pha maceu ics
A icle
3D P in ing o a Mul i-Laye ed Polypill Con aining
Six D ugs Using a No el S e eoli hog aphic Me hod
Pamela Robles-Ma inez 1, Xiaoyan Xu 1, Sa ah J. T en ield 1, A hee Awad 1,
Al a o Goyanes 2,3 , Richa d Tel o d 4, Abdul W. Basi 1,2,* and Simon Gais o d 1,2,*
1Depa men o Pha maceu ics, UCL School o Pha macy, Uni e si y College London, 29–39 B unswick
Squa e, London WC1N 1AX, UK; [email p o ec ed] (P.R.-M.); [email p o ec ed] (X.X.);
[email p o ec ed] (S.J.T.); a hee [email p o ec ed] (A.A.)
2FabRx L d., 3 Romney Road, Ash o d TN24 0RW, UK; [email p o ec ed]
3Depa amen o de Fa macología, Fa macia y Tecnología Fa macéu ica, R +D Pha ma G oup (GI-1645),
Uni e sidade de San iago de Compos ela, 15782 San iago de Compos ela, Spain
4
School o Chemis y and Fo ensic Sciences, Uni e si y o B ad o d, Richmond Road, B ad o d BD7 1DP, UK;
R.Tel o [email p o ec ed]
*Co espondence: [email p o ec ed] (A.W.B.); s.gais o [email p o ec ed] (S.G.)
Recei ed: 7 May 2019; Accep ed: 3 June 2019; Published: 11 June 2019


Abs ac :
Th ee-dimensional p in ing (3DP) has demons a ed g ea po en ial o mul i-ma e ial
ab ica ion because o i s capabili y o p in ing bespoke and spa ially sepa a ed ma e ial
con o ma ions. Such a concep could e olu ionise he pha maceu ical indus y, enabling he
p oduc ion o pe sonalised, mul i-laye ed d ug p oduc s on demand. He e, we de eloped a no el
s e eoli hog aphic (SLA) 3D p in ing me hod ha , o he i s ime, can be used o ab ica e mul i-laye
cons uc s (polypills) wi h a iable d ug con en and/o shape. Using his echnique, six d ugs,
including pa ace amol, ca eine, nap oxen, chlo amphenicol, p ednisolone and aspi in, we e p in ed
wi h di e en geome ies and ma e ial composi ions. D ug dis ibu ion was isualised using Raman
mic oscopy, which showed ha whils sepa a e laye s we e success ully p in ed, se e al o he d ugs
di used ac oss he laye s depending on hei amo phous o c ys alline phase. The p in ed cons uc s
demons a ed excellen physical p ope ies and he di e en ma e ial inclusions enabled dis inc d ug
elease p o iles o he six ac i es wi hin dissolu ion es s. Fo he i s ime, his pape demons a es he
easibili y o SLA p in ing as an inno a i e pla o m o mul i-d ug he apy p oduc ion, acili a ing a
new e a o pe sonalised polypills.
Keywo ds:
h ee-dimensional p in ing; ixed-dose combina ions; addi i e manu ac u ing; 3D
p in ed d ug p oduc s; p in le s; able s; pe sonalized medicines; mul iple-laye dosage o ms;
s e eoli hog aphy; a polyme isa ion
1. In oduc ion
Mul iple d ug he apies ha e gained inc easing a en ion in heal hca e because o imp o ed
ea men ou comes o diseases wi h complex pa hologies, such as HIV-1 in ec ion, hype ension,
ube culosis and ype II diabe es melli us [
1
–
4
]. Despi e his, polypha macy (in ol ing he
adminis a ion o i e o mo e medicines) is a guably he mos p essing p esc ibing issue, linked o
inc easing a es o non-adhe ence and pa ien con usion due o he high pill bu den and complex
adminis a ion equi emen s [
5
]. Such challenges can be o e come by u ilising ixed-dose combina ions
(FDCs) o polypills, whe eby mo e han one d ug is inco po a ed in o he same d ug p oduc [
6
–
8
].
Indeed, he comme cially a ailable polypill (Polycap
TM
), which con ains en e ic-coa ed aspi in,
amip il, sim as a in, a enolol, and hyd ochlo o hiazide, has been shown o be e ec i e in educing
mul iple ca dio ascula isk ac o s [9,10].
Pha maceu ics 2019,11, 274; doi:10.3390/pha maceu ics11060274 www.mdpi.com/jou nal/pha maceu ics
Pha maceu ics 2019,11, 274 2 o 16
The main ba ie o he widesp ead in oduc ion o FDCs, howe e , lies in hei manu ac u ing
and lack o lexibili y in dosing. Con en ional powde compac ion ypically p oduces homogeneous
able s con aining ixed s eng hs on a la ge comme cial scale, an app oach ha is wholly unsui able o
he apies ha equi e lexibili y in dosing o d ug combina ion(s). Fo example, i a pa ien equi es a
change in dose and/o d ug whils main ained on a FDC, o en he ea men will ha e o be wi hd awn
and he pa ien would be ini ia ed on sepa a e dosage o ms [
11
]. Fu he mo e, na ow he apeu ic
index d ugs o hose ha equi e equen dose i a ions a e unsui able o FDC egimens [
12
,
13
].
In he e a o pe sonalised medicine, i is clea ha a no el pla o m ha enables a lexible p ocess o
ailo ed dosing and d ug combina ions is equi ed [14–17].
