M ingzhu Ye,Hongling Dun,Lixi Yo,Yicheng Fng,Xioyu Zhng,Ling Dong,Feifei Yng,Xinggng Yng,?,W eisn Pn ,?

a Shenyang Pharm aceu tical University,103W enhua Road,Shenyang 110016,China

b Zhejiang University of Technology,18 Chaow ang Road,Zhejiang 310014,China

Keywords:Long acting PLGA M icrospheres Elevated tem peratu res Ko rsm eyer-Peppas equation Weibu ll fun ction

A B S T R A C TThe ob jec t of the study w as to deve lop a qu ick and rep roducib le accelerated in vitro re lease m ethod to p red ict and deduce the function of the real tim e(37°C)re lease fo r long ac ting PLGA m icrospheres.The m ethod cou ld be described in severa l steps.First,the release of them icrospheresw ere stu d ied using the sam p le and separatem ethod at 37°Cw ith no rm a l orbita l shaking and elevated tem peratu res w ith m agnetic stirring tofu rther accelerate the release.Secon d,them ost sim ilar p rof ile at elevated tem peratu resw ith the rea l tim e release w as chosen w ith the help of the n va lue in the f itted Ko rsm eyer-Peppas Fun ction.Th ird,the W eibu ll fun ction and conversion ratio w ere used to dedu ce the fun ction of rea l tim e release accord ing to the chosen p rof ile atelevated tem peratu res.The key poin t in th is study w as to p rovide a qu ick and p recise m ethod to p red ict the rea l tim e release fo r long acting p rogesterone PLGAm icrospheres.So the elevated tem peratu res coup led w ith m agnetic stirring w ere used to accelerate the release fu rther,and w hen there havem any sim ilar re lease p rof iles w ith the real tim e release at elevated tem peratu res,releasing tim e at elevated temperatu res and the R2 of the f ina l deduced fun ction w ill be used to help choosing the m ost sim ilar re lease p rof ile w ith the real tim e release.Fou r d ifferen t types of p rogesterone PLGA m icrospheres w ere used to verify them ethod,and a ll the deduced func tion co rrelated w ell w ith the real tim e releases,for R2=0.9912,0.9781,0.9918 and 0.9972,respectively.

1. Introduction

Po ly(lactic-co-glyco lic acid)(PLGA)m icrospheres as a trad itiona l sustained re lease system s have been deep ly stud ied and there a lready have been m any p rodu cts on the m arket[1-3].One of the genera lities of the p roducts is that each of them has a long du ration of action,w h ich ranges from w eeks to severa l m on ths[4-7].How ever,th is genera lity not on ly greatly im p roved patien ts'com p lian ce[8]bu t a lso b rings m any“h igh risk”[9,10],because these m icrospheres con tain a large am oun t of active d rugs,and any unexpected changes in the release w ill lead to toxicity or severe side effects[11].So,there is a great need to assu re the release p rof ile of these m icrospheres[9].How ever,studying the release p rof iles w ill be both tim e-consum ing and expensive,especia lly in the p relim inary stages of form u lation developm en t and the quality con tro l phases[12],because it w ill take a lo t of tim e tof in ish the release and any m ino r changesm ay greatly in f luen ce the re lease p rof iles.Therefo re,a qu ick and easy m ethod to p red ict the d rug release in a short tim e w ou ld be very help fu l[13,14],and th is w as a lso recom m ended by the Am erican Association of Pha rm aceu tica l Scien tists/In ternationa l Pharm aceu tica l Federation(AAPS/FIP)[15].

Itw as w ell-know n tha t there w erem any pa ram eters,su ch as tem peratu re,p resen ce of tem peratu re,p resen ce of enzym es and su rfactan ts and pH,cou ld acce lerate the rate of PLGA hyd ration and degradation or enhan ce the d rug d iffusion to acce lerate the d rug re lease[16].Stirring rate cou ld a lso in crease the re leasing rate in a w ay.So in th is study,the temperatu res coup led w ith m agnetic stirring w ere used tofu rther accelerate the releasing rate.

Neverthe less,du ring the past few years,there have been m any articles ta lk ing abou t e levated tem peratu re acce lerated re lease to p red ict the rea l tim e release.Fo r exam p le,Janagam et al.[17]and Shen et al.[18],they both use elevated tem peratu res to accelerate the m icrospheres re lease,the sam e accum u lated releases w ere ach ieved at d ifferen t tem peratu res.W hen the great co rrelation betw een the p rof ile of acce lerated release and the real tim e releasew as obtained,the accelerated re lease p rof ile cou ld be used to p red ict the rea l tim e release.How ever,how to p red ict and how to rea lly describe the rea l tim e re lease in a m o re clearly w ay w ere no tm en tioned.

