Assessing drug effect from distributional data: A population approach with application to Duchenne Muscular Dystrophy treatment.

BACKGROUND AND OBJECTIVE: In Duchenne Muscular Dystrophy (DMD) treatment, muscle fiber size can be considered as an indicator of muscle health and function. In particular, the statistical distribution of fibers cross-sectional areas (CSAs) has been used as quantitative efficacy endpoint. For each patient, assessment of treatment effect relies on the comparison of pre- and post-treatment biopsies. Since biopsies provide "distributional data", i.e. empirical distributions of fibers CSA, the comparison must be carried out between the empirical pre- and post-treatment distributions. METHODS: Here, distributional fiber CSA data are analyzed by means of a hierarchical statistical model based on the population approach, considering both the single patient and the population level. RESULTS: The proposed method was used to assess the histological clinical effects of Givinostat, a compound under study for DMD treatment. At the single patient level, a two-component Gaussian mixture adequately represents pre- and post-treatment distributions of log-transformed CSAs; drug effect is described via a dose-dependent multiplicative increase of muscle fiber size. The single patient model was also validated via muscle composition data. At the patient population level, typical model parameters and inter-patient variabilities were obtained. CONCLUSIONS: The proposed methodological approach completely characterizes fiber CSA distributions and quantifies drug effect on muscle fiber size, both at the single patient and at the patient population level. This approach might be applied also in other contexts, where outcomes measured in terms of distributional data are to be assessed.

Short-term variations in bone remodeling biochemical markers: cyclical etidronate and alendronate effects compared.

Bone-remodeling markers have been proposed to monitor antiosteoporotic therapy, as substantial changes in these markers usually occur in a relatively short time interval. In this study we have evaluated the short term effects of two bisphosphonates on bone-remodeling markers with the aim of 1) defining the shortest reliable time interval after which markers should be measured, and 2) comparing the effects of different bisphophonates. To do so, 74 postmenopausal women with a lumbar spine t score of at least -1 were randomly allocated to 4 different treatments: calcium carbonate (500 mg/day; n = 18), 5 mg/day alendronate (A5; n = 18), 10 mg/day alendronate (A10; n = 20), and cyclical etidronate (CE; n = 18). Serum and 24-h urine samples were collected at baseline and 14, 28, 56, and 84 days after the beginning of therapy. Type I collagen N-terminal (NTx) and C-terminal (CTx) telopeptides and total deoxypyridinoline (tDPD) were measured in urine and normalized for urinary creatinine excretion. Osteocalcin and bone alkaline phosphatase in serum were measured. Alendronate (at both doses) and CE significantly decreased bone-remodeling markers, whereas calcium carbonate did not. Bone resorption markers reduction reached a plateau 14 (A10) or 28 (A5 and CE) days after the beginning of treatment, whereas osteocalcin and bone alkaline phosphatase were significantly reduced at 56 (A10) and 84 (CE) days. The global effects of alendronate and CE on NTx and CTx (calculated as the area under the curve) were significantly different from those of calcium (P < 0.05), but were not significantly different from each other. The percent change from baseline obtained with tDPD, NTx, or CTx during bisphosphonate treatment were significantly different (P < 0.05), but this difference disappeared when the variability in the calcium carbonate group was taken into account. In conclusion, this study shows that 1) etidronate and alendronate induce a significant and rapid reduction in bone-remodeling markers; 2) the changes in NTx, CTx, and tDPD urinary excretions reach a plateau after 2-4 wk of treatment; and 3) short term treatments with CE or alendronate induce similar changes in the urinary excretion of NTx and CTx.

Evidence that fluoride therapy increases trabecular bone density in a peripheral skeletal site.

We measured the spinal bone density (SBD) and femoral condyle bone density (FCD) in normal and osteoporotic females (n = 219) both before and during fluoride therapy. SBD and FCD in untreated osteoporotics were significantly lower (P < 0.05) than those in the age-matched controls. SBD and FCD were correlated in the untreated (r = 0.62; P < 0.0001) as well as in the fluoride-treated osteoporotics (r = 0.42; P < 0.0001). SBD and FCD were significantly increased (P < 0.05) in response to fluoride therapy. The average rates of increase in FCD and SBD were similar (1.3 +/- 1.3 vs. 1.24 +/- 1.4 mg/cc.month). We conclude that the osteogenic action of fluoride is not limited to the axial skeleton. An increase in trabecular bone density also occurs at peripheral weight-bearing sites such as the femoral condyle.

