Dental Implant Survival in Vascularized Bone Flaps: A Systematic Review and Meta-Analysis

Abstract

Background:

Maxillofacial reconstruction with vascularized bone restores facial contour and provides structural support and a foundation for dental rehabilitation. Routine implant placement in such cases, however, remains uncommon. This study aims to determine dental implant survival in patients undergoing vascularized maxillary or mandibular reconstruction through a systematic review of the literature.

Methods:

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, the literature was queried for implant placement in reconstructed jaws using MeSH terms on PubMed, Embase, and Cochrane platforms. Weighted implant survivals were calculated for the entire cohort and sub-cohorts stratified by radiotherapy. Meta-analyses were performed to estimate effect of radiation on implant osseointegration.

Results:

Of 3,965 publications identified, 42 studies were reviewed, including 1,084 patients with 3,636 dental implants. Weighted implant survival was 92.2% at a median follow-up of 36 months. Survival was 97.0% in 269 implants placed immediately in 60 patients versus 89.9% in 1,897 delayed implants placed in 597 patients, with follow-up of 14 and 40 months, respectively. Dental implants without RT exposure had better survival than those exposed to radiation (95.3 vs. 84.6%; p<0.01) at median follow-up of 36 months. Meta-analyses showed radiation significantly increased the risk of implant failure (risk ratio [RR]: 4.74, p<0.01) and suggested that implants placed prior to radiotherapy trended towards better survival (88.9% vs. 83.4%, p=0.07, RR: 0.52; p=0.14).

Conclusions:

Overall implant survival was 92.2%; however, radiotherapy adversely impacted outcomes. Implants placed before radiation may demonstrate superior survival than implants placed after.

INTRODUCTION

Dental rehabilitation through osseointegrated implant placement significantly improves quality of life (QoL) ^(1–4) by restoring maxillofacial functions, including mastication and speech. In recent decades, osseointegrated dental implants have become the mainstay for their potential to restore dental continuity in a single procedure⁽⁵⁾. Osseointegrated implant placement can be immediate or delayed, followed by delivery of implant-supported/retained prosthetic teeth. Immediate implantation has advantages, such as better access to the bone, ease of determining the interdental relationship, and shorter time to complete dental rehabilitation⁶. However, implant placement is more commonly performed as a secondary, delayed procedure to avoid damaging the bone flap and to allow for a healed, stable substrate. Existing studies demonstrate disparate approaches to implant placement with variable results, underscoring the need for a systematic, rigorous review to summarize the findings.

In oncology, osseointegrated implant placement in the context of radiotherapy (RT) is an additional concern. RT adverse events include hot spots and osteoradionecrosis; however, radiation’s impact on osseointegrated implant survival is unclear. While some surgeons suggest unchanged implant survival in native bone versus irradiated flaps in head and neck (H/N) cancer patients⁷, others demonstrate lower implant survival rates⁸. There is also a lack of evidence regarding implant placement success, before or after RT, in the reconstructed maxilla and mandible.

This study aims to provide an exhaustive summary of existing evidence regarding immediate or delayed osseointegrated dental implant survival in vascularized bone flaps for maxillary and mandibular reconstruction. Furthermore, this study seeks to better understand whether implant placement before or after RT impacts implant survival. Lastly, this study reviews the patient-reported outcomes following vascularized bone flap and dental implant placement.

METHODS

Search Methodology for Studies

A comprehensive electronic literature review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines^{9, 10}. The study was exempted from institutional board review. PubMed, Embase via the Elsevier platform, and Cochrane provided by Wiley Online Library were queried for articles published in English from 1992–2017 using MeSH terms for maxillary and mandibular resection followed by reconstruction with a vascularized free flap and dental implantation. Keywords and subject headings were searched using Boolean operators (AND, OR). Related terms were also incorporated to ensure relevant papers were retrieved. A manual search of references was performed to find relevant articles not detected in the electronic bibliographic search.

Selection Process

Results of the search were imported into an EndNote database. Publications were screened where two investigators (HP, IP) independently reviewed titles and abstracts for duplicates and poor fit within the focus of the systematic review. If at least one reviewer coded the title to continue to the next round, the other investigator independently reviewed the full text article and classify the articles based on the eligibility criteria. Both investigators independently assessed all articles deemed eligible for full-text review. Discrepancies were discussed with a third investigator (JAN) until consensus was reached. In case of multiple articles by the same author(s), assessment of the most recent article was done to avoid duplication of the cases (Figure 1).

