Application of Machine Learning Methods for the Development of Antidiabetic Drugs.

Diabetes is a chronic non-communicable disease caused by several different routes, which has attracted increasing attention. In order to speed up the development of new selective drugs, machine learning (ML) technology has been applied in the process of diabetes drug development and opens up a new blueprint for drug design. This review provides a comprehensive portrayal of the application of ML in antidiabetic drug use.

Prospective study of histomorphometry, biochemical bone markers and bone densitometric response to pamidronate in ß-thalassaemia presenting with osteopenia-osteoporosis syndrome.

This study aimed to evaluate bone remodelling disorders in thalassaemia by using pamidronate (PD) infusion with or without hormone replacement therapy (HRT) as a diagnostic-therapeutic tool. In this prospective study, 24 adult thalassaemia major (TM) and 10 thalassaemia intermedia (TI) patients received either PD and HRT or HRT only (controls) for 3 years. Eugonadal patients with TI had PD only. Bone remodelling was assessed by dual energy X ray absorptiometry (DXA scan), type 1-collagen biochemical bone markers (BBM) and histomorphometry of iliac crest biopsy before and after PD. As a group, thalassaemics had a significant improvement in spinal and femoral bone mineral density Z scores following PD (P < 0·01) compared to the controls. Although BBM were comparable pre-therapy, they were significantly lower in the PD cohort (P < 0·001) compared to the control group. All patients had osteopenia, diminished osteoid formation and bone volume on histomorphometry pre-therapy with high turnover bone disease (HTO) in TM and low-turnover disease (LTO) in TI. In TM, bone volume improved significantly, whereas TI patients showed little or no response to PD. In conclusion, histomorphometry data suggest that TM patients have a distinct pathology of high turnover bone disease compared to TI patients, who have low-turnover disease.

Oral, rectal, and vaginal pharmacokinetics of misoprostol.

OBJECTIVE: To compare the pharmacokinetic profiles of orally, rectally, and vaginally administered misoprostol tablets in pregnant women. METHODS: Women between 7 and 14 completed weeks of gestation were recruited and randomly assigned to be given 400 microg misoprostol orally, rectally, or vaginally 3 hours before surgical termination of pregnancy. Blood samples were obtained at 0, 7.5, 15, 30, 45, 60, 90, 120, and 240 minutes and later analyzed for plasma concentrations of misoprostol free acid (the principle metabolite). RESULTS: Vaginal misoprostol was present in the circulation longer than oral misoprostol and had a greater area under curve at 240 minutes (P <.001). Rectal misoprostol had a similar pattern but a much lower area under curve at 240 minutes. Oral misoprostol had a significantly greater peak plasma concentration and a shorter duration to maximum concentration than either rectal or vaginal misoprostol (both P <.001). CONCLUSION: Oral misoprostol tablet is also absorbed by the rectal and vaginal routes. Misoprostol administered in early pregnancy has route-dependent pharmacokinetics and is absorbed best when administered vaginally. LEVEL OF EVIDENCE: I