Clinical performances of galactosyl hydroxylysine, pyridinoline, and deoxypyridinoline in postmenopausal osteoporosis.

We have previously shown that galactosyl hydroxylysine (GHYL), pyridinoline (PYD), and deoxypyridinoline (DPD) have a better accuracy and discriminate power than hydroxyproline in distinguishing postmenopausal osteoporotic women from premenopausal controls. In this study, we evaluated the clinical performances of GHYL, PYD, and DPD, alone or in combination, in distinguishing postmenopausal osteoporotic women (OPBD, n = 26) from age-matched controls (CBD, n = 19). The diagnosis of osteoporosis was based upon the bone density (BD) of the lumbar spine measured by quantitative computed tomography (CBD: BD > 108 mg/cm3; OPBD: BD < 70 mg/cm3). Urinary excretion of GHYL, PYD, and DPD were measured by HPLC, and all data were expressed as the molar ratio with the creatinine excretion (GHYL/CR, PYD/CR, and DPD/CR). The clinical performances were tested by: Z score analysis (Z), Receiver Operated Characteristic curve analysis (%Acc) and logistic-regression analysis of the posterior probabilities for prediction from a logistic model (LOGIST). GHYL/CR, PYD/CR, and DPD/CR were significantly increased in OPBD compared with CBD. The clinical performances were similar for the three assays, with slightly better performances for GHYL/CR (GHYL/CR: Z = 3.14, %Acc = 70 +/- 8, LOGIST P = 0.01; PYD/CR: Z = 2.19, %Acc = 67 +/- 8, LOGIST P = 0.051; DPD/CR: Z = 2.13, %Acc = 65 +/- 8, LOGIST P = 0.06). None of the possible combinations of the three assays yielded better clinical performances than GHYL/CR alone. In conclusion, this study further confirms the validity of GHYL, PYD, and DPD as markers of bone resorption.

Age-related decreases in insulin-like growth factor-I and transforming growth factor-beta in femoral cortical bone from both men and women: implications for bone loss with aging.

We determined the skeletal content of insulin-like growth factor-I (IGF-I) and transforming growth factor-beta (TGF beta) in human bone as a function of age, using 66 samples of femoral cortical bone obtained from 46 men and 20 women between the ages of 20-64 yr. We found a linear decline in the skeletal content of IGF-I (nanograms per mg protein) with donor age (r = -0.43; P < 0.001) in the total population. The skeletal content of TGF beta also decreased with age (i.e. 1/TGF beta vs. age; r = 0.28; P < 0.02) for the total population. We did not observe any difference in the skeletal growth factor content between male and female donors. IGF-I content, when analyzed by decade divisions of age, showed a reduction between the 20- to 29-yr-old and the 50- to 59-yr-old subjects (P < 0.02). The loss rate of IGF-I was 1.56 ng/mg protein.yr, corresponding to a net loss of 60% of skeletal IGF-I between the ages of 20-60 yr. The loss rate of TGF beta was 0.03 ng/mg protein.yr, corresponding to a net loss of 25% of the skeletal TGF beta between the ages of 20-60 yr.

Truncation of the amino terminus of PTH alters its anabolic activity on bone in vivo.