Figure 1:

Patient selection for the systematic review

Inclusion criteria comprised studies involving segmental resection of the maxilla or mandible followed by reconstruction using bone flap(s) or composite flaps—osteocutaneous, osteomyocutaneous, or osteoseptomyocutaneous—implanted with the dental fixture(s). Studies involving single case reports, non-human subjects, and inadequate information about implant survival were excluded. Patients who experienced flap failure or morbidity and mortality due to recurrent malignancy were excluded from analysis.

Study Design

The primary outcome was implant survival at last available follow-up in the overall cohort of all selected studies. The secondary outcomes were weighted implant survival based on RT status (i.e., radiation vs. no radiation) and its timing (i.e., radiation before or after implant placement).

Implant survival was defined as in situ, immobile, or stable on clinical exam, free of complications such as infection, inflammation and suppurations, mobility, discomfort, ongoing pathological process, peri-implantitis, neuropathies, persistent paraesthesia, peri-implant radiolucency, or significant peri-implant bone absorption.

Data Extraction

Demographics and clinical information (age, sex, indications for surgery, defect location, flap type, reconstruction location, implant placement, survival) were manually extracted and summarized. Implant-related variables included total implants in the flaps, timing of implantation, average follow-up time after implant placement, and implant survival.

Statistical Analysis

The quality of non-randomized studies was assessed using the Methodological Index for Non-Randomized Studies (MINORS)¹¹. The average number of implants per patient was calculated (total number of patients who received implants from the total number of implants placed in the flaps). Weighted implant survival was calculated for the entire cohort and for sub-cohorts stratified by no radiation, pre-RT, and post-RT.

Weighted average survival (x) was calculated as the sum of the product of total implants (weight, w_i) times the survival (x_i) for each study, divided by the sum of the weights. Chi-square statistics were used to analyze differences in implant survival. All tests were two-sided, and p < 0.05 was used to determine significance. All other analyses were performed using SPSS 24.0 (IBM Corp., Armonk, NY)

Meta-analysis and forest plots were generated for studies with sufficient data for both intervention (total implant placed) and outcome (implant failure/osseointegration). A random or fixed model meta-analysis was performed to compare risk of implant failure in irradiated versus non-irradiated flaps. Outcomes were reported as risk ratios (RRs) with 95% confidence interval. Heterogeneity across the studies was assessed using Q statistic, and p=0.1 was considered significant. The I² statistic was used to quantify heterogeneity. Sensitivity analyses were conducted using sequential study omission to explore potential explanation of heterogeneity and impact on the overall results. Publication bias was investigated by exploring funnel plots where appropriate. All analyses were performed using Comprehensive Meta-Analysis software, version 3 (Biostat, Englewood, NJ).

RESULTS

Of 3965 records identified, 42 studies published from 1992–2017 were included (Figure 1)^(6–47). Studies included one randomized controlled trial, six prospective cohorts, and 35 retrospective cohort studies or case series (Table 1). The MINORS value ranged from 56%–94%.

Table 1:

