Praziquantel amorphous solid dispersions: whether the polymer, the enantiomer or the process decides release and crystallization in gastric, fasted and fed media

Background and motivation

Schistosomiasis is a tropical disease caused by trematode worms of the genus Schistosoma. The World Health Organization (WHO) estimates that more than 250 million people needed preventive treatment in 2021 (refs 1–3). Praziquantel (PZQ) is the first-line drug for all forms of the disease and is on the WHO Model List of Essential Medicines for adults and children (refs 3, 4). PZQ is a Biopharmaceutics Classification System (BCS) class II drug: it permeates well but dissolves poorly, about 0.4 mg/mL in water at 25 °C, so it dissolves slowly in the gastrointestinal tract (refs 5–7). It also undergoes extensive first-pass metabolism, which converts it rapidly into inactive metabolites (refs 6, 8–10). The result is a high dose, usually 20 mg/kg three times a day at intervals of 4 to 6 h, or a single 40 mg/kg dose for preventive chemotherapy (refs 11, 12), and tablets that are large and hard to swallow, in particular for children (Section 1).

A second problem is chirality. PZQ has one stereocentre, and the innovator product Biltricide and its generics contain a 1:1 crystalline racemate (refs 12, 13). Only the (R)-enantiomer carries the antischistosomal activity. The (S)-enantiomer has little or no activity, but it gives a bitter taste and side effects such as abdominal pain, fever, nausea and vomiting (refs 8, 14–16; Meister et al. 2014, ref 57, on the activity of the enantiomers and metabolites; Meyer et al. 2009, ref 15, on taste as a reason to switch). A product that contains only (R)-PZQ is expected, in the authors' words in Section 4.3, to remove most of these drawbacks and to halve the tablet size or the overall dose. Isolating the enantiomer costs more; the authors argue that if a formulation also raised bioavailability, "the dose could be further reduced" and offset that cost.

The formulation tool chosen here is the amorphous solid dispersion (ASD), in which the drug is dispersed in the amorphous state in a polymer carrier. Because no crystal lattice has to be broken, the apparent solubility, the dissolution rate and often the bioavailability rise, and the hydrophilic polymer improves wetting (refs 17–23). Dissolution of an ASD can produce a supersaturated solution, which raises the flux of drug across the intestinal membrane; the polymer's task is then to keep the drug in solution by slowing nucleation and crystal growth (refs 50, 51). PZQ ASDs had already been made by solvent evaporation (SE), fusion, co-precipitation, co-grinding, spray drying and hot-melt extrusion (HME) (refs 24–34), and racemic PZQ ASDs raised bioavailability in preclinical models (Liu et al. 2018, ref 24; refs 27, 58). Most of those studies, however, tested dissolution only in simple aqueous media. Media that simulate the fasted and fed gut are known to be informative for poorly soluble drugs (Klein 2010, ref 35), but for PZQ they had been applied only to a crystalline formulation (Eason et al. 2022, ref 36). There were also few studies of release from ASDs of the active (R)-enantiomer.

The direct predecessor is an earlier study from the same group (Polyzois et al. 2024, ref 42). It made ASDs of racemic and enantiopure PZQ and found that hydroxypropyl methylcellulose acetate succinate (HPMCAS) and hydroxypropyl methylcellulose (HPMC) both delayed PZQ crystallization, that HPMC was less effective than the MF or LF grades of HPMCAS, and that (R)-PZQ had a shorter nucleation induction time than the racemate from supersaturated solution. These two polymers also gave promising release and stability at 30% drug loading (DL), which is why the present study keeps both polymers and that loading (Section 3.2). Two further results frame the choice of media. Elkhabaz et al. 2021 (ref 56) found that for posaconazole and atazanavir, crystallization induction times with HPMCAS present were much longer in buffer than in fasted-state simulated intestinal fluid, that is, bile salts and lecithin weakened the polymer as an inhibitor. Castro et al. 2000 (ref 49) found that both a high-fat and a high-carbohydrate meal raise PZQ bioavailability, the carbohydrate meal more.

Against this background, the paper sets out to measure how chirality affects PZQ release from ASDs in aqueous media of different pH and in fasted-state simulated gastric fluid (FaSSGF), fasted-state simulated intestinal fluid (FaSSIF) and fed-state simulated intestinal fluid version 2 (FeSSIF-V2). At the same time it varies the polymer (HPMCAS-MF versus HPMC E5 LV) and the preparation process (SE versus HME), all at 30% DL, and it follows crystallization during release by polarized light microscopy (PLM) and powder X-ray diffraction (PXRD). The study is a factorial screen of in vitro release; it does not measure nucleation rates, in vivo exposure, taste or storage stability.