I isin hisniche ha h ee-dimensionalp in ing(3DP)o e ssigni ican p omiseasa ans o ma i e
echnology [
18
–
23
]. Th ee-dimensional p in ing is an addi i e manu ac u ing echnique ha ab ica es
objec s om a compu e -aided design (CAD) ile in a laye -by-laye manne [
24
–
28
]. Owing o i s
lexibili y, 3DP allows he combina ion o mul iple ma e ials in a single dosage o m wi h di e en
geome ies [
29
–
33
]. Since he d ugs a e physically sepa a ed, i is possible o adjus doses and elease
p o iles indi idually as well as o co- o mula e d ugs ha may po en ially in e ac [
12
,
34
–
38
]. Indeed,
p e ious s udies ha e 3D p in ed polypills con aining pa ace amol and ca eine wi h a ying designs
(mul i-laye ed and DuoCaple ), enabling speci ic elease p o iles o be a ained depending on he
posi ion o he d ug in he caple , independen o d ug solubili y [
39
]. Khaled. e al. ab ica ed a 3D
p in ed polypill con aining i e d ugs ha we e eleased in wo di e en p o iles [
40
]. The same g oup
inco po a ed h ee di e en d ugs wi hin a single 3D p in ed able using a semisolid ex usion-based
p in e , each o which ha e a dis inc elease p o ile depending on hei spa ial loca ion [41].
The mos widely used 3DP echnique in pha maceu icals is used deposi ion modelling (FDM),
which in ol es he use o d ug-loaded polyme ilamen s as eeds ock ha a e hea ed and deposi ed
laye -by-laye [
42
–
47
]. Howe e , mos comme cially a ailable FDM p in e s can only p in wi h a
limi ed numbe o ilamen s, and hence, enabling a maximum deposi ion o a limi ed numbe o
spa ially sepa a ed d ugs [
35
]. S e eoli hog aphic (SLA) 3DP is an al e na i e echnology hi he o
ela i ely unexplo ed o pha maceu ical applica ions. I wo ks by using a lase o pho ocu e a liquid
esin, comp ising a pho opolyme isable monome and a pho oini ia o ha upon exposu e o ligh
ini ia es polyme isa ion o he monome [
48
]. S e eoli hog aphic 3DP o e s some key ad an ages o e
o he 3DP echnologies including a oidance o he mal deg ada ion [
28
,
49
], imp o ed esolu ion and
highe accu acy [
50
], and is also conside ed a as e me hod han FDM o selec i e lase sin e ing (SLS)
3DP p ocesses [51]. Fu he in o ma ion abou he SLA p ocess has been desc ibed elsewhe e [52].
By blending a d ug in o he esin, SLA 3DP has p e iously been used o make able s [
52
] and
hyd ogels [
53
] and i s e sa ili y has allowed he explo a ion o how geome ic pa ame e s in luence
d ug elease kine ics [
54
]. Thus a , howe e , no such wo k has demons a ed he abili y o SLA
o p oduce polypills, likely due o he di icul y in p in ing spa ially-sepa a ed laye s. A pa icula
challenge ela es o he so wa e and ha dwa e o comme cially a ailable SLA p in e s, which does no
allow o mul i- esin p in ing.
In his wo k, o he i s ime, we ha e o e come his limi a ion o SLA p in ing by de eloping
an SLA p in e ha is capable o p in ing mul i-laye ed able s. We exempli y i s use by p in ing
a polypill 3D-p in ed able (p in le ) con aining six di e en model d ugs (pa ace amol, nap oxen,
ca eine, aspi in, p ednisolone, and chlo amphenicol), some o which a e commonly adminis e ed
oge he o imp o e hei e icacy. The spa ial sepa a ion o he d ugs was de e mined wi h Raman
mic oscopy and he modi ica ion o d ug elease a es upon changing polypill geome y (cylind ical
and ing shapes) and excipien addi ion was e alua ed using dissolu ion es s. C i ically, his wo k has
gene a ed a new SLA pha maceu ical p in ing p ocess, e olu ionising he manu ac u e o polypills
and ea men pa hways o pa ien s.
Pha maceu ics 2019,11, 274 3 o 16
2. Ma e ials and Me hods
The model d ugs pa ace amol (MW =151.2 g/mol), ace ylsalicylic acid (MW =180.2 g/mol),
nap oxen (MW =252.2 g/mol), chlo amphenicol (MW =323.1 g/mol), and ca eine (MW =194.2 g/mol)
we e pu chased om Sigma–Ald ich L d. (Gillingham, UK) and p ednisolone (MW =360.4 g/mol)
was pu chased om Se e n Bio ech L d. (Kidde mins e , UK).
Polye hyleneglycoldiac yla e(PEGda, a e ageMW 575 g/mol)anddiphenyl(2,4,6- ime hylbenzoyl)
phosphine oxide (TPO) we e pu chased om Sigma-Ald ich L d. (Gillingham, UK). The sal s o
p epa ing he bu e dissolu ion media we e pu chased om VWR In e na ional L d., Poole, UK. All
ma e ials we e used as ecei ed.
2.1. 3D P in ing
PEGda was used as he pho opolyme isable monome and TPO as he pho oini ia o (PI). The
composi ions o he o mula ions a e shown in Table 1.
Table 1. Composi ions (% w/w) o he ini ial esins o p in ing.
Ma e ial
Fo mula ion Type I (% w/w) Type II (% w/w) Type III (% w/w)
PEGda 89 89 44.5
PEG300 - - 44.5
TPO 1 1 1
D ug 10 10 10
Each o mula ion was p epa ed by dissol ing he d ug and he PI in liquid PEGda and PEG300
when applicable. The componen s we e added in o a beake unde cons an s i ing un il comple e
dissolu ion o he powde s in he polyme (s). Then each solu ion was pou ed in o a esin ay
o p in ing.