Summary of Included Studies and Their Cohorts’ Characteristics

Study	Time of Surgery	Study Design	Total No. of Patients with Flap(s)	Average Age, years (min–max)	Sex		Proportion of Malignant Cases	Defect Location: Mandible/Maxilla/Both
					M	F
Zlotolow et al., 1992	1990–1992	Case series	7	45 (17–65)	4	3	Mixed	Mandible
Sclaroff et al., 1994	1991–1994	Retrospective cohort	22	54 (29–79)	16	6	Majority malignant	Mandible
Barber et al., 1995	nr	Prospective pilot	5	nr	nr	nr	All malignant	Mandible
Roumanas et al., 1997	nr	Retrospective cohort	20	(25–78)	10	10	Majority malignant	Mandible
Gürlek et al., 1998	1988–1994	Retrospective cohort	20	47 (27–74)	9	11	Majority malignant	Mandible
Chang et al., 1998	1994–1997	Prospective study	12	36 (17–65)	5	7	None	Mandible
Foster et al., 1999	1985–1996	Retrospective cohort	75	49 (12–82)	47	28	Majority malignant	Mandible
Schliephake et al., 1999	1988–1994	Retrospective cohort	44	nr	32	12	All malignant	Mandible
Kovacs et al., 2000	1991–1998	Retrospective cohort	11	nr	nr	nr	Mixed	Mandible
Shaw et al., 2005	1987–2002	Retrospective cohort	81	58 (15–80)	49	32	All malignant	Both
Teoh et al., 2005	1986–2001	Retrospective cohort	24	42 (7–81)	10	14	Majority malignant	Mandible
Kramer et al., 2005	1998–2001	Prospective study	16	47 (20–71)	13	24	Majority malignant	Both
Garrett et al., 2006	1997–2001	Prospective study	46	60 (19–83)	22	24	Majority malignant	Mandible
Gbara et al., 2007	1992–1994	Retrospective cohort	30	61 (36–75)	18	12	Majority malignant	Mandible
Smolka et al., 2008	1992–2000	Retrospective cohort	56	56 (41–79)	43	13	All malignant	Mandible
Hundepool et al., 2008	1995–2005	Retrospective cohort	70	53 (6–82)	40	30	All malignant	Both
Raoul et al., 2009	1996–2007	Retrospective cohort	30	46 (19–72)	18	12	Majority malignant	Both
Virgin et al., 2010	2001–2008	Retrospective cohort	168	60 (nr)	111	57	All malignant	Mandible
Salinas et al., 2010	1994–2002	Retrospective cohort	44	nr	25	19	Majority malignant	Mandible
Bodard et al., 2011	nr	Retrospective cohort	23	46 (17–66)	17	6	nr	Mandible
Barrowman et al., 2011	1992–2007	Retrospective cohort	31	51 (20–76)	18	13	All malignant	Both
Buddula et al., 2011	1987–2008	Retrospective cohort	48	61 (33–92)	29	19	All malignant	Both
Chang et al., 2011	2005–2007	Retrospective cohort	10	41 (23–65)	7	3	None	Mandible
Dholam et al., 2011	nr	Retrospective cohort	12	nr	nr	nr	All malignant	Both
Fenlon et al., 2012	nr	Prospective study	41	nr	nr	nr	nr	Both
Meloni et al., 2012	2009–2014	Prospective study	10	52 (nr)	6	4	Mixed	Both
Fierz et al., 2013	2004–2007	Retrospective cohort	46	54	31	15	All malignant	Both
Ferrari et al., 2013	1998–2008	Retrospective cohort	14	50 (15–63)	8	6	Mixed	Mandible
Qu et al., 2013	2006–2010	Retrospective cohort	33	38 (19–66)	20	13	None	Mandible
Jacobsen et al., 2014	1997–2005	Retrospective cohort	33	52 (20–69)	17	16	Majority malignant	Mandible
Wang et al., 2015	2006–2008	Retrospective cohort	19	40 (28–55)	12	7	None	Mandible
Bodard et al., 2015	nr	Retrospective cohort	26	54 (19–72)	17	9	nr	Mandible
Hakim et al., 2015	1993–2012	Retrospective cohort	198	52 (nr)	nr	nr	Majority malignant	Both
Fang et al., 2015	nr	Retrospective cohort	74	47 (19–75)	nr	nr	All malignant	Mandible
Burgess et al., 2017	2009–2015	Retrospective cohort	59	51 (18–77)	35	24	All malignant	Both
Barber et al., 2016	2001–2009	Retrospective cohort	114	54 (nr)	64	50	All malignant	Both
Ch’ng et al., 2016	2005–2011	Retrospective cohort	246	nr	nr	nr	All malignant	Both
Kumar et al., 2016	2012–2014	Randomized controlled trial	33	40 (nr)	26	8	Mixed	Mandible
Jackson et al., 2016	2005–2014	Retrospective cohort	46	62 (17–80)	31	15	Majority malignant	Mandible
Wu et al., 2016	2004–2011	Retrospective cohort	36	52 (nr)	25	11	All Malignant	Both
Burgess et al., 2017	2009–2015	Retrospective cohort	59	51 (18–77)	35	24	All malignant	Both
Sozzi et al., 2017	1998–2012	Retrospective cohort	22	nr (12–70)	12	10	Mixed	Both
Kniha et al., 2017	nr	Retrospective cohort	28	58 (nr)	15	13	Majority malignant	Both

nr, not reported; M, male; F, female, Majority malignant

Patient Population and Flap Details

Among the 42 studies, a total of 2042 patients underwent vascularized bone reconstruction following segmental resection of maxilla, mandible, or both; median age was 54 years (range: 6–92; Table 1). Among studies reporting patient’s sex, 61% were male (n=897) and 39% were female (n=580).

Twenty-nine studies included over 50% of patients with oncologic reconstructions, with 14 exclusively examining this population. Four studies included patients exclusively with reconstruction for benign disease(13–16) The majority of the studies included patients with mandibular reconstruction. Twenty-six studies utilized the free fibula flap (FFF) for reconstruction (Table 2). Over one-third of the study groups also used other donor sites, including the iliac crest (n=14)^{(6–8, 14, 17–26)}, scapula (n=5)^{(8, 19, 22, 25, 27)}, and radial artery free flaps (RAFF; n=4)^(25–28).