Background concepts

Amorphous and crystalline solubility. The crystalline solubility is the equilibrium concentration of drug over its crystals. The amorphous solubility is higher: it is the highest concentration the drug can reach in solution before a second, drug-rich liquid or amorphous phase separates out. The authors measure it with the method of Ilevbare and Taylor (2013, ref 39), in which a concentrated drug solution in an organic solvent is added slowly to the medium while a probe watches light scattering at a wavelength the drug does not absorb; the onset of a sharp rise in scattering marks the amorphous solubility. The ratio of the two solubilities tells how much supersaturation an ASD can offer at most. Table 2 of the paper reports it as A/C (card equation E1):

Here $S_{\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{r}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{u}\mathrm{s}}$ and $S_{\mathrm{c}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{e}}$ are the amorphous and crystalline solubility in the same medium at 37 °C, both in mg/mL, and A/C is dimensionless. In plain terms: A/C is the number of times more drug the amorphous form can hold in solution than the crystal can, which is the largest supersaturation an ASD can create before a drug-rich phase forms. For the racemate at pH 1.6, for example, 1.41 / 0.19 gives 7.4, the value printed in Table 2.

Non-sink release and percent release. In a non-sink test more drug is added than the crystalline solubility can hold, so the solution can become supersaturated and crystals can form. Here ASD powder is added to give a target PZQ concentration of either 0.667 mg/mL (printed in the Methods as 667 µg/mL) or 1.5 mg/mL, and "% release" is the measured concentration as a percentage of that target. The higher target lies inside the measured range of amorphous solubility (1.17–1.74 mg/mL across media), so at 1.5 mg/mL the polymers are tested close to the maximum supersaturation; the paper describes this target as "close to PZQ amorphous solubility". A curve that rises and then falls means drug has left solution, which the authors read as crystallization.

Two-stage (pH-shift) tests. A two-stage test mimics transit from stomach to intestine: 1 h in a gastric medium, then a concentrate converts it to an intestinal medium for a second hour. It matters most for polymers whose solubility depends on pH.

The two polymers. HPMCAS is a cellulose ester that carries acetyl and succinoyl groups (Figure 1c). The succinoyl groups are carboxylic acids: in acid they are barely ionized and the polymer does not dissolve, and the MF grade used here dissolves only from about pH 5.5–6 (refs 43, 44). HPMCAS is therefore an enteric polymer. Its glass transition temperature ($T_g$) is about 120 °C (ref 66). HPMC E5 LV is a cellulose ether with only methyl and hydroxypropyl groups (Figure 1d). It has no ionizable groups and dissolves at any pH; its $T_g$ is about 154 °C (ref 65). Both polymers are widely used crystallization inhibitors (refs 52, 53).

Biorelevant media. Simulated fluids contain bile salt and lipid at levels that mimic the gut. Bile salt and phospholipid form mixed micelles that can take up a lipophilic drug and raise its solubility (refs 45–47). They can also weaken a polymer as a crystallization inhibitor, as the posaconazole and atazanavir work cited above showed (ref 56). Table 1 of the paper gives the compositions, reproduced below.

Table A. Composition of the simulated media (from Table 1 of the paper). The amount of FeSSIF-V2 powder is printed as 9.76 g in Table 1 but as 9.60 g per 900 mL buffer, made up to 1 L, in Section 2.2.1; the card leaves it unresolved (note X3).

Racemate and enantiomer. (R)-PZQ and (S)-PZQ differ only in the arrangement at one stereocentre (Figure 1a, 1b). The racemate is a 1:1 mixture, and the racemic solid and the pure enantiomer form different crystals, as their different melting points and PXRD patterns show (Figures S11, S14). The racemate melts at 136–142 °C and (R)-PZQ at 113–115 °C (cited values, refs 61, 63), and the measured $T_g$ values are about 37 °C and about 34 °C (Figure S14). Their amorphous solubilities are similar in most media, but their crystalline solubilities differ: crystalline (R)-PZQ is more soluble than the racemate crystal in every medium in Table 2. Note by the AI writer (not discussed in the paper): a lower melting point usually goes with a higher crystalline solubility, so this ranking is what one would expect from the melting data.

Detecting crystals. Under crossed polarisers in PLM, crystals show bright interference colours (birefringence), while amorphous particles and dissolved polymer stay dark. In PXRD a crystalline solid gives sharp peaks at characteristic angles (2θ), an amorphous solid only a broad halo. Both techniques are used qualitatively in this paper.

Methods in detail

The work splits into four contributions. The first is a solubility map: amorphous and crystalline solubility of both drug forms in nine media, which fixes the ceiling and the floor for every release test. The second is the preparation of six ASDs at 30% DL by two routes, SE and HME, with checks that they are X-ray amorphous. The third is a release matrix of single- and two-stage non-sink tests at two target concentrations in aqueous, fasted-state and fed-state media. The fourth is the detection of crystals during and after release by PLM and PXRD. The flow below shows how the parts connect.

