Crushing tablets or sprinkling capsules: Implications for clinical strategy and study performance based on BE studies of rivaroxaban and deferasirox

Abstract

Administration of oral medicinal products as crushed tablets or open capsules is an important delivery option for patients suffering from dysphagia. To obtain full interchangeability of generics with the original products, demonstration of bioequivalence (BE) between both products administered as crushed tablets/open capsules was required for poorly soluble product by European Medicines Agency (EMA) at the time of development of our rivaroxaban and deferasirox generic products. We present the results of two BE studies with modified administration of these products, which compared relative bioavailability between generic and reference products. In the rivaroxaban study, the test product was administered as a capsule sprinkled on and mixed with applesauce, whereas the reference tablet was crushed and administered with applesauce under fed conditions. In the deferasirox study, both treatments were administered as crushed tablets under fasting conditions. Both studies applied a two‐way crossover design and were conducted after a single‐dose in healthy volunteers. The 90% confidence interval of the geometric mean ratio area under the analyte concentration versus time curve, from time zero to the time of the last measurable analyte concentration and maximum measured analyte concentration over the sampling period of the test to reference ratio were 103.36–110.37% and 97.98–108.45% for rivaroxaban, respectively, and 96.69–107.29% and 94.19–109.45% for deferasirox, respectively. Thus, the BE criteria (80.00–125.00%) were met in both studies which demonstrated that bioavailability was not affected when the test and reference products were administered in the form of crushed tablet/open capsule. These results support the argument of redundancy of crushed product studies for poorly soluble drugs, which is in line with the currently revised position of the EMA on this topic.

Study Highlights.

• WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

Bioavailability for some products may be impacted by tablet crushing. Therefore, the bioavailability of the product administered as the crushed product is compared to the administration of the intact product during the development of the original product. It is questionable if the differences in pharmaceutical properties between original and generic products which are bioequivalent in the form of intact product, may result in different bioavailability when administered as a crushed product.

• WHAT QUESTION DID THIS STUDY ADDRESS?

Is bioavailability comparable between generic and original products when administered as crushed tablet/open capsule? Does it affect the interchangeability of both products?

• WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

The studies provide practical advice how to design and perform crushed tablet studies to ensure smooth study conduct, process standardization, and reliability of the data. The results of the studies bring additional proof that bioequivalence (BE) studies with crushed tablets/ open capsules between original and generic products are redundant in clinical development of generic products.

• HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

The results of the BE studies with rivaroxaban and deferasirox, which compared bioavailability between original and generic products administered as open capsule/crushed tablet, support the current position of the European Medicines Agency that such comparison is not necessary in the clinical development of generics, regardless of Biopharmaceutics Classification System classification.

INTRODUCTION

Dysphagia

Dysphagia is the medical term for swallowing difficulties and may represent a significant medical issue for administration of solid dosage forms (DFs). The prevalence of dysphagia reported in the literature is characterized by a broad range of epidemiological data, with elderly residents in hospital or nursing homes being at the highest risk. Dysphagia affects ~12–30% of community‐dwelling elderly and 53–60% of elderly nursing home residents. ¹ , ² Dysphagia is often related to natural features of aging, such as hyposalivation or weakening of swallowing muscles, however, it is often present in conjunction with primary underlying diseases or injury. Swallowing difficulties are present in patients suffering from Parkinson's disease (80% of patients), ³ multiple sclerosis (43%), ⁴ dementia (72.4–85.9%), ⁵ , ⁶ and head or neck cancer (31.5%). ⁶ Dysphagia also occurs as a consequence of stroke (55.4%), ⁶ traumatic brain injury (up to 93%), ⁷ , ⁸ heartburn and regurgitation symptoms (48%), ⁹ pneumonia (44.6%), ⁶ or intake of drugs, such as anticholinergics or chemotherapeutics. Psychological aversion to swallowing solid DFs may also play a crucial role in some patients. Apart from daily inconveniences associated with food intake, dysphagia may result in malnutrition or dehydration. It represents a substantial medical issue in therapy with oral solid DFs, often leading to low adherence to the treatment and poor therapy outcomes.

