In e na ional Jou nal o
En i onmen al Resea ch
and Public Heal h
A icle
In luence o Connec ion Type and Pla o m Diame e
on Ti anium Den al Implan s Fa igue: Non-Axial
Loading Cyclic Tes Analysis
Ana I. Nicolas-Sil en e 1, Eugenio Velasco-O ega 2,*, I an O iz-Ga cia 2,
Al a o Jimenez-Gue a 2, Lo e o Monsal e-Guil 2, Raul Ayuso-Mon e o 3, Ja ie Gil 4
and Jose Lopez-Lopez 5
1Pe iodon al and Implan Su ge y, CEIR Campus Ma e Nos um, School o Den is y, Uni e si y o Mu cia,
30008 Mu cia, Spain; [email p o ec ed]
2Comp ehensi e Den is y o Adul s and Ge odon ology, Facul y o Den is y, Uni e si y o Se ille,
41009 Se ille, Spain; [email p o ec ed] (I.O.-G.); [email p o ec ed] (A.J.-G.);
[email p o ec ed] (L.M.-G.)
3Facul y o Den is y, Uni e si y o Ba celona, 08007 Ba celona, Spain; [email p o ec ed]
4Chai man o Bioenginee ing Ins i u e o Technology, Uni e si a In e nacional de Ca alunya,
08017 Ba celona, Spain; xa ie [email p o ec ed]
5Facul y o Den is y, Se ice o he Medical-Su gical A ea o Den is y Hospi al, Uni e si y o Ba celona,
08007 Ba celona, Spain; [email p o ec ed]
*Co espondence: [email p o ec ed]
Recei ed: 1 No embe 2020; Accep ed: 30 No embe 2020; Published: 2 Decembe 2020


Abs ac :
Two-pieces den al implan s mus p o ide s abili y o he implan -abu men -in e ace.
The connec ion ype and pla o m diame e could in luence he biomechanical esis ance and s ess
dis ibu ion. This s udy aims o e alua e he a igue o di e en ypes o connec ions, ex e nal and
in e nal, and di e en pla o m diame e s. Th ee implan designs wi h he same leng h we e used:
(a) ex e nal hexagon/na ow pla o m; (b) in e nal double hexagon/na ow pla o m; (c) in e nal
oc agon/ egula pla o m. A a igue es was de eloped o es ablish he numbe o cycles needed
be o e ac u e. A 30
º
oblique load wi h a sinusoidal unc ion o a igue a a equency o 15 Hz and
10% s ess a ia ion was applied o each sys em. The a igue load limi (FLL) o design (a) was 190 N,
being he nominal-cu a u e-momen (NCM) =1.045; FLL =150 N, wi h a NCM =0.825 o (b), and
FLL =325 N, wi h a NCM =1.788 o (c). The pla o m diame e a ec s he FLL, ob aining lowe FLL
on a na ow pla o m. The connec ion ype in e e es wi h he implan walls’ wid h, especially in
na ow implan s, making in e nal connec ions mo e uns able a his le el. Long- e m clinical s udies
o assess he es o a ion’s success a e and su i al a e manda o y.
Keywo ds: connec ion ype; den al implan s; design; a igue es ; pla o m diame e
1. In oduc ion
Two-piece den al i anium implan s ha e been widely used o single- oo h eplacemen s up o
ull-a ch ehabili a ion [
1
]. The implan -abu men -in e ace (IAI) [
2
] has o esis dynamic o ces and
be s able o wi hs and unc ional loads and o educe sc ew loosing [
3
]. To main ain he s abili y o
he IAI, di e en implan connec ion geome ies ha e been de eloped, which can be summa ized in
wo b oad g oups: ex e nal and in e nal connec ions. In e ms o design, in he ex e nal connec ions,
bo h he implan index and he p os he ic abu men index a e loca ed abo e he le el o he implan
pla o m. In con as , in he in e nal connec ions, bo h s uc u es a e loca ed inside he implan ’s body,
below he le el o he implan pla o m [4].