Table 2:

Implant Survival in Patients Who Underwent Dental Rehabilitation with Microvascular Flaps

Study	Flap Type(s)	No. of Pts. with Implants in Flap(s)	Total No. of Implants in Flap(s)	Average No. of Implants in Flap(s) per Pt.	Average or median Follow-up, months	Median Implant Survival, %
Zlotolow et al., 1992	FFF	7	23	3.3	13	100.0
Sclaroff et al., 1994	FFF, DCIA	22	114	5.2	18	88.0
Barber et al., 1995	FFF	5	20	4.0	14	100
Roumanas et al., 1997	FFF	19	54	2.8	26	98.1
Gürlek et al., 1998	FFF, DCIA	20	60	3.0	47	91.7
Chang et al., 1998	FFF	8	34	4.3	14	100.0
Foster et al., 1999	FFF, DCIA	14	71	5.1	18	99.0
Schliephake et al., 1999	FFF, DCIA, SFF	21	83	4.0	36	100.0
Kovacs et al., 2000	DCIA	11	41	3.7	48	97.6
Shaw et al., 2005	FFF, DCIA, RAFF	35	126	3.6	48	73.9
Teoh et al., 2005	FFF	24	81	3.4	52	94.0
Kramer et al., 2005	FFF, DCIA	16	51	3.2	24	96.0
Garrett et al., 2006	FFF	17	58	3.4	20	94.8
Gbara et al., 2007	FFF	30	121	4.0	12	96.7
Smolka et al., 2008	FFF	30	108	3.6	50	92.0
Hundepool et al., 2008	FFF	18	69	3.8	18	97.0
Raoul et al., 2009	FFF	30	105	3.5	76	96.2
Virgin et al., 2010	FFF, RAFF	4	18	4.5	27	100.0
Salinas et al., 2010	FFF	44	114	2.6	41	82.4
Bodard et al., 2011	FFF	23	75	3.3	28	80.0
Barrowman et al., 2011	FFF, DCIA, SFF	31	32	1.0	18	95.7
Buddula et al., 2011	FFF, DCIA, SFF	48	59	1.2	36	90.2
Chang et al., 2011	FFF	10	25	2.5	39	100.0
Dholam et al., 2011	FFF	12	35	2.9	18	71.2
Fenlon et al., 2012	FFF, DCIA	41	145	3.5	36	87.6
Meloni et al., 2012	FFF	10	51	5.1	48	94.6
Fierz et al., 2013	FFF, RAFF, SFF	22	46	2.1	54	82.6
Ferrari et al., 2013	FFF	14	57	4.1	71	91.9
Qu et al., 2013	DCIA	25	81	3.2	26	100.0
Jacobsen et al., 2014	FFF	23	99	4.3	67	81.0
Wang et al., 2015	FFF	19	51	2.7	36	100.0
Bodard et al., 2015	FFF	26	80	3.1	72	97.5
Hakim et al., 2015	FFF	37	119	3.2	85	91.6
Fang et al., 2015	FFF	74	192	2.6	60	90.1
Barber et al., 2016	FFF, DCIA	30	82	2.7	36	87.8
Ch’ng et al., 2016	FFF	54	243	4.5	37	92.8
Kumar et al., 2016	FFF	33	104	3.2	12	99.0
Jackson et al., 2016	FFF	46	183	4.0	30	93.4
Wu et al. 2016	FFF, DCIA	22	126	5.7	120	93.2
Burgess et al., 2017	FFF, DCIA, RAFF, SFF	59	199	3.4	24	93.6
Sozzi et al., 2017	FFF	22	92	4.2	94	98.0
Kniha et al., 2017	FFF, DCIA	28	109	3.9	36	100.0

FFF, free fibula flap; BIFF, bone-impacted fibula free flap; DCIA, deep circumflex iliac artery; RAFF, radial artery free flap; SFF, scapula free flap

Thirteen studies reported additional procedures aimed to enhance bone height/quality or soft tissue management techniques^{(7, 11, 16, 21, 28–36)}. These included double-barrel fibula, vertical distraction osteogenesis, bone grafting, bone-impacted FFF, or sub-periosteal dissection with denture-guided epithelial regeneration.