All p in le s we e ab ica ed using a Fo m 1+SLA 3D p in e (Fo mlabs Inc., Some ille, MA,
USA). The p in e was equipped wi h a 405 nm lase able o ab ica e objec s wi h a esolu ion o 300
µ
m
and a laye hickness o 25
µ
m, 50
µ
m, 100
µ
m o 200
µ
m. P in le s (a cylinde —10 mm diame e
and 3 mm heigh , o a ing—10 mm diame e and 6 mm heigh ) we e designed in Au oCAD®® 2017
(Au odesk Inc, San Ra ael, CA, USA) and expo ed as a s e eoli hog aphic ile (.s l) (Figu e 1) o he
P e o m So wa e .2.3.3 OpenFL, (Fo mlabs Inc., Some ille, MA, USA).
Pha maceu ics 2019, 11, x FOR PEER REVIEW 3 o 16
we e pu chased om Sigma–Ald ich L d. (Gillingham, UK) and p ednisolone (MW = 360.4 g/mol)
was pu chased om Se e n Bio ech L d. (Kidde mins e , UK).
Polye hylene glycol diac yla e (PEGda, a e age MW 575 g/mol) and diphenyl (2,4,6-
ime hylbenzoyl) phosphine oxide (TPO) we e pu chased om Sigma-Ald ich L d. (Gillingham,
UK). The sal s o p epa ing he bu e dissolu ion media we e pu chased om VWR In e na ional
L d., Poole, UK. All ma e ials we e used as ecei ed.
2.1. 3D P in ing
PEGda was used as he pho opolyme isable monome and TPO as he pho oini ia o (PI). The
composi ions o he o mula ions a e shown in Table 1.
Table 1. Composi ions (% w/w) o he ini ial esins o p in ing.
Fo mula ion
Ma e ial Type I (% w/w) Type II (% w/w) Type III (% w/w)
PEGda 89 89 44.5
PEG300 - - 44.5
TPO 1 1 1
D ug 10 10 10
Each o mula ion was p epa ed by dissol ing he d ug and he PI in liquid PEGda and PEG300
when applicable. The componen s we e added in o a beake unde cons an s i ing un il comple e
dissolu ion o he powde s in he polyme (s). Then each solu ion was pou ed in o a esin ay o
p in ing.
All p in le s we e ab ica ed using a Fo m 1+ SLA 3D p in e (Fo mlabs Inc., Some ille, MA,
USA). The p in e was equipped wi h a 405 nm lase able o ab ica e objec s wi h a esolu ion o 300
μm and a laye hickness o 25 μm, 50 μm, 100 μm o 200 μm. P in le s (a cylinde —10 mm diame e
and 3 mm heigh , o a ing—10 mm diame e and 6 mm heigh ) we e designed in Au oCAD®® 2017
(Au odesk Inc, San Ra ael, CA, USA) and expo ed as a s e eoli hog aphic ile (.s l) (Figu e 1) o he
P e o m So wa e .2.3.3 OpenFL, (Fo mlabs Inc., Some ille, MA, USA).
Figu e 1. 3D designs o he p in le s. Type I: Cylinde (le , 10 mm diame e and 3 mm heigh ), Types
II and III: Ring ( igh , 10 mm diame e and 6 mm heigh ).
The p in le s we e ab ica ed keeping he o de o he d ugs in he polypill unchanged, ha ing
he d ugs wi h he highe wa e solubili y (pa ace amol and ca eine) in he inne laye s, whe eas he
d ugs wi h he lowes wa e solubili y (nap oxen and p ednisolone) we e p in ed in he ou e laye s
(Figu e 1). Th ee o ms o polypill we e p in ed:
 Type I: Cylinde shape
 Type II: Ring shape
 Type III: Ring shape wi h a soluble ille (PEG 300)
The Fo m 1+ p in e is designed o p in homogeneous objec s. To ab ica e p in le s wi h
di e en d ugs in disc e e laye s i is necessa y o pause p in ing in o de o change he esin
o mula ion in he p in ing ay. Hence, he use o an applica ion p og amming in e ace (OpenFL
Figu e 1.
3D designs o he p in le s. Type I: Cylinde (
le
, 10 mm diame e and 3 mm heigh ), Types II
and III: Ring ( igh , 10 mm diame e and 6 mm heigh ).
The p in le s we e ab ica ed keeping he o de o he d ugs in he polypill unchanged, ha ing
he d ugs wi h he highe wa e solubili y (pa ace amol and ca eine) in he inne laye s, whe eas he
d ugs wi h he lowes wa e solubili y (nap oxen and p ednisolone) we e p in ed in he ou e laye s
(Figu e 1). Th ee o ms o polypill we e p in ed:
•Type I: Cylinde shape
•Type II: Ring shape
•Type III: Ring shape wi h a soluble ille (PEG 300)
Pha maceu ics 2019,11, 274 4 o 16
The Fo m 1+p in e is designed o p in homogeneous objec s. To ab ica e p in le s wi h di e en
d ugs in disc e e laye s i is necessa y o pause p in ing in o de o change he esin o mula ion in he
p in ing ay. Hence, he use o an applica ion p og amming in e ace (OpenFL e sion o P eFo m
so wa e) was equi ed o enable he 3D p in e o be manually communica ed wi h.
The OpenFL e sion o he so wa e P eFo m was used o allow pausing o p in ing and aising
o he build pla o m o enable swi ching o he esin ay. Once he esin ay was changed, he build
pla e was lowe ed o i s p e ious posi ion and p in ing was esumed. The equi ed numbe o laye s
(6 blocks o laye s o 0.5 mm o he cylinde able s and 6 blocks o laye s o 1 mm o he ing-shaped
p in le s) was hen easily p in ed, wi h a deionised wa e inse o he p in ed objec be ween esins o
a oid c oss-con amina ion. A e his, he pla o m was e u ned o i s p e ious posi ion o p in he
nex block o laye s un il he polypill was comple ed.