Contribution 1: A solubility map that sets the ceiling for every release test

① Problem. Release from an ASD cannot be interpreted without knowing how much drug the medium can hold. The crystalline solubility is the floor to which a supersaturated solution falls if crystals form; the amorphous solubility is the ceiling above which a drug-rich phase separates. Both could shift with pH, with bile salt and with chirality, and each shift would change the meaning of a release curve.

② What they did. For the amorphous solubility, 50 mL of medium was held at 37 °C in a water-jacketed beaker and stirred at 400 rpm, and a 125 mg/mL DMSO stock of either drug form was pumped in at 50–200 µL/min. Light scattering at 400 nm, where PZQ does not absorb, was recorded with a 1 mm probe every 5–10 s, and the concentration at the onset of a rapid rise in scattering was taken as the amorphous solubility (Section 2.2.6). For the crystalline solubility, excess crystals were stirred at 37 °C and 300 rpm for 48 h, the solids were removed by ultracentrifugation (40 000 rpm, 37 °C, 15 min), and the supernatant was diluted and assayed by HPLC at 220 nm against a 1–300 µg/mL standard curve (Section 2.2.7). Nine media were used: buffers at pH 1.6, 3.0, 4.5, 5.0, 5.8 and 6.5, and FaSSGF, FaSSIF and FeSSIF-V2. Racemate and enantiomer were compared by F-test and unpaired t-test at the 95% confidence level (Section 2.2.10). The values carry standard deviations, but the number of replicates behind them is not printed.

③ Why it works. The scattering onset marks the point at which the solution can no longer hold more drug as a single phase, so it measures the ceiling directly in each medium. Measuring the crystalline solubility in the same nine media gives the A/C ratio, the maximum supersaturation, medium by medium. The six buffers spanning pH 1.6 to 6.5 test ionization, the three biorelevant media test micellar solubilization at rising bile salt levels, and the paired drug forms test chirality. If the ceiling hardly moves between two media, a difference in release between them must come from the polymer or from crystallization rather than from solubility.

④ How it differs from earlier work. The aqueous values agree with earlier reports for racemic and enantiopure PZQ (refs 28, 40, 41; Section 3.1). What is new is the complete grid: both drug forms, both solid states and nine media including fed-state fluid, all at 37 °C.

Contribution 2: Six ASDs made by solvent evaporation and hot-melt extrusion

① Problem. A fair comparison needs ASDs of the same DL, particle size and amorphous state. To test whether a solvent-free, continuous route could serve a product, the compositions also had to be made by HME, which works only inside a temperature window: hot enough for the drug to melt or dissolve in the polymer, not so hot that it degrades. Because (R)-PZQ melts more than 20 °C below the racemate, the window is not the same for the two forms.

② What they did. For SE, drug and polymer were dissolved in dichloromethane–methanol 1:1 (v/v) at 30% DL, dried on a rotary evaporator with a 50 °C water bath and then overnight in a vacuum oven at room temperature (Section 2.2.2). Both drugs were made with both polymers, giving four SE ASDs. For HME, a 5 mL Xplore PME extruder with conveying screws ran at 20 rpm, at 150 ± 1 °C for the racemate and 120 ± 1 °C for (R)-PZQ, with HPMCAS-MF only (Section 2.2.3). All products were cryomilled and passed through a 60 mesh (250 µm) sieve. HPMC was not extruded because its SE ASDs crystallized more during release, and (R)-PZQ material for extrusion was limited (Section 4.4). The raw drugs were examined by a DSC heat–cool–reheat cycle (cooling to −75 °C at 20 °C/min, heating at 10 °C/min), and all starting materials, physical mixtures and ASDs by PXRD from 4° to 40° 2θ. The authors judge the extrusion temperatures against a rule from the HME literature (ref 64), which Section 4.4 states in words (card equation E2):

Here $T_{\mathrm{H}\mathrm{M}\mathrm{E}}$ is the processing temperature, $T_{m,\mathrm{d}\mathrm{r}\mathrm{u}\mathrm{g}}$ the melting point of the drug and $T_{g,\mathrm{p}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{m}\mathrm{e}\mathrm{r}}$ the glass transition temperature of the polymer, all in °C. In plain terms: the barrel should be above the drug's melting point and 15 to 60 °C above the polymer's glass transition, and the second condition can be relaxed if the drug plasticizes the polymer (refs 67–69).

