Narrow Therapeutic Index Drugs Research Articles

Real-life Bioequivalence of Tacrolimus in Patients With Living-Related Donor De Novo Renal Transplantation: An Observational Study

Tacrolimus is a narrow therapeutic index drug. As a result, regulatory agencies worldwide recommend stringent bioequivalence evaluation criteria for approval of generics. Despite this, the professional transplantation societies have raised concerns over the safety and efficacy of generic substitutions. We conducted this pragmatic real-life bioequivalence study to assess the effect of generic substitutions of tacrolimus. This was an observational study including recipients of renal transplantation who were considered for generic medication substitution. Transplanted organs were from living-related donors and were performed at least 1 month before the study. Time of administration of the drug, time of dosing with respect to meals, and time of blood sample collection were controlled; however, the lot number of the generic drugs was not controlled. The participants were allowed to use their usual supplies irrespective of the lot number. Concentration (C0) was quantified by liquid chromatography with tandem mass spectrometry after the generic substitution from ABC brand to XYZ brand. The average C0 ± SD with generic ABC was 11.09 ± 4.26 ng/mL and generic ABC was 9.7 ± 4.12 ng/mL. Though there was no statistically significant difference observed between the concentrations, when the individual patient data was examined, 2 patients were found to have a very high concentration of tacrolimus and at least 7 patients fell below the therapeutic range. These derangements called for retitration with the new generic tacrolimus (40%). The results of our study suggest that generic-to-generic substitutions should be carried out very carefully in a closely observed setting in patients with renal transplants. The strength of our study is that it matched the real clinical practice setting as much as possible unlike a bioequivalence study. Therefore, we recommend repeating C0 at least 3 times over a period of 7 to 10 days with a generic substitution to prevent untoward consequences.

Tacrolimus Trough Variability- Predictors and Outcomes in Pediatric Heart Transplantation

Tacrolimus is a narrow therapeutic index drug, requiring consistent levels to maximize solid organ transplant success while reducing adverse effects. High tacrolimus trough level variability (TTLV) is a significant predictor of late rejection in both kidney and liver transplant. We sought to describe the prevalence of high TTLV in our pediatric heart transplant (HT) population, and identify whether this variability is associated with patient specific factors and poor outcomes. Single center, retrospective study of HT patients from 1/2002 - 12/2018. Multi-organ recipients and retransplants were excluded. Records were reviewed for demographics, tacrolimus trough levels (TTLs) and outcomes (rejection, coronary vasculopathy, need for re-transplantation, death). Variability was calculated as the TTL standard deviation (SD) using each patient's TTLs from 1 year post-HT to last follow-up. High TTLV was defined as SD ≥3. Statistical analysis included Kruskal-Wallis, Chi-squared, and Wilcoxon rank sum testing. A total of 119 patients were included (male, 57%) with pre-HT diagnoses of congenital heart disease (57%) and cardiomyopathy (42%). The median age at HT was 2.7 yrs (IQR 0.8, 12.2) with a follow-up time post HT of 5.7 yrs (IQR 3.3, 9.9). A total of 6,344 TTLs were included, with a median of 33 TTLs/pt (IQR 14, 76). The median SD for the entire cohort was 3.1 (IQR 2.3, 4.0) with 55% of the population having a SD ≥3. Levels obtained from 1-5 yrs of age were found to have significantly greater variability (median SD 3.1 [IQR 2.4, 4.1]; p-value< 0.01) than 6-12 yrs (SD 2.5) and 12+ yrs (SD 2.5). After dichotomizing patients for higher variability (SD ≥ 3), there were no significant associations with patient specific factors including: sex, pre-HT diagnosis, pre-HT clinical course (acute vs chronic), age at HT, time with HT and Hollingshead Socioeconomic Score. High TTLV was associated with increased rejection treatment (74% vs 44%, p<0.01), higher frequency of coronary vasculopathy (23% vs 7%, p=0.02), and death (20% vs 6%, p=0.02). TTLV is high in pediatric HT recipients, especially in those < 5 yrs. Patients with a SD ≥3 experienced a higher frequency of poor outcomes including rejection, coronary vasculopathy and death.