Tablet crushing/capsule opening as alternatives for patients with dysphagia

For patients with dysphagia who are indicated for pharmacological therapy with oral solid DFs, seeking a product containing the same active pharmaceutical ingredient (API) in an alternative oral DF (such as liquids, lozenges, dispersible or effervescent tablets, etc.) or a product with alternative routes of administration (such as injections, transdermal patches, or inhalers) may be a solution. However, many products are available only in oral solid DFs and alternatives are not available. Tablet crushing or capsule opening represents an alternative solution how to deal with dysphagia and ensure proper pharmacological therapy. For convenient administration, the powder generated by tablet crushing/capsule opening is typically mixed with food (porridge, apple puree, yogurt, jam, etc.) or beverages. The administration via invasive methods, such as nasogastric or gastric tubes, is reserved for more severe cases of dysphagia.

Risk/benefits for tablet crushing/capsule opening

Apart from delivery of proper pharmacological therapy to patients with dysphagia, tablet crushing may be applied to alter pharmacokinetic (PK) properties of the drug. Due to omitting of the disintegration phase in the stomach, absorption may be accelerated, and effective plasma concentrations reached faster. This may be beneficial in pain treatment where a rapid onset of action is desired. For example, prasugrel acts more rapidly against platelet activation and aggregation when administered as a crushed tablet compared to an intact tablet. ¹⁰ The PKs may be impacted also in an undesired way, therefore certain medicines should not be crushed. For example, crushing of modified‐release formulations leads to destroying of gradual release mechanism of the product and results in overdose, which may be fatal in medicines such as morphine. ¹¹ Crushing of enteric coated formulations leads to destruction of the functional gastro‐resistant layer, which may result into decomposition of the drug by stomach acid (proton pumps inhibitors ¹² and pancreatin ¹³ ), not reaching a proper place of action (sulphasalazine ¹⁴ ), irritation of the stomach (acetylsalicylic acid ¹⁵ ) or esophagus (alendronate ¹⁶ ), unacceptable taste (quinine ¹⁷ and pseudoephedrine ¹⁸ ), or stability issues due to light‐sensitivity (nifedipine ¹⁹ ). Inhalation of the powder generated by tablet crushing represents an occupational issue for personnel in long‐term care units, therefore any carcinogenic or teratogenic drugs, hormones, and corticosteroids should not be crushed.

Regulatory environment regarding crushed tablets

In 2018, the European Medicines Agency (EMA) issued a position which generally required the demonstration of bioequivalence (BE) between test and reference products, if the reference product allowed for the possibility to administer the tablet crushed/disintegrated and if the test product was also intended for this additional mode of administration. The rationale was that the bioavailability of an API may be altered if the product is crushed/disintegrated (and dispersed in food). As the change in bioavailability may be formulation/product‐specific, tablet crushing could result in different bioavailability in a test and reference product, even if the BE between these products has been already shown with intact products. The requirement for the additional study could be waived if the API belonged to Biopharmaceutics Classification System (BCS) classes 1 or 3 (good solubility) and product behaved appropriately in in vitro testing (comparative evaluation of excipients, multimedia in vitro dissolution, and disintegration testing).

However, in March 2019, the EMA published a revised position, stating that BE does not need to be demonstrated between test and reference products for administration as crushed tablets. The EMA recognized that it is highly unlikely that the change in bioavailability will be different between crushed test and reference products once BE has been successfully demonstrated with the intact/non‐dispersed tablet.

The two studies presented in this publication testing generic prototypes of rivaroxaban and deferasirox were performed at the time when the original position of the EMA was in place. Both APIs are poorly soluble drugs and belong to BCS class 2, thus BE studies for crushed/disintegrated tablets against reference products were required if the generic product aimed for full interchangeability with the reference product. The absence of this alternative method of administration in the SmPC of generics would be a disadvantage for the product compared to the reference and other generic products on the market.

Rivaroxaban/deferasirox introduction

Rivaroxaban is a highly selective direct factor Xa inhibitor which acts via interruption of the intrinsic and extrinsic pathway of the blood coagulation cascade, inhibiting both thrombin formation and development of thrombi. ²⁰ For patients who are unable to swallow whole tablets, the rivaroxaban tablet may be crushed and mixed with water or apple puree immediately prior to use and administered orally. After the administration of a crushed 15 mg or 20 mg tablet, the dose should be immediately followed by food, whereas the lower strengths 2.5 mg and 10 mg can be taken with or without food. ²⁰ The clinical program for generic rivaroxaban consisted of two BE studies on the intact formulations, one on the 10 mg strength under fasting conditions and one on the 20 mg strengths under fed conditions.