The p in le s we e p in ed di ec ly on he build pla o m a oom empe a u e wi hou suppo s.
2.2. P in le Dimensions
The p in le s we e weighed and measu ed (wid h and heigh ) using a digi al callipe (0.150 mm
PRO-MAX, Fowle , mod S 235 PAT). The measu emen s we e pe o med in iplica e.
2.3. Raman Spec oscopy and Mapping
Samples we e moun ed and ocused using a 50
×
objec i e on a Renishaw RA802 Pha maceu ical
Analyse equipped wi h a 785 nm lase ope a ing a 50% powe (ca. 100 mW a sample). Spec al
a ays we e acqui ed wi h 26,000 spec a eco ded o e he su ace o he sample using a s ep size o
50 µm in he x- (10.15 mm) and y- (6.5 mm) dimensions.
P ocessing was pe o med wi h Renishaw WiRE so wa e using wo app oaches: (i) di ec classical
leas -squa es (DCLS) componen ma ching o e e ence 3D p in s o he pu e d ugs in he p in ing
ma ix and (ii) using di ec classical leas -squa es (DCLS) componen ma ching o e e ence spec a
ex ac ed om each o he 6 laye s o he p in ed polypill laye s.
Highe spa ial esolu ion maps we e acqui ed ac oss he nap oxen, aspi in, and pa ace amol
laye s using he same basic acquisi ion pa ame e s, wi h an inc eased spa ial esolu ion achie ed by
acqui ing spec al a ays wi h ca. 30,000 spec a eco ded o e a sec ion o he sample (1 mm along x
and 3 mm along y) using a s ep size o 10 µm.
2.4. X- ay Powde Di ac ion (XRPD)
X- ay powde di ac ion pa e ns o pu e d ugs and indi idual p in ed discs (23
×
1 mm) we e
eco ded using a Rigaku MiniFlex 600 (Rigaku, The Woodlands, TX, USA) wi h a Cu K
α
X- ay sou ce
(
λ
=1.5418 Å) and accompanying so wa e Mini lex Guidance Ve sion 1.2.01. The in ensi y and ol age
applied we e 15 mA and 40 kV. The angula ange o da a acquisi ion was 3–40
◦
2
θ
, wi h a s ep size o
0.02◦a a speed o 2◦min−1.
2.5. De e mina ion o D ug Concen a ion in he Polypills
P in le s we e c ushed using a mo a and pes le wi h 50 mL o e hanol o enhance ex ac ion o
poo ly wa e -soluble d ugs, his solu ion was hen aken o 1 L wi h deionised wa e and cons an ly
s i ed du ing 24 h. Samples o he solu ions we e il e ed h ough a 0.45
µ
m il e (Millipo e L d.,
Dublin, I eland) and he amoun o d ug in solu ion was de e mined using HPLC (Hewle Packa d
1050 Se ies HPLC sys em, Agilen Technologies, Cheadle, UK).
The alida ed HPLC assay consis ed o a s a iona y phase o an Eclipse 5
µ
m C18 column,
4.6 mm ×150 mm
(Agilen , San a Cla a, CA, USA) and a mobile phase wi h a g adien elu ion sys em
o o ho-phospho ic acid, pH =2.7 (A) and ace oni ile (B) a 25
◦
C. The g adien sys em consis ed o ;
0–7.5 min linea change om A–B (87:13 / ) o A–B (50:50 / ) and kep un il 8.5 min; hen 8.5–9.5 min
linea change o he ini ial condi ions, A–B (87:13 / ). The low a e was kep a 1.5 mL/min and he
injec ion olume was 20
µ
L. The eluen was sc eened a a wa eleng h o 263 nm. The e en ion imes
Pha maceu ics 2019,11, 274 5 o 16
o he d ugs we e as ollows: pa ace amol, 2 min; ca eine, 2.6 min; aspi in, 5.2 min; chlo amphenicol,
5.8 min; p ednisolone, 6.15 min; and nap oxen, 9.4 min (so a o al elu ion ime o 10 min). All
measu emen s we e made in duplica e.
2.6. Dynamic D ug Dissolu ion Tes ing Condi ions
D ug dissolu ion p o iles o he p in le s we e ob ained wi h a USP II appa a us (Model PTWS,
Pha ma es , Ge many). The p in le s we e placed in 750 mL o 0.1 M HCl o 2 h o simula e he
gas ic compa men , and hen ans e ed in o 950 mL o modi ied Hanks (mHanks) bica bona e
physiological medium o 35 min (pH 5.6 o 7.4); and hen in modi ied K ebs bu e (1000 mL) (pH
7 o 7.4 and hen o 6.5). The modi ied Hanks bu e -based dissolu ion medium (136.9 mM NaCl,
5.37 mM KCl, 0.812 mM MgSO
4·
7H
2
O, 1.26 mM CaCl
2
, 0.337 mM Na
2
HPO
4·
2H
2
O, 0.441 mM KH
2
PO
4
,
4.17 mM NaHCO
3
) o ms an in si u modi ied K eb’s bu e by addi ion o 50 mL o p e-K ebs solu ion
(400.7 mM NaHCO3and 6.9 mM KH2PO4) o each dissolu ion essel [55,56].
