The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 773

Long-Term Implant Survival in the

Grafted Maxillary Sinus:
A Systematic Review

Massimo Del Fabbro, BSc, PhD1 Maxillary sinus augmentation has

Stephen S. Wallace, DDS2 been shown to be the most pre-
Tiziano Testori, MD, DDS3 dictable of the preprosthetic sur-
gical techniques used to enhance
The predictability of maxillary sinus augmentation has been extensively reported.
bone volume for the placement of
Procedural outcomes, most often measured as implant survival rates, have
dental implants in previously com-
customarily used inclusion criteria that included a minimum 1-year loading
time. The inclusion criteria of this review extended the minimum postloading promised sites.1 Multiple evidence-
evaluation to 3 years to determine if the previously reported short-term survival based reviews have demonstrated
rates are maintained. An electronic search of the literature was performed and that sinus floor augmentation re-
retrieved articles were screened using specific inclusion criteria, paramount of sults in an overall implant survival
which was a minimum of 3 years of follow-up. The search revealed 18 articles
rate well beyond 90%.2–10 Most
for the lateral window approach (6,500 implants in 2,149 patients) and 7 for
clinical studies included in these
the transalveolar approach (1,257 implants in 704 patients). Overall, implant
survival after a minimum of 3 years loading was 93.7% and 97.2% for the reviews, and therefore the reviews
lateral window and transalveolar approaches, respectively. Of importance is themselves, included many cases
the fact that 80% of failures occurred within the first year and 93.1% of the with a relatively short follow-up
failures occurred within 3 years. The risk of implant failure after 3 years can now period, namely 1 to 3 years. Since
be directly calculated as the overall risk of failure after 3 years (6.3%) × the
implant treatment is purported to
incidence of late failures (6.9%), thus equaling 0.43%. This review discredits
be a long-term solution to partial
the theory that studies of a lower level of evidence report inflated results
when compared with prospective randomized controlled clinical trials. (Int J or complete edentulism, success
Periodontics Restorative Dent 2013;33:773–783. doi: 10.11607/prd.1288) should be evaluated over a time
frame longer than 1 year, as it is
1  esearcher and Head, Section of Oral Physiology, Research Center for Oral
R
with other implant therapies.
Health, Department of Biomedical, Surgical and Dental Sciences, Università The primary purpose of this
degli Studi di Milano, IRCCS Istituto Ortopedico Galeazzi, Milano, Italy. systematic review was to investi-
2Associate Clinical Professor, Columbia University, Department of Periodontics, New York,
gate implant survival rates based
New York, USA.
3Head, Section of Implant Dentistry and Oral Rehabilitation, Research Center for on clinical studies with a follow-up
Oral Health, Department of Biomedical, Surgical and Dental Sciences, Università degli of a minimum 3 years. A further
Studi di Milano, IRCCS Istituto Ortopedico Galeazzi, Milano, Italy. aim was to assess whether implant
Correspondence to: Dr Massimo Del Fabbro, Department of Biomedical, Surgical and survival was affected by several of-
Dental Sciences, Università degli Studi di Milano, IRCCS Istituto Ortopedico Galeazzi, ten confounding variables. Specifi-
Via Riccardo Galeazzi 4, 20161 Milano, Italy; fax: +39 02 50319960; cally researched were: type of graft
email: [email protected].
material, implant surface texture,
©2013 by Quintessence Publishing Co Inc. simultaneous or staged implant