A rrhen ius equation w as a lso w ide ly used to p red ict the rea l tim e release,and the equation w as k=A·exp(-Ea/(RT)),w here k is the rate constan t,A is the p re-exponen tia l facto r,Eais the energy ofactivation.R is the gas constan t,and T is the abso lu te tem peratu re.Fo r exam p le,Zo ln ik et a l.[19],in their study,the activation energy of the m icrospheres w ere ca lcu lated based on re lease data at the e leva ted tem peratu res(53°C,60°C,70°C),and the calcu lated activation energy w as used to calculate the rate constan t at rea l tim e release(37°C),they found the p red icted rate constan t w as in agreem en t w ith the rea l rate constan t at 37°C,so they announ ced that these e levated tem peratu re(53°C,60°C,70°C)can be used to p red ict the real tim e re lease.In articlesm en tioned above,one of them ost impo rtan t questions shou ld be d iscussed.The release rate of the m icrospheres cou ld no t be the sam e from the beginn ing to the end,so the rate constan t of a period cou ld not describe the re lease p rof ile in tegra lly.How ever,there are still som e articles using the sam e m ethod[12],they ca lcu lated the re lease rate separately to describe the rea l tim e re lease.In th is w ay,it seem s m ore accu rate.Bu t it still can't describe the rea l tim e re lease in tegra lly and a lso can't describe it in a clear and easy w ay.There fo re,in th is study,aim ing at the sho rtcom ings,an easy and qu ick m ethod w as developed.

Here,p rogesteronew as chosen as am ode ld rug and the trad itiona lly so lven t evapo ration m ethod w as used to m ake the p rogesterone-PLGA m icrospheres.The elevated tem peratu res coup led w ith m agnetic stirringw ere used to greatly accelerate the re leasing rate,com pared w ith the re leasing rate a t 37°C w ith o rbita l shaking.Re lease characteristics at d ifferen t cond itions w ere stud ied.And the data w ere deep ly analyzed by o rigin 8.0.The Ko rsm eyer-Peppas equation and W eibu ll fun ction w ere a lso used to describe the re lease p rof iles and to dedu ce the fun ction of the rea l tim e release.

2. M ateria ls and methods

2.1. M aterials

Po ly(d,l-lactide-co-glyco lide)w ith a lactide:g lyco lide ratio 65:35,noted as PLGA RG653H(inheren t viscosity=0.321 d l/g in CHCl3at 25°C)w as kind ly donated from Evon ik,Ltd.Po ly(d,llactide-co-glyco lide)w ith a lactide:g lyco lide ratio 65:35,no ted as PLGA 65:35(inheren t viscosity=0.386 d l/g in CHCl3at25°C),and(d,l-lactide-co-glyco lide)w ith a lactide:glyco lide ratio 75:25,no ted as PLGA 75:25(inheren t viscosity=0.401 d l/g in CHCl3at 25°C)w as pu rchased from Jinan Daigang Biom ateria lCom pany,Ltd.(Shandong,Ch ina).Progesterone(99.0%)w as pu rchased from Beierka Bio logical Med icine Co Ltd(W uhan,Ch ina).Po ly(viny l a lcoho l)(PVA,MW 30000)w as pu rchased from Ku raray Ch ina Co,Ltd.HPLC grade ethy l fo rm ate and m ethy lene ch loride w ere ob tained from Yuw ang Com pany(Shandong,Ch ina).

2.2. M ethods

2.2.1. Preparation ofm icrospheres

Trad itiona lly so lven t evapo ration m ethod w as used to m ake the p rogesterone-PLGA m icrospheres.And in th is m ethod,ethy l form ate[20,21]w as used to substitu tem ethy lene d ich loride to m ake environm en ta lly friend ly[22].

First,20m g p rogesterone and 200m g PLGA RG653H(0.321 d l/g,Evon ik)w ere d isso lved in 3.2m l ethy l fo rm ate,and then they w ere pou red in to 8m l of 1%PVA aqueous solu tion.A fter the add ition,the em u lsion w as con tinuously stirred for 220m in.A fter that,the m icrospheres w ere co llected by f iltration and w ashed w ith d istilled w ater fo r 3 tim es and d ried overn igh t un der vacuum[22].