③ Why it works. Racemic PZQ is thermally stable at its melting point and decomposes significantly only above 200 °C (ref 62). For (R)-PZQ the authors assume it "should likely" decompose at a similar temperature; this was not tested. The drug $T_g$ values, about 37 °C and about 34 °C, lie far below that of HPMCAS (about 120 °C), so in the authors' words "plasticization of both polymers would occur" and X-ray amorphous extrudates "could be prepared at reasonable temperatures". The arithmetic of the window bears this out for the racemate: 150 °C is above its melting range of 136–142 °C and 30 °C above the HPMCAS $T_g$, inside the 15–60 °C band. For (R)-PZQ, 120 °C is above its melting range of 113–115 °C but equal to the HPMCAS $T_g$, which is outside the band; the authors cover this case with the plasticization argument, without measuring the $T_g$ of the blends. All six ASDs came out X-ray amorphous (Figure S11). One small discrepancy concerns the melting point itself: the DSC endotherm of (R)-PZQ in Figure S14 peaks at about 111 °C, slightly below the cited 113–115 °C, and no measured melting point is printed in the text (card note X19).

④ How it differs from earlier work. Earlier PZQ ASDs were each made by one route (refs 24–34). Here the same composition is made by SE and by HME and tested in the same media, so the process effect can be read directly, and the active enantiomer is extruded as well as the racemate.

Contribution 3: A release matrix of single- and two-stage tests in three families of media

① Problem. An enteric polymer such as HPMCAS releases little in the stomach, so a single-stage test at intestinal pH may overstate how a product behaves after a gastric stage. Gastric pH varies between people and with food, so one acid start is not enough. Bile salts can raise solubility but can also weaken a polymer as an inhibitor, so buffer results may not carry over to fasted or fed fluid.

② What they did. All tests used a USP II paddle apparatus (Hanson Vision G2 Classic 6) at 37 °C and 100 rpm for 2 h, in triplicate, with a final volume of 50 mL (Section 2.2.8). Drug concentration was followed in situ by fibre-optic UV (Pion Rainbow; 5 mm probe in aqueous and fasted media, 2 mm in fed media) with second-derivative baseline correction. Single-stage tests used 50 mL of PBS pH 6.5, FaSSIF or FeSSIF-V2. Two-stage aqueous tests ran 1 h in 45 mL of HCl solution at pH 1.6, 47 mL of PBS at pH 3.0 or 48 mL of acetate buffer at pH 5.0, after which concentrated PBS at pH 7.3 brought the volume to 50 mL at pH 6.5. The fasted protocol ran 1 h in 45 mL of FaSSGF and then added 5 mL of tenfold-concentrated FaSSIF. The fed protocol ran 1 h in 45 mL of acetate buffer at pH 4.5 and then added 5 mL of tenfold-concentrated FeSSIF-V2 at pH 12, giving FeSSIF-V2 at pH 5.8. Each protocol was run at targets of 0.667 and 1.5 mg/mL; the (R)-PZQ extrudate went only through single-stage PBS pH 6.5 and FeSSIF-V2 at 1.5 mg/mL. The paper does not say whether first-stage percentages are calculated on the 45–48 mL gastric volume or on the final 50 mL (card note X16).

③ Why it works. In situ UV gives a continuous concentration–time curve without sampling or filtration, so a rise followed by a fall, the signature of crystallization from supersaturation, is visible as it happens. Starting at three acid pH values tests how sensitive the enteric hold-back is to gastric pH. The two targets separate release well below the amorphous solubility (0.667 mg/mL, roughly half of it or less) from release close to it (1.5 mg/mL), where the polymer is tested hardest.

④ How it differs from earlier work. Earlier PZQ ASD studies tested dissolution mostly in simple aqueous media (Section 1). This study adds three biorelevant protocols, several gastric pH starts and two targets, for both drug forms. The authors state that, "to the best of our knowledge", no earlier study had tested PZQ ASDs in biorelevant media.

Contribution 4: Finding crystals during and after release

① Problem. A falling release curve suggests crystallization but does not prove it, and particles left over in an HPMCAS test could be crystals or undissolved ASD. The paper needs independent evidence of crystals, and controls showing that the polymer alone does not produce the same signal.

② What they did. Aliquots were taken at 5, 60 and 120 min during release at the 1.5 mg/mL target only and imaged on a Nikon Eclipse E600 polarized light microscope with a 20× objective (Section 2.2.5). Neat polymers were run through the same protocols as controls. HPMC ASDs were not imaged in FaSSGF or FaSSIF, because they had already shown crystals in aqueous media (Section 3.3). After release, solutions were filtered through 0.45 µm filters and the residues analysed by PXRD (Section 3.4.2). HPMCAS residues from aqueous and fasted media were not X-rayed, because those samples showed neither birefringence nor a concentration drop.