Impact of switching to a heat-not-burn tobacco product on CYP1A2 activity

BackgroundCigarette smoking induces cytochrome P450 1A2 (CYP1A2) expression and activity, while smoking cessation normalizes the levels of this enzyme. The aim of this publication is to summarize the data on CYP1A2 gene expression and activity in preclinical and clinical studies on the Tobacco Heating System (THS), currently marketed as IQOS® with HEETs®, and to summarize the potential effects on CYP1A2 to be expected upon switching to reduced-risk products (RRPs). MethodsWe summarized PMI’s preclinical and clinical data on the effects of switching from cigarette smoking to THS. ResultsData from four preclinical mouse and rat studies showed that, upon either cessation of cigarette smoke exposure or switching to THS exposure, the upregulation of CYP1A2 observed with exposure to cigarette smoke reverted close to fresh-air levels. Data from four clinical studies yielded similar results on CYP1A2 activity within a time frame of five days. Furthermore, the effects of switching to THS were similar to those seen after smoking cessation. ConclusionsBecause smoking cessation and switching to either electronic cigarettes or THS seem to have similar effects on CYP1A2 activity, the same measures taken for patients treated with narrow therapeutic index drugs that are metabolized by CYP1A2 and who quit smoking should be recommended for those switching to RRPs.

A Systematic Review of Population Pharmacokinetics of Carbamazepine

Carbamazepine is a narrow therapeutic index drug requiring therapeutic drug monitoring, while population pharmacokinetics is an approach that can aid individualized dosing regimens and to date, several population pharmacokinetic studies of carbamazepine have been conducted. This systematic review aims to summarize the factors influencing carbamazepine pharmacokinetics and model methodologies, and to identify any knowledge gaps, which may then be used to inform future studies. PubMed and Scopus databases were systematically searched from the date of their inceptions to August 2019. All population pharmacokinetic studies of carbamazepine performed in humans using a nonlinear mixed-effect modeling approach were extracted from these databases and included in this review. Twenty-three articles were included. A one-compartment pharmacokinetic model was employed in most of these studies. While body size, carbamazepine dose, co-medications, age, gender, race, and CYP1A2 polymorphism were identified as significant predictors of carbamazepine clearance, weight was the only significant predictor of the volume of carbamazepine distribution. Exponential and additive relationships were the most frequently used models when analyzing respectively inter-individual and residual variability, and the magnitude of inter-individual variability on carbamazepine clearance ranged from 1.5% to 44.5%. Seventeen of the studies contained a model evaluation, and of these, an external evaluation was conducted in ten. This review highlights the significant predictors of carbamazepine pharmacokinetics that have been identified. However, since information regarding the relationship between carbamazepine pharmacokinetic variability and its pharmacodynamics is lacking, future research relevant to this issue may be required.

Is Bioequivalence a Sufficient Measure of Equivalence?

This article reviews the U.S. Food and Drug Administration (FDA) regulation of generic medications—specifically, the use of bioequivalence to compare generic and brand prescriptions. New or “brand” drugs are subjected to extensive review by the FDA before they can be marketed to the public. Generics, which are posited to be identical to brands, are subject to a less extensive review process and must prove only that the generic is the “bioequivalent” (BE) of the brand drug. Generic medications are important because they comprise almost 80% of prescriptions filled in the United States and cost 80% to 85% less than brand drugs, playing a crucial role in patients’ access to cost-effective treatments. However, there is dissension about whether they can be interchanged with brand drugs without any consequences for the patient, especially for Narrow Therapeutic Index (NTI) drugs, which have precise dosage requirements. The regulatory designation of bioequivalence also has implications for doctor–patient relationships, patient outcomes, and patient legal rights. This article aims to establish that there is insufficient evidence to conclude whether using bioequivalence is adequate to determine whether two drugs can be considered equivalent given the medical and legal implications that flow from deeming two drugs equivalent.