2.2.2. Encapsu lation efficiency of the progesterone PLGA m icrospheres

Abou t 20m g of d ried m icrospheresw ere com p letely d isso lved in 3m lm ethy lene d ich loride,and then 18m lm ethano l w as pou red in to it to p recip itate PLGA.The f ina l suspension w as f iltered th rough a 0.22μm m em brane.The am oun t of p rogesterone in the f iltrate w as ana lyzed by HPLC.The Diam ond C18(4.6m m×25 cm,5μm)w as used as an ana ly tica l co lum n w ith am obile phase consisting ofm ethano l,aceton itrile and w ater(45:30:25,v/v)at a f low ra te of 1m l/m in.the UV detecto r w as set at 254 nm,and the encapsu lation eff icien cy(EE)and d rug load ing(DL)w as def ined below.

EE(%)=(Ma/Mo)×100%,w here“Ma”m eans the actua l am oun t of p rogesterone loaded in the m icrospheres,“Mo”m eans the theoretical p rogesterone loaded in the m icrospheres.

DL(%)=(n/m)×100%,w here“n”m eans the am oun t of p rogesterone in the m icrospheres,“m”m eans the w eigh t of the p rogesteronem icrospheres.

2.2.3. The characteristics ofprogesteronem icrospheres

The d iam eters of the p rogesterone-PLGA m icrospheres w ere m easu red by Sym patec Laser Particle Size Ana lyzer.The m icrospheres w ere also observed by scann ing electron m icroscope(SEM).(Hitach iS-3400;Hitach iHigh Techno logies,Kyoto,Japan).

The d ifferen tia l scann ing ca lorim eter(Sh im adzu,Japan)w as used to determ ine the glass tran sition tem peratu re(Tg)of the m icrospheres[23].The sam p les of them icrospheres w ere sea led in a lum inum pans w ith p inho le lids,and w ere heated to 150°C at 20°C/m in and rap id ly coo led to-10°C.The second cycle w as used to determ ine the glass transition tem peratu re[13,23].

2.2.5. Data analysis

The accum u lated releases of p rogesterone-PLGA m icrospheres at d ifferen t tem peratu res w ere deep ly analyzed using the Origin 8.0.In add ition,the Ko rsm eyer-Peppas equation and W eibu ll fun ction w ere used to describe the release p rof ile,and to he lp revea l the re lationsh ip betw een the rea l tim e release p rof ile and accelerated release p rof ile,so as to p red ict the rea l tim e release.

As fo r biodegradab le erosion system s,d rugs w ere d ispersed in the po lym er even ly,and the Korsm eyer-Peppas equation cou ld be used to exp lain the release m echan ism s sim p ly.Fo r Ko rsm eyer-Peppas equation[26,27]Mi/M=k·tn,w here the Mi/M w as the fraction of d rug released at tim e t,k w as the rate constan t,and n w as the release exponen t related to the releasing m echan ism.Form icrospheres,the correlations betw een the n va lue and release m echan ism s are:n＜0.43,non-f ick iand iffusion,and n=0.43,Fick iand iffusion,and 0.43＜n＜0.85,anom a lous d iffusion,and n＞0.85,case II d iffusion[17].

And W eibu ll Fun ction w as used to describe the release p rof iles fo r its w idely use in m icrospheres,and it w as a lso recomm end[28]by Am erican Association of Pharm aceu tical Scien tists/In ternationa l Pharm aceu tica l Federation(AAPS/FIP)w o rkshop.The W eibu ll fun ction[29-31]m en tioned here is:Y=1-exp(-(a·t)b),w here Y m eans the percen t of the d rug release from the m icrospheres,and a is related to the apparen t rate constan t,and b is co rrespond ing to the shape of the release p rof ile[29-31].W eibu ll Fun ction[30]is su itab le for sigm oida l d rug re lease p rof iles,and it is used to m odel release p rof iles[30]that show zero to m in im a l in itial bu rst release,zero to m in im a l d iffusion-m ed iated re lease rate[30]o r erosion-dom inated p rocess coup led w ith m in im a l d iffusive release rates[30].

Tab le 1-The DL,EE and Tg of the m icrospheres.

In add ition,the detailed m ethod of the elevated tem peratu res to p red ict rea l tim e re lease from p rogesterone-PLGA m icrospheresw as deep ly d iscussed in the resu lts and d iscussion part.

2.2.6. Reproducibility testing

The rep roducibility of th is m ethod to p redict the real tim e release from p rogesterone-PLGA m icrospheres w as investigated.The PLGAs w h ich have the d ifferen t inheren t viscosity,d ifferen t lactide/glyco lide ratio,and pu rchased from d ifferen t com pan ies w ere used to m ake p rogesterone-PLGA m icrospheres.No t on ly the ethy l fo rm ate w as used as the o rgan ic phase tom ake them icrospheres,bu t a lso them ethy lene ch lo ride w as used.Differen t d rug load ing rates w ere obtained by changing the gram s of p rogesterone and PLGA in the p reparations.