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
774

insertion, residual crestal bone mum of 20 sinus floor elevations bone in combination with
height, use of a biologic barrier performed, root-form implants bone substitute, or 100%
membrane, timing of implant fail- placed, multiple interventions (eg, bone substitute. The group
ure, duration of follow-up, publi- simultaneous ridge augmentation) relative to 100% autogenous
cation date, and level of evidence not performed, implant survival bone was further divided ac-
(study design) of included studies. rates clearly specified or calculable cording to the graft form into
from reported data, details on at block graft, particulate graft,
least two of the investigated vari- or a combination of block and
Method and materials ables provided, and a minimum particulate graft. The site of
follow-up of at least 36 months of harvesting the autogenous
Search method functional loading or mean follow- graft was not considered for
up of at least 60 months reported. the analysis.
A search was conducted for ar- In instances in which a decision • Type of implant surface:
ticles published up to October could not be made on the basis of machined or textured.
2010 using the following electronic the title and abstract alone, the full • Implant placement with re-
databases: MEDLINE, EMBASE, text was obtained and examined. spect to grafting procedure:
and the Cochrane Central Regis- For all articles included at this simultaneous or delayed.
ter of Controlled Trials. Custom- stage, the full text was obtained
ary key words were used, alone and carefully examined indepen- Extracted data also included
and in combination, to search the dently by two reviewers. Studies the presence or absence of a bio-
databases. The search was limited were excluded if they did not meet logic barrier membrane, residual
to studies involving human sub- the above inclusion criteria or were crestal bone height, and important-
jects. No restrictions pertaining not of a clinical nature (eg, reviews, ly, the timing of implant failures.
to language or study design were technical reports). If multiple pub- The primary outcome measure for
applied. lications existed relative to con- this review was implant survival at
Additionally, a hand-search was secutive phases of the same study, 36 months.
performed of the main international only the most recent data (longest
journals in the fields of implant den- follow-up) were considered.
tistry, periodontics, and oral and Data analysis
maxillofacial surgery from 1986 to
October 2010. For these journals, Data extraction The data relative to the lateral
the online section dedicated to ar- window and the transalveolar ap-
ticles ”in press” was also searched, The characteristics of the included proaches were kept separate.
when available. Finally, the refer- studies were examined indepen- The implant was used as the unit
ence lists of the most relevant pa- dently by two reviewers, and the of analysis. Comparison between
pers and reviews were checked for publications were grouped by groups for each variable consid-
possible additional studies. study design: randomized con- ered was made using implant sur-
trolled trials (RCT), controlled trials, vival at 36 months. Conventional
case series, or retrospective stud- nonparametric tests such as the
Selection criteria ies. The extracted data were then Pearson chi-square were used for
tabulated according to the follow- data analysis. Odds ratios (ORs) and
The titles and abstracts of articles ing parameters: 95% confidence intervals (CIs) were
identified from the electronic also calculated. The significance
searches were screened using the • Type of graft material: 100% level was set at P = .05. For each
following inclusion criteria: mini- autogenous bone, autogenous group and subgroup, data were

The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 775

synthesized using the weighted as the unit of analysis, while very bovine HA. The survival rate using
mean along with 95% CI. Data few also provided patient-based 100% bone substitute was 96.25%
analysis was performed using a results. The mean follow-up of the (95% CI, 94.02% to 98.49%)
statistical software package (SPSS, included studies ranged from 36 to (Table 3).
IBM). 125 months, with 4 studies report- Four articles were excluded
ing data on more than 10 years of because separate outcomes in
function.25–27,33 terms of implant survival for any
Results type of graft material used were
not provided.12,20,21,23
The initial search provided a to- Lateral approach The overall implant survival us-
tal of 1,250 articles reporting on ing 100% autogenous bone was
maxillary sinus augmentation in Graft material significantly lower than that using
combination with dental implant In the group using 100% autoge- 100% bone substitute (P < .001;
placement. The final selection ac- nous bone, the overall implant sur- OR, 4.42; 95% CI, 3.78 to 5.06).
cording to the follow-up duration vival was 85.32% (95% CI, 78.69%
provided 25 studies (24 articles), to 91.95%). In the subgroup using Implant surface texture
of which 18 reported data on the only block grafts, the survival rate Implants with a machined surface
lateral approach11–28 (Table 1) and was 79.46%, while in the subgroup displayed an overall survival rate of
7 reported data on the transal- using a combination of block and 81.0% (95% CI, 72.81% to 89.18%)
veolar approach22,27,29–33 (Table 2). particulate grafts, the implant for 250 patients and 1,005 im-
Two articles reported on both ap- survival was significantly higher plants placed, while implants with
proaches22,27 and another reported (98.85%; OR, 22.23; 95% CI, 9.16 a rough surface displayed an over-
on both a retrospective study and to 35.3) (P < .001). In a very small all survival rate of 96.57% (95% CI,
a prospective controlled trial.28 number of cases among those in- 94.50% to 98.64%) for 1,855 pa-
In summary, 1 RCT (4%), 3 con- cluded, 100% particulate autograft tients and 5,276 implants placed.
trolled trials (12%), 5 case series was used, with an implant survival The difference was significant
reports (20%), and 15 retrospec- of 89.53% (Table 3). (P < .001; OR, 6.61; 95% CI, 5.88
tive studies (64%) were included. In the group using grafts made to 7.34) (Table 4). This analysis did
The articles were published over a of bone substitutes in combination not take account of the degree of
14-year period, from 1997 to 2010 with autogenous bone (compos- roughness, the type of coating, or
for the lateral approach and from ites), only two articles (both using the procedure adopted to roughen
2008 to 2010 for the transalveolar autogenous bone with hydroxyap- the implant surface. One study
approach. Overall, 6,500 implants atite [HA]) could be classified.13,19 (44 patients and 219 implants)
were placed in 2,149 patients for Implant survival for composite could not be included in the analy-
the lateral approach and 1,257 grafts was 87.70% (95% CI, 85.71% sis because the implant type and
implants in 704 patients for the to 89.69%) (Table 3). surface were not specified.20
transalveolar approach. The over- In the group using 100% bone Machined implants were used
all implant survival was 93.71% substitute, the following materials more often in combination with au-
(range, 75.57% to 100.00%; 95% were used alone or in combina- togenous bone (65.1%) than with
CI, 90.20% to 97.22%) and 97.2% tion: anorganic bovine bone (ABB), bone substitute materials, either
(range, 94.30% to 100.00%; 95% β–tricalcium phosphate, porcine alone (0.6%) or mixed with auto-
CI, 95.47% to 98.97%) for the lat- bone, absorbable HA, ABB + de- grafts (21.8%). On the contrary,
eral and transalveolar approaches, mineralized freeze-dried bone al- rough implants were used more
respectively. All articles provided lograft (DFDBA), and DFDBA + frequently with nonautogenous
outcomes considering the implant absorbable HA + nonabsorbable bone replacement grafts.