Healthy User Bias in Comparative Safety Studies for Brand-Name vs. Generic Products: The Example of Warfarin.

Warfarin was selected as a case study to examine confounding when comparing a product across different manufacturers because it is a narrow therapeutic index drug with prevalent beliefs for brand-name superiority. Medicare beneficiaries aged ≥65years with atrial fibrillation and an incident outpatient warfarin prescription from July 2006 through July 2015 were included in the study population (N=746,098). Substantial imbalances were observed between brand-name warfarin and generics for (i) clinical comorbidity, (ii) socioeconomic status, (iii) prescriber specialty, (iv) recent ambulatory and emergent care, (v) drug adherence, (vi) pharmacy setting (e.g., retail, mail-order), and (vii) risk scores for bleeding and thrombosis. Patients receiving brand-name warfarin were healthier than patients receiving generic manufactured warfarin. Utilization of generic warfarin products also differed by geographic region and pharmacy setting. Manufacturer-level comparative-safety studies for causal inference should carefully consider the presence of these imbalances and their potential for introducing healthy user bias.

Empirical evidence of risk penalties for NTI Drugs

Drug innovations can outpace regulatory drug approval processes designed to control risk, creating heterogeneous risks among approved drugs. This paper estimates a price model for risky narrow therapeutic index (NTI) drugs. The traditional generic approval process has been criticized as insufficient to guarantee therapeutic equivalence in NTI drugs, leading to higher risks of toxicity or ineffectiveness when a patient switches from a brand-name version of an NTI to a generic version, or between generic versions. Using data from the Medical Expenditure Panel Survey, this paper finds evidence of a significant price penalty for NTI drugs. The paper also finds a smaller gap between brand-name and generic prices for NTI drugs than for non-NTI drugs, consistent with costly switching. An analysis of drug consumption bundles also supports this theory. These results show that despite the many information asymmetries and agency issues in the pharmaceutical market, there is evidence of sensitivity to risk in price and consumption behavior.

Decision Science for Generic Drug Development and Review.

Decision Science for Generic Drug Development and Review.

Additive manufacturing of personalized orodispersible warfarin films

Warfarin is a narrow therapeutic index drug that requires personalized dosing which is currently not achieved by the marketed products. Further, pediatric and geriatric patients may face swallowing problems with solid oral dosage forms. To face these issues, the aim of the present study was to investigate semisolid extrusion 3D printing for production of warfarin containing orodispersible films. Extrusion 3D printing was successfully utilized to produce transparent, smooth and thin, yet flexible and strong orodispersible films containing therapeutic doses of warfarin (3.9–7.4 mg). Excellent linearity (R2 = 0.9996) between the designed sizes of the films and the drug contents was achieved indicating semisolid extrusion 3D printing as a promising way to produce orodispersible warfarin films with personalized doses. This one-step-process utilizes disposable syringes hindering the printing material to be in contact with the printing equipment making it a potential method for on-demand in-hospital compounding.

Bioequivalence Guidelines Requirements for Orally Administered Generics (IR Products) in Gulf Cooperation Council Countries, European Union and United States of America

The importance of the bioequivalence requirements is increasing in the GCC (Gulf Cooperation Council) region pertaining towards the growth of generic drug market. GCC bioequivalence guidelines describes the requirements for bioequivalence study as per this region regulatory requirements. This review article mainly focuses on bioequivalence study requirements in the Gulf Cooperation Council countries, European union countries and the United States of America that needs to be fulfilled in order to successfully submit a generic application. This comparison includes the bioequivalence approaches such as study design, sample size, study condition, pharmacokinetic parameters, statistical analysis, narrow therapeutic index drugs, highly variable drug products and BCS based bio waiver requirements. The rationale for this article is to give prompt overview on the bioequivalence requirements in these region and in comparison with well-established regulatory such as USFDA and EMA. In addition, it also highlights the possibilities for harmonization.