Differen t re lease m ed ium s in clud ing“0.5%SDS-pH 7.4 PBS-0.1%NaN3”and“1%SDS-pH 7.4 PBS-0.1%NaN3”w ere used to do the rep roducibility testing.W hen use these release m ed ium,it shou ld be no ted that:f irst,10m g of the m icrosphere w ere suspen ded in 10m l release m ed ium;second,at each tim ing poin t,the sam p les w ere cen trifuged at 4500 rpm fo r 5m in,then 9m l of the sam p les w as rem oved and 9m l fresh m ed ium w as added to m ain tain the sink cond ition.

In a ll,the p red icting m ethod of ethy l form atep rogesterone-PLGA(65:35,Daigang)m icrospheres,m ethy lene ch lo ride-p rogesterone-PLGA(65:35,Daigang)m icrospheres and m ethy lene ch lo ride-p rogesterone-PLGA(75:25,Daigang)w ill be d iscussed in the nex t part.

3. Resu lts and d iscussion

3.1. Physicochem ical properties ofprogesterone m icrospheres

The physicochem ica l p roperties of p rogesterone-PLGAm icrospheresw ere listed in Tab les 1-3.The SEM im agesw ere show n in Fig.1.

Tab le 2-The pa rticle size d istribu tion of the m icrospheres.

Tab le 3-The in trinsic viscosity of PLGA and the re lease m ed ium of the m icrosphe res.

Tab le 4-The resu lts of Ko rsm eyer-Peppas Equation fo r the re lease of ethy l fo rm ate-p rogesterone-PLGA(RG653H,Evon ik)m icrospheres(Tg=45.3±0.61).

3.2. In vitro release testing

The p rogesterone-PLGA m icrospheres (noted as ethy l form ate-p rogesterone-PLGA(Evon ik)m icrospheres)w ere m ade by the so lven t evaporation m ethod,and the o rgan ic phase w as ethy l form ate.The in trinsic viscosity of PLGA w as 0.321 d l/g.Fig.2 show s the in vitro release of them icrospheres at 37°C and e levated tem peratu res.

3.3. Da ta analysis

The release cu rves at d ifferen t tem peratu resw ere f itted by the Korsm eyer-Peppas fun ction[32],and resu lts w ere show n in Tab le 4.A t 37°C,n=0.3162,it seem s that the p rogesterone released from the PLGA m icrospheres fo llow ed the non-Fick iand iffusion.How ever,as a poo r w ater so lub le d rug,p rogesterone usually can't fo llow the non-Ficikandiffusion.And th is question w ill be deep ly d iscussed here.The m echan ism of d rug re lease from PLGA w as very com p lex.There have tw o m ain release m echan ism s,w h ich are d iffusion and degradation/erosion[33].The release rate is often said to be d iffusioncon tro lled in itia lly[30,33,34]and degradation/erosion contro lled du ring the f ina l stage of the release period[30,33,34].Here,som e of the re lease m echan ism s of d ifferen t stages w ere listed in the Tab le 5.And in Fig.3,the relation sh ip betw een release m echan ism s and d ifferen t releasing p rocesses w as deep ly described.So the sim p ly use of Ko rsm eyer-Peppas equation to determ ine the re lease m echan ism s w as a bit arbitrary.Fu rtherm ore,the Korsm eyer-Peppas equation shou ld be used to describe the p rof ile w hen the d rugs w ere d ispersed in the po lym ers even ly.In sho rt,the Ko rsm eyer-Peppas equation cou ld no t be to ta lly app rop riate to describe the rea l release m echan ism,especially w hen the release p rof ile has a sign if ican t in itia l bu rst part.And w hen the in itia l bu rst of the re lease p rof ile at 37°C w as excluded,the fun ction w ill be:y=0.04519t0.4634,R2=0.9822,and n=0.4634,w h ich m eans the m echan ism of the re lease p rof ile at 37°C w as anom a lous transpo rt.

It a lso im p lies n va lues in these fun ction s w h ich describe the release p rof ilesw ith the large in itialbu rstatelevated temperatu res cou ld no t tota lly describe the rea l re lease m echan ism at e levated tem peratu res.And it shou ld be noted that in th is study,the stirring rate at e levated tem peratu res w as faster than that at 37°C to obtain a quicker release rate.Th is a lso exp lained w hy the n va lue changes w ith tem peratu re irregu larity.