Deferasirox is an orally active chelator that is highly selective for iron (III) and is indicated for the treatment of chronic iron overload in patients with beta thalassemia major or other anemias. ²¹ For patients who are unable to swallow whole tablets, tablets may be crushed and administered by sprinkling the full dose onto soft food, for example, yogurt or apple sauce (pureed apple). The dose should be immediately and completely consumed, and not stored for future use. ²¹ One BE study on the intact deferasirox formulation was conducted on the highest strength of 360 mg (Table 1).

TABLE 1.

Mean PK parameters and results of rivaroxaban and deferasirox BE studies based on plasma rivaroxaban and deferasirox levels.

PK parameter	Arithmetic means (±SD)		T/R ratio (%)	90% CI	Intra‐subject CV (%)
	Test	Reference
Rivaroxaban
BE study, rivaroxaban, 20 mg: open capsule (test) versus crushed tablet (reference)
C max (ng/mL)	248.39 (±41.89)	242.16 (±48.91)	103.08	97.98–108.45	13
AUC0−t (ng·h/mL)	2787.40 (±658.95)	2611.97 (±624.09)	106.81	103.36–110.37	8
T max (h) a	4.50 (1.00–12.00)	4.25 (0.50–8.00)	–	–	–
BE study, rivaroxaban, 20 mg: whole capsule (test) versus whole tablet (reference)
C max (ng/mL)	354.13 (92.74)	376.11 (98.47)	94.22	88.82–99.95	15
AUC0−t (ng·h/mL)	3049.39 (709.04)	2913.22 (885.68)	106.62	101.78–111.69	12
T max (h) a	4.50 (3.00–10.07)	3.75 (0.50–6.00)	–	–	–
Deferasirox
BE study, deferasirox, 360 mg; crushed tablet (test) versus crushed tablet (reference)
C max (ng/mL)	15,868 (±3971)	15,693 (±3874)	101.54	94.19–109.45	18
AUC0−t (ng·h/mL)	126,745 (±39,504)	124,122 (±35,736)	101.85	96.69–107.29	12
T max (h) a	3.17 (2.33–8.00)	2.84 (2.00–4.50)	–	–	–
BE study, deferasirox, 360 mg; whole tablet (test) versus whole tablet (reference)
C max (ng/mL)	17,715 (3489)	17,181 (4405)	104.55	96.37–113.43	18
AUC0−t (ng·h/mL)	150,118 (40517)	156,664 (39678)	95.62	90.76–100.74	11
T max (h) a	2.67 (1.00–5.00)	2.35 (1.02–5.00)	–	–	–

Abbreviations: AUC_0−t , area under the analyte concentration versus time curve, from time zero to the time of the last measurable analyte concentration; BE, bioequivalence; CI, confidence interval; C _max, maximum measured analyte concentration over the sampling period; CV, coefficient of variation; PK, pharmacokinetic; T/R, test/reference; T _max, time of the maximum measured analyte concentration over the sampling period.

^a Median (range).

The clinical development of rivaroxaban and deferasirox generic products was extended by one additional BE study with open capsule/crushed tablets per each product to fulfill EMA regulatory requirements at that time. A generic formulation of rivaroxaban was developed in the form of a hard capsule, therefore bioavailability of the powder from an opened capsule was compared against crushed film‐coated tablet (FCT) of original product Xarelto in the BE study. The generic deferasirox product was developed in the same DF as the original product Exjade and thus, bioavailability of two crushed FCTs was compared.

METHODS

The studies were conducted between October and December 2017 at Pharma Medica Research Inc. (PMRI), Canada. The clinical study was carried out in accordance with the Declaration of Helsinki, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use–Good Clinical Practice, study protocol, PMRI standard operating procedures, and all applicable laws and regulations. The protocol and informed consent forms were reviewed and approved by Health Canada and by Institutional Review Board Advarra (Canada).