In vitro studies of parathyroid hormone (PTH) structure and function have suggested that the anabolic effect of PTH on bone requires the presence of amino acid residues 28-34 (domains for protein kinase C activation and mitogenic activity), but not amino acid residues 1-7 (adenylate cyclase activation domain). We have tested this hypothesis with in vivo studies of human PTH (hPTH) analogs. Serum biomarkers and selected histomorphometric parameters of bone formation and resorption were assessed in adult, female, Sprague-Dawley rats following 19 daily injections of vehicle, 10 micrograms/kg body weight (bw) of hPTH(1-38), or a dose range of 10, 40, and 100 micrograms/100 g bw of hPTH(2-38) or hPTH(3-38). Treatment with hPTH(1-38) increased serum osteocalcin, the percentage of osteoblast surface, percentage of osteoid surface, percentage of bone volume, trabecular thickness, and bone formation rate, while it decreased the percentage of osteoclast surface. The hPTH(2-38) fragment exhibited 10%-25% of the in vivo anabolic activity of hPTH(1-38), while it had no effect on the percentage of osteoclast surface. The hPTH(3-38) fragment exhibited no biological activity on bone. In contrast, serum INS-PTH (intact-N-terminal specific PTH) levels were similarly and significantly increased above control in rats treated with hPTH(1-38), hPTH(2-38), or hPTH(3-38) at the same dose. This preliminary finding suggests that the differential activity of these peptides on bone is not due to differences in the circulating level of immunoreactive PTH (intact and amino-terminal fragments of PTH from endogenous and exogenous sources) several hours after PTH injection. However, we can draw no conclusion regarding the relative clearance rates of these peptides. Last, because hPTH(3-38) was without any detectable biological activity on rat bone in vivo, its mitogenic activity was confirmed in two osteoblast-like cell lines. In summary, the anabolic effect of hPTH(1-38) on bone in vivo was (1) diminished by removal of amino acid residue 1, and (2) abolished by the removal of amino acid residues 1 and 2. Although these findings suggest that the therapeutic benefits of exogenous PTH administration may depend upon activation of not only protein kinase C, but also adenylate cyclase, they do not rule out a differential PTH response due to other causes, e.g., metabolic inactivation.

Antinociceptive activity of eel calcitonin, injected into the inflamed paw in rats.

This study examined the possible peripheral activity of eel calcitonin in the modulation of the response to noxious pressure on inflamed paws in rats (Randall and Selitto test). The intraplantar injection of eel calcitonin (20-200 ng/rat) but not the subcutaneous administration (200 ng and 2 micrograms/rat, s.c.), was able to significantly inhibit hyperalgesia induced by intraplantar injection of carrageenin. The development of oedema on the other hand was not inhibited. The intraplantar administration of eel calcitonin (200 ng/rat) in a non-inflamed paw did not modify paw pressure thresholds. Eel calcitonin (200 ng/rat, intraplantar, i.pl.) was also able to elicit an antinociceptive effect on formalin-induced hyperalgesia, both when the peptide was injected before or after (60 min) formalin. This effect, at difference with morphine (80 micrograms/rat, i.pl.), was not blocked by naloxone (10 micrograms/rat, i.pl.). These results demonstrate the local antinociceptive effect of eel calcitonin in inflammatory pain and might indicate a new way of using calcitonin in the control of pain.

Effect of fluvastatin on lipids and fibrinolysis in coronary artery disease.

This randomized, double-blind, placebo-controlled study shows that 20-week fluvastatin treatment induces beneficial changes in the lipid panel and a shift in the fibrinolytic pathway toward activation through a decrease in tissue plasminogen activator antigen. Fluvastatin treatment causes no variation in lipoprotein(a) circulating levels.

Treatment with pertussis toxin does not prevent central effects of eel calcitonin.

To determine whether or not the CNS inhibitory activity of eel calcitonin (eCT) on adenylyl cyclase is the endocellular mechanism underlying the antinociceptive effect of the peptide, as shown for morphine analgesia, we administered Bordetella pertussis toxin (PTX) by intracerebroventricular (ICV) injection (0.5 microgram/rat) to block the receptor-mediated inhibition of adenylyl cyclase. In PTX-treated rats there was no change in eCT (2.5 micrograms/rat, ICV)-induced antinociceptive activity (hot-plate test) nor in eCT (100 ng/rat, ICV) inhibition of gastric acid secretion (Shay test) whereas morphine (5 micrograms/rat, ICV) analgesia was significantly reduced. In vitro studies showed no reduction of eCT binding in the CNS of rats treated with PTX in vivo. Moreover, PTX treatment did not change the inhibitory effect of eCT on adenylyl cyclase in isolated membranes from rat striatum in contrast with opiates (DAME and morphine) whose effects were lost. As PTX is known to inactivate the guanidine binding inhibitory protein Gi, these data suggest that a G protein, distinct from the Gi protein involved in the coupling of opiate receptors into a functional response, could be responsible for regulating the intracellular pathways resulting in eCT-induced antinociceptive effect and inhibition of gastric acid secretion.

