| Popis: |
Hidrofilne ogrodne tablete na osnovi hipromeloze predstavljajo enega izmed najpogosteje uporabljenih dostavnih sistemov za doseganje podaljšanega sproščanja ZU. Bolniki izkazujejo boljšo sprejemljivost jemanja zdravil s podaljšanim sproščanjem na račun manj pogostih odmerjanj. Ključni kritični atribut kakovosti (CQA) tablet s podaljšanim sproščanjem je profil sproščanja, saj je neposredno povezan z varnostjo, kakovostjo in učinkovitostjo zdravil. Zaradi tega moramo imeti v proizvodnji in kontoli kakovosti kontrolno strategijo za dokazovanje ustreznega sproščanja na vsaki izdelani seriji tablet kot končni specifikacijski parameter. Vpliv na sproščanje moramo preučiti že v fazi razvoja zdravila s pristopom načrtovanja kakovosti (QbD), kjer moramo najprej identificirati kritične atribute materialov (CMA) za razvoj formulacije ter kritične procesne parametre (CPP) pri razvoju tehnološkega postopka. Identifikacijo CMA in CPP ter njihov vpliv preverimo z ustreznim sistematičnim pristopom načrtovanja eksperimentov (DoE) ter primerno statistično obdelavo podatkov. Na podlagi teh rezultatov določimo tudi specifikacijske meje za CMA ter tehnološke meje za CPP, da bomo v proizvodnem merilu lahko proizvajali izdelek, ki bo ustrezal končni specifikaciji in zahtevam za varnost, učinkovitost in kakovost. Vzporedno moramo razviti tudi primerno kontrolno strategijo (analitsko metodo), s katero bomo preverjali ustreznost sproščanja ZU iz hidrofilnih ogrodnih tablet. Pri pregledu literature smo ugotovili, da je znanstvenih objav o ogrodnih tabletah na osnovi hipromeloze veliko. Pri vseh teh številnih objavah pa smo pogrešali celovit in sistematičen pristop določevanja vplivov fizikalno-kemijskih lastnosti hipromeloze, izbire tehnološkega postopka ter procesnih parametrov na profil sproščanja. Zato je bil naš namen doktorske disertacije ravno to – preveriti vse naštete vplive in jih povezati v zaključeno celoto. Pri tem smo v cwlotnwm eksperimentalnem delu uporabili enako sestavo tablet, ki je vključevala hipromelozo tipa 2208 (HPMC K4M), ter karvedilol kot slabo topno modelno ZU, zato se ves nadaljnji tekst nanaša izključno na našo izbrano formulacijo. V našem eksperimentalnem delu smo se najprej osredotočili na metodo direktnega tabletiranja (DC), kjer smo preverili, kako fizikalno-kemijske lastnosti oziroma s funkcionalnostjo povezane lastnosti (FRC) hipromeloze vplivajo na sproščanje karvedilola. Preučevali smo naslednje FRC parametre hipromeloz: velikost delcev, stopnja substitucije in viskoznost. Uporabili smo osem različnih serij hipormeloz, ki so tvorile ustrezno velik eksperimentalni prostor študij za FRC parametrov. S primerjavo najhitrejšega in najpočasnejšega profila smo dokazali, da sta si profila signifikantno različna (faktor primerljivosti – f2 = 46.7). Rezultate sproščanj smo statistično obdelali, tako da smo za vsako časovno točko kreirali dva modela – enega z metodo multiple linearne regresije (MLR) in enega z metodo delnih najmanjših kvadratov (PLSR). Tako MLR kot PLSR modeli imajo dobro napovedno moč, kar smo dokazali z več statističnimi metodami, hkrati pa smo ujemanje modelov validirali z dodatnimi tremi serijami hipromeloze, ki niso bile vključene v kreiranje modelov. Pri primerjavi rezultatov generiranih modelov se je izkazalo, da so PLSR modeli bolj primerni od MLR modelov. V prvem delu profila sproščanja (do 5. ure) ima dominanti vpliv na sproščanje velikost delcev hipromeloze. Večji kot so delci, hitrejše je sproščanje, kar si razlagamo z manjšo kontaktno površino večjih delcev in s tem počasnejšim nabrekanjem ter posledično hitrejšo difuzijo ZU iz tablete. Stopnja hidroksipropoksi substitucije (%HP) postane signifikantno pomemben parameter (višje vrednosti %HP zvišajo profil sproščanja) po približno 5. uri, ko dominantni vpliv velikosti delcev upade in ima do zaključka sproščanja največji vpliv na profil sproščanja. To podpirajo tudi rezultati za navidezno viskoznost, ki postane pomembna pri približno 7. urah sproščanja. Višje vrednosti navidezne viskoznosti pričakovano znižajo hitrost sproščanja. Hidrofilne ogrodne tablete