③ Why it works. Birefringence separates crystals from amorphous particles and dissolved polymer, and the neat-polymer controls rule out polymer artefacts in the images. PXRD peaks at the angles of the neat drug identify the crystals as PZQ. Where all three are available, each claim of crystallization rests on three independent signals: a concentration drop, birefringent particles that the polymer control does not show, and drug peaks in the residue.

④ How it differs from earlier work. The earlier study measured nucleation induction times of PZQ with the polymers (ref 42). Here crystallization is followed inside the release test, in each medium. In exchange, no induction times or growth rates are measured, so how fast crystals form is inferred from the shape of the release curves.

Experimental analysis

Setup in brief. There are six ASDs, all at 30% DL, named drug–polymer–process: (R,S)-PZQ-HPMCAS SE, (R)-PZQ-HPMCAS SE, (R,S)-PZQ-HPMC SE, (R)-PZQ-HPMC SE, (R,S)-PZQ-HPMCAS HME and (R)-PZQ-HPMCAS HME. Release values are means of triplicates as a percentage of the target. Values at 0.667 mg/mL, the fed-state values and the pH 3.0 and 5.0 shift values are read from Figures 2, 3, 4 and S1 and are approximate. One gap in the source matters throughout: the captions of Figures 2 and 3 define panels C and D for the 1.5 mg/mL tests in aqueous and fasted media, and the text cites them, but only panels A and B are printed. Those results exist only as values in the text (card note X1).

Main result 1: the solubility ceiling hardly depends on pH, bile salt or chirality. Table B reproduces Table 2 of the paper.

Table B. Amorphous and crystalline solubility of PZQ at 37 °C in mg/mL (mean ± SD) and their ratio A/C (from Table 2 of the paper).

Across the six buffers the amorphous solubility stays within 1.17–1.43 mg/mL for the racemate and 1.20–1.40 mg/mL for (R)-PZQ. With a calculated (not measured) pKa of 1.14, PZQ is neutral from about pH 3 upward (Figure S13), which in the authors' words "suggests limited ionization-dependent solubility changes". The biorelevant media raise the values only a little; they are highest in FeSSIF-V2 (amorphous 1.74 and 1.70 mg/mL), "probably" because of its 10 mM sodium taurocholate, and the authors read the modest gain as a sign that PZQ "has limited interaction with micellar components". From this they suggest that the known positive food effect of PZQ "could depend on factors other than an increase in drug solubility" (Section 4.1).

Chirality matters little for the ceiling: by t-test the amorphous solubilities differ significantly only in PBS pH 6.5 (1.17 versus 1.39 mg/mL). Crystalline (R)-PZQ is more soluble than the racemate in every medium, so its A/C is lower (4.0–5.8 against 5.6–7.7). Two printed statements do not match the table. The discussion speaks of amorphous solubility "approximately 8- and 6-fold" above the crystalline value, but the highest tabulated A/C values are 7.7 and 5.8, so "6-fold" exceeds every (R)-PZQ entry. The ranges in Section 3.1 (amorphous 1.17–1.54 and 1.20–1.67 mg/mL, crystalline racemate from 0.17 mg/mL) also differ from the table (upper values 1.74 and 1.70, lower value 0.16). The card uses the table values (note X4).

Main result 2: HPMCAS holds the drug back in acid; HPMC releases at any pH. Table C summarizes the tests at 0.667 mg/mL, where all curves are printed.

Table C. Release (% of target) at 0.667 mg/mL. Ranges span the three HPMCAS ASDs ((R,S)- and (R)-PZQ SE, (R,S)-PZQ HME) and the two HPMC SE ASDs (read from Figures 2A–B, 3A–B, 4A–B and S1A–B, approximate). In two-stage tests the 60 min value is the end of the gastric stage, whose volume basis is not stated.

In every two-stage test the HPMCAS ASDs release little in the first hour (the text gives "no more than 30%" from pH 1.6 and at most 25% from pH 3.0 or 5.0), while the HPMC ASDs release almost all their drug in the gastric medium whatever its pH. The authors explain the hold-back "due to the low solubility of HPMCAS MF" below about pH 6.0, through the "minimal ionization of succinoyl groups" (Sections 3.2.1, 4.2); polymer dissolution itself was not measured. After the switch, HPMCAS catches up to about 92–103% when the test started at pH 1.6 or in FaSSGF, but only to about 62–77% when it started at pH 3.0 or 5.0. The paper does not discuss this difference.

In single-stage PBS pH 6.5 the HPMCAS ASDs were complete within 40–100 min, while the HPMC ASDs ended at about 82–84%; (R)-PZQ-HPMC SE peaked at 86% at about 25 min and lost 1% by 120 min, "likely due to drug crystallization". In FaSSIF all ASDs released about 93–102% within about 30 min and stayed level, with "no relevant differences" from PBS pH 6.5 except a slight gain for HPMC. In FeSSIF-V2 every ASD released less (result 4).