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
776

Table 1 Lateral sinus floor augmentation

Study No. of No. of No. of Implant

Article Year type Graft material patients sinuses implants survival rate (%)
Block and Kent11 1997 Re ABG 33 53 173 88.40
Block et al12 1998 Re ABG/ABG + BS 16 27 73 95.90
Kaptein et al13 1998 Re ABG + BS 88 132 388 88.14
Buchmann et al14 1999 CT ABG 50 75 167 100.00
Johansson et al15 1999 Re ABG 39 39 131 75.57
Lekholm et al16 1999 Re ABG 55 82 280 81.07
Becktor et al17 2002 Re ABG 81 120 329 79.64
Valentini and 2003 Re BS 59 78 187 94.65
Abensur18
Hallman and 2004 CS ABG + BS 20 30 108 86.11
Zetterqvist19
Boyne et al20 2005 RCT ABG/BS 44 88 219 83.11
Peleg et al21 2006 CT ABG/ABG + BS/ 731 731 2,132 97.94
BS
Jurisic et al22 2008 Re BS 28 28 40 100.00
Yamamichi et al23 2008 Re ABG/ABG + BS/ 188 188 466 96.35
CS BS 69 69 159 92.50
BS
Sbordone et al24 2009 Re ABG 28 39 70 95.71
Garlini et al25 2010 Re BS 26 27 47 100.00
Scarano et al26 2010 Re BS 113 153 264 94.70
Tetsch et al27 2010 Re ABG/BS 461 461 1,207 97.10
Zijderveld et al28 2009 CT ABG/BS 20 24 60 98.33
Re = retrospective; CT = clinical trial; CS = case series; RCT = randomized clinical trial; ABG = autogenous bone graft; BS = bone substitute; NR = not reported.

Time of implant placement 97.02%). The difference was signifi- Residual ridge height
The overall survival rate for im- cant (P < .001; OR, 1.69; 95% CI, Three studies did not report the
plants placed in a simultaneous 1.45 to 1.93) (Table 5). height of the residual ridge at the
procedure was 95.95% (95% CI, Four studies accounting for 641 time of surgery.11,16,17 For those that
91.22% to 100.67%), while the patients and 2,035 implants were could be evaluated for this parame-
overall survival rate for implants not classifiable because both pro- ter, a threshold height of 5 mm was
placed in a two-stage procedure cedures were adopted but sepa- considered. The implant survival
was 93.34% (95% CI, 89.66% to rate data were not provided.16,17,20,27 after at least 36 months of follow-

The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 777

it was not used. The difference was

significant (P < .001; OR, 2.426;
95% CI, 2.09 to 2.75) (Table 6).
Implant Residual ridge Implant Mean
placement height (mm) surface Range (mo)
Study design
Immediate NR Rough 36–132 mo 70.0 The overall implant survival report-
Immediate 5.1 ± 2.2 Rough 63–126 mo 75.0 ed in the 13 retrospective stud-
ies was 89.39% (range, 75.57%
Immediate/delayed <5/>5 Rough ≥ 70 mo 55.0
to 100.00%; 95% CI, 84.83% to
Immediate <5 Machined/rough > 36– > 60 mo 60.0 93.96%), while for the 6 prospec-
Immediate <5 Machined NR 36.0 tive studies, it was 96.42% (range,
83.11% to 100.00%; 95% CI,
Immediate/delayed NR Machined NR 36.0
90.93% to 101.91%). The differ-
Immediate/delayed NR Machined 22–105 mo 64.2 ence was significant (P < .001; OR,
Immediate/delayed <5/>5 Machined/rough 2.3–9.4 y 78.0 3.19; 95% CI, 2.85 to 3.53).