In . J. En i on. Res. Public Heal h 2020,17, 8988; doi:10.3390/ije ph17238988 www.mdpi.com/jou nal/ije ph
In . J. En i on. Res. Public Heal h 2020,17, 8988 2 o 12
The ex e nal connec ions a e usually p o ided o an ou e hexagon whose unc ion is o p o ide
o a ional o que con ol du ing implan placemen [
5
] and an i- o a ional con ol be ween he implan
index and he p os he ic abu men index. Se e al s udies indica e ha his ype o connec ion is less
a o able o s ess dis ibu ion and has lowe s abili y when compa ed o he in e nal connec ion [
6
].
The in e nal connec ion can p esen di e en designs depending on i s geome ic ea u es and
can be di ided in o an in e nal hexagon, in e nal oc agon, ilobed sys em, o mo se ape connec ion
be ween o he s [
7
,
8
]. This ype o design inc eases he implan -abu men con ac a ea and imp o es
he dis ibu ion and dissipa ion o o ces, p o iding be e s abili y [
9
], bu is he in e nal conical
connec ion ( ape Mo se connec ion) he one ha shows he mos in ima e ela ion be ween he implan
and he abu men , p o iding he mos excellen s abili y and bac e ial seal [10,11].
The p esence o di e en design ea u es (angles, channels, s aigh walls, cones, ubes) in he
di e se connec ions ypes p e en s o a ion be ween he componen s o he sys em [
12
]. I s s uc u al
in eg i y is c ucial o long- e m s abili y [
13
], and some ac o s could induce de o ma ion when he
sys em is o e loaded, o e - o quing o non-axial o ces a e p esen ed [
14
]. The hickness o he implan
walls is a ele an ac o since, some imes, he design o he an i- o a ional componen s inside he
connec ion o ces o lea e walls excessi ely hin, especially in na ow implan s [15,16].
The use o na ow implan s is widely documen ed in pa ien s wi h de icien bone c es al wid h
in which, o some eason (inc eased healing ime, cos , o pa ien mo bidi y), he applica ion o
ho izon al bone egene a i e echniques is no indica ed [
17
–
19
]. The mechanical s eng h o i anium
na ow implan s is some imes no enough o suppo he dynamic o ces. The implan sys em does no
o e long- e m in eg i y o he connec ion complex, ep esen ing a signi ican isk o ac u es [
20
,
21
].
Some aspec s o he di e en connec ion con igu a ions, such as biomechanical esis ance and s ess
dis ibu ion, a e c ucial [
22
]. No much is known abou implan a igue de ailed by he ype o
connec ion. Hence, he objec i e o he p esen s udy was o e alua e he a igue o di e en kinds o
connec ions, ex e nal and in e nal, and di e en pla o m diame e s, o es ablish which ype o design
suppo ed highe alues. Ou null hypo hesis was ha indexa ion design and pla o m diame e
in luences i anium implan a igue in he long- e m.
2. Ma e ials and Me hods
2.1. Den al Implan s
Fi y- ou i anium den al implan s om h ee di e en implan sys ems we e compa ed in his
s udy. The cha ac e is ics o each implan g oup a e summa ized in Table 1and each implan design is
exposed in Figu e 1:
-
G oup I (n =19): Su gimplan CE: i anium g ade 5 den al implan wi h hexagon ex e nal
connec ion (pla o m: 3.5 mm, leng h: 12 mm) (Galimplan SLU, Sa ia, Lugo, Spain)
-
G oup II (n =18): Su gimplan CI Double Hexagon: i anium g ade 5 den al implan wi h
double hexagon in e nal connec ion (pla o m: 3.5 mm, leng h: 12 mm) (Galimplan SLU, Sa ia,
Lugo, Spain)
-
G oup III (n =17): Su gimplan CI Oc agonal: i anium g ade 5 den al implan wi h oc agonal
in e nal connec ion (pla o m: 4.0 mm, leng h: 12 mm) (Galimplan SLU, Sa ia, Lugo, Spain).