Fig.1-The SEM of the p rogesterone-PLGA m icrospheres.(A)Ethy l fo rm ate-p rogesterone-PLGA(RG653H,Evon ik)m icrospheres,(B)Ethy l fo rm ate-p rogesterone-PLGA(65:35,Daigang)m icrospheres,(C)M ethy lene ch lo ride-p rogesterone-PLGA(65:35,Daigang)m icrospheres,(D)M ethy lene ch lo ride-p rogesterone-PLGA(75:25,Daigang)m icrospheres.

Fig.2-In v itro re lease p rof iles of the ethy l fo rm ate-p rogesterone-PLGA(RG653H,Evon ik)m icrospheres at d ifferen t tem peratu res.

The in itial bu rst release in the release p rof ile at 37°C w as because of the low inheren t viscosity of the choosing PLGA(RG653H,0.321 d l/g,Evon ik)and the so lven t evaporation m ethod.An d the bu rst release w as m ostly because of the accum u lation of the loaded d rug at the su rface of the m icroparticles [35].In add ition,the release p rof ile of p rogesterone-PLGA(RG653H,0.321 d l/g,Evon ik)m icrospheres at 37°C exh ibited sim ilar shape and in itia l bu rst w ith o ther literatu res[35-37]w h ich also described the release p rof ile of p rogesterone-PLGA m icrospheres.

As am atter of fact,Peppas and Sah lin[38]have a lready dedu ced fun ctions to describe the d iffusion and erosion m echan ism separately.

When the va lue of R/Fw as larger,the erosion m echan ism s con tribu ted larger.W hen the va lue of R/F w as sm a ller,the d iffusion m echan ism con tribu ted larger[38].The Peppas and Sah lin fun ction cou ld be m uch m o re app rop riate to describe the com p lex re lease m echan ism.

Neverthe less,in th is study,the Ko rsm eyer-Peppas equation w as stillused to describe the release p rof iles,the n value here w as no t used to iden tify the re lease m echan ism bu t to describe the shape of the release p rof ile,and to p red ict the rea l tim e release.Sin ce the n va lue in the Korsm eyer-Peppas equation cou ld be used to clearly describe the shape of the relapse p rof ile,the n va lues at 43°C and 47°C w ere c losest to 37°C w h ich w ere show n in Tab le 4,it m eans the p rof ile at 43°C and 47°C exh ibited the sim ilar re lease p rof ile w ith the p rof ile at 37°C.So the release p rof ile at 43°C and 47°C cou ld be used to p red ict the rea l tim e release(37°C).The n va lue at 52°C w as no t m en tioned in Tab le 4,because 52°C w as h igher than Tg(45.3±0.61)of the m icrospheres.W hen the tem peratu re w as near o r above the Tg,the chain of the po lym er w ill be m ore m obility so the d rugs w ill be released from the po lym er in a qu icker w ay.In add ition,the h igh m agnetic stirring rate used in th is study a lso in creased the releasing rate from the po lym er.So,at 52°C,the release w as f inished in a very short tim e.Few available tim ing poin ts cou ld be f itted by the Ko rsm eyer-Peppas equation and W eibu ll fun ction.Th is w as a lso the reason that part of the n va lue w as no t m en tioned in Tab les 6-8 in the“rep roducibility testing”part.

Tab le 5-Exp lanation of the o rigin s of the phases observed du ring d rug re lease[33].

Fig.3-The com p lex p ictu re of physico-chem ica l p rocesses tak ing p lace w ith in PLGA m atrices,lead ing to d rug re lease.The in f luen ce of p rocesses on d rug release and on o ther p rocesses is illustrated by arrow s.Note that som e arrow s poin t in both d irections.(M od if ied from Fig.2 in Re feren ce[33].).

Tab le 6-The resu lts of Ko rsm eyer-Peppas Equation fo r the re lease of ethy l fo rm ate-p rogesterone-PLGA(65:35,Daigang)m icrospheres(Tg=53.2±1.1).

Tab le 7-The resu lts of Ko rsm eyer-Pep pas Equation fo r the re lease ofm ethy lene ch lo ride-p rogesterone-PLGA(65:35,Daigang)m icrospheres(Tg=52.9±0.99).

Tab le 8-The resu lts of Ko rsm eyer-Pep pas Equation fo r the re lease ofm ethy lene ch lo ride-p rogesterone-PLGA(75:25,Daigang)m icrospheres(Tg=57.4±0.87).