Strategy for the Prediction of Steady-State Exposure of Digoxin to Determine Drug–Drug Interaction Potential of Digoxin With Other Drugs in Digitalization Therapy

Digoxin, a narrow therapeutic index drug, is widely used in congestive heart failure. However, the digitalization therapy involves dose titration and can exhibit drug-drug interaction. Ctrough versus area under the plasma concentration versus time curve in a dosing interval of 24 hours (AUC0-24h) and Cmax versus AUC0-24h for digoxin were established by linear regression. The predictions of digoxin AUC0-24h values were performed using published Ctrough or Cmax with appropriate regression lines. The fold difference, defined as the quotient of the observed/predicted AUC0-24h values, was evaluated. The mean square error and root mean square error, correlation coefficient (r), and goodness of the fold prediction were used to evaluate the models. Both Ctrough versus AUC0-24h (r = 0.9215) and Cmax versus AUC0-24h models for digoxin (r = 0.7781) showed strong correlations. Approximately 93.8% of the predicted digoxin AUC0-24h values were within 0.76-fold to 1.25-fold difference for Ctrough model. In sharp contrast, the Cmax model showed larger variability with only 51.6% of AUC0-24h predictions within 0.76-1.25-fold difference. The r value for observed versus predicted AUC0-24h for Ctrough (r = 0.9551; n = 177; P < 0.001) was superior to the Cmax (r = 0.6134; n = 275; P < 0.001) model. The mean square error and root mean square error (%) for the Ctrough model were 11.95% and 16.2% as compared to 67.17% and 42.3% obtained for the Cmax model. Simple linear regression models for Ctrough/Cmax versus AUC0-24h were derived for digoxin. On the basis of statistical evaluation, Ctrough was superior to Cmax model for the prediction of digoxin AUC0-24h and can be potentially used in a prospective setting for predicting drug-drug interaction or lack of it.

Legal and economic identification and assessment of pharmacy substitution in narrow therapeutic index drugs, on the example of epileptic medications in Poland

Legal and economic identification and assessment of pharmacy substitution in narrow therapeutic index drugs, on the example of epileptic medications in Poland

Individualization of Dosage Regimen-an Approach Towards Improved Clinical Outcomes

Dosage regimen plays the most important role in obtaining the desired therapeutic outcomes. The focus perpetually is on the design of doses, their intervals and duration to acquire the optimum response. Many factors like age, gender, weight, body surface area, pathological state play a role in formulating these regimens and each factor is given importance in deciding regimen that gives the optimum response. The experiments gradually reveal the correct regimens after repeatedly being worked upon. Different parameters such as Cmax, Tmax, are calculated and estimated in order to decide by which dose and its interval the therapeutic level is maintained in the blood. Therapeutic drug monitoring is one of the most important aspects in dosage regimen where the narrow therapeutic index drugs are given and designed in such a way as to get the therapeutic response with minimized toxicity.

Individualized in vitro and in silico methods for predicting in vivo performance of enteric-coated tablets containing a narrow therapeutic index drug

The efficacy of narrow therapeutic index (NTI) drugs is closely related to their plasma concentration-time profile. Particularly for these compounds interindividual variability of gastrointestinal (GI) parameters relevant to in vivo drug release may result in fluctuations of the plasma concentration. The present study focused on assessing the influence of individual GI pH- and transit profiles on drug release of enteric valproate tablet formulations by means of individualized in vitro dissolution experiments. After initial experiments simulating GI passages in average healthy adults, a novel in vitro dissolution model was used to simulate individual GI pH- and transit profiles with physiologically relevant dissolution media. Based on the dissolution profiles obtained in these experiments, individual in silico plasma profiles were generated and compared to fasted in vivo data applying a mean Euclidean distance approach. Simulated individual gastric residence time was identified as crucial parameter determining the onset of absorption, whereas the shape of the plasma profile is mainly influenced by individual valproate pharmacokinetics. The novel in vitro and in silico methods used in this study are promising tools for estimating in vivo drug release and plasma concentration in individual subjects and thus may contribute to a prospective risk assessment for NTI formulations.