Evidence for different classes of calcitonin binding sites in rat CNS: an autoradiographic study with carbo-calcitonin.

The distribution of binding sites in the rat CNS for a synthetic analogue of eel calcitonin, [Asu1-7]eel calcitonin (carboCT) was investigated. This distribution was compared to that for the natural peptide to see whether a modified molecule would also reveal different classes of binding sites for CT. The regional distribution of 125I-carboCT binding in coronal sections of rat CNS was examined by an in vitro autoradiographic technique. Non-specific binding was assessed after addition of excess cold carboCT or eel CT and the results showed that carboCT binding is specific and that it is displaced equally by cold carboCT and by eel CT. There was dense labelling in the nucleus accumbens, in the tractus striohypothalamicus, in the anterior and posterior part of the hypothalamus except for the nucleus ventromedialis, in the amygdala, in the pars medialis of the reticular formation, in the nucleus ruber, in the periventricular gray and in the raphe magnus. Grains were less dense in the hypothalamus lateralis, in the substantia nigra and in the nucleus interpeduncularis. In contrast to eel CT, carboCT did not bind in the spinal cord, nor did carboCT prevent eel CT binding in this area, whereas it was able to prevent it in the brain. These results are consistent with the existence of different classes of binding sites for CT in rat brain and in spinal cord, and indicate that the substitution of the S-S bond with a C-C bond in the eel CT molecule makes the peptide more selective for one class of binding sites.

Effects of age on binding sites for calcitonin gene-related peptide in the rat central nervous system.

The binding site distribution of calcitonin gene-related peptide (CGRP) was studied in the central nervous system of aged rats (22 months old) and compared with that of young rats (2 months). The regional distribution of [125I]Tyr-rat CGRP binding in coronal sections of young and old rat CNS was examined by an in vitro autoradiographic technique. The results, showed that in aged rats there was a marked reduction in CGRP binding, without any change in binding affinity, in the hippocampus, the nucleus rhomboideus, the nucleus arcuatus, the colliculus superior, the substantia grisea centralis and the spinal cord. In the cortical areas, the amygdala, the caudatus putamen and the accumbens binding was not modified. In the cortex cerebellaris CGRP binding was strikingly greater in the aged rats. The increase in binding might be a consequence of an adaptive process due to a decline of the peptide synthesis with age and is suggestive of a role for CGRP in the cerebellum functions.

Effect of unmodified eel calcitonin on gastric acid secretion and gastric ulcers in the rat.

A new peptide, eel calcitonin (eCT), synthesized to the sequence of calcitonin (CT) extracted from the ultimo branchial bodies of eels, is now on the market in some countries for clinical use in the treatment of Paget's disease of bone and in the prophylaxis or treatment of certain osteoporoses. We have now studied the effect of eCT on the inhibition of gastric acid secretion and protection against experimentally induced gastric ulcers in rats as other CTs have previously been shown to have such effects. EelCT shows a dose dependent inhibition of gastric acid secretion, both volume and concentration of acid - dose range 100-900 ng/Kg injected subcutaneously. Central administration is also effective at a dose of 100 ng/rat. There is no published evidence for a direct effect of CTs in the stomach and there is considerable speculation on possible mediating pathways. Somatostatin may be involved in the inhibitory effect of eCT on gastric acid secretion because when administered both centrally or peripherally eCT is ineffective in rats depleted of somatostatin by pretreatment with cysteamine. However, other mechanisms must also be involved as eCT has no preventive effect against gastric ulcers induced by ethanol but has a high index of protection against ulcers induced by cold restraint stress or indomethacin.

Association between amygdala reactivity and a dopamine transporter gene polymorphism.