Main result 3: at the high target, HPMC loses dissolved drug and HPMCAS does not, except in fed fluid. Table D gathers the 1.5 mg/mL results per formulation.

Table D. Release (% of target) at 1.5 mg/mL. Rows marked "text" come from Sections 3.2.1–3.2.2 and Section 4.3 because Figures 2C, 2D, 3C and 3D are not printed; the other rows are read from Figures S1C–D and 4C–D (approximate).

The contrast between the polymers is clearest here. Outside FeSSIF-V2, every HPMC curve at 1.5 mg/mL reaches a maximum of about 65–82% and then levels off or falls; no HPMCAS curve falls outside the fed-state media. Microscopy and diffraction back up the falls in aqueous media. PLM of (R)-PZQ-HPMC SE shows few particles at 5 min and birefringent needles and aggregates at 60 and 120 min in all four aqueous protocols (Figure 5). For (R,S)-PZQ-HPMC SE the text reports rectangular birefringent particles after 60 min, and the images show interference colours at 120 min (Figure S9), while neat HPMC gives only occasional particles without interference colours (Figure S8). PXRD of the residues shows weak racemate peaks at 7.8°, 15.0°, 19.8° and 25.2° 2θ (Figure 7A) and clear (R)-PZQ peaks at 8.3°, 14.9°, 17.4°, 19.6°, 21.8°, 24.3° and 27.3° 2θ (Figure 7B). No HPMCAS sample in any medium showed birefringence (Figures S2–S7).

The authors conclude that HPMC E5 LV was "not sufficiently effective in preventing PZQ crystallization", and that the absence of a concentration decrease with HPMCAS is "suggesting" that it inhibits crystallization better, "consistent with HPMCAS MF being a better nucleation inhibitor" (Section 4.2). For the fasted media the evidence is thinner: HPMC ASDs were not imaged in FaSSGF or FaSSIF, and no fasted-state residues are shown, so crystallization there rests on the release drops alone (card note X9). Release at 1.5 mg/mL was higher in FaSSIF than in PBS pH 6.5, most clearly for the racemate with HPMCAS (more than 85% against 71%), which the authors relate "likely" to the slightly higher solubility in FaSSIF (1.49 against 1.17 mg/mL for the racemate).

Main result 4: fed-state fluid defeats both polymers. FeSSIF-V2 has the highest amorphous solubility of all media, yet every ASD released less there than in the other single-stage media: about 68–80% at 0.667 mg/mL and about 34–52% at 1.5 mg/mL in single-stage tests (Figure 4A, 4C). In the fed two-stage test, the HPMC ASDs released about 103% (0.667 mg/mL) or about 69–74% (1.5 mg/mL) in acetate buffer at pH 4.5 and then fell after the switch, to about 79–83% or about 30–36%. The HPMCAS ASDs rose only to about 60–63% or about 26–30% (the text gives approximately 64% and 32%).

With HPMC, PLM shows birefringent crystals from 5 min in single-stage FeSSIF-V2 (Figure 6) and after the switch in the two-stage test (Figure S10), and PXRD shows clear PZQ peaks for both drug forms (Figure 7C). The text reports no crystals during the pH 4.5 stage, but the 60 min panel of (R)-PZQ-HPMC SE in Figure S10 shows birefringent aggregates; the paper does not say whether that aliquot was taken before or after the medium change at 60 min (card note X8).

With HPMCAS, no particle was birefringent, and the authors take the dark particles left in the fed two-stage tests to "appear to consist of undissolved ASD" (Section 3.3). PXRD shows small peaks in the two-stage residues and only halos in the single-stage SE residues (Figure 7D); among the extrudates, the (R)-PZQ residue after FeSSIF-V2 shows sharp peaks and the racemate residue almost none (Figure S12). The authors conclude that the higher bile salt and lecithin content of FeSSIF-V2 "promoted crystallization", which HPMCAS could not inhibit as well as in buffer, "in broad agreement" with the posaconazole and atazanavir results (ref 56). Their discussion says this was observed "through PLM and PXRD analyses" for both polymers; for HPMCAS the evidence is PXRD only (card note X7).

Main result 5: (R)-PZQ gains over the racemate only in simple buffer, and crystallizes faster. With HPMCAS, (R)-PZQ outperformed the racemate in two conditions named by the authors, both at 1.5 mg/mL: single-stage PBS pH 6.5 (approximately 88% against 71% for SE and 78% for HME) and the pH 1.6 to 6.5 shift (91% against 73% and 69%). The authors write that this "is likely attributable to the slightly higher amorphous solubility of (R)-PZQ in PBS pH 6.5" (1.39 against 1.17 mg/mL), and the maximum concentrations reached, approximately 1.4 and 1.1 mg/mL, sit close to those two solubilities. The paper cites "Table 1" for these solubility data; they are in Table 2 (card note X5). In fasted media the two forms behaved alike, which the authors put down to bile salts raising PZQ solubility.