Effect of Flying at High Altitude on Early Exposure of Paracetamol in Humans.

Pharmacokinetics of paracetamol (APAP) was studied on-board during an air flight and compared to those on ground after 500 mg oral dose in 20 healthy human volunteer in parallel design study. Saliva samples were obtained every 15 min up to 2 h after dosing. Pharmacokinetic parameters were calculated by non compartmental analysis and one compartment models using Winnonlin program V5.2. Results have showed that on-board to ground ratios for area under curves AUC0→1, AUC0→2, time to reach maximum saliva concentration Tmax, absorption rate constant Ka and maximum saliva concentration Cmax were 0.62, 0.38, 1.01, 0.81 and 0.79 respectively. Effective membrane permeability coefficients were optimized by Nelder-Mead algorithm using Simcyp program V13. This showed similar rate of absorption and early exposure up to one hour, and lower bioavailability after 1 h on-board. This can be explained by the increased liver blood flow at high altitude that led to increased liver metabolism on-board. However, APAP elimination parameters were not calculated due to short sampling time. This suggests a need for dose adjustment on-board during long air flights, especially for narrow therapeutic index drugs with flow limited metabolism.

Do older adults know the purpose of their medications? A survey among community-dwelling people

This study assessed the level of knowledge among community-dwelling older individuals about the purpose of medications and the characteristics that influenced that knowledge. We focused on drugs frequently used and/or often involved in iatrogenic events. This cross-sectional survey included 2690 community-dwelling older adults, aged 68years and above, who reported using at least one drug of interest, in the canton of Vaud, Switzerland. Participants reported drugs that they currently used and described their purpose. A good knowledge level was defined as knowledge of the exact purpose or identification of the anatomical system or organ targeted by the drug. We performed a multivariate logistic regression analysis to identify factors associated with a good knowledge level for all drugs of interest. Sampling weights were employed to provide representative estimates. On average, patients had good knowledge of 80.6% of the drugs reported. The highest knowledge levels were demonstrated for non-steroidal antiinflammatory drugs, antidiabetics, analgesics, and endocrinological drugs and the lowest for platelet aggregation inhibitors, minerals, anticoagulants, and other narrow therapeutic index drugs. Overall, 66% of participants had good knowledge of the purpose of all the drugs of interest. Polypharmacy and receiving help with drug management were negatively associated with good knowledge (adjusted OR4-5 drugs 0.45; 95% CI 0.29-0.71, adjusted OR≥ 6 drugs 0.20; 95% CI 0.13-0.31, and adjusted ORhelp 0.42; 95% CI 0.18-0.99). This study showed that education about drugs was lacking among patients that received multiple drugs, particularly patients that used anticoagulants and antiplatelet inhibitors.

Are Clinical Therapeutic Protein-Drug Interaction Studies Needed When Disease Modification Is Driving the Potential Interaction?