Essential for detection of relevant external stimuli and for fear processing, the amygdala is under modulatory influence of dopamine (DA). The DA transporter (DAT) is of fundamental importance for the regulation of DA transmission by mediating reuptake inactivation of extracellular DA. This study examined if a common functional variable number tandem repeat polymorphism in the 3' untranslated region of the DAT gene (SLC6A3) influences amygdala function during the processing of aversive emotional stimuli. Amygdala reactivity was examined by comparing regional cerebral blood flow, measured with positron emission tomography and [(15)O]water, during exposure to angry and neutral faces, respectively, in a Swedish sample comprising 32 patients with social anxiety disorder and 17 healthy volunteers. In a separate US sample, comprising 85 healthy volunteers studied with blood oxygen level-dependent functional magnetic resonance imaging, amygdala reactivity was assessed by comparing the activity during exposure to threatening faces and neutral geometric shapes, respectively. In both the Swedish and the US sample, 9-repeat carriers displayed higher amygdala reactivity than 10-repeat homozygotes. The results suggest that this polymorphism contributes to individual variability in amygdala reactivity.

From trial and error to trial simulation III: a framework for interim analysis in efficacy trials with antidepressant drugs.

Clinical trials with antidepressant drugs often fail to detect drug effect, even with drugs that are known to be efficacious. In a previous publication, we showed that a model-based approach is required to address some of the existing challenges in the design of clinical trial protocols. Here, we illustrate how the implementation of an interim analysis (IA) may help to identify studies that are headed for failure, early in the trial before completion of treatment. In contrast to traditional IA procedures, an adaptive Bayesian approach is proposed to optimize the timing of analysis and decision criteria for futility and efficacy, taking into account enrollment rate and treatment response at intermediate visits in the trial. Validation procedures involving re-enrollment of patients confirmed the performance of the method. Our findings reveal that optimization of the timing and decision criteria at the interim stage is critical for the accuracy of the conclusions about treatment efficacy or futility.

Biochemical markers of bone metabolism in the assessment of osteoporosis.

Several biochemical markers are available for the study of bone metabolism. The characteristics of these markers are reviewed and their potential application in osteoporosis is examined. It is concluded that biochemical markers are potentially useful in identifying those at risk of developing osteoporosis, selecting appropriate treatment, monitoring therapy, and studying the pathogenesis of osteoporosis.

Galactosyl hydroxylysine and deoxypyridinoline: a methodological comparison.

Galactosyl hydroxylysine and deoxypyridinoline are at present the most promising markers of bone resorption. Various studies have indeed shown that these two markers discriminate with high accuracy subjects with different rates of bone turnover and that their accuracies and discriminate power are very similar. The aim of this paper is to compare the practicality and the reproducibility of the HPLC galactosyl hydroxylysine and deoxypyridinoline assays. In summary, this review shows that the galactosyl hydroxylysine and deoxypyridinoline HPLC assays differ mainly in the need, in using deoxypyridinoline, for an acid hydrolysis and a preextraction of the urine samples. This implies two major problems for deoxypyridinoline: 1) more time is required due to the cumbersome preanalytical procedures; and 2) a lower reproducibility. Our data, in fact, show that both the intra-assay and inter-assay coefficient of variation of the deoxypyridinoline assay are almost 100% higher than those of the galactosyl hydroxylysine assay.

Detection of pyridinium cross-links in human bile.

This study evaluated whether pyridinium cross-links, which are positively charged, besides renal clearance are also cleared by the liver into bile. In 13 human bile samples tested, we were able to detect both pyridinoline (PYD) and deoxypyridinoline (DPD) in small amounts which were estimated to be about 1-2% of the amount usually found in urine. To further evaluate the amount of pyridinium cross-links excreted through bile, we studied the stability of these compounds at the alkaline pH of bile. No effect on their stability was detected over a 6-hour incubation. The origin of these molecules in bile and the significance of this finding in the use of PYD and DPD as bone resorption markers are discussed.

High prevalence of hypovitaminosis D among free-living postmenopausal women referred to an osteoporosis outpatient clinic in northern Italy for initial screening.