Implant Placement for Dental Rehabilitation

Among 2042 patients with free-flap–based reconstruction, 53% (n=1084) underwent dental rehabilitation with 3636 total implants (Table 2). Implant types are described in Table, Supplemental Digital Content 1, which demonstrates types of implants used in each study. Most studies (85.7%) included patients with reconstructions following oncologic resections before or after RT; 14.3% studied patients with benign H/N lesions.

Implant Survival Outcomes

Implant survival, entire cohort

Among 1084 patients with implants, the overall median implant survival was 94.7% and weighted survival was 92.2%, with median follow-up of 36 months (Table 2).

Immediate versus delayed implant placement

Twelve studies reported on 151 patients who received 582 implants at initial flap reconstruction (immediate implant placement)(9, 11, 14, 15, 18, 19, 23, 31, 35, 37–39). However, due to limited data regarding implant success, a meta-analysis comparing success rates of immediate versus delayed placement was not possible. Only four studies(8, 13, 15, 39) reported on osseointegration outcome for implants in patients who underwent immediate placement (total patients, n=60; total implants, n=269; see Table, Supplemental Digital Content 2, which demonstrates survival outcomes in patients with immediate implant placement). Weighted survival was 97.0%, with median follow-up of 14 months.

Twenty-three studies(13, 7–9, 16, 17, 20, 21, 25–27, 29, 30, 36, 39–45) reported on implant survival in 597 patients with delayed implant placement (n=1897 implants). Weighted implant survival was 89.9% at a median follow-up of 40 months (see Table, Supplemental Digital Content 3, which demonstrates survival outcomes in patients with delayed implant placement).

Impact of radiation on implant survival

Nine studies reported on implant survival in non-irradiated flaps in over 118 patients with 827 implants(6, 23, 35, 38–40, 43, 46, 47) Weighted survival was 94.9% at a median follow-up of 36 months (Table 3). Among RT patients, implant survival was lower than in non-RT patients (proportion of successful implants: 633/748 vs. 788/827; survival rate: 84.6% vs. 95.3%; p<0.01; Tables 3–5).

Table 3:

Survival Outcome in the No Radiation (Control) Implant Placement Group

Study	No. of Patients	No. of Implants	No. of Implants Failed	Average or Median Follow-up, months	Median Implant Survival, %
Sclaroff et al., 1994	6	34	2	18	94.1
Teoh et al., 2005	17	56	0	17	100.0
Gbara et al., 2007	18	72	0	12	100.0
Salinas et al., 2010	nr	63	6	41	90.5
Fenlon et al., 2012	29	110	3	36	97.3
Jacobsen et al., 2014	nr	86	9	64	89.5
Ch’ng et al., 2016	32	177	9	37	94.9
Jackson et al., 2016	nr	156	9	30	94.2
Sozzi et al., 2017	16	73	1	94	98.6
Total	>118	827	39	36 (12–94) *	95.3 **

nr, not reported

^*Median follow-up

^**Weighted Survival

Table 5:

Survival Outcomes in the Post-irradiation Implant Placement Group

Study	Total No. of Patients	Total No. of Implants	No. of Failed Implants	Average or Median Follow-up, months	Implant Survival, %
Barber HD et al., 1995 a	5	20	0	14	100.0
Foster et al., 1999 b	3	15	0	18	100.0
Shaw et al., 2005 b	12	45	15	48	66.7
Teoh et al. 2005 a	6	25	5	13	80.0
Raoul et al., 2009 b	6	18	1	90	94.4
Salinas et al., 2010 a	22	51	14	41	72.6
Bodard et al., 2011 b	5	13	3	81	76.9
Barrowman et al., 2011 a	nr	15	5	18	66.7
Buddula et al., 2011 b	48	59	8	36	86.4
Jacobsen et al., 2014 b	3	13	8	67	38.5
Fierz et al., 2013 b	12	20	6	54	70.0
Ch’ng et al., 2016 b	nr	43	8	37	81.4
Hakim et al., 2015 b	16	48	5	85	89.6
Kumar et al., 2016 b	8	24	2	12	91.7
Jackson et al., 2016 b	19	26	5	30	80.8
Wu et al., 2016	22	126	9	120	93.2
Sozzi et al., 2017 b	5	8	1	94	87.5
Summary statistics	>192	569	95	41 (12–120) *	83.4 **

nr, not reported

^aHyperbaric oxygen was given before and or after radiotherapy.

^bNo information reported about the receipt of hyperbaric oxygen.

^*Median follow-up

^**Weighted Survival

Impact of radiation timing on implant survival

Six studies included over 38 patients who received 179 implants before RT(6, 23, 35, 38–40) At a median follow-up of 24 months, 20 implants failed, resulting in an overall weighted implant survival of 88.8% (Table 4).