One condition does not fit the authors' list: after the pH 3.0 shift at 1.5 mg/mL, (R)-PZQ-HPMCAS SE also ended above the racemic SE ASD (about 58% against about 46%, with wide error bars), a gap of similar size (card note X13). Note by the AI writer (not discussed in the paper): all aqueous two-stage tests end in PBS pH 6.5, so the solubility explanation predicts an (R)-PZQ advantage in each of them; the pH 3.0 data fit that prediction, and the pH 5.0 data show a smaller gap (about 58% against about 51%).

With HPMC the picture reverses. (R)-PZQ-HPMC released as fast as or faster than the racemate, but at 1.5 mg/mL it lost dissolved drug in every medium, whereas the racemate lost it only in fasted and fed media (Table D). The authors take this to "indicate" that (R)-PZQ "has a higher tendency to crystallize from highly supersaturated solutions", in line with its shorter induction time in the earlier study (ref 42). Their reasons are offered as possibilities: identical stereochemistry "can" give more efficient packing and faster nucleation, the counter-enantiomer in the racemate "can act as an 'impurity'", bile salts "may" have added to the effect in FaSSIF, and the residues are named as (R,S)-PZQ anhydrous Form A and (R)-PZQ hemihydrate. In FeSSIF-V2 the authors report "no significant difference" between the forms with either polymer, but no statistical test was applied to release, and the HPMC pair ends at about 34% for (R)-PZQ against about 48% for the racemate (card note X11).

Main result 6: the two processes give broadly similar release. For the racemate with HPMCAS, SE and HME end close together in PBS pH 6.5 (about 101% against about 104%) and FaSSIF (about 102% against about 101%) at 0.667 mg/mL, in the FaSSGF to FaSSIF test (about 99% against about 98%), and in PBS pH 6.5 at 1.5 mg/mL (71% against 78%). The (R)-PZQ extrudate released more slowly than its SE counterpart in PBS pH 6.5 but reached the same approximately 88%.

Differences appear elsewhere: after the pH 3.0 shift at 0.667 mg/mL the racemic SE ASD ends at about 77% and the HME ASD at about 62% (first hour about 23% against about 9%); after the pH 5.0 shift at 1.5 mg/mL, about 51% against about 43%; and in the first hour at pH 4.5, about 22% against about 12%. In FeSSIF-V2 at 1.5 mg/mL the extrudates end slightly lower (about 42% and 44% against 47% and 52%), which the authors say "can be attributed to different extents of drug crystallization" seen by PXRD. Their conclusion is that the process had "no discernable impact" on release. The data support that for single-stage tests at intestinal pH and for the fasted protocol, but not for every pH-shift test, where gaps reach about 15 points (card note X12).

Controls and what they rule out. Three controls strengthen the crystallization evidence. The as-made ASDs show only halos (Figure S11), so crystals found after release formed during the test rather than surviving preparation. The physical mixtures show drug peaks in the presence of polymer, so PXRD can detect crystalline PZQ in these matrices. The neat-polymer PLM controls show no interference colours, so birefringence in the ASD samples comes from the drug.

Some things are not controlled. Polymer dissolution at each pH was not measured, so the enteric explanation of the gastric hold-back is inferred from the known properties of HPMCAS. The racemate was 95%+ pure and the enantiomer 99.9% pure, so every chirality comparison also carries a purity difference of up to about 5% (card note X17). And a small peak at 31.5° 2θ in residues from the pH 5.0 runs, which the authors assign to NaOH without a reference pattern, coincides with a small sharp peak of neat HPMC at about 31.5° (card note X15).

Discussion

Where the findings apply. The results hold for 30% DL powders below 250 µm, one grade each of HPMCAS and HPMC, a 50 mL USP II test for 2 h and two targets, all in vitro. Within that frame the polymer effect is robust: it appears in every protocol, at both targets, and for HPMC in aqueous and fed media it is backed by three kinds of evidence. The chirality and process effects are smaller, depend on the medium, and rest partly on text values without curves or statistics. The authors' account of how the outcome is decided can be drawn as one chain.

What remains unresolved. Every rate in this chain is inferred from release curves: no induction times or growth rates were measured, with or without bile salts, so the mechanisms behind the HPMC failure, the (R)-PZQ effect and the FeSSIF-V2 effect rest on curve shapes and cited work. The crystal forms are not identified; the HPMC residues from FeSSIF-V2 carry strong peaks at about 6.2° and 8.2° 2θ that are not prominent in the neat racemate (Figure 7C), so the fed medium may produce a different solid (card note X14). Why HPMCAS ASDs started at pH 3.0 or 5.0 recover less after the switch than those started at pH 1.6 is not discussed. The authors themselves name one open question: "Further studies are needed to understand the stability of supersaturated PZQ solutions in fed-state simulated media" (Section 4.2). The source of the positive food effect, which the authors think "could depend on factors other than" solubility, is also left open.