Fig.4-(A)The re lease p rof iles of the ethy l fo rm ate-p rogesterone-PLGA(RG653H,Evon ik)m icrospheres at 37°C and 43°C,and the fun ction w h ich w as dedu ced from the re lease p rof ile at 43°C.(B)The re lease p rof iles of the ethy l fo rm ate-p rogesterone-PLGA(RG653H,Evon ik)m icrospheres at 37°C and 47°C,and the fun ction w h ich w as dedu ced from the re lease p rof ile at 47°C.

The com parison of the re lease p rof ile at 37°C and 43°C w ere show n in Fig.4A.A lthough the tw o p rof iles have d ifferen t x-ax is sca les,they share the sim ilar shape w h ich m eans they have the sim ilar functions.W ith th is am azing f ind ing,w e just need to convert the x-ax is sca les to deduce the fun ction of the rea l tim e release.So the p rof ile of the 43°Cw as used to p red ict and deduce the fun ction of the re lease p rof ile of p rogesterone PLGA m icrospheres at 37°C.

In the Fig.4A,it cou ld be seen clearly that the m icrospheres re leased abou t 40%at 2 h at 43°C and need abou t 30 h to reach the sam e am oun t at 37°C,th is w as the conversion ratio of the x-ax is betw een the release p rof ile at 37°C and 43°C.An d the conversion ra tio w as 2:30=1:15.The p rof ile at 43°C w ere f itted by the W eibu ll fun ction as Y=1-exp(-(0.08828t)0.4345),R2=0.9322.By using of the conversion ratio,the fun ction of the re lease p rof ile at 37°C w as dedu ced as Y=1-exp(-(0.08828t/15)0.4345),R2=0.9921.

It shou ld be no ted that,in bo th the Korsm eyer-Peppas equation and the Weibu ll function,the value of the Mi/M w as from 0 to 1,bu t in the f igu res,the y-axism eans the accum ulated re lease(100 Mi/M),so a ll the deduced fun ctions w ere m u ltip lied by 100%to show in the f igu res.

As m en tioned above,the re lease p rof ile of the m icrospheres at 47°C cou ld a lso be used to dedu ce the fun ction of the rea l tim e release.The com parison of the release p rof ile betw een 37°C and 47°C,and the correlation betw een the deduced fun ction and the fun ction of rea l tim e release w ere show n in Fig.4B.

The W eibu ll fun ction of the release p rof ile at 47°C w as:Y=1-exp(-(0.8362t)0.4627),R2=0.9830.From Fig.4B,the m icrospheres re leased abou t 50%of its contentsw ith in 0.5 h at47°C,and it need abou t 71.5 h to reach the sam e am oun t at 37°C,and th is w as the conversion ratio of the x-axis betw een the re lease p rof ile at 37°C and 47°C.Sin ce the conversion ratio w as 0.5:71.5=1:143,the dedu ced fun ction of the release p rof ile at 37°C w as Y=1-exp(-(0.8362t/143)0.4627),R2=0.9912.Bu t th inking abou t the releasing tim e at 43°C and 47°C,the release p rof ile at 47°Cw as a great choice to dedu ce the fun ction of the rea l tim e re lease.

In sum m ary,w ith th is k ind ofm ethod,the shape of the rea l tim e release cou ld be p redicted and deduced successfu lly and the m ethod w as sim p le.A fter a series in vitro re lease at elevated tem peratu res w ere done,the m ost sim ilar p rof ile w ith the rea l tim e re lease p rof ile w as p icked,and the conversion ratio w as used to deduce the release p rof ile.

3.4. Reproducibility testing

The physicochem ica l p roperties of these p rogesteronem icrospheres m ade by d ifferen t types of PLGA w ere elaborately listed in Tables 1-3.An d the SEM im ages w ere show n in Fig.1.

3.4.1. Data analysis of ethyl form ate-progesterone-PLGA(65:35,Daigang)m icrospheres

For ethy l form ate(organ ic phase)-p rogesterone-PLGA(65:35,0.386 d l/g,Daigang),the in vitro release p rof ile at d ifferen t tem peratu res w ere show n in Fig.5A,and described by the Korsm eyer-Peppas equation w ere listed in Tab le 6.The n va lue of these fun ctions show ed that the release p rof ile at 55°Cw as the m ost sim ilar to the release p rof ile at 37°C.

The W eibu ll fun ction w as used to describe the release p rof ile a t 55°C as Y=1-exp(-(0.6425t)0.6533),R2=0.8408.In the Fig.5B,the conversion ratio betw een the tw o release p rof ile at x-ax is w as 0.5:150=1:300.Therefo re,the dedu ced fun ction of the rea l tim e releasew as:Y=1-exp(-(0.6425t/300)0.6533),R2=0.9781.