Study treatments and design

Rivaroxaban study

The open‐label, single‐dose, randomized, two‐period, two‐treatment, two‐sequence, crossover study evaluated the comparative bioavailability between test rivaroxaban capsules 20 mg, sprinkled on and mixed with applesauce, and the reference Xarelto 20 mg, FCT, crushed and administered with applesauce, both after a single‐dose in healthy subjects under fed conditions. The test formulation was manufactured by Sanofi – Synthelabo Ltd., India, and Zentiva, k.s., Czech Republic in GMP facilities. The reference product Xarelto 20 mg, FCT (Bayer Pharma AG, Germany) was purchased from the German market. The washout between drug administrations was 7 days. Subjects entered the clinical facility at least 10 h predose and were housed until at least 24 h postdose. Subjects were required to return to the clinical site for one remaining blood sample at 48 h postdose. Both products were administered after an overnight fast of at least 10 h, as per randomization code. The test capsule was opened, sprinkled on and mixed with applesauce. The reference FCT was crushed and administered with applesauce. Both treatments were followed by 240 mL (±5 mL) of room temperature potable water. Immediately after drug dosing, a standardized high‐fat, high‐calorie breakfast (HFHCB) was administered. Subjects were required to consume the entire meal within 30 min of the drug administration time. Subjects then fasted for at least 4 h after completing the HFHCB. Standardized meals were provided throughout confinement. Water was restricted from 1 h prior to drug administration until 1 h postdose, except for the water administered with the drug. Access to water was otherwise freely available to subjects. PK blood samples were collected into tubes containing K₂ EDTA prior to dosing (0‐h) and at 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 7, 8, 10, 12, 16, 24, and 48 h postdose. The primary study end points considered for BE assessment were the PK parameters maximum measured analyte concentration over the sampling period (C _max) and the area under the analyte concentration versus time curve, from time zero to the time of the last measurable analyte concentration (AUC_0−t ). The evaluation of BE was based upon plasma concentrations of the parent compound rivaroxaban. The design of the BE study with rivaroxaban on intact DFs is provided in Supplementary Material S1.

Deferasirox study

The open‐label, single‐dose, randomized, two‐period, two‐treatment, two‐sequence, crossover study evaluated the comparative bioavailability between test deferasirox 360 mg FCT and the reference Exjade 360 mg FCT, after a single‐dose, crushed and administered with applesauce in healthy subjects under otherwise fasted conditions. The test formulation was manufactured by Zentiva, k.s., Czech Republic in GMP facilities. The reference product Exjade 360 mg, FCT (Novartis Europharm Limited, Ireland) was purchased from the German market. The washout between drug administrations was at least 7 days. Subjects entered the clinical facility at least 10 h predose and have been housed until at least 24 h postdose. Subjects were required to return to the clinical site for three remaining blood samples. Both products were administered after an overnight fast of at least 10 h, as per randomization code. Both FCTs were crushed and administered with ~15 mL of applesauce followed by 240 mL (±5 mL) of room temperature potable water. Subjects then fasted for at least 4 h after dosing. Standardized meals were provided throughout confinement. Water was restricted from 1 h prior to drug administration until 1 h postdose, except for the water administered with the drug. Access to water was otherwise freely available to subjects. PK blood samples were collected prior to dosing (0‐h) and at 0.5, 1, 1.5, 2, 2.33, 2.67, 3, 3.33, 3.67, 4, 4.5, 5, 6, 8, 10, 12, 16, 24, 36, 48, and 72 h after drug administration. Subjects were required to return to the clinic for the 36, 48, and 72‐h samples. The primary study end points considered for BE assessment were the PK parameters C _max and AUC_0−t . The evaluation of BE was based upon plasma concentrations of the parent compound deferasirox. The design of the BE study with deferasirox on intact FCT is provided in Supplementary Information S1.

Study subjects

Healthy, non‐smoking, male and female subjects, 18 years of age or older with body mass index greater than or equal to 18.5 and less than or equal to 30.0 kg/m² were included in both studies. Subject eligibility was assessed based on medical history, demographic data collection, vital signs measurements (blood pressure, pulse rate, respiration rate, and temperature), electrocardiogram, laboratory tests (biochemistry, hematology, serology, immunohematology, and urinalysis), and physical examination at screening within 21 days prior to the first drug administration. Subjects were ineligible to participate in the study if they had known history or presence of clinically significant disease or condition which, in the opinion of the investigator, would jeopardize the safety of the subject or impact the validity of the study results. Subjects could not have participated in a clinical if having donated blood greater than or equal to 50 mL and less than 500 mL within 30 days prior to drug administration or greater than or equal to 500 mL within 56 days prior to drug administration. Use of prescription or over‐the‐counter medicines, herbal/natural products, nutritional supplements, and vitamins was restricted 14 days prior to drug administration. Grapefruit and grapefruit‐containing products were prohibited within 7 days prior to dosing, whereas products containing alcohol, caffeine, and xanthine were not to be used within 48 h prior to dosing. Subjects underwent alcohol and drugs of abuse tests prior to both study period. Forty and thirty‐four subjects were included in rivaroxaban and deferasirox BE studies, respectively.