What it suggests for formulation. The authors conclude that (R)-PZQ-HPMCAS "appears to be a potential formulation" to replace the racemic product, avoiding the side effects and bitter taste of (S)-PZQ and lowering the pill burden, and that HME is "a promising technique" for PZQ ASDs. The AI writer's reading of the data adds four points. First, the case for (R)-PZQ rests on halving the dose and removing the bitter enantiomer, not on better release: its in vitro gain is small and appears in simple buffer only. Second, because (R)-PZQ lost dissolved drug faster with HPMC, it needs a strong inhibitor even more than the racemate does. Third, HPMCAS trades gastric release for stability at intestinal pH, and the lower recovery after the pH 3.0 and 5.0 starts suggests that a less acidic stomach could slow release further. Fourth, a product to be taken with food would need a stronger inhibitor or another excipient. The obvious next experiments are induction times of both forms with each polymer in buffer and in FeSSIF-V2, identification of the residue crystal forms, and in vivo and storage data for the (R)-PZQ ASDs.

Limitations

Limitations stated by the authors. The paper has no limitations section, and the authors state few limits. They treat PXRD as qualitative only, explaining that intensities "should not be attributed only to crystalline concentration" and that some peak shifts come from "sample non-uniformity in the aluminum cavity holder" (Section 3.4.2). They tested the (R)-PZQ extrudate in only two conditions because material was limited, and did not extrude HPMC (Section 4.4). They call for further study of supersaturated PZQ in fed-state media (Section 4.2), and they make the dose-reduction argument conditional: "If bioavailability enhancement could be achieved" (Section 4.3).

The AI writer's own judgement. The card rates the paper as moderately solid overall. The polymer contrast is well supported, but several points weaken the chirality and process claims. (1) The 1.5 mg/mL curves in aqueous and fasted media are not printed (Figures 2C, 2D, 3C, 3D), so the main chirality result in PBS pH 6.5 exists only as text values (X1). (2) No statistical test was applied to any release comparison; the t-test covers solubility only, so "no significant difference" in FeSSIF-V2 is qualitative, and it sits beside an HPMC pair at about 34% against about 48% (X11). (3) The (R)-PZQ advantage with HPMCAS is not confined to the two conditions the authors name, since the pH 3.0 shift shows a gap of similar size (X13). (4) The drugs differ in purity, 95%+ against 99.9%, so each chirality comparison is confounded with impurity content, which is relevant to nucleation (X17).

(5) HPMC crystallization in fasted media rests on release drops only, and HPMCAS crystallization in FeSSIF-V2 on PXRD only (X9, X7, X10). (6) The named crystal forms are not identified by the data shown (X14). (7) "No discernable impact" of process holds only in some media (X12), and the statement that HPMCAS ASDs release at least as much as HPMC ASDs in single-stage FeSSIF-V2 fails for the racemate at 0.667 mg/mL, where the HPMCAS ASDs end at about 70% and 68% against about 80% for HPMC (X18). (8) The volume basis of first-stage release is not stated, and several values exceed 100% (X16). Reporting slips that do not change the conclusions: the Figure 2B legend labels the green (R)-PZQ-HPMCAS series as HPMC (X2), and Section 3.2.2 prints the low target once as 0.667 µg/mL instead of mg/mL (X6).

Conclusion

This paper screened six 30% DL praziquantel ASDs through single- and two-stage release in aqueous, fasted-state and fed-state media. The amorphous solubility, about 1.2–1.7 mg/mL, barely moves with pH, bile salt or chirality, so the differences in release come mainly from the polymer. HPMCAS-MF holds release back in acid but keeps PZQ dissolved at pH 6.5 and in FaSSIF; HPMC releases at any pH but lets PZQ crystallize at the high target, and (R)-PZQ more readily than the racemate. FeSSIF-V2 caused crystallization and poor release with both polymers. (R)-PZQ-HPMCAS beat the racemate only in simple pH 6.5 media at the high target, which the authors link to its slightly higher amorphous solubility there, and SE and HME gave broadly similar release.

The authors' recommendation, (R)-PZQ-HPMCAS as a candidate to replace the racemic product, is worded as a possibility ("appears to be a potential formulation"), and the evidence behind it is in vitro release alone. Its value lies in the dose and taste benefits of removing (S)-PZQ, provided a strong inhibitor holds the faster-crystallizing enantiomer in solution, and provided the fed-state failure can be solved.