It shou ld be d iscussed that,in the W eibu ll fun ction of the re lease p rof ile at 55°C,the va lue of R2w as a little sm a ll.From the Fig.5B,the shape of the re lease p rof ile at 55°C looksm ore like a line than“S”shape.So,w hen using theWeibu ll function to describe the p rof ile,the va lue of R2w ou ld be a little sm a ll.Bu t it can still describe the p rof ile at 55°C.W hen the temperatu re w asm u ch h igher than the Tg,the re leasing rate w ill be greatly in creased,and th is w ill leads to huge bu rst release at the f irst tim ing poin t,and the re lease p rof ile in tegra lly w ill looks like a straigh t line,so as to lose the oppo rtun ity to exh ibit the sim ilar shape w ith the release p rof ile of the rea l tim e release.An d th ism ethod can'tbe used to deduce the function of the rea l tim e re lease.Th is w as a lso the reason that som e of the n va lues w ere notm en tioned in Tab les 4,6-8.So itw as no t the h igher tem peratu re the better.The choosing e levated tem peratu res shou ld be used to exh ibit the shape of the release p rof ile.

Fig.5-(A)In vitro re lease p rof iles of the ethy l fo rm ate-p rogesterone-PLGA(65:35,Daigang)m icrospheres at d iffe ren t tem pe ratu res.(B)The re lease p rof iles of the ethy l form ate-p rogesterone-PLGA(65:35,Daigang)m icrospheres at 37°C and 55°C,and the fun ction w h ich w as dedu ced from the re lease p rof ile at 55°C.

3.4.2. Data analysis ofm ethylene ch loride-progesterone-PLGA(65:35,Daigang)m icrospheres

The m ethy lene ch lo ride(o rgan ic phase)-p rogesterone-PLGA(653H,0.386 d l/g,Daigang)w as a lso m ade to help test the rep roducibility of the p red ictingm ethod.The in vitro release p rof ile at d ifferen t tem peratu res w ere show n in Fig.6A and described by the Ko rsm eyer-Peppas equation w ere listed in Tab le 7.The n va lue of these fun ctions show ed that the re lease p rof ile at 51°Cw as them ost sim ilar to the release p rof ile at 37°C.

Fig.6-(A)In vitro re lease p rof iles of the m ethy lene ch lo ride-p rogesterone-PLGA(65:35,Daigang)m icrospheres at d ifferen t tem peratu res.(B)The re lease p rof iles of the m ethy lene ch loride-p rogesterone-PLGA(65:35,Daigang)m icrospheres at 37°C and 51°C,and the fun ction w h ich w as dedu ced from the re lease p rof ile at 51°C.

Fig.7-(A)In vitro re lease p rof iles of the m ethy lene ch lo ride-p rogesterone-PLGA(75:25,Daigang)m icrospheres at d ifferen t tem peratu res.(B)The release p rof iles of the m ethy lene ch lo ride-p rogesterone-PLGA(75:25,Daigang)m icrospheres at 37°C and 60°C,and the fun ction w h ich w as dedu ced from the re lease p rof ile at 60°C.

TheW eibu ll Fun ction w as used to describe the re lease p rof ile at 51°C,as Y=1-exp(-(0.5393t)0.6195),R2=0.9796.In Fig.6B,the conversion ratio betw een the tw o re lease p rof ile at x-axis w as 0.5:80=1:160.There fo re,the deduced fun ction of the rea l tim e re lease w as:Y=1-exp(-(0.5393t/160)0.6195,R2=0.9918.

3.4.3. Data analysis ofm ethylene chloride-progesterone-PLGA(75:25,Daigang)m icrospheres

The in vitro release p rof iles at d ifferen t tem peratu res w ere show n in Fig.7A.The Ko rsm eyer-Peppas equation w as used to describe the release p rof iles and w as listed in Table 8.The n va lue of these fun ctions show ed that the re lease p rof ile at 60°C w as the m ost sim ilar to the re lease p rof ile at 37°C.Acco rd ing to the Fig.7B,the conversion ratio betw een the tw o release p rof iles at x-axis w as 1:400.TheW eibu ll Function of the release p rof ile at 60°C w as:Y=1-exp(-(0.6394t)0.7850),R2=0.9819.Therefore,the deduced fun ction of the rea l tim e release w ou ld be:Y=1-exp(-(0.6394t/400)0.7850,R2=0.9972.It a lso exh ibited a great co rrelation betw een the deduced fun ction and the p rof ile of the rea l tim e re lease.