The o mula ions we e es ed in he small in es inal en i onmen o 3.5 h (pH 5.6 o 7.4), ollowed
by pH 6.5 ep esen ing he colonic en i onmen [
55
,
57
,
58
]. The medium is p ima ily a bica bona e
bu e in which bica bona e (HCO
3−
) and ca bonic acid (H
2
CO
3
) co-exis in equilib ium, along wi h
CO
2
(aq) esul ing om dissocia ion o he ca bonic acid. The pH o he bu e is con olled by an
Au o pH Sys em
™
[
59
,
60
], which consis s o a pH p obe connec ed o a sou ce o ca bon dioxide gas
(pH- educing gas), as well as o a supply o helium (pH-inc easing gas), con olled by a con ol uni .
The con ol uni is able o p o ide a dynamically adjus able pH du ing es ing (dynamic condi ions)
and o main ain a uni o m pH alue o e he o he wise uns able bica bona e bu e pH.
The paddle speed o he USP-II was ixed a 50 pm and he es s we e conduc ed a
37 ±0.5 ◦C
(n=3). Sample o he dissolu ion media (1 mL) was wi hd awn and he d ug concen a ion was
de e mined by HPLC using he me hod desc ibed abo e.
2.7. De e mina ion o Swelling Ra io (SR) o Indi idual Laye s
Th ee-dimensional p in ed blocks o laye s o each o mula ion we e quickly insed wi h
deionised wa e hen blo ed wi h il e pape o emo e any uncu ed liquid o mula ion and wa e on
he su ace immedia ely ollowing ab ica ion, hen hey we e weighed (W
i
). The cylinde s we e hen
placed in o 0.1 M HCl o 2 h, hen ans e ed o modi ied Hanks (mHanks) bica bona e physiological
medium o 22 h a 37
◦
C o simula e he dissolu ion es condi ions. A speci ic ime poin s he excess
wa e was ca e ully wiped o and he laye s we e weighed (W
s
). The SR was calcula ed using he
ollowing equa ion:
SR =
Ws
Wi
(1)
3. Resul s and Discussion
3.1. 3D P in ing P ocess
Fo he i s ime, i was possible o modi y a comme cial SLA 3D p in e in o de o ab ica e a
se ies o polypill p in le s con aining six d ugs and in unique geome ies (Type I: cylind ical and Types
II and III: ing-shaped; Figu e 2). The comme cially a ailable Fo m 1+p in e has he unc ionali y o
only c ea e homogeneous objec s composed o single esins, making i impossible o he p oduc ion o
p in ed dosage o ms con aining spa ially-sepa a ed ac i e ing edien s. In o de o achie e mul i- esin
p in ing, i was iden i ied ha he p in e would need o be paused, he build pla o m aised, and
he esin ay emo ed and eplaced wi h a new esin o mula ion. Al hough he 3D p in e P eFo m
so wa e does ha e he unc ionali y o pause p in ing a any poin du ing he ab ica ion p ocess, he
build pla o m cu en ly emains in he same posi ion (whe e ei he he objec o he pla o m i sel a e
wi hin he esin ay), physically obs uc ing he change o he esin ay o he ma e ial wi hin i .

Pha maceu ics 2019,11, 274 6 o 16
In o de o o e come his challenge, we e-designed he p in e so wa e o enable a con olled
aising and lowe ing o he build pla o m once p in ing was paused, acili a ing manual changing
o he esin in he p in e ay. To achie e his, he Fo m 1+p in e so wa e was manually modi ied
using he OpenFL e sion o P eFo m so wa e, which is an applica ion p og amming in e ace o
he Fo m 1 and Fo m 1+Fo mLabs 3D p in e s. An applica ion p og amming in e ace is a g oup o
unc ions, commands, p o ocols, and objec s ha allows p og amme s o c ea e so wa e o in e ac
wi h an ex e nal sys em ( he 3D p in e in his case) wi hou ha ing o w i e a code om sc a ch.
He e, he so wa e was modi ied o include command inpu s ha enabled he ollowing six s eps o be
ca ied ou : (1) he esin o mula ion was p in ed using SLA; (2) he p in ing p ocess was paused upon
laye comple ion; (3) he build pla o m was aised, enabling esin ay emo al; (4) he esin ay was
eplaced which included a di e en esin o mula ion; (5) he build pla e was lowe ed o i s p e ious
posi ion and; (6) p in ing was esumed o c ea e he nex o mula ion laye .
In his way, mul iple polypill p in le s could be easily ab ica ed in 30 min, wi h he o de o d ugs
in he laye s con olled by he esin o mula ion in he ank a any pa icula poin . The cus omised
p in se ings (wi h six lase passes o he i s laye o ensu e adhesion and wo o he es ) used
allowed he success ul p oduc ion o p in le s di ec ly on he build pla o m, achie ing good adhesion
wi hou signi ican ly a ec ing he dimensions. C ucially, his app oach a oids ma e ial was age and
po en ial dose a ia ion compa ed wi h o he me hods ha u ilise suppo s o adhesion o he build
pla o m ha need o be emo ed and disca ded pos -p in ing.
Pha maceu ics 2019, 11, x FOR PEER REVIEW 6 o 16
unc ions, commands, p o ocols, and objec s ha allows p og amme s o c ea e so wa e o in e ac
wi h an ex e nal sys em ( he 3D p in e in his case) wi hou ha ing o w i e a code om sc a ch.
He e, he so wa e was modi ied o include command inpu s ha enabled he ollowing six s eps o
be ca ied ou : (1) he esin o mula ion was p in ed using SLA; (2) he p in ing p ocess was paused
upon laye comple ion; (3) he build pla o m was aised, enabling esin ay emo al; (4) he esin
ay was eplaced which included a di e en esin o mula ion; (5) he build pla e was lowe ed o i s
p e ious posi ion and; (6) p in ing was esumed o c ea e he nex o mula ion laye .