Dear Editor, We read with great interest the article by Coutant and Hall1 recently published in The Journal of Clinical Pharmacology regarding disease-drug interactions in inflammatory states. This article provides a comprehensive review of the impact of inflammatory diseases on drug clearance via cytokine-cytochrome P450 (CYP)-mediated disease-drug interactions, with a conclusion that in most cases inflammatory disease–drug interactions are not expected to be of clinical concern. The information in this article raises some important questions regarding therapeutic protein–drug interaction (TP-DI) assessments for therapeutic proteins being developed for the treatment of inflammatory diseases. This topic is of particular interest given that the updated 2017 Food and Drug Administration guidance2 does not include discussion on therapeutic proteins and that the Food and Drug Administration is currently seeking public comments on the optimal regulatory framework in which to assess TP-DIs.3 We would like to point out several key items that should be taken into consideration when evaluating the TP-DI potential for therapies targeting proinflammatory cytokines. First, though comparisons of cytokine levels between patients and healthy subjects may be useful for understanding extreme differences in drug clearance between inflammatory disease and healthy status,1 such comparisons may overestimate the magnitude of difference in drug exposure caused by TP-DIs mediated through disease/cytokine modulation by therapeutic proteins. In fact, most of the disease-mediated TP-DIs due to normalization/suppression of proinflammatory cytokines have shown at most only mild to moderate effects on drug exposures and have rarely led to a dose adjustment for the victim drug (AbbVie data on file for risankizumab and for other related TP-DIs4-8). Second, dedicated TP-DI studies are typically conducted in patients, without the full course of the proposed therapeutic protein treatment for therapeutic benefit and disease modification. Such studies put patients at risk for developing antidrug antibodies, which could negatively affect future treatment. Therefore, the scientific benefit gained from conducting TP-DI studies does not necessarily outweigh the risks. Third, drug interactions mediated through disease/cytokine modification raise different considerations than other types of drug interactions. As pointed out by Coutant and Hall,1 the net effect of treatment in this case is reduced cytokine levels, leading to partial restoration of CYP enzyme and drug transporter activities, potentially resulting in mild reduction of victim drug exposures. As such, reduced efficacy, rather than toxicity, is the likely outcome. There are only a few sensitive CYP substrates with a narrow therapeutic index for which a small change in exposure would fall outside the desired therapeutic range, and these should be closely monitored anyway. Fourth, similar to therapeutic proteins, small-molecule drugs have the same potential to cause drug interactions if they reduce cytokine levels via disease modification, yet such assessments for small-molecule drugs are neither performed nor required by regulatory guidelines, indicating the lack of necessity or clinical relevance of such evaluations. The ultimate goal of drug interaction studies is to determine the magnitude of change in drug exposure that can be expected clinically so that labeling can be put into place. However, based on the points discussed above, when changes in drug exposure are driven by disease modification, conventional drug interaction studies are often not clinically informative. It is our opinion that dedicated TP-DI studies are generally not needed, especially for the following 2 scenarios: (1) For new therapeutic proteins that target pathways previously studied for other therapeutic proteins, TP-DI studies are not needed if the previously studied therapeutic protein has demonstrated little or no TP-DI potential; and (2) if a therapeutic protein is being developed for multiple indications and the therapeutic protein has been shown to have little or no TP-DI potential in the disease with the most severe inflammatory burden, TP-DI studies are not needed in any new indications that have a less severe inflammatory burden. For a novel target pathway, data on cytokines that have known effects on CYPs or transporters can be collected before and after therapeutic protein treatment in phase 2/3 studies to guide the TP-DI risk assessment, and, if needed, a TP-DI substudy can be incorporated into the phase 2/3 studies to evaluate the impact of the therapeutic protein on narrow therapeutic index drugs. We believe that continuing to default to dedicated TP-DI studies for therapeutic proteins targeting inflammatory diseases is not supported by available scientific evidence, and drug development standards need to be revised accordingly. The authors are employees and shareholders of AbbVie.

High tacrolimus trough level variability is associated with rejections after heart transplant.