It shou ld a lso be no ted that the d ifferen t d rug load ing rates and particle size d istribu tion s w ere ob tained by using d ifferen t types of PLGA and organ ic phases.Th is w as because it w as d iff icu lt to change one of these factors w ith the sim ilar p reparing m ethod.Fo r exam p le,in o rder to in crease the d rug load ing ra te,the en capsu lation eff icien cy usua lly decreased and m aybe other factors like partic le size d istribu tion w illalso changes.So in th is study,d ifferen t types of PLGA and d ifferen t o rgan ic phases w ere used to ob tain the d ifferen t d rug load ing rate,particle size d istribu tion and even using d ifferen t release m ed ium s.How ever,in th is w ay,m any d ifferen t factors w ere m ixed and tested to help verify the rep roducibility of the p red icting m ethod.

Luckily,from these resu lts,even w ith d ifferen t kinds of d rug load ing rate,particle size d istribu tion,d ifferen t re lease m ed ium and d ifferen t p repa ring m ethod w ith d ifferen t types of PLGA,the shape of the re lease p rof iles w ere sim ila r,and there w ere m any other literatu res[35-37]deep ly stud ied the p rogesterone-PLGAm icrospheresw ith m any d ifferen t parameters,and they a ll exh ibited the sim ilar shape of the re lease p rof ile as m en tioned in th is study.In a ll,the rep rodu cibility of the p red icting m ethod w as great,and th is m ethod cou ld be a po ten tia l good choice to p red ict the rea l tim e release.

How ever,the methodstill has som e sho rtcom ings and it cou ld be seen in f igu res describing the co rre lation betw een the deduced fun ction and the rea l tim e re lease.The f ina l pa rt of the p rof ile of the rea l tim e re lease d idn't show perfect co rrelation w ith the deduced fun ction.In the very f inal part,the m icrospheres w ere a lready rup tu red because of the bu lk erosion,so it cou ld not release d rugs in the con tro lled m anner w h ich w as the sam e w ith the fo rm er part.

In add ition,the Weibu ll fun ction cou ld on ly be used to describe erosion-dom inated p rocess coup led w ith m in im a l d iffusive release rates.So,the m ethod in th is study cou ldn't perfectly describe the very f ina l part of the p rof iles,bu t it can successfu l describe the p rof ile in general.

4. Con clu sion

Them ethod of e levated tem peratu res coup led w ith m agnetic stirring to p red ict rea l tim e re lease from long acting p rogesterone PLGA m icrospheres cou ld be described brief ly in severa l steps.First,ob tain release p rof iles at 37°C w ith o rbita l shak ing and elevated tem peratu res w ith m agnetic stirring.Secon d,the Ko rsm eyer-Peppas equation w as used to dep ict the release p rof iles at d ifferen t tem peratu res to help f ind ing the m ost sim ilar re lease p rof iles from the elevated tem peratu res w ith the p rof ile of rea l tim e release.Th ird,the W eibu ll fun ction w as used to describe the chosen elevated tem peratu re,and the fun ction of the real tim e release(37°C)w as calcu lated w ith the he lp ofW eibu ll fun ction and conversion ratio.W ith the use of th ism ethod,the fun ction of the rea l tim e release cou ld be easily deduced acco rd ing to the release p rof ile at elevated tem peratu res.

Differen t types of PLGA(RG653H(Evon ik),65:35(0.386,Daigang),75:25(0.401,Daigang))and d ifferen t o rgan ic phases(ethy l fo rm a te and m ethy lene ch loride)w ere used to m ake the p rogesterone PLGAm icrospheresw ith d ifferen t d rug loading rate,particle size d istribu tion and even d ifferen t re lease m ed ium s w ere used to help verify the p red icting m ethod.A ll the dedu ced fun ctions exh ibited a great correlation w ith the p rof ile of the rea l tim e re lease in th is study w ith R2=0.9912,0.9781,0.9918 and 0.9972,respectively.

How ever,th is e levated tem peratu res p red icting m ethod has lim itation s.Because the Weibu ll fun ction w as on ly su itab le fo r sigm oida l d rug re lease p rof iles,and it w as used to m ode l release p rof iles that show ed zero to m in im a l in itia l bu rst release,zero tom in im a ld iffusion-m ed iated re lease rate or erosion-dom inated p rocess coup led w ith m in im ald iffusive release rates[30].So the very f ina l part of the rea l tim e re lease w as no t perfectly described by the deduced fun ction.Bu t th is m ethod w as stilla great choice to p red ictand deduce the function of rea l tim e release in genera l.It a lso p rovided an insp iration in p red icting the re lease p rof ile of long acting m icrospheres.

Declaration of in terest

No con f lict of in terest ex its in the m anuscrip t,and the artic le is app roved by a ll the au tho rs.