In his way, mul iple polypill p in le s could be easily ab ica ed in 30 min, wi h he o de o
d ugs in he laye s con olled by he esin o mula ion in he ank a any pa icula poin . The
cus omised p in se ings (wi h six lase passes o he i s laye o ensu e adhesion and wo o he
es ) used allowed he success ul p oduc ion o p in le s di ec ly on he build pla o m, achie ing
good adhesion wi hou signi ican ly a ec ing he dimensions. C ucially, his app oach a oids
ma e ial was age and po en ial dose a ia ion compa ed wi h o he me hods ha u ilise suppo s o
adhesion o he build pla o m ha need o be emo ed and disca ded pos -p in ing.
Figu e 2. Polypill p in le (a) Type I (cylinde shape) and (b) Type II ( ing shape). The Type III
o mula ion was isually iden ical o Type II, and hence, has no been included he e. The scale is in
cm.
3.2. Physical Cha ac e is ics
3.2.1. D ug Dis ibu ion and Solid-S a e Cha ac e is ics
Raman spec oscopy has p e iously been used o e alua e he spa ial dis ibu ion and phase o
d ugs wi hin able s, and as such, was used he e o map a c oss-sec ional su ace o a mul i-laye ed
3D- p in ed la polypill [61]. P ocessing o he a ays using DCLS componen ma ching o p oduce
alse colou ep esen a ions o dis ibu ion shows he p esence o he six d ugs wi hin he six laye s
o he polypill (Figu e 3), highligh ing he success in u ilising SLA o p in sepa a e esin o mula ions
wi hin sepa a e compa men s.
Figu e 3. Visual imaging o a Type I polypill, using (a) op ical ligh mic oscopy and (b) Raman
mapping. The images show he spa ial sepa a ion o laye s.
Figu e 2.
Polypill p in le (
a
) Type I (cylinde shape) and (
b
) Type II ( ing shape). The Type III
o mula ion was isually iden ical o Type II, and hence, has no been included he e. The scale is in cm.
3.2. Physical Cha ac e is ics
3.2.1. D ug Dis ibu ion and Solid-S a e Cha ac e is ics
Raman spec oscopy has p e iously been used o e alua e he spa ial dis ibu ion and phase o
d ugs wi hin able s, and as such, was used he e o map a c oss-sec ional su ace o a mul i-laye ed
3D-p in ed la polypill [
61
]. P ocessing o he a ays using DCLS componen ma ching o p oduce
alse colou ep esen a ions o dis ibu ion shows he p esence o he six d ugs wi hin he six laye s o
he polypill (Figu e 3), highligh ing he success in u ilising SLA o p in sepa a e esin o mula ions
wi hin sepa a e compa men s.
Howe e , de ailed in e oga ion o indi idual Raman spec a wi hin he mapped a eas leads us o
no e ha he e is e idence o “di usion” o ce ain d ugs (nap oxen, aspi in, and pa ace amol) be ween
he laye s which was no an icipa ed h ough isual examina ion o he whi e ligh mic oscopic image
which shows a dis inc bounda y be ween each. Fu he mapping ac i i ies we e pe o med, ocusing
on he h ee laye s con aining nap oxen, aspi in, and pa ace amol. These spec al a ays we e acqui ed
wi h a signi ican ly highe esolu ion o e a educed a ea, i.e., 1 mm in xby 3 mm in he s-dimensions o
be e unde s and he dis ibu ion o d ug in his a ea (Sec ion 2.3). Figu e 4a–c show DCLS p ocessing
Pha maceu ics 2019,11, 274 7 o 16
o hese a ays using pu e p in ed d ug e e ences (i.e., con aining one d ug plus p in ing ma ix)
o demons a e his di usion e ec be ween hese h ee laye s. I is easonably clea o see ha he
p inciple d ug con en is wi hin he laye con aining ha d ug, bu he e is e idence o he d ug
di using in o he nex laye s wi h an a enua ing signal, i.e., a diminishing concen a ion.
Pha maceu ics 2019, 11, x FOR PEER REVIEW 6 o 16
unc ions, commands, p o ocols, and objec s ha allows p og amme s o c ea e so wa e o in e ac
wi h an ex e nal sys em ( he 3D p in e in his case) wi hou ha ing o w i e a code om sc a ch.
He e, he so wa e was modi ied o include command inpu s ha enabled he ollowing six s eps o
be ca ied ou : (1) he esin o mula ion was p in ed using SLA; (2) he p in ing p ocess was paused
upon laye comple ion; (3) he build pla o m was aised, enabling esin ay emo al; (4) he esin
ay was eplaced which included a di e en esin o mula ion; (5) he build pla e was lowe ed o i s
p e ious posi ion and; (6) p in ing was esumed o c ea e he nex o mula ion laye .
In his way, mul iple polypill p in le s could be easily ab ica ed in 30 min, wi h he o de o
d ugs in he laye s con olled by he esin o mula ion in he ank a any pa icula poin . The
cus omised p in se ings (wi h six lase passes o he i s laye o ensu e adhesion and wo o he
es ) used allowed he success ul p oduc ion o p in le s di ec ly on he build pla o m, achie ing
good adhesion wi hou signi ican ly a ec ing he dimensions. C ucially, his app oach a oids
ma e ial was age and po en ial dose a ia ion compa ed wi h o he me hods ha u ilise suppo s o
adhesion o he build pla o m ha need o be emo ed and disca ded pos -p in ing.
Figu e 2. Polypill p in le (a) Type I (cylinde shape) and (b) Type II ( ing shape). The Type III
o mula ion was isually iden ical o Type II, and hence, has no been included he e. The scale is in
cm.