In . J. En i on. Res. Public Heal h 2020,17, 8988 3 o 12
Table 1. Implan cha ac e is ics dis ibu ed by g oups.
G oup G oup I G oup II G oup III
n 19 18 17
Connec ion Type Hexagon Ex e nal
Connec ion
Double Hexagon
In e nal Connec ion
Oc agonal In e nal
Connec ion
Diame e 3.5 3.5 4.0
Leng h 12 12 12
Ma e ial Ti anium G ade 5 Ti anium G ade 5 Ti anium G ade 5
Figu e 1. Implan design o each expe imen al g oup.
2.2. Fa igue Tes
A a igue es was pe o med o ob ain he numbe o cycles be o e ac u e. The maximum
and minimum o ce applied was eco ded o each sample. The assays we e pe o med wi h a
se o-hyd aulic es ing machine (MTS 858 Mini Bionix II, MTS, Minneapolis, MN, USA) equipped
wi h a load cell MTS 661.19F-01 o 5 kN.
The sphe ical membe o he load applica ion was made o i anium g ade 5 (Figu e 2).
In . J. En i on. Res. Public Heal h 2020,17, 8988 4 o 12
Figu e 2. Sphe ical membe o he load applica ion design de ails.
The implan s we e ixed 30
◦
angula ed wi h he axis z o he load cell (Figu e 3). They we e
loaded wi h a sinusoidal unc ion o a igue a a equency o 15 Hz and 10% s ess a ia ion. The e o
du ing loading measu emen s was less han 5 N, and he maximum loading applied o he implan was
a ound 80% o he alue o he implan ailu e load, ob ained by a s a ic es unde he same geome ic
condi ions as a igue es s, ollowing ISO 14801:2008 ecommenda ions [
23
]. All es s we e ca ied ou
unde s able en i onmen al condi ions wi h a empe a u e o 25 ◦C and ela i e humidi y o 60%.
Figu e 3. Load cell o e he sample in he es ing machine.
In . J. En i on. Res. Public Heal h 2020,17, 8988 5 o 12
2.3. Scanning Elec on Mic oscopy (SEM) Analysis
The ac u e samples we e obse ed by SEM a 10 kV using a Neon 40 Focused Ion Beam Scanning
Elec on (FIB-SEM) mic oscope (Ca l Zeiss NTS GmbH, Obe kochen, Ge many).
2.4. S a is ical Analysis
S a is ically signi ican di e ences among he h ee g oups we e assessed using SPSS 18.0 so wa e
(SPSS Inc., Chicago, IL, USA). Di e ences be ween g oups we e analyzed by Analysis o Va iance
(ANOVA), and a mul iple compa ison Fishe es was applied. The le el o signi icance was es ablished
a a p- alue o 0.05.
3. Resul s
The ailu e mode was simila in all expe imen al g oups, including la ge de o ma ions a he
implan neck a ea. The implan neck ac u e ook place mos o he cases be ween he i s and
second h eads.
3.1. Hexagon Ex e nal Connec ion
The minimum and maximum load exp essed in New ons (N) applied o all he samples o he
hexagon ex e nal connec ion g oup was 190 N and 400 N, espec i ely. The cycles applied be o e
ac u e we e be ween 3074 and 5,000,000. The cyclic load diag am ob ained om he esul s o he es
is shown in Figu e 4.
Figu e 4.
Cyclic load diag am o hexagon ex e nal connec ion ob ained om he esul s o he es
showed in Table 2.

In . J. En i on. Res. Public Heal h 2020,17, 8988 6 o 12
Table 2. Summa y o he esul s ob ained in each expe imen al g oup.