Delayed <5 Machined NR 60.0 Failure analysis

The timing of failure was reported
Delayed <6 NS NR 36.0 for 389 of 409 failed implants. A
higher proportion of failures oc-
Immediate 1–7 Rough 24–108 mo 69.0
curred during the healing phase
and within the first year of load-
Immediate/delayed 4–9 Rough NR 36.0 ing (Fig 1). Taken together, these
Immediate/delayed ≤ 4 / 5–8 Machined/rough >7y 84.0 two periods accounted for 80.5%
Immediate/delayed ≤ 4 / 5–8 Machined/rough > 36 mo 36.0 of the total implant failures, while
another 12.6% occurred in the
Delayed ≤3/>3 Rough NR36 mo 36.0 second and third years of loading.
Immediate 7.4 (5–9.5) Rough 4–11 y 72.0 Considering the overall failure rate,
the risk of implant failure occurring
Delayed 2–3 Rough 5–10 y 60.0
after 3 years in function amounts to
NR 5.2 Rough 1–15 y 52.5 0.43%.
Delayed 5.4–7 Rough 55–73 mo 60.0
Publication date
With the article by Valentini and
Abensur18 published in 2003 serv-
ing as the first “late” study, the
implant survival rate reported in
up was 95.21% (2,941 implants) absorbable barrier membrane the studies published from 1997
and 95.23% (3,021 implants) for re- covering the graft.20,21,23,26 A mem- to 2002 was 85.66%, while in stud-
sidual ridge heights of < 5 mm and brane was not used in 10 included ies published since 2003, it was
≤ 5 mm, respectively. The differ- articles, while 4 articles did not 96.21%. The difference was signifi-
ence was not significant. specify whether a membrane was cant (P < .001; OR, 4.25; 95% CI,
used.14,16,17,24 The overall implant 3.81 to 4.69) (Table 7).
Use of a barrier membrane survival was 97.12% when a mem-
Four studies reported use of an brane was used and 93.29% when

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
778

Table 2 Transalveolar sinus floor augmentation

Implant
No. of No. of survival rate
Article Year Study type Graft material patients implants (%)
Fermegard and Astrand29 2009 Re None 36 53 94.34
Calvo-Guirado et al30 2010 CS BS 30 60 96.67
Crespi et al31 2010 CS BS 20 30 100
Nedir et al32 2010 CS None 17 25 100
Tetsch et al27 2010 Re None 522 983 97.25
Bruschi et al33 2010 Re None 46 66 95.45
Jurisic et al22 2008 Re None 33 40 100
RE = retrospective; CS = case series; BS = bone substitute; NR = not reported.

Table 3 Graft materials

No. of No. of No. of Implant Weighted mean %

Material Form articles patients implants survival rate (%) follow-up (mo) machined
100% autogenous Block 3 175 740 79.46 100.0
bone graft Particulate 2 42 191 89.53 0.0
Block + particulate 3 86 261 98.85 13.9
Total ABG 8 303 1,192 85.32 52.8 65.1
Combined ABG + BS 2 108 496 87.70 56.1 21.8
100% bone substitute 7 759 1,922 96.25 54.9 0.6
ABG = autogenous bone graft; BS = bone substitute.

Table 4 Implant surface

No. of No. of No. of Implant Weighted mean

Surface articles patients implants survival rate (%) follow-up (mo)
Machined 7 250 1,005 81.00 52.9
  Older studies 4 186 776 80.41 49.1
  Newer studies 3 64 229 82.97 65.9
Rough 13 1,855 5,276 96.57 63.1
  Older studies 4 176 765 90.98 61.2
  Newer studies 9 1,679 4,511 97.52 63.4
Older = before reference no. 18.

The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 779

placed without the addition of

grafting material, while 90 implants
in 50 patients used graft material
Follow-up
(see Table 2). The implant survival
Implant Residual ridge Implant Range Mean calculated for the two groups was
placement height (mm) surface (y) (mo) 97.78% (mean weighted follow-
Immediate 6.3 ± 0.3 Rough NR 36.0 up, 36 months) and 97.17% (mean
weighted follow-up, 57.55 months)
Immediate 5–8 Rough NR 36.0
for studies with and without graft
Delayed 8.2 Rough NR 36.0 materials, respectively. No signifi-
Immediate 5.4 ± 2.3 Rough NR 60.0 cant difference was found (P = .78;
OR, 0.81; 95% CI, 0.21 to 1.41).
Immediate 8.2 Rough 1–15 55.0
The mean residual ridge height
Delayed <3 Rough 5–16 125.0 was greater than 5 mm in all stud-
Immediate 4–9 Rough NR 36.0 ies except for that by Bruschi et
al,33 in which only cases with less
than 3 mm of residual ridge height
were treated.