lahko izdelamo s pomočjo različnih metod aglomeracije. Da bi preučili vpliv različnih metod aglomeracij na profil sproščanja, smo v naslednjem sklopu eksperimentov smo izdelali tablete z metodama suhega granuliranja t.j. briketiranje (DGS) in kompaktiranje (DGRC), z metodama vlažnega granuliranja t.j. granuliranje v hitrovrtečem mešalniku (HS) in vrtinčnoslojnem granulatorju (FBD) ter direktnega tabletiranja (DC). Vrednotili smo naslednje CQA-je zmesi za tabletiranje: pretočni čas, nasipni kot, nasipni in stresalni volumen, Carr-ov inedeks, Huasnerjevo razmerje, distribucija velikosti delcev (PSD), kompresibilnost (Heckel-ov model), kompaktibilnost ter naslednje CQA-je tablet: poroznost, profil sproščanja in analiza teksture gela. Zmesi za tabletiranje, izdelane z metodama vlažnega granuliranja, izkazujejo boljše pretočne in stisljivostne lastnosti kot zmesi za tabletiranje, izdelane z metodama suhega granuliranja. Od proizvodnih metod je najmanj primerna metoda DC, saj ima zmes za tabletiranje, izdelana z DC, najslabše pretočne lastnosti in prav tako najslabše kompresibilnostne ter kompaktibilnsotne lastnosti. Izbira metode granuliranja pomembno vpliva na tvorbo gelskega sloja in profil sproščanja iz končnih hidorfilnih ogrodnih tablet, kar posledično vodi tudi v signifikantno različne profile sproščanja med HS in DC (f2 = 44.8) ter med HS in DGRC (f2 = 48.3). Izbira metode granuliranja ima največji vpliv na profil sproščanja. Tablete, izdelane z metodo DC, izkazujejo najhitrejši profil sproščanja. Sledijo jim tablete, izdelane s suhim granuliranjem, in nato še tablete, izdelane z vlažnim granuliranjem. Ugotovili smo, da sta PSD granulata in poroznost tablet sta pomembna faktorja, ki vplivata na razlike v profilih sproščanja. Nazadnje smo se osredotočili na proces vlažnega granuliranja. S spreminjanjem procesnih parametrov granuliranja in tabletiranja (količina vode kot granulirne tekočine in tip šobe pri granuliranju ter trdnost tablet pri tabletiranja) smo preučevali njihov vpliv na profil sproščanja karvedilola. Izvedli smo eksperimentalni rezultate statistično ovrednotili rezultate s PLS modeliranjem. V eksperimentalni načrt zadnjega dela naloge smo vključili še vpliv izbranih fizikalno-kemijskih lastnosti hipromelozet.j. stopnje substitucije, navidezne viskoznosti in porazdelitve velikosti delcev hipromeloze. Dodatno smo kot neodvisno spremenljivko v eksperimentalni načrt vključili tudi porazdelitev velikosti granul. Rezultati PLS analize kažejo, da imajo količina vode in tip šobe pri granuliranju ter trdnost tablet večji vpliv na sproščanje kot fizikalno-kemijske lastnosti hipromeloze. Pri primerjavi profilov sproščanja smo tako dobili tudi signifikantno razlikovanje (najnižji f2 =41.1), kadar smo uporabili različne nivoje faktorjev procesnih parametrov ob isti vgrajeni seriji hipromeloze. Tip šobe in količina vode vplivata na gostoto in poroznost granul, kar se odrazi tudi na porazdelitvi velikosti granul. S PLS modeliranjem smo ugotovili, da večji delež frakcije granul z velikostjo nad 500 μm upočasni hitrost profila sproščanja, večji delež frakcije granul z velikostjo pod 500 μm pa hitrost profila sproščanja zviša. Ugotovili smo torej, da preko izvedbe procesa granuliranja vplivamo na lastnost granul, te pa imajo nato nadaljnji vpliv na profil sproščanja. Trdnost tablet je prav tako pomemben faktor, ki vpliva na sproščanje karvedilola. Rezultati eksperimentalnega dela doktorske disertacije prikazujejo celovit in sistematičen pristop k razumevanju vplivov fizikalno-kemijskih lastnosti hipromeloze, izbire tehnološkega postopka in procesnih parametrov na profil sproščanja hidrofilnih ogrodnih tablet na osnovi hipromeloze. Kot inovativen način preverjanja vplivov smo uporabili multivariantne statistične metode, kjer smo za vsako posamezno časovno točko kreirali svoj model in tako dobili vpogled, kako se ti vplivi dinamično spreminjajo skozi celotno časovno skalo profila sproščanja. To nam omogoča boljše razumevanje procesov v hidrofilnih ogrodnih tabletah. Matrix tablets based on hypromellose represent one of the most versatile delivery