations are not able to prevent polyspermic fertilization and highlight the importance of using several approaches to characterize the genetic complexity of molar tissues and reproductive wastage. Altogether, our previous and current data show the association of NLRP7 mutations with several types of hydatidiform moles and with triploid spontaneous abortions.