Table 5 Timing of implant insertion with respect to grafting

Discussion
Implant Weighted
No. of No. of No. of survival mean follow- From 2003 to 2010, 10 systematic
articles patients implants rate (%) up (mo) reviews were published relative to
Simultaneous 10 1,049 3,084 95.95 59.9 implant outcomes following maxil-
lary sinus augmentation.1–10 Most if
Delayed 8 459 1,381 93.34 59.9
not all of these reviews used a min-
imum 1 year of loading or a mean
loading period of over 1 year,
Table 6 Use of a barrier membrane with some cases less than 1 year,
as part of their inclusion criteria. If
Implant Weighted data were given regarding implant
No. of No. of No. of survival mean follow-
articles patients implants rate (%) up (mo) survival for periods longer than 1
year, it was either as part of a da-
Membrane 3 1101 3,021 97.12 56.5
tabase with a longer mean survival
No membrane 10 790 2,414 93.29 68.8 period or as a speculation made on
a straight-line continuation of the
short-term failure rates contained
in their databases.5 The purpose
of this review was to perform a sys-
Transalveolar approach sisted of four retrospective stud- tematic review with an actual mini-
ies22,27,29,33 and three case series.30–32 mum inclusion criterion of 3 years
Of the seven included studies, The mean weighted follow-up was of loading. Note that to increase
two reported results longer than 56.01 months. The majority of im- the size of the database, studies
10 years.27,33 Included studies con- plants (1,167 in 654 patients) were with a mean loading period of at

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
780

Fig 1    Trend of implant failures over time.

60 The healing period before prosthetic load-
ing ranged between 4 and 12 months.

50
Implant Failure (%)

40

30

20

10

0
Healing 1 2 3 4 5 6 7 >7
Period (y)

Table 7 Main features of studies categorized by publication date

No. of 100% 100% No. of No. of Implant

articles Prospective Retrospective ABG BS patients implants survival rate (%)
Before 2003 7 1 6 5 0 362 1,541 85.66
After 2003 10 5 5 3 7 1,787 4,959 96.21
ABG = autogenous bone graft; BS = bone substitute.

least 60 months were also includ- One of the major problems en- surfaces in conjunction with au-
ed, even though a small proportion countered in reviews of this type togenous bone grafts. The review
of the cases reported on loading is the presence of confounding by Pjetursson et al5 recalculated
times of less than 3 years. variables, ie, situations in which the overall failure data (estimated
The articles included in this re- more than one variable may affect annual failure, 3.48%; estimated
view showed substantial differenc- the outcome of a procedure. It is 3-year survival rate, 90.1%) sepa-
es in residual crestal bone height, sometimes difficult to surmount rating the machined-surface im-
type of implants placed, graft ma- this problem because of factors plant data (6.86% and 81.4%,
terial used, success/survival criteria, such as inadequate data report- respectively) and rough-surfaced
duration of follow-up, study design ing, changes in surgical techniques implant data (1.19% and 96.1%,
and objectives, inclusion and ex- over time, lack of prospective trials respectively). The confoundance
clusion criteria, data reporting, and with only one variable, or the reali- is the result of an unduly large
method of statistical analysis. The ty that eliminating studies with only proportion of machined implants
unit of analysis in the present re- one variable would result in a very placed in grafts with 100% autog-
view was the implant and not the small database. Confounding vari- enous bone. When machined-sur-
patient. The analysis was based on ables have been noted in previous face implants were excluded from
implant survival, since few articles reviews,2,3,5 especially with regard the data, the annual failure rates
reported on implant success (with to the confounding relationship be- for 100% bone replacement grafts,
nonuniform criteria). tween the use of machined implant composites, and 100% autogenous