Method of administration

Rivaroxaban study

In each study period, the test capsule (sprinkled on and mixed with applesauce) or reference tablet (crushed and administered with applesauce) were administered. Due to different DFs of both treatments, preparation and dose administration differed for test and reference products. Administration procedures of test and reference products are depicted in Figures 1 and 2, respectively. The full description of the administration procedures is available in Supplementary Material S1.

FIGURE 1.

Administration procedure for rivaroxaban capsules 20 mg.

FIGURE 2.

Administration procedure for reference Xarelto 20 mg, film‐coated tablets.

Deferasirox study

In each study period, either Deferasirox 360 mg, FCT or Exjade 360 mg, FCT, crushed and mixed with applesauce, were administered. As the DFs were the same, a single procedure for preparation and dose administration was designed for both treatments. The administration procedure is depicted in Figure 3. The full description of the administration procedure is available in Supplementary Material S1.

FIGURE 3.

Administration procedure for deferasirox, film‐coated tablets.

Sample processing and bioanalysis

Blood samples were collected in pre‐chilled, labeled 4 mL (rivaroxaban)/6 mL (deferasirox) blood collection tubes containing K₂ EDTA as the anticoagulant. Predose samples were collected extra in 10 mL K₂ EDTA tubes within 60 min prior to dosing. The samples were maintained in an ice‐water bath throughout sample collection and until further processing. Within 45 min of whole blood collection, whole blood samples were centrifuged at ~4°C for ~10 min at (approximately 1500 x g). The centrifugation had to be completed within 45 min from whole blood collection. Plasma was divided into two approximately equal aliquots in labeled polypropylene tubes and stored in the freezer at −25 ± 10°C. Then, the samples packed in dry ice were shipped to the analytical facility in two separate aliquots. Plasma samples were assayed for rivaroxaban/deferasirox using the respective proprietary validated high performance liquid chromatography method with tandem mass spectrometry detection developed by the Contract Research Organisation (CRO). The methods were validated within the analytical range 1–500 ng/mL for rivaroxaban and 150–30,000 ng/mL for deferasirox according to the GLP principles.

Pharmacokinetic and statistical analysis

The following PK parameters were estimated for rivaroxaban/ deferasirox using a noncompartmental approach in SAS: C _max, time of the maximum measured analyte concentration over the sampling period (T _max) and AUC_0−t (as calculated by the linear trapezoidal method). The actual postdose sample collection times were used in the PK analysis. Statistical analysis was performed using the PROC GLM procedure from SAS (version 9.4). Analysis of variance (ANOVA) was performed on log‐transformed AUC_0−t and C _max parameters. The significance of the sequence, period, treatment, and subject (sequence) effects (all fixed) was tested. The least‐squares‐means, the differences between the treatments least‐squares‐means, and the corresponding standard errors of these differences were estimated for log‐transformed AUC_0−t and C _max parameters. Based on these statistics, the ratios of the geometric means for treatments and the corresponding 90% confidence intervals (CIs) were calculated. Evaluation of T _max was based on a nonparametric approach using Wilcoxon's rank sum test. The intra‐subject coefficient of variation (ISCV) was estimated based on the formula $\sqrt{e^{\mathrm{MSE}}-1}$, where MSE is the mean square error obtained from the ANOVA model of the ln‐transformed parameters. Statistical analyses were generated using SAS (version 9.4) using the GLM procedure. The 90% CI of the geometric mean ratio plasma rivaroxaban/deferasirox AUC_0−t and C _max of the test to reference products should be between the standard BE criteria 80.00 and 125.00%. An assessment of safety was based primarily on the frequency and severity of adverse events (AEs).