Tacrolimus, the major immunosuppressant after heart transplant (HTx) therapy, is a narrow therapeutic index drug. Hence, achieving stable therapeutic steady state plasma concentrations is essential to ensure efficacy while avoiding toxicity. Whether high variability in steady state concentrations is associated with poor outcomes is unknown. We investigated the association between tacrolimus trough level variability during the first year post-HTx and outcomes during and beyond the first postoperative year. Overall, 72 patients were analyzed for mortality, of whom 65 and 61 were available for rejection analysis during and beyond the first year post-HTx, respectively. Patients were divided into high (median >28.8%) and low tacrolimus level variability (<28.8%) groups. Mean tacrolimus levels did not differ between the groups (12.7 ±3.4 ng/mL vs 12.8 ±2.4 ng/mL, P=.930). Patients in the high variability group exhibited higher long-term rejection rate (median total rejection score: 0.33 vs 0, P=.04) with no difference in rejection scores within the first year post-HTx. Multivariate analysis showed that high tacrolimus trough level variability was associated with >8-fold increased risk for any rejection beyond the first year post-HTx (P=.011). Mortality was associated only with cardiovascular complications (P=.018), with no effect of tacrolimus through level variability.

A “Large-N” Content Uniformity Process Analytical Technology (PAT) Method for Phenytoin Sodium Tablets

Accurate assessment of tablet content uniformity is critical for narrow therapeutic index drugs such as phenytoin sodium. This work presents a near-infrared (NIR)-based analytical method for rapid prediction of content uniformity based on a large number of phenytoin sodium formulation tablets. Calibration tablets were generated through an integrated experimental design by varying formulation and process parameters, and scale of manufacturing. A partial least squares model for individual tablet content was developed based on tablet NIR spectra. The tablet content was obtained from a modified United States Pharmacopeia phenytoin sodium high-performance liquid chromatography assay method. The partial least squares model with 4 latent variables explained 92% of the composition variability and yielded a root mean square error of prediction of 0.48% w/w. The resultant NIR model successfully assayed the composition of tablets manufactured at the pilot scale. For one such batch, bootstrapping was applied to calculate the confidence intervals on the mean, acceptance value, and relative SD for different sample sizes, n = 10, 30, and 100. As the bootstrap sample size increased, the confidence interval on the mean, acceptance value, and relative SD became narrower and symmetric. Such a ‘large N’ NIR-based process analytical technology method can increase reliability of quality assessments in solid dosage manufacturing.

Effect of processing parameters and controlled environment storage on the disproportionation and dissolution of extended-release capsule of phenytoin sodium

Phenytoin sodium (PS) is a narrow therapeutic index drug that dictates maintenance of narrow plasma concentration of the drug. This requires a good quality product that meets regulatory specifications throughout shelf-life as well as during the usage period. Quality of the drug product may change if not develop in a scientific fashion which may have ramifications on clinical outcome. The focus of the study was to understand the effect of process variables and storage conditions on the disproportionation and phase transformation of the drug, and dissolution of extended-release capsules of PS. Four formulations were prepared to contain either powder blend or granules of PS with commonly used excipients. The granules were prepared in a high shear mixture by granulating with water, ethanol or mixture thereof. The capsules were stored in pharmacy vials for 4-weeks at 40 °C/75% RH and 30 °C/65% RH. Formulations were characterized by spectroscopies (FTIR, NIR, and chemical imaging), X-ray powder diffraction (XRPD), dissolution and assay. Granules filled capsule did not meet USP dissolution specifications and there was a significant decrease in dissolution on exposure to stability conditions. Dissolution decreased from 77.56 ± 3.83% to 42.34 ± 4.31%, and 63.96 ± 6.12 to 46.53 ± 2.91 in physical mixture and water granulated formulations, respectively, after four weeks exposure at 40 °C/75% RH. Spectroscopic (NIR, FTIR, and chemical imaging) data indicated disproportionation of the drug during the storage period. Moreover, XRPD data explained a decrease in dissolution in the initial granules filled capsules as well as temperature and humidity exposed samples. It indicated the phase transformation of the drug and/or excipients, disproportionation and/or interactions. These changes further accelerated on exposure to stability conditions. In conclusion, our studies indicated the significant effect of process variables and stability conditions on the critical quality attributes of PS extended-release capsule that has the potential impact on the clinical performance of the product.