The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 781

grafts were similar (1.13%, 1.10%, spective, randomized controlled variable. This has proven itself
and 1.27%, respectively), revealing clinical trials are the highest stan- to be a difficult task, since not all
that the variable most responsible dard of evidence, it is also true that studies have reported this data.
for a higher failure rate was in fact these studies are relatively few in Further, while crestal bone height
the machined implant surface. The number, and waiting for them to be limitations or ranges may be part
validity of this approach is con- published may withhold valuable of the inclusion/exclusion criteria
firmed in this review. information from clinicians for long for a study, most do not report the
This review also considered periods of time. The data from this initial ridge height for the failed
the date of publication as a poten- review report quite the opposite, implants. The study by Peleg et
tial confounding variable and ad- with the implant survival rate from al,21 which had the greatest sample
dressed this by dividing the data 13 retrospective studies noted to size among the included studies,
into studies published before and be 89.39%, while 6 prospective reported long-term survival follow-
after the randomly chosen date of studies reported a rate of 96.42%. ing simultaneous implant place-
2003, arbitrarily placing the study The difference was significant ment for three ranges of crestal
by Valentini and Abensur18 in the (P < .001; OR, 3.19; 95% CI, 2.85, bone heights. Implant survival
latter group. This was the first long- 3.53). One could hypothesize that for crestal bone height ranges of
term study that used 100% bone well-designed prospective studies 1 to 2 mm, 3 to 5 mm, and > 5 mm
replacement grafts and rough-sur- often estimate the treatment effect were reported as 95.9%, 98.5%,
face implants. The dramatic differ- in an ideal situation since they are and 98.4%, respectively, for an
ence in implant survival between normally carried out with a stand- overall implant survival of 97.9%.
the old and new studies, as seen ardized protocol, selective inclu- This study accounted for 49.6%
in Table 7 (85.66% and 96.21%, sion criteria, experienced surgical of cases presenting with less than
respectively), highlights how evi- teams, and controlled variables. 5 mm of residual ridge height that
dence-based decision making has Conversely, retrospective studies were considered in the present re-
enhanced procedural outcomes may more closely reflect every day view and markedly contributed to
in maxillary sinus grafting. None clinical practice, with much larger the increased survival rate for this
of the earlier studies used 100% variability in study parameters subgroup. In the present review,
bone substitute. All implants used when compared with prospective no significant difference was found
with autogenous blocks in the early clinical trials. between survival rates of implants
studies had a machined surface. The overall survival rate for placed in ridges of at least 5 mm
This latter association displayed implants placed in a simultaneous compared with those inserted in
the lowest implant survival rate, procedure was higher than that ridges of less than 5 mm (95.23%
confirming findings from previous for delayed placement (95.95% vs [3,021 implants] and 95.21% [2,941
systematic reviews.2,5,6 93.34%, respectively). This differ- implants], respectively).
Another variable considered ence was significant (P < .001; OR, It may be of value to view the
for the first time in this review was 1.69; 95% CI, 1.45 to 1.93). Previ- reported results on simultaneous
the effect of study design on the ous reviews have reported 1-year vs delayed placement with con-
reported outcomes of implant sur- survival rates that either show no sideration of residual crestal bone
vival. This analysis was performed statistical difference2,3,5–7 or a dif- heights. While not always true,19
to determine if there was truth to ference in favor of simultaneous simultaneous placement is usually
the perception that studies of a placement.10 Other reviews did not performed when there is a mini-
lower level of evidence (eg, retro- specifically evaluate this aspect.4,8,9 mum of 4 to 5 mm of crestal bone
spective, case series) have a ten- Previous reviews have been present. It is conceivable that a
dency to present inflated positive criticized for not considering re- portion of the failures for delayed
outcomes. While it is true that pro- sidual ridge height as a potential placement (implants placed in