3.2. Physical Cha ac e is ics
3.2.1. D ug Dis ibu ion and Solid-S a e Cha ac e is ics
Raman spec oscopy has p e iously been used o e alua e he spa ial dis ibu ion and phase o
d ugs wi hin able s, and as such, was used he e o map a c oss-sec ional su ace o a mul i-laye ed
3D- p in ed la polypill [61]. P ocessing o he a ays using DCLS componen ma ching o p oduce
alse colou ep esen a ions o dis ibu ion shows he p esence o he six d ugs wi hin he six laye s
o he polypill (Figu e 3), highligh ing he success in u ilising SLA o p in sepa a e esin o mula ions
wi hin sepa a e compa men s.
Figu e 3. Visual imaging o a Type I polypill, using (a) op ical ligh mic oscopy and (b) Raman
mapping. The images show he spa ial sepa a ion o laye s.
Figu e 3.
Visual imaging o a Type I polypill, using (
a
) op ical ligh mic oscopy and (
b
) Raman mapping.
The images show he spa ial sepa a ion o laye s.
Pha maceu ics 2019, 11, x FOR PEER REVIEW 7 o 16
Howe e , de ailed in e oga ion o indi idual Raman spec a wi hin he mapped a eas leads us
o no e ha he e is e idence o “di usion” o ce ain d ugs (nap oxen, aspi in, and pa ace amol)
be ween he laye s which was no an icipa ed h ough isual examina ion o he whi e ligh
mic oscopic image which shows a dis inc bounda y be ween each. Fu he mapping ac i i ies we e
pe o med, ocusing on he h ee laye s con aining nap oxen, aspi in, and pa ace amol. These
spec al a ays we e acqui ed wi h a signi ican ly highe esolu ion o e a educed a ea, i.e., 1 mm in
x by 3 mm in he s-dimensions o be e unde s and he dis ibu ion o d ug in his a ea (Sec ion 2.3).
Figu e 4a–c show DCLS p ocessing o hese a ays using pu e p in ed d ug e e ences (i.e., con aining
one d ug plus p in ing ma ix) o demons a e his di usion e ec be ween hese h ee laye s. I is
easonably clea o see ha he p inciple d ug con en is wi hin he laye con aining ha d ug, bu
he e is e idence o he d ug di using in o he nex laye s wi h an a enua ing signal, i.e., a
diminishing concen a ion.
Figu e 4. Raman mapping o a Type I polypill ac oss he nap oxen, aspi in, and pa ace amol laye s
wi h an inc eased spa ial esolu ion (30,000 spec a ac oss 1 mm in x by 3 mm in y. (a) Shows he
pa ial di usion o pa ace amol in o he adjacen laye s; (b) shows he pa ial di usion o aspi in in o
he adjacen laye s; and (c) shows he pa ial di usion o nap oxen in o he adjacen laye s.
Con e sely, ca eine and p ednisolone we e localised solely wi hin hei espec i e laye s, wi h
no e idence o any di usion. Fu he mo e, hese d ugs ha e appea ed o ac as a ba ie o di usion
o he o he laye s, e.g., he e is no e idence o he pa ace amol di using in o he ca eine laye ,
whe eas i does di use in o he aspi in laye . E idence o his di e en ial di usion e ec is p esen ed
in Figu e 5, whe e dis ibu ion o each d ug is e alua ed by plo ing DCLS ma ch ac oss he y-
dimension o he mapped polypill.
Figu e 4.
Raman mapping o a Type I polypill ac oss he nap oxen, aspi in, and pa ace amol laye s
wi h an inc eased spa ial esolu ion (30,000 spec a ac oss 1 mm in xby 3 mm in y. (
a
) Shows he pa ial
di usion o pa ace amol in o he adjacen laye s; (
b
) shows he pa ial di usion o aspi in in o he
adjacen laye s; and (c) shows he pa ial di usion o nap oxen in o he adjacen laye s.
Con e sely, ca eine and p ednisolone we e localised solely wi hin hei espec i e laye s, wi h no
e idence o any di usion. Fu he mo e, hese d ugs ha e appea ed o ac as a ba ie o di usion o
he o he laye s, e.g., he e is no e idence o he pa ace amol di using in o he ca eine laye , whe eas i
does di use in o he aspi in laye . E idence o his di e en ial di usion e ec is p esen ed in Figu e 5,
whe e dis ibu ion o each d ug is e alua ed by plo ing DCLS ma ch ac oss he y-dimension o he
mapped polypill.
Pha maceu ics 2019,11, 274 8 o 16
Pha maceu ics 2019, 11, x FOR PEER REVIEW 8 o 16
Figu e 5. D ug dis ibu ion p o iles in he Y-dimension o he polypill showing he di usion be ween laye s in he pa ace amol, aspi in, and nap oxen laye s, wi h a igh
dis ibu ion in he ca eine and p ednisolone laye s.
Figu e 5.
D ug dis ibu ion p o iles in he Y-dimension o he polypill showing he di usion be ween laye s in he pa ace amol, aspi in, and nap oxen laye s, wi h a
igh dis ibu ion in he ca eine and p ednisolone laye s.
Pha maceu ics 2019,11, 274 9 o 16
This phenomenon was hypo hesised o be due o he phase o he d ugs wi hin he p in ed
polypill; pos -p in ing, he laye s con aining pa ace amol, nap oxen, aspi in, and chlo amphenicol
we e isually clea wi h a glassy appea ance and he p ednisolone and ca eine o mula ions we e
whi e (opaque), which was an ini ial indica o o di e ences in solid-s a e cha ac e is ics (Figu e 2).
These indings we e u he in e oga ed using XRPD (Figu e 6).