systems for achieving sustained release of active pharmaceutical ingredient (API). Patients express better compliance when using sustained release tablets due to less frequent taking medicines. A key critical quality attribute (CQA) for sustained release tablet is drug dissolution profile because it is directly related to safety, efficiency and quality of medicine. Therefore an appropriate control strategy to proof suitability of dissolution as a test in control of release specification on every produced tablet batch is mandatory. The influence of different factors on dissolution rate must be already studied at the development phase with quality by design (QbD) approach. At first critical material attributes (CMA) should be identified during the formulation development and after that critical process parameters (CPP) for technological process development. Identification of CMA and CPP and their influence should be evaluated with design of experiment (DoE) approach, including appropriate statistical evaluation. Based on this results specification limits for CMA and technological limits for CPP is set in order to manufacture a product in a large production scale that comply to release specification limits. In parallel a suitable control strategy (analytical method) should be developed to check appropriate drug dissolution. There is a plenty of scientific publications on matrix tablets based on hypromellose. Nevertheless a comprehensive and systematic approach for determination of influence of hypromellose characteristics, selection/performance of technological process and process parameters on drug dissolution profile is missing. Hence, aim of our experimental work was exactly this – to evaluate all the influences and to connect them into entirety. For that reason the same composition of tablets were used. Formulation comprises hypromellose type 2208 (HPMC K4M) and carvedilol as model drug with low solubility. All findings in our experimental work refers on our specific formulation. The first focus of our experimantal work was direct compression (DC) method where influence of physico-chemical properties or functionality related characteristics (FRC) of hypromellose on drug dissolution profile was examined. The following FRC were studied: particle size distribution (PSD), degree of substitution and viscosity. Eight batches of hypromellose were used creating a suitably large design space for FRC. Dissolution profiles of carvedilol were constructed through entire 12-hour time period. It was demonstrated that the fastest and the slowest dissolution profile are not similar (similarity factor - f2 = 46.7). Drug dissolution results were statistically evaluated. For each time point two models were generated – one with multiple linear regression (MLR) and one with partial least square regression (PLSR). All MLR and PLSR models have good predictive power which was proved using different statistical tools. These were also additionally confirmed with validation of models using three extra hypromellose batches which were not included in models creation. When comparing results from generated models it can be seen that PLSR models are more appropriate than MLR models. In the first part of drug dissolution profile (until 5th hour) the PSD of hypromellose has dominant influence on dissolution of carvedilol. Larger particles results in faster dissolution. Larger particles have smaller contact surfaces leading to slower gel formation and quicker diffusion of dissolved drug from the tablets. Hydroxypropoxy substitution (%HP) becomes significantly important parameter approximately after 5th hour (when PSD influence subside) and have the most important role on dissolution till the end of the drug dissolution profile. This finding show that after 5th hour hydrataion of hypromellose and acting on molecular level has a dominant role. Results for apparent viscosity support this thesis because viscosity becomes important at approximately 7th hour. Higher values of apparent viscosity as expected decrease dissolution profile. Matrix tablets can be manufactured using various