Implan
Type
Minimum
Load (N)
Maximum
Load (L)
Minimum
Cycles
Maximum
Cycles
Fa igue
Load Limi
(FFL) (N)
Nominal
Cu a u e
Momen
(N.m)
Hexagon
ex e nal
connec ion
190 400 3074 5,000,000 190 1.045
Double
hexagon
in e nal
connec ion
150 400 1583 5,000,000 150 0.825
Oc agonal
in e nal
connec ion
325 550 3555 5,000,000 325 1.788
The a igue load limi (F
FL
, acco ding o ISO 14801:2008) was F
FL
=190 N, being he nominal
cu a u e momen (N.m) =1.045.
3.2. Double Hexagon In e nal Connec ion
The minimum and maximum load exp essed in New ons (N) applied o all he samples o he
double hexagon in e nal connec ion g oup was 150 N and 400 N, espec i ely. The cycles applied
be o e ac u e we e be ween 1583 and 5,000,000. The cyclic load diag am ob ained om he esul s o
he es is shown in Figu e 5.
Figu e 5.
Cyclic load diag am o double hexagon in e nal connec ion g oup ob ained om he esul s
o he es .
The a igue load limi (F
FL
, acco ding o ISO 14801:2008) was F
FL
=150 N, being he nominal
cu a u e momen (N.m) =0.825.
3.3. Oc agonal In e nal Connec ion
The minimum and maximum load exp essed in New ons (N) applied o all he samples o he
oc agonal in e nal connec ion g oup and he cycles applied be o e ac u e we e 325 N and 550 N
In . J. En i on. Res. Public Heal h 2020,17, 8988 7 o 12
espec i ely. The cycles applied be o e ac u e we e be ween 3555 and 5,000,000. The cyclic load
diag am ob ained om he esul s o he es is shown in Figu e 6.
Figu e 6.
Cyclic load diag am o he oc agonal in e nal connec ion g oup ob ained om he esul s o
he es .
The a igue load limi (F
FL
, acco ding o ISO 14801:2008) was F
FL
=325 N, being he nominal
cu a u e momen (N.m) =1.788.
A summa y o he esul s o he h ee expe imen al g oups is shown in Table 2.
The lack o e en ion be ween he abu men and den al implan was assessed as a ailu e.
The ac u e mechanism s a s by abu men sc ew loosening p oducing c acks on he su ace ha g ow
wi h he load cycles and la e ac u e, bu no due o des uc ion o he implan neck o shoulde .
Analysis o ac u ed sc ews by SEM e ealed ha he mode and he egion o ac u e we e he same o
he di e en sys ems s udied. The ac u e su aces we e simila o all implan s co esponding o he
connec ion zone and ac u ed he body o he implan , acco ding o he indica ions o he in e na ional
s anda ds o ac u e a igue beha io [24] o he den al implan s wi h good mechanical beha io .
S a is ically, he hexagonal ex e nal connec ion p esen ed a lowe a igue limi load wi h s a is ical
di e ences signi icance han he double hexagonal in e nal connec ion (p<0.012) and also in ela ion
o he oc agonal in e nal connec ion (p<0.003). When bo h in e nal connec ions a e compa ed,
he oc agonal connec ion p esen s a highe limi a igue load han he double hexagonal connec ion
wi h s a is ical di e ences signi icance (p<0.004).
The s ia ions om he ac og aphy can be obse ed in Figu e 7, whe e he c ack g ows om he
su ace specimens and om 10-mm benea h he su ace. In all cases, we obse ed he same mo phology
o ac u e. The equiaxed g ains can be obse ed, and in hei aces, he ma ks o he c ack in he
p opaga ion p ocess o ac u e.
In . J. En i on. Res. Public Heal h 2020,17, 8988 8 o 12
Figu e 7.
SEM images a a magni ica ion o
×
500 and
×
1500 showing he s ia ions om he ac og aphy.