Pharmacokinetic results

Forty subjects were enrolled in the rivaroxaban study and 38 subjects completed the study and were included in PK dataset. Two subjects were withdrawn prior to dosing in period 2. The demographic baseline characteristics are depicted in Table 2. The 90% CIs of the geometric mean ratio rivaroxaban AUC_0−t and C _max of the test to reference products were 103.36–110.37% and 97.98–108.45%, respectively, thus the BE criteria were met. A significant treatment effect was detected by ANOVA for AUC_0−t parameter. Despite the statistical significance, the 90% CI of the test to reference ratio was entirely contained within the 80.00–125.00% BE range and thus deemed acceptable from the regulatory perspective. ANOVA did not detect a significant difference in any of the PK parameters for period and sequence effects. ISCV for C _max was 13%, which was slightly lower than expected (16%). T _max was comparable between study treatments.

TABLE 2.

Demographic baseline characteristics of rivaroxaban and deferasirox studies.

Demographic characteristic	Crushed tablet rivaroxaban study (N = 40)	Whole tablet rivaroxaban study (N = 40)	Crushed tablet deferasirox study (N = 34)	Whole tablet deferasirox study (N = 28)
Age, years, mean (SD)	43 (10)	38 (11)	44 (10)	45 (11)
Sex, n (%)
Male	27 (67.5)	11 (27.5)	17 (50.0)	12 (42.9)
Female	13 (32.5)	29 (72.5)	17 (50.0)	16 (57.1)
Race, n (%)
Asian	11 (27.5)	5 (12.5)	7 (20.6)	3 (10.7)
Black or African American	10 (25.0)	17 (42.5)	5 (14.7)	10 (35.7)
White	19 (47.5)	18 (45.0)	22 (64.7)	15 (53.6)
Ethnicity, n (%)
Hispanic/Latino	8 (20.0)	11 (27.5)	10 (29.4)	5 (17.9)
BMI (kg/m2), mean (SD)	25.8 (2.9)	25.4 (2.7)	26.3 (2.8)	25.5 (2.5)

Abbreviation: BMI, body mass index.

In the deferasirox study, 34 subjects were enrolled, and 32 subjects completed the study. Two subjects discontinued from the study prior to period 2 dosing. The 90% CIs of the geometric mean ratio deferasirox AUC_0−t and C _max of the test to reference products were 96.69–107.29% and 94.19–109.45%, respectively, thus within the BE range. ANOVA did not detect a significant difference in any of the PK parameters for period, sequence, and treatment effects. ISCV observed for C _max (18%) was similar to the estimated ISCV based on in‐house data and literature (19%).

The study medications were well‐tolerated by the healthy volunteers that participated in both studies. All AEs were mild in severity and resolved. No serious AEs were reported.

The summary of the PK and statistical analyses for the plasma concentrations of rivaroxaban and deferasirox, are presented in Table 1. The mean plasma rivaroxaban and deferasirox concentration‐time profiles are shown in Figure 4.

FIGURE 4.

Mean plasma concentration‐time profiles of (a) deferasirox study crushed product, and (b) deferasirox study whole product, (c) rivaroxaban study crushed product, and (d) rivaroxaban study whole product.

DISCUSSION AND CONCLUSION

We provide results from four bioequivalence studies: two studies were performed to assess the relative bioavailability of crushed and dispersed formulations containing either rivaroxaban or deferasirox, and two studies assessed the relative bioavailabilities of the respective whole/intact formulations. Content of the rivaroxaban capsule, sprinkled and administered with applesauce, exhibited equivalent rate and extent of absorption to the reference product, which was crushed and administered with applesauce in healthy subjects after a single oral dose under fed conditions. The rate of absorption as shown on the C _max was slightly higher (103.08% test/reference ratio), in contrast to the C _max ratio in the whole tablet study (94.22%). However, these differences are considered insignificant as the respective CIs overlap. The extent of exposure as shown by the AUC_0−t ratio remained identical in the two studies (106.81% and 106.62%). Observed mean C _max was decreased by 30–36% in the crushed tablet study (248.39 ng/mL) when compared to the study with whole product (354.13 ng/mL), which may indicate higher distribution of powder in food and a more gradual release to the intestine within multiple contraction waves of the stomach. In contrast, the whole product may not be fully disintegrated in the stomach and can be released to the intestine in larger fractions, potentially within one contraction wave, resulting in higher rate of absorption in comparison to the crushed product. Mean AUC_0−t values were decreased by ~10% (crushed tablet study vs. whole tablet study). As bioavailability of rivaroxaban is almost complete (80–100%), ²² this finding may indicate minimal losses during crushing procedure and study drug administration. Alternatively, this difference could be caused by interindividual differences in the study subjects when comparing the two separate populations.