agglomeration processes. To evaluate the influence of different agglomeration processes on dissolution profile the next set of experiments tablets were processed using dry granulation method (i.e. compacting (DGRC) and slugging (DGS)) wet granulation method (i.e. granulation in high shear mixer (HS) and in fluid bed granulator (FBD) and also direct compression method. The following CQAs of compression mixtures were evaluated: flow time, angle of repose, bulk and settled volume, Carr index, Hauser ratio, PSD (using sieve analysis), compressibility (using Heckel plot), compactibility, and the following CQA of tablets: porosity, drug dissolution profile and gel layer texture analysis. Compression mixtures produced by wet granulation method express better flow and compressibility propreties in comparison to compression mixtures produced by dry granulation method. At least appropriate method is DC because of the worst flowing, compressibility and compactibility characteristics. The selection of agglomeration method importantly influence on gel layer formation and dissolution profile from matrix tablets, which consequently reflects also in significantly different drug release profiles between HS and DC (f2 = 44.8) and between HS and DGRC (f2 = 48.3). The biggest influence on dissolution profile has the selection of granulation method. Tablets made by DC have the fastest drug dissolution profile following by tablets mad by dry granulation method. The slowest profile express tablets produced by wet granulation method. PSD of granulate and tablet porosity are important factors that have influence on drug dissolution profile. On the last chapter wet granulation in high shear mixer was examined in detail. It was investigated how the variations in process parameters of granulation and tabletting reflects on dissolution profile of carvedilol. Design of experiments and statistical evaluation using PLS models was done. Investigated independent variables – factors were the following: quantity of water as granulation liquid and type of nozzle in granulation process and tablet hardness at tableting process. Additionally the influence of hypromellose characteristics were evaluated using two extra batches of hypromellose. Therefore in PLS models as independent variables also hypromellose characteristics were included: PSD, degree of substitution, apparent viscosity. PSD of granulate was also incorporated in PLS model creation as independent variables. PLS results demonstrate that granulation liquid quantity and type of nozzle during granulation and tablet hardness have bigger influence on dissolution profile than hypromellose characteristics. Comparing dissolution profiles, significant difference between profiles were observed (the lowest f2 =41.1) when different levels of process parameters were used but the same incorporated hypromellose batch. Type of nozzle and granulation liquid quantity influence on density and porosity of granules which also reflects on PSD of granulate. Using PLS modeling it was established that higher proportion of granules with size above 500 μm decreases dissolution profile, and higher proportion of granules with size below 500 μm increases dissolution profile. It was concluded that the granulation process parameters have an effect on granule propreties, which has further effect on dissolution profile. Tablet hardness proved to be also important factor that have influence on dissolution profile of carvedilol. Results of experimental work of doctoral thesis show comprehensive and systematic approach to understanding of hypromellose characteristics, selection/performance of technological process and process parameters on drug dissolution profile of matrix tablets based on hypromellose. Multivariate statistical methods were used as innovative manner of evaluating parameters. For each time point own model was generated and in that way we get insight how are these influences dynamically changed through entire scale of dissolution profile. The results of our experimental work will contribute to better understanding of processes in matrix tablets. |