Table 4:

Survival Outcome in the Pre-irradiation Implant Placement Group

Study	Total No. Patients	Total No. of Implants	No. of Implants Failed	Average or Median Follow-up, months	Implant Survival, %
Sclaroff et al., 1994 a	16	80	0	18	100.0
Teoh et al., 2005 a	1	5	0	15	100.0
Fenlon et al., 2012 b	12	35	15	36	57.1
Ch’ng et al., 2016 b	nr	23	3	37	87.0
Jackson et al., 2016 b	8	25	2	30	92.0
Sozzi et al., 2017 b	1	11	0	18	100.0
Summary statistics	>38	179	20	24 (15–37) *	88.8 **

nr, not reported

^aNo one received hyperbaric oxygen.

^bNo information reported about the receipt of hyperbaric oxygen.

^*Median follow-up

^**Weighted Survival

Seventeen studies included over 192 RT patients with secondary, delayed implant placement (n=569)(12, 8, 10, 18, 22, 26, 27, 32, 33, 35, 36, 38–40, 42, 43, 47). Of those, 95 implants failed at a median follow-up of 41 months. Overall weighted implant survival was 83.4% (Table 5).

Comparing weighted implant survival among the irradiated groups, implant survival tended to be superior among the patients who received RT following implant placement than among those who received RT before implant placement (88.8% vs. 83.4%; p=0.07).

Meta-analysis

A pooled analysis of seven studies^{(23, 35, 38–40, 43, 47)} that included 270 implants placed before or after RT in 77+ patients versus 721 implants in 95+ control patients revealed that RT significantly increased the risk of implant failure (RR=4.74; p<0.01), regardless of timing, when compared with controls (Figure 2).

Figure 2:

Forest plot comparing implant failure in patients with implant placement before or after radiotherapy versus no radiotherapy (i.e., controls). Events: Failure=implant failure; RT=radiotherapy. Heterogeneity: Q=10.2, p=0.12; I-square=41%; Tau-square=0.20. Overall effect: Z=5.68, risk ratio= 4.74, p<0.01. Random-effects model was used to conduct the meta-analysis. Size of each data marker corresponds to the weight of the study assigned by the model.

To study RT timing’s impact on implant survival, an analysis of four studies^{(35, 38–40)} compared 64 implants placed before RT in 10+ patients versus 102 implants placed after RT in 30+ patients. Implants in the former cohort had a reduced risk of failure, though this was not statistically significant (RR: 0.52; p=0.14; Figure 3).

Figure 3:

Forest plot comparing implant failure in patients who underwent pre-RT implant placement versus post-RT implant placement. Events: Failure=implant failure; RT=radiotherapy. Heterogeneity: Q=0.56, p=0.91; I-square=0%; Tau-square=0.00. Overall effect: Z= −1.49, risk ratio=0.52, p=0.14. Random-effects model was used to conduct the meta-analysis. Size of each data marker corresponds to the relative weight of the study assigned by the model.

QoL analysis

Ten studies reported QoL outcomes after reconstruction^{(9, 17, 28–31, 41, 42, 44, 45)}. Three used standardized measurement tools: European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (EORTC-QLQ30)⁽⁴⁴⁾ and its H/N module (EORTC-QLQ H&N35)^{(9, 44)}, and Oral Health Impact Profile-14 (OHIP-14)(30) None of the studies used overlapping methodology precluding direct comparisons. Nonetheless, these studies generally reported improvement in aesthetic and social aspects as well as function with dental implant placement.

DISCUSSION

Even though vascularized bone flaps begin the restoration of orofacial form and function^{(48, 49)}, reconstruction is incomplete until viable dentition is restored. In this systematic review, osseointegrated dental implants in vascularized flaps demonstrated a weighted survival of 92.2% at an average follow-up of about three years. Immediate implant placement had superior survival (97.0%), compared with delayed placement (89.9%), with a median follow-up of 14 and 40 months, respectively. Advanced technology in virtual surgical planning (i.e. computer-aided design and manufacturing), may facilitate immediate dental implant placement, allow for accurate positioning, all while accelerating dental restoration(5, 50). If implants are not placed during flap reconstruction, many oncologic patients may not undergo future dental restoration, regardless of the need for RT. Lastly, from a cost perspective, payment for implant placement may be bundled with the initial reconstruction, reducing the overall cost compared with the delayed procedure, often billed as an out-of-pocket expense. Overall, immediate dental implant placement is a feasible technique that can lead to better outcomes and may be more cost effective. Given the challenges of oncologic resection, reconstruction and subsequent treatment, this systematic review presents the literature foundation for immediate dental implant placement at the time of initial reconstruction, with virtual planning and protocols to enable early dental restoration even in the setting of RT(50).