4. Discussion
This expe imen al s udy aimed o e alua e he in luence o he connec ion ea u es and pla o m
diame e in he a igue esponse o i anium g ade 5 den al implan s. Th ee di e en implan sys ems
we e assessed, one na ow implan sys em (3.5 mm pla o m) wi h an ex e nal hexagonal connec ion,
one na ow implan sys em (3.5 mm pla o m) wi h double hexagon in e nal connec ion, and one
egula pla o m (4.0 mm) implan sys em wi h an oc agonal in e nal connec ion.
The implan -abu men in e ace geome y is an in luencing ac o o he ansmission o s ess
a ound he implan [25].
This expe imen al es is a eliable me hod o de e mine he e ec o di e en pa ame e s,
such as connec ion design o pla o m diame e , on implan dynamic ailu e s eng h. The same
company manu ac u ed he h ee expe imen al g oups and he h ee sphe ical membe s a ached o
he connec ion, using he same i anium g ade 5. This ac is one o he s eng hs o he p esen s udy.
In mos s udies, au ho s compa e ailu e s eng h be ween di e en implan s and di e en abu men
in e aces, wi h di e en shapes, su ace cha ac e is ics, and ma e ial p ope ies, and he compa abili y
is comp omised [26].
In addi ion o he ac o s men ioned abo e, some ac o s inhe en in he hos may a ec he
dis ibu ion o s ain and s ess in bone and implan s. A s udy de eloped by Oli ei a e al. concluded
ha he densi y o he medulla bone and he hickness o he co ical bone also a ec he dis ibu ion
o s ain and s ess, nega i ely a ec ing he dec ease in medulla bone densi y [27].
Lo Giudice e al. demons a ed ha he bone p epa a ion could also a ec he bone quali y
showing be e esul s in os eo omies pe o med wi h ul asonic ips and concluding ha he use o
he piezosu ge y p ese es he bone mo phology and dec eases he p esence o mic o ac u es [
28
].
The ma ginal bone loss a ound implan s is also in luenced by he ac he implan is placed in na i e
bone o placed in g a ed issues. Galindo-Mo eno e al. ound in a e ospec i e coho s udy ha
implan s placed in g a ed issues showed mo e ma ginal bone loss han implan s placed in p is ine
bone [
29
]. The ype o connec ion also a ec s he ma ginal bone s abili y, being he ex e nal connec ions
s ongly associa ed wi h an inc eased ma ginal bone loss, no only in he i s wel e mon hs bu o e
ime [29].
Also, bone quali y and c es al bone loss can be in luenced by o he ac o s. No only does he
neck shape, mic o h eads, o su ace ex u e a ec c es al bone s abili y, bu he implan -abu men
connec ion appea s o be a signi ican ac o on pe i-implan c es al bone le el [
30
]. The abu men
heigh also has an impo an ole, as demons a ed in an
in i o
s udy de eloped by Spina o e al.
They sugges ed ha he sho e he abu men heigh , he g ea e he ma ginal bone loss, especially in
cemen - e ained p os hesis [31].
In . J. En i on. Res. Public Heal h 2020,17, 8988 9 o 12
Se e al s udies ha e es ed den al implan s using s a ic loading, while o he s use cyclic loads [
32
].
Mos o he e iews ocus on implan design bu does no men ion a igue as a complex ailu e
mode [
22
]. A ew s udies ha e conside ed he e ec o he implan diame e on a igue pe o mance,
concluding ha na ow implan s ailed o show ypical a igue beha io , which migh be a ibu ed o
he implan design [
33
]. The inconsis en a igue beha io obse ed o na ow implan s could esul
om ac o s like no ches, den s, o machining ma kings. To da e, no s udies ha e been published
abou a ac u e mode analysis o suppo his assump ion [33].
Ca nei o e al. [34] de eloped an in i o s udy e alua ing he ac u e esis ance o in e nal and
ex e nal hexagon in egula and na ow implan s, concluding ha i anium is a ma e ial ha p esen s
no clea e idence o he exac poin be ween he plas ic and elas ic limi s. No signi ican educ ion o
he blending elas ic limi was ound be ween na ow and egula in e nal connec ions.