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
782

grafted sinuses that presented with Taken together with the fact that  4. Graziani F, Donos N, Needleman I,
1 to 3 mm of crestal bone) are a re- there is no ideal proven therapy for Gabriele M, Tonetti M. Comparison of
implant survival following sinus floor
sult of poor grafting or inadequate this inflammatory disease, it must augmentation procedures with implants
time intervals between grafting be considered a major risk factor in placed in pristine posterior maxillary
bone: A systematic review. Clin Oral Im-
and implant placement/loading. late implant failure. plants Res 2004;15:677–682.
Studies by Avila et al34 and Soardi   5. Pjetursson BE, Tan WC, Zwahlen M, Lang
et al35 have shown that significant NP. A systematic review of the success
of sinus floor elevation and survival of
time is necessary for graft matura- Conclusions implants inserted in combination with
tion, especially in wide sinuses. sinus floor elevation. J Clin Periodontol
2008;35(suppl):216–224.
The data present an even more This review extended the 1-year  6. Del Fabbro M, Rosano G, Taschieri S.
positive long-term survival rate survival rate of previous reviews Implant survival rates after maxillary si-
(93.7%) than that reported or spec- to 3 years postloading. The im- nus augmentation. Eur J Oral Sci 2008;
116:497–506.
ulated in previous reviews. An es- plant survival rate remained high at  7. Nkenke E, Stelzle F. Clinical outcomes
timated annual failure rate, which 93.7%. This review confirmed data of sinus floor augmentation for implant
placement using autogenous bone or
assumes that the annual implant from earlier reviews with regard to bone substitutes: A systematic review.
failure rate is constant over time,5 implant surfaces, graft materials, Clin Oral Implants Res 2009;20(suppl 4):
would mean that given enough use of membranes, and timing of 124–133.
  8. Jensen SS, Terheyden H. Bone augmen-
time, all implants will fail. The use implant placement. Further, it ad- tation procedures in localized defects in
of such a straight-line continua- dressed the effect of study design the alveolar ridge: Clinical results with
different bone grafts and bone-substi-
tion of reported 1-year failure rates on implant survival outcomes and tute materials. Int J Oral Maxillofac Im-
cannot be supported by the long- demonstrated the need for true or plants 2009;24(suppl):218–236.
term data reported. The studies in real-time follow-ups to report sur-   9. Esposito M, Grusovin MG, Rees J, et al.
Interventions for replacing missing teeth:
this review reported that 80% of vival data more accurately. Augmentation procedures of the maxil-
failures occur within the first year, lary sinus. Cochrane Database Syst Rev
2010;(3):CD008397.
and 93.1% of failures occur within 10. Del Fabbro M, Bortolin M, Taschieri S,
3 years. The risk of implant failure Acknowledgment Rosano G, Testori T. Implant survival in
after 3 years can now be directly maxillary sinus augmentation. An updat-
The authors reported no conflicts of interest ed systematic review. J Osteol Biomat
calculated as the overall risk of fail- 2010;1:69–79 [erratum 2010;1:186].
related to this study.
ure after 3 years (6.3%) × the in- 11. Block MS, Kent JN. Sinus augmenta-
tion for dental implants: The use of au-
cidence of late failures (6.9%), thus togenous bone. J Oral Maxillofac Surg
equaling a very low 0.43%. References 1997;55:1281–1286.
It is quite likely that failures oc- 12. Block MS, Kent JN, Kallukaran FU,
  1. Aghaloo TL, Moy PK. Which hard tissue Thunthy K, Weinberg R. Bone mainte-
curring after 3 years are the result augmentation techniques are the most nance 5 to 10 years after sinus grafting. J
of factors unrelated to the place- successful in furnishing bony support for Oral Maxillofac Surg 1998;56:706–714.
implant placement? Int J Oral Maxillofac 13. Kaptein MLA, de Putter C, de Lange
ment of implants into grafted si- GL, Blijdorp PA. Survival of cylindrical
Implants 2007;22(suppl):49–70.
nuses. Certainly one can speculate   2. Wallace SS, Froum SJ. Effect of maxillary implants in composite grafted maxil-
on the possibility of implant frac- sinus augmentation on the survival of en- lary sinuses. J Oral Maxillofac Surg
dosseous dental implants. A systematic 1998;56:1376–1380.
ture in the heavy load-bearing area review. Ann Periodontol 2003;8:328–343. 14. Buchmann R, Khoury F, Faust C, Lange
of the posterior maxilla and fail-  3. Del Fabbro M, Testori T, Francetti L, DE. Peri-implant conditions in periodon-
Weinstein R. Systematic review of surviv- tally compromised patients following
ures resulting from peri-implantitis. maxillary sinus augmentation. A long-
al rates for implants placed in the grafted
The incidence of peri-implantitis maxillary sinus. Int J Periodontics Restor- term post-therapy trial. Clin Oral Im-
has been reported at disturbingly ative Dent 2004;24:565–577. plants Res 1999;10:103–110.

high rates in studies by Lindhe and

Meyle36 (12% to 40% of sites) and
Koldsland et al37 (11.3% to 47.1%).