Among RT patients, the implant survival rate was lower than for no-RT patients (84.6% vs. 95.3%, p<0.01). The meta-analysis strengthened this finding, demonstrating a significantly greater risk of implant failure in patients with RT after implant placement or in patients with implants placed in an irradiated flap (RR: 4.74; p<0.01; Figure 2). Radiation may cause peri-implant soft tissue inflammation, potentially impairing osseointegration of the implants by either directly jeopardizing bone vascularity or indirectly creating peri-implant pseudo heat pockets resulting from backscatter exposure, all of which impedes osteoinduction^{(9, 36, 46)}. Although it is not surprising that implant survival in non-irradiated flaps is superior to that in irradiated flaps, the debate has shifted to focus on timing of implant placement with respect to RT delivery(19, 51)

In this systematic review, RT stratified analysis revealed a superior survival for implants placed before RT versus those placed after although this was not statistically significant (p=0.07), possibly due to lack of power. The meta-analysis comparing risk of implant failure between pre-RT and post-RT patients showed a similar trend (RR=0.52; p=0.14). Implant placement before RT allows for osseointegration promoted by maximum flap vascularity but includes the potential development of hot spots that cause local bone erosion. On the other hand, insufficient vascularity may impair osseointegration in post-RT implant placement. In the latter scenario, some authors suggest implant placement 12 months after RT, which may allow for improved bone perfusion and implant survival^{(52, 53)}. This is supported in part by physiological studies that demonstrate a continuous loss of capillaries for up to 6 months post-RT⁽⁵⁴⁾ and over 70% reduction in bone regeneration immediately after RT, with a recovery towards baseline at 1-year⁽⁴⁷⁾. It is noteworthy in this review, patients waited 6–12 months after RT for secondary implant placement. However, the impact of absent dentition for such a protracted period likely has functional and health-related QoL impact. The difference in the follow-up time between the RT cohorts should be noted—patients with the implants placed before RT had shorter follow-ups, a bias that naturally favors survival in this cohort. Several studies also presented patients treated with hyperbaric oxygen exposure during RT did yet did not demonstrate any significant unidirectional pattern in the implant survival or treatment benefit(12, 22, 26, 43).

Ch’ng et al in 2016 presented a comprehensive study with a global examination of implant placement in head and neck reconstruction(38). This study found progressive implant loss as time increased and demonstrated that most implants lost in the native mandible and maxilla typically occurred in the first 36 months. Conversely, implants placed in FFF had increased loss beyond 60 months, with survival analysis demonstrating only a 40% survival beyond the 72-month timepoint. This finding may be due to survival bias – a type of selection bias that selects on those that have survived past a certain point (in this case 60 months). Osteoradionecrosis, as a late process occurring following RT, can certainly play a role in late losses in general. The authors discuss loss of all implants in a patient with a FFF and osteoradionecrosis. Continued examination of long-term outcomes is needed to better understand this implant survival process(38).

Few studies reported functional and QoL measures using a validated standardized questionnaire^{(9, 30, 41, 44)}. Therefore, it is important to properly assess post-rehabilitation oral functions, which play a crucial role in health-related QoL(1) Gürlek et al. used a subjective questionnaire and found that 75% of patients reported excellent cosmetic outcomes and were able to return to work, be socially active, eat a regular diet with minimal restriction in public, and have normal speech⁽¹⁷⁾. Dholam at al. used the EORTC QLQ-H&N35⁽⁴⁴⁾, one of the most commonly used questionnaires to assess QoL in patients undergoing treatment for H/N cancer. Fang et al. used OHIP-14, a questionnaire evaluating oral dysfunction-related QoL following surgery, where most patients reported no problems with their oral health at 1-year follow-up⁽³⁰⁾. These results highlight the limited utilization of validated patient-reported outcome measures for patients undergoing reconstruction following mandibular or maxillary defects and necessitates further research(55) Recently, the FACE-Q patient-reported outcome measure has been described in the H/N patient population(56) and may be another measure to assess postoperative QoL in dental reconstructive patients. While this current study focuses on survival and implant osteointegration, the main clinical outcome of interest in this population is dentoalveolar restoration. Implants are critical to this goal, but if the goal is not attained for any number of reasons, a patient who has integrated implants is, in essence, no different from one without. Implants are the foundation and enable the end goal, yet certainly from the patient perspective, are of little relevance without a final prosthesis.