In ou s udy, he ailu e due o he bending elas ic limi was obse ed a 190 N wi h ex e nal
hexagon na ow implan s e sus he 150 N esul ed in he in e nal hexagon na ow implan s. This esul
could be because, in he in e nal connec ion, he indexa ion ea u es a e ubica ed inside he implan ’s
body, lea ing hinne walls han ex e nal connec ions. Besides, an impo an cause o he high a igue
li e o he oc agonal in e nal connec ion is he size o he esis an sec ion. The double hexagonal
in e nal connec ion and ex e nal sys em p esen a highe alue o he a ea han he in e nal.
Ou es simula ed he clinical si ua ion when he s ess concen a ion esul ing om occlusal
o ces leads o mic o ac u es and bone loss a ound he implan , leading o mobili y and ac u e o he
implan [35,36]. On he o he hand, ou esul s showed an inc ease om 150 N o 325 N in he elas ic
limi by inc easing he pla o m diame e in he in e nal connec ion om 3.5 o 4 mm. This di e ence
be ween he pla o m diame e was no signi ican in he esul s showed by Ca nei o e al., al hough
hey ound a mo e subs an ial numbe o c acks in he na ow implan s han he egula .
Tole ances o manu ac u ing a e he main eason o he loose- i o he componen s and equi ed
he manu ac u e o imp o e he i . In hese si ua ions (loose- i ), he possibili y o ho izon al mo emen
and o a ion be ween sc ew and implan and lowe han he o ces o igh en i , mic omo emen s
could ha e led o a p og essi e unsc ewing o he abu men sc ew unde condi ions dynamic loading.
The mos cause o he high a igue li e o he ex e nal connec ion is he size o he esis an sec ion.
The ex e nal sys em p esen s a lowe alue o he su ace han he in e nal. This ac p oduces a wo se
load dis ibu ion. This eason explains he di e ences in mechanical p ope ies. Besides, he ole ances
in he in e nal connec ions a e be e , and his good inishing p o okes a highe a igue limi o he
in e nal connec ion sys em [37–39].
Each implan -abu men in e ace has i s ad an ages and disad an ages. Acco ding o
Maeda e al. [40],
he ex e nal hexagon in e ace has bene i s such as sui abili y o he wo-s age
me hod, p o ision o an an i- o a ion mechanism, e ie abili y, and compa ibili y among di e en
sys ems. Howe e , inc eased sc ew loosening, componen ac u e, and di icul y in sea ing abu men s
in deep subgingi al issues a e p oblems commonly expe ienced wi h ex e nal hexagon connec o s [
41
].
The ad an ages o he in e nal hexagon ollowing Maeda a e ease in abu men connec ion,
sui abili y o one s age implan ins alla ion, highe s abili y and sui abili y o single- oo h es o a ion,
highe esis ance o la e al loads due o he lowe cen e o o a ion, and be e o ce dis ibu ion.
The mas ica o y loading a an e io egions is a iable wi h a mean alue o 286 N, s.d. 164 N,
while he pos e io a ea shows a mean alue o 579 N, s.d. 235 N [
42
]. Those da a showed a high
subjec a iabili y so ha he use o na ow implan s is ecommended jus o he an e io egion.
In he pos e io a ea, i is manda o y o use a wide pla o m.
The p esen s udy esul s suppo he accep ance o he null hypo hesis es ed since he e was
a di e ence in he maximum o ce suppo ed in na ow implan s (in e nal o ex e nal connec ions)
and egula pla o m implan s. Di e en ypes o connec ions also p esen ed di e ences in he a igue
load limi . Clinical s udies a e manda o y o es he s abili y o he di e en connec ions e alua ed,
assessing he success a e and su i al o he p os hesis in an e io and pos e io ee h.