To establish the prevalence of hypovitaminosis D among free-living postmenopausal women referred to an osteoporosis outpatient clinic in Northern Italy, we evaluated 25-hydroxyvitamin D (25(OH)D) levels in 570 postmenopausal women who had been consecutively referred to our clinic in the 12 months beginning October 1995. Parathyroid hormone (PTH), serum calcium (Ca), creatinine (Cr) and osteocalcin (OC), urinary calcium (Ca24h) and creatinine (Cr24h), and the bone mineral density of the lumbar spine (LBMD) and femur (FBMD) were also measured. 1,25-Dihydroxyvitamin D (1,25(OH)2D) concentrations were measured in 23 women. All women had normal electrolyte serum concentrations and kidney function. Mean +/- SD 25(OH)D concentration was 18.3 +/- 8.3 ng/ml. A significant (p < 0.001) seasonal variation was seen for both 25(OH)D and PTH. Women were divided into two groups based on their vitamin D status: low vitamin D status (25(OH)D < 12 ng/ml, n = 161, 28%) and normal vitamin D status (25(OH)D > or = 12 ng/ml, n = 409, 72%). Hypovitaminosis D was found in 38.5% of all the women in the time period December-May and in 12.5% in the other half-year; among women > 70 years old 51% had hypovitaminosis D in the time period December-May and 17% in the other half-year. PTH was significantly (p < 0.05) increased, and Ca24h, OC and FBMD significantly (p < 0.05) decreased in women with hypovitaminosis D. 1,25(OH)2D positively correlated with 25(OH)D (p < 0.0001), but did not correlate with PTH, age or creatinine clearance. In conclusion, hypovitaminosis D is an important, underestimated problem in Italian free-living postmenopausal women referred to an outpatient osteoporosis clinic.

Collagen type I of rat cortical and trabecular bone differs in the extent of posttranslational modifications.

This study sought to evaluate whether the architecture of the matrix of cortical and trabecular bone is exactly the same. For this purpose we analyzed the extent of some posttranslational modifications of type I collagen, which is the major component of bone matrix. Ten female and 10 male 100-day-old rats were sacrificed and the content of hydroxylysine, glycosylated hydroxylysine, and pyridinium cross-links of collagen from cortical and trabecular bone was determined. The amount of each compound was expressed as a molar ratio with hydroxyproline. The collagen posttranslational modification pattern appears to be the same in both sexes but with a higher extent of differences in females compared with males. Comparing cortical and trabecular bone, the former contains a higher amount of hydroxylysine residues whereas in the latter, glycosylation of hydroxylysine is higher and pyridinium cross-link concentration is lower. Moreover, an inverse linear relationship between glycosylated hydroxylysine and pyridinium cross-links concentration was established, both for female (r = -0.455, P = 0.04) and male rats (r = -0.426; P = 0.06). This paper discusses what these findings may mean in functional terms.

17 beta-Estradiol and tamoxifen prevent the over-glycosylation of rat trabecular bone collagen induced by ovariectomy.

We have recently demonstrated that ovariectomy in the rat causes over-glycosylation of collagen which is restricted to trabecular bone. In order to obtain further evidence, we studied whether estrogen or tamoxifen treatment prevented over-glycosylation of trabecular bone collagen. Forty one-hundred-day-old female rats were subjected to ovariectomy (n = 30) or sham-operation (n = 10). Starting the day of the operation, sham-operated rats were treated with vehicle, while ovariectomized rats were divided into three groups and treated with vehicle (n = 10), estrogen (n = 10) or tamoxifen (n = 10). Five rats from each group were sacrificed at 115 and 145 days of age. Femurs and tibiae were separated into cortical and trabecular bone, demineralized, hydrolyzed and analyzed by HPLC for hydroxylysine glycoside and hydroxyproline content. Hydroxylysine glycoside content was expressed as a molar ratio with hydroxyproline. The results can be summarized as follows: 1) cortical bone collagen glycosylation did not vary among the different groups; 2) over-glycosylation of trabecular bone collagen observed in the ovariectomized rats was prevented by the administration of either 17 beta-estradiol or tamoxifen. These data demonstrated that estrogens affect glycosylation of trabecular bone collagen.