The International Journal of Periodontics & Restorative Dentistry

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 783

15. Johansson B, Wannfors K, Ekenbäck 23. Yamamichi N, Itose T, Neiva R, Wang HL. 31. Crespi R, Capparè P, Gherlone E. Os-
J, Smedberg JI, Hirsch J. Implants and Long-term evaluation of implant survival teotome sinus floor elevation and simul-
sinus-inlay bone grafts in a 1-stage pro- in augmented sinuses: A case series. Int taneous implant placement in grafted
cedure on severely atrophied maxillae: J Periodontics Restorative Dent 2008; biomaterial sockets: 3 years of follow-up.
Surgical aspects of a 3-year follow-up 28:163–169. J Periodontol 2010;81:344–349.
study. Int J Oral Maxillofac Implants 1999; 24. Sbordone L, Toti P, Menchini-Fabris G, 32. Nedir R, Nurdin N, Vazquez L, Szmukler-
14:811–818. Sbordone C, Guidetti F. Implant success Moncler S, Bischof M, Bernard J-P. Os-
16. Lekholm U, Wannfors K, Isaksson S, in sinus-lifted maxillae and native bone: teotome sinus floor elevation technique
Adielsson B. Oral implants in combina- A 3-year clinical and computerized to- without grafting: A 5-year prospec-
tion with bone grafts. A 3-year retro- mographic follow-up. Int J Oral Maxillo- tive study. J Clin Periodontol 2010;37:
spective multicenter study using the fac Implants.2009;24:316–324. 1023–1028.
Brånemark implant system. Int J Oral 25. Garlini G, Redemagni M, Donini M, 33. Bruschi GB, Crespi R, Capparè P, Gher-
Maxillofac Surg 1999;28:181–187. Maiorana C. Maxillary sinus elevation lone E. Transcrestal sinus floor elevation:
17. Becktor JP, Eckert SE, Isaksson S, Keller with an alloplastic material and im- A retrospective study of 46 patients up
EE. The influence of mandibular denti- plants: 11 years of clinical and radiologic to 16 years. Clin Implant Dent Relat Res
tion on implant failures in bone-grafted follow-up. J Oral Maxillofac Surg 2010; 2010 Oct 26 [epub ahead of print].
edentulous maxillae. Int J Oral Maxillo- 68:1152–1157. 34. Avila G, Wang HL, Galindo-Moreno P, et
fac Implants 2002;17:69–77. 26. Scarano A, Piattelli A, Assenza B, et al. al. The influence of the bucco-palatal dis-
18. Valentini P, Abensur DJ. Maxillary sinus Porcine bone used in sinus augmenta- tance on sinus augmentation outcomes.
grafting with anorganic bovine bone: tion procedures: A 5-year retrospective J Periodontol 2010;81:1041–1050
A clinical report of long-term results. clinical evaluation. J Oral Maxillofac Surg 35. Soardi CM, Spinato S, Zaffe D, Wang
Int J Oral Maxillofac Implants 2003;18: 2010;68:1869–1873. HL. Sinus floor augmentation with min-
556–560. 27. Tetsch J, Tetsch P, Lysek DA. Long-term eralized human bone allograft in sinuses
19. Hallman M, Zetterqvist L. A 5-year pro- results after lateral and osteotome tech- of different size: An histologic and his-
spective follow-up study of implant- nique sinus floor elevation: A retrospec- tomorphometric analysis. Clin Oral Im-
supported fixed prostheses in patients tive analysis of 2190 implants over a time plants Res 2011;22:560–566.
subjected to maxillary sinus floor augmen- period of 15 years. Clin Oral Implants 36. Lindhe J, Meyle J. Peri-implant disease:
tation with an 80:20 mixture of bovine hy- Res 2010;21:497–503. Consensus report of the Sixth European
droxyapatite and autogenous bone. Clin 28. Zijderveld SA, Schulten EAJM, Aartman Workshop on Periodontology. J Clin
Implant Dent Relat Res 2004;6:82–89. IHA, ten Bruggenkate CM. Long-term Periodontol 2008;35(suppl):282–285.
20. Boyne PJ, Lilly LC, Marx RE, et al. De changes in graft height after maxillary si- 37. Koldsland OC, Scheie AA, Aass AM.
novo bone induction by recombinant nus floor elevation with different grafting Prevalence of peri-implantitis related to
human bone morphogenetic protein-2 materials: Radiographic evaluation with severity of the disease with different de-
(rhBMP-2) in maxillary sinus floor aug- a minimum follow-up of 4.5 years. Clin grees of bone loss. J Periodontal 2010;
mentation. J Oral Maxillofac Surg 2005; Oral Implants Res 2009;20:691–700 81:231–238.
63:1693–1707. 29. Fermergård R, Åstrand P. Osteotome si-
21. Peleg M, Garg AK, Mazor Z. Predictabil- nus floor elevation without bone grafts—
ity of simultaneous implant placement in A 3-year retrospective study with Astra
the severely atrophic posterior maxilla: A Tech implants. Clin Implant Dent Relat
9-year longitudinal experience study of Res 2012;14:198–205.
2132 implants placed into 731 human si- 30. Calvo-Guirado JL, Gómez-Moreno G,
nus grafts. Int J Oral Maxillofac Implants López-Marí L, Ortiz-Ruiz AJ, Guardia-
2006;21:94–102. Muñoz J. Atraumatic maxillary sinus el-
22. Jurisic M, Marxovic A, Radulovic M, evation using threaded bone dilators for
Brkovic BMB, Sándor GK. Maxillary sinus immediate implants. A three-year clini-
floor augmentation: Comparing osteo- cal study. Med Oral Patol Oral Cir Bucal
tome with lateral window immediate and 2010;15:e366–e370.
delayed implant placements. An interim
report. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 2008;106:820–827.

Volume 33, Number 6, 2013

© 2013 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.