The C _max test/reference ratio of deferasirox was similar between crushed and whole tablet study, while AUC_0−t test/reference ratio was slightly lower (95.62%) following crushed tablet administration. In general, all mean PK parameters displayed lower absolute values in the crushed tablet study, which may be attributed to intersubject differences between the subject groups enrolled in the two studies.

Both these studies give insight into the practicalities of dosing procedures when administering oral solid DFs altered to facilitate swallowing. As swallowing issues are predominant in the elderly population and in specific indications, pharmaceutical products must be tested in such regimens to demonstrate that such alterations (namely tablet crushing or capsule opening and dispersion in a beverage or soft food) do not influence the rate and extent of exposure. The first important decision is which tablet crushing system to use as there are many systems available and they differ among countries. The rivaroxaban study utilized the Silent Knight Pill Crusher with disposable vessels and bags for crushing the reference tablet, whereas the opened capsule needed no additional device, and the deferasirox study used the Pill Crusher PharmaSystems for both treatments. The choice of crushing system is important as the critical factors to control in such studies are sufficient crushing of a pill, standardization of the crushing and administration process, and minimization of loss of crushed material. These factors were taken into consideration when designing the crushing/administration process (Figures 2 and 3) in order not to increase study variability. Specifically, the Silent Knight Pill Crusher would have fulfilled the selection criteria for the tablet‐crushing device but could not be used in the deferasirox study because of the large size of the tablets, which could result in rupture of the pouch and high likelihood of loss of crushed material. In the end, as both studies displayed relatively lower or similar ISCV of the C _max and AUC parameters when compared to whole‐formulation studies, it is clear that the tablet‐crushing and administration measures were sufficiently robust and can serve as a template for future design of crushed tablet studies. In addition, these studies used two real‐world methodologies for crushing, which is not standard in pharmaceutical development where tablet crushing can often be done via mortar and pestle. ²³ Administration procedures for both studies has to be developed specifically for each study, taking into consideration the selected crusher and test medication. It was important that the study staff was appropriately trained, and that time pressure was avoided during the crushing procedure.

The need to perform crushed‐tablet studies for novel treatments developed as tablets/capsules remains critical for patients with swallowing difficulties; however, the need to perform such studies as part of generic development has been shown as not necessary. Even the early regulatory requirements admitted that crushed tablet BE studies are not needed for well‐soluble compounds (BCS classes 1 and 3) due to the fact that formulation characteristics no longer play a role once the tablet is crushed and in vivo behavior is dictated by API characteristics, which in the case of well‐soluble compounds results in bioequivalent rate and extent of absorption. Hence, no additional crushed tablet studies are required for well‐soluble compounds. The studies described above demonstrate that the rate and extent of absorption were not affected when the test and reference products were administered in the form of crushed tablet/open capsule, thus showing that crushed tablet BE trials for poorly soluble compounds are also practically redundant. Even though BCS 2 or 4 class API characteristics could be significantly different between original and generic products, such differences would already be detected in the intact tablet BE study. These results are in line with the current position of the EMA which states that “it is highly unlikely that the change in bioavailability will be different between test and reference, once BE has been shown between test and reference with the intact or non‐dispersed tablet.” ²⁴

AUTHOR CONTRIBUTIONS

J.S. and J.B. wrote the manuscript, designed the research, performed the research, and analyzed the data. T.H. designed the research.

FUNDING INFORMATION

This study was entirely sponsored by Zentiva, k.s., Prague, Czech Republic.

CONFLICT OF INTEREST STATEMENT

The authors are current or former employees of Zentiva, k.s, Prague, Czech Republic.

Supporting information

Data S1.

Sus J, Bosak J, Hauser T. Crushing tablets or sprinkling capsules: Implications for clinical strategy and study performance based on BE studies of rivaroxaban and deferasirox. Clin Transl Sci. 2024;17:e13752. doi: 10.1111/cts.13752