Indeed, ou o he d ugs (aspi in, pa ace amol, nap oxen, and chlo amphenicol) we e ound
o be in he amo phous phase due o he absence o sha p peaks in he XRPD spec a (Figu e 6b,d–
espec i ely). Con e sely, p ednisolone and ca eine we e ound o be p esen in he c ys alline phase
(Figu e 6a,c espec i ely). Speci ically, se e al c ys alline peaks we e ound pos -p in ing o ca eine
(a 12.6, 27.2, and 28.0 2
θ
) and o p ednisolone, one c ys alline peak was obse ed a 16.4 2
θ
. In bo h
cases, consis en peak shi s o ~+1 2
θ
was appa en , which was a ibu ed o he s ess–s ain in luence,
o he change in heigh p esen a ion, o a p in ed disc e sus he aw powde .
P e ious s udies ha e highligh ed ha amo phous d ug ma e ials ha e a highe p opensi y o
di use ac oss polyme ic ma ices [
62
]. As such, i is likely ha in his s udy he amo phous d ugs
(aspi in, pa ace amol, nap oxen, and chlo amphenicol) a e di using ac oss he laye s mo e eadily
compa ed wi h he c ys alline d ugs (ca eine and p ednisolone), which emain in hei espec i e
laye s. S abilising d ugs in hei amo phous phase as a solid dispe sion is a ou able o low solubili y
d ugs due o he po en ial o an inc ease in d ug solubili y and bioa ailabili y.
3.2.2. P in le Dimensions and Weigh Va ia ion
In o de o e alua e he e ec o d ug addi ion on he esin p in abili y, he consis ency in weigh
and dimensions o he polypill p in le s was e alua ed ( a ge dimensions: 3 mm
×
10 mm o he
cylinde s and 6 mm
×
10 mm o he ings) (Table 2). In gene al, all he o mula ions yielded sligh ly
wide diame e s han hei co esponding a ge s, anging om 10.73 mm o 11.07 mm. In gene al,
heigh and weigh a ia ion we e highe o Type I cylind ical p in le s compa ed wi h Type II and III
ing-shaped p in le s. This a iabili y in mass could be due o he mul iple ac o s bo h om he liquid
o mula ion and he se ings o he p in e . The numbe o lase passes o each p in ed laye and he
lase powe di ec ly a ec he cu ing dep h, and hence, he p ope ies o he p in ed laye [
63
]. Hence,
he pa ame e s need o be op imised o each esin ype. I should also be no ed ha he di e ence
in he a ge and eal dimensions could be adjus ed by simply scaling he elec onic objec . HPLC
was used o e alua e d ug con en o he polypills pos -p in ing. D ug loading was ound o ange
be ween 85–104%, which is wi hin he accep able ange o con en uni o mi y (85–115%) se by he
B i ish Pha macopoeia.
Table 2. Dimension and weigh da a o he polypills.
Type I
Wid h (mm) ±%CV Heigh ±SD (mm) Weigh ±SD (mg)
10.99 ±1.0 2.81 ±9.8 329 ±13.6
Type II
Wid h (mm) ±%CV Heigh ±SD (mm) Weigh ±SD (mg)
11.07 ±0.1 6.12 ±0.03 501.13 ±6.3
Type III
Wid h (mm) ±%CV Heigh ±SD (mm) Weigh ±SD (mg)
10.73 ±0.18 6.12 ±0.02 553 ±8.9
Pha maceu ics 2019,11, 274 16 o 16
58.
Goyanes, A.; Ha on, G.B.; Me chan , H.A.; Basi , A.W. Gas oin es inal elease beha iou o modi ied- elease
d ug p oduc s: Dynamic dissolu ion es ing o mesalazine o mula ions. In . J. Pha m.
2015
,484, 103–108.
[C ossRe ]
59.
Me chan , H.A.; F os , J.A.; Basi , A.W. Appa a us and Me hod o Tes ing Medicamen s. U.S. Pa en
15/392,113, 20 Ap il 2014.
60.
Me chan , H.A.; Goyanes, A.; Pa asha , N.; Basi , A.W. P edic ing he gas oin es inal beha iou o
modi ied- elease p oduc s: U ili y o a no el dynamic dissolu ion es appa a us in ol ing he use o
bica bona e bu e s. In . J. Pha m. 2014,475, 585–591. [C ossRe ]
61.
T en ield, S.J.; Goyanes, A.; Tel o d, R.; Wilsdon, D.; Rowland, M.; Gais o d, S.; Basi , A.W. 3D p in ed d ug
p oduc s: Non-des uc i e dose e i ica ion using a apid poin -and-shoo app oach. In . J. Pha m.
2018
,549,
283–292. [C ossRe ]
62.
Yonemochi, E.; Sano, S.; Yoshihashi, Y.; Te ada, K. Di usi i y o amo phous d ug in solid dispe sion. J.
The m. Anal. Calo im. 2013,113, 1505–1510. [C ossRe ]
63.
Co cione, C. De elopmen and Cha ac e iza ion o No el Pho opolyme izable Fo mula ions o
S e eoli hog aphy. J. Polym. Eng. 2014,34, 85–93. [C ossRe ]
64.
Rao, V.M.; Haslam, J.L.; S ella, V.J. Con olled and comple e elease o a model poo ly wa e -soluble
d ug, p ednisolone, om hyd oxyp opyl me hylcellulose ma ix able s using (SBE)7m-
β
-cyclodex in as a
solubilizing agen . J. Pha m. Sci. 2001,90, 807–816. [C ossRe ] [PubMed]
65.
Di Colo, G.; Baggiani, A.; Zambi o, Y.; Mollica, G.; Geppi, M.; Se a ini, M.F. A new hyd ogel o he ex ended
and comple e p ednisolone elease in he GI ac . In . J. Pha m. 2006,310, 154–161. [C ossRe ] [PubMed]
©
2019 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access
a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion
(CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).