Implant osseointegration may vary according to bone flap source, flap height and density, discrepancy of the transplanted flap, and the quality of peri-implant soft tissue. Most studies in this review used FFF, a mainstay of mandibular reconstruction due to its ease in harvesting, elevating with skin, muscle, and crural septum, and osteotomizing at various points, allowing for adaptation. Deep circumflex iliac artery (DCIA) flap was the second most common flap. The disadvantage of DCIA flaps is its cutaneous component is often unreliable due to the absence of skin perforators arising from the DCIA(57) Although this study did not assess overall survival based on flap type, Fenlon et al., reported equal survival among all flap types, including FFF, DCIA flap, RAFF, and rib(23) Bone height correction techniques, such as double-barrel fibula,⁽¹³⁾ vertical distraction osteogenesis, and bone-only grafts, also improved implant stability and allowed for optimal dental occlusion to facilitate movements required for chewing, swallowing, or speech. However, at this institution, double-barreling the fibula may adversely impact final restoration because of too much bone height and inability to place the final prosthesis. Barber et al. reported greater 5-year survival of implants in bone-impacted FFF (96%), in which morselized fibula is used to fill the medullary cavity of FFF, compared with the traditional FFF (77 %; p <0.01)⁷. However, there was no difference in survival between the irradiated and non-radiated groups (83.3% vs. 84.6%; p=0.70). One study demonstrated additional soft tissue management procedures (i.e., vestibuloplasty with fibromucosal, palatal mucosal, or skin graft and sub-periosteal dissection with denture-guided epithelial regeneration⁽³³⁾) had improved implant stability and survival.

The current findings should be interpreted in light of the strengths and limitations of the literature. This review encompasses all English publications over two decades and is comprehensive; however, there may be temporal bias due to recent trends in implant types and surgical developments. The data is subject to the heterogeneity of the included studies, data collection methods, and follow-up time. As longer follow-ups tend to lead to higher implant failure rates, the results may be affected by survival bias. As reported by Ch’ng et al., most implants were lost after 60 months of follow-up(38). In this review, the median follow-up was 36 months; only eight studies reported 92.1% survival at median follow-up of 60 months or longer. implant type and prosthesis design were not studied separately; these are factors that could affect implant outcome³⁹. Implant loss over time would be ideally examined using Kaplan Meyer time to even analysis techniques, however unfortunately due to the lack of granular data this was not possible. Furthermore, it is critical to consider the interactions of multiple implants placed in an individual patient. Given that prostheses require more than a single implant for support, as well as the fact that loss of an implant could render a full construct unusable, the concept of implant interaction is critical. A better analysis may be a clustered survival analysis utilizing more robust statistical techniques, though the current data did not enable such to be performed(58–60). Lastly, given the heterogeneity of studies, we were unable to determine with appropriate accuracy differences between maxillary implants and mandibular implants. Prospective studies should be conducted to further investigate these aspects of dental implant placement and prosthesis design following oromaxillofacial reconstruction, especially in patients undergoing H/N cancer surgery.

CONCLUSIONS

Osseointegrated dental implant placement into vascularized bone flaps is a successful technique that improves function and health-related QoL and should be strongly considered in patients undergoing segmental mandibular or maxillary reconstruction. For the oncologic patient, the data support greater implant survival rates when placed before RT. The success of immediate implant placement can be leveraged with using emerging technologies (virtual surgical planning), which allow for accurate positioning while accelerating the process of dental restoration.

Supplementary Material

Supplemental Digital Content 1

Supplemental Digital Content 1. Table shows Types of Implants Per Study.

Supplemental Digital Content 2

Supplemental Digital Content 2. Table shows Survival Outcome in Patients Who Underwent Immediate Dental Implant Placement.

Supplemental Digital Content 3

Supplemental Digital Content 3. Table shows Survival Outcome in Patients Who Underwent Delayed Dental Implant Placement.

Abbreviations

DCIA

deep circumflex iliac artery

EORTC

European Organization for Research and Treatment of Cancer

FFF

free fibula flap

H/N

head and neck

OHIP

Oral Health Impact Profile

MINORS

Methodological Index for Non-randomized Studies (MINORS)

RAFF

radial artery free flap

RR

risk ratios

RT

radiotherapy

QLQ-H&N 35

Quality of Life Questionnaire-Head and Neck

QoL

quality of life