Survival rate and risk factors of implants in patients with head and neck neoplasms: a systematic review and meta-analysis Authors: Yingman Liu, Chunhao Yang, Yafan Song, Yinxia Sun, Hongru Wang, Yue Zhang, Haoning Tang, Yanfeng Li Published: 2026 ABSTRACT PurposeThe purpose of this meta-analysis was to evaluate the survival rate of oral implants and investigate potential risk factors for implant therapy in HNN patients.Materials and methodsAn extensive search was conducted in four electronic databases from their inception to April 2024. Study selection, data extraction, and risk of bias assessment were performed independently by two reviewers. Survival rates were calculated and analyzed using R software. Meta-analyses were performed using fixed-effects or random-effects models based on heterogeneity assessment (random-effects model used when I² > 50% and/or Q-test P < 0.10) to calculate the risk ratio (RR) for implant survival. although retrospectively registered in Prospective Register of Systematic Reviews (PROSPERO), the study protocol was developed prior to data synthesis and adhered to the PRISMA guidelines.ResultsForty-two studies (2,448 patients, 10,164 implants) with heterogeneous follow-up durations (range: 1–12 years) were included. The pooled aggregate implant survival rate was 88% (95% CI: 85–90%), derived from single-arm analyses and representing crude survival proportions rather than time-specific estimates. Significant heterogeneity was observed among studies (I² = 94.9%; P < 0.0001). The meta-analyses revealed the following results: (1) Implant survival was significantly lower in irradiated bone compared to non-irradiated bone (18 studies, RR for survival = 0.95; 95% CI = 0.91–0.99; P = 0.03). (2) Implant survival was significantly lower in grafted bone compared to native bone (15 studies, RR for survival = 0.92; 95% CI = 0.88–0.97; P = 0.0008). (3) implant survival rates were significantly higher in patients who underwent hyperbaric oxygen therapy (HBO) compared to those who did not (5 studies, RR for survival = 1.47; 95% CI = 1.06–2.03; P = 0.02). (4) implants placed in the anterior region had significantly higher survival compared to those in the posterior region (6 studies, RR for survival = 1.07; 95% CI = 1.02–1.12; P = 0.008).ConclusionThe aggregated evidence suggests an overall implant survival rate of 88% (95% CI: 85–90%), though with substantial heterogeneity (I² = 94.9%). Radiation therapy and bone grafting were associated with statistically significant but modest reductions in implant survival. Conversely, implant placement in the anterior region and adjunctive HBO showed potential survival benefits. HBO showed potential benefit (RR = 1.47), but this finding was based on only five studies with considerable heterogeneity (I² = 91%) and requires confirmation in larger and randomized trials.Trial registrationThe investigation was retrospectively registered in the International PROSPERO under the application identifier CRD42024604320.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12903-026-08304-8. Introduction Head and Neck Neoplasms (HNN) refer to a broad spectrum of pathologically diverse tumors occurring in the upper gastrointestinal or upper respiratory tract [1] ranking as the sixth most common type of cancer globally. According to GLOBOCAN data, 946,456 new cases and 482,001 deaths were reported in 2022 [2, 3]. The primary treatment often involves extensive surgical resection, which can result in severe anatomical and functional deficits of the maxillofacial region [4]. With the development and advancement of oral healthcare, implant technology has become widely used for the reconstruction of dentition defects or loss. Studies have reported 5-year and 10-year survival rates of dental implants at 97% and 90%, respectively [5]. For HNN patients, dental implant-supported rehabilitation remains a challenge. Due to the unique characteristics of tumor patients, the survival rate of implants is influenced by various factors. Achieving osseointegration within the bone is the first and most crucial step for successful dental implantation. Therefore, any factors that alter the microenvironment of the surrounding soft and hard tissues may affect implant survival [6, 7]. Potential influencing factors include patient-related factors, iatrogenic factors, and implant-related factors. A comprehensive understanding of these potential influencing factors can help clinicians avoid or mitigate the risk of implant failure. However, a comprehensive analysis of factors affecting implant success in HNN patients is lacking. Crucially, the relative importance and effect size of specific risk factors are poorly quantified and often contradictory across studies. Therefore, this study employs a systematic review and meta-analysis to address two specific aims: First, to establish a pooled estimate of the implant survival rate in HNN patients. Second, and more critically, to quantitatively synthesize evidence on the impact of 14 predefined factors on implant survival. These factors were conceptually stratified into three categories: (1) Patient-related prognostic factors (gender, smoking status, cancer stage); (2) Treatment-related modifiers (radiotherapy, chemotherapy, hyperbaric oxygen therapy, bone grafting, soft tissue grafting); and (3) Anatomical and procedural variables (implant site, bone graft source, implant timing, prosthesis type). This framework allows for clinically interpretable risk stratification, though we acknowledge that the observational nature of included studies limits true prognostic modeling. Therefore, we use the term ‘risk factor’ to denote statistical associations rather than established causal prognostic indicators. By providing a quantitative synthesis of prognostic evidence, this work aims to clarify which factors significantly influence implant success, thereby informing preoperative counseling, surgical decision-making, and postoperative care protocols for this complex patient population. Methods Search strategy The researchers in this paper searched four electronic databases (PUBMED, EMBASE, Cochrane Central Register of Controlled Trials, and Web of Science,) from their inception to April 2024. The search was restricted to English-language publications due to resource constraints for translation. This may introduce language bias. Grey literature was searched via OpenGrey, ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform to identify unpublished or ongoing studies. Conference abstracts from the International Congress of Oral Implantologists and European Association for Osseointegration were also screened. Additionally, the reference lists of all included studies and relevant reviews were manually screened to identify additional eligible records. The search strategy was constructed around the PICOS tool (Table 1): (P) Population: Patients diagnosed with HNN who underwent dental implant therapy. (I)Intervention: dental implant treatment and rehabilitation. Specific exposures included predefined risk factors (e.g., radiotherapy, bone grafting). (C)Comparator: For comparative analyses, the comparator groups were defined based on the absence of the specific exposure or a different state of the factor (e.g., non-irradiated bone vs. irradiated bone). (O) Outcomes: the survival rate and influential factors of dental implants in HNN patients; (S) Study design: randomized controlled trials and observational studies. The search strategy combined terms for the population (HNN) and intervention (dental implant), using both MeSH terms and free-text keywords. The full strategy for PubMed is provided in Supplementary Table 1. Table 1Summary of the applied PICO criteriaStudy characteristicsCriteriaP-PopulationHNN Patients treated with implant-based restorationI-InterventionDental implant treatment and rehabilitation.Specific exposures included predefined risk factors (e.g., radiotherapy, bone grafting)C-ComparatorFor comparative analyses, the comparator groups were defined based on the absence of the specific exposure or a different state of the factor (e.g., non-irradiated bone vs. irradiated bone)O-OutcomesPrimary: survive rates of dental implants in HNN patients.Secondary: risk factors of dental implants in HNN patients.S-Study typeRandomized controlled trials; longitudinal cohort studies; other observational studies (case-control, cross-sectional) Table 1Summary of the applied PICO criteriaStudy characteristicsCriteriaP-PopulationHNN Patients treated with implant-based restorationI-InterventionDental implant treatment and rehabilitation.Specific exposures included predefined risk factors (e.g., radiotherapy, bone grafting)C-ComparatorFor comparative analyses, the comparator groups were defined based on the absence of the specific exposure or a different state of the factor (e.g., non-irradiated bone vs. irradiated bone)O-OutcomesPrimary: survive rates of dental implants in HNN patients.Secondary: risk factors of dental implants in HNN patients.S-Study typeRandomized controlled trials; longitudinal cohort studies; other observational studies (case-control, cross-sectional) Summary of the applied PICO criteria Dental implant treatment and rehabilitation. Specific exposures included predefined risk factors (e.g., radiotherapy, bone grafting) Primary: survive rates of dental implants in HNN patients. Secondary: risk factors of dental implants in HNN patients. Inclusion criteria (1) Population: Human subjects diagnosed with HNN (benign or malignant) who underwent surgical resection and subsequent rehabilitation with dental implants; (2) Study Design: Randomized controlled trials, prospective cohort studies, retrospective cohort studies, and case-control studies; (3) Outcome: The study must have reported the implant survival rate and/or analyzed associations between specific factors and implant survival/failure. The minimum follow-up period must be at least 12 months; (4) Report: Sufficient data must be available to extract or calculate relevant effect estimates (e.g., number of implants/patients, survival/failure events by group). Exclusion criteria (1) Case reports, technical notes, in vitro studies, animal studies, reviews, and editorials; (2) Duplicate publications (the most comprehensive or recent version was retained). Study selection The literature was screened and excluded using EndNote, a literature management software. The entire selection process was conducted independently by two researchers through two sequential phases: Phase 1 (Title/Abstract Screening), aimed at rapidly excluding clearly irrelevant records (e.g., non-clinical studies, reviews, conference abstracts, protocols, or correspondence); and Phase 2 (Full‑Text Assessment), during which the full texts of the remaining potentially eligible records were retrieved and rigorously evaluated against the predefined inclusion and exclusion criteria to determine final inclusion. Throughout this process, both researchers independently screened the literature, comparing their findings. If they reached consensus, the studies were included; if discrepancies arose, they were discussed and resolved by a third researcher. Data extraction A standardized, two-part data extraction process was performed independently by two reviewers. Part 1: An eight-item form was used to extract descriptive data from all included studies: author, country, publication year, mean patient age, sample size (patients), total number of implants, overall survival rate, and follow-up period. Part 2: For studies included in comparative analyses, a separate detailed form was used to extract outcome-specific data. This included, for each comparison group: (1) The number of implants. (2) Key variables relevant to the analysis (e.g., radiation dose, implant timing, flap type—when reported). Handling of Missing Data and Multi-arm Studies: Missing Data: For data essential to meta-analysis (e.g., exact failure counts, radiation dose), corresponding authors were contacted via email. If no response was received, the study was excluded from that specific analysis but retained in the systematic review. Multi-arm Studies: For studies with multiple relevant groups, data for each pairwise comparison of interest were extracted separately. To avoid double-counting participants from a shared control group in any single meta-analysis, the control group sample size was appropriately divided according to standard recommendations. For overall survival pooling, the survival rate at the longest available follow-up was extracted for each study. For studies reporting survival at multiple time points, the longest follow-up was used to maximize observation time, acknowledging that this introduces variability. Sensitivity analyses stratified by follow-up duration (≤ 3 years, 3–5 years, > 5 years) were performed to assess stability of estimates (Fig. 2A). Risk of bias assessment Two researchers independently assessed the risk of bias based on the Newcastle-Ottawa Scale (NOS) criteria for evaluating the quality of observational studies, specifically cohort and case-control studies. The NOS evaluates studies across three predefined domains: Selection (maximum 4 stars), Comparability (maximum 2 stars), and Outcome (maximum 3 stars), with a total possible score of 9 stars. NOS scores ≥ 6 was classified as methodologically acceptable studies with lower risk of selection and reporting bias [8, 9]. Any disagreements between the reviewers were resolved through discussion or by consultation with a third senior researcher. The distribution of NOS scores across study designs is provided in Supplementary Table 2. Data analysis The meta-analysis was conducted using Review Manager (RevMan) version 5.4. For count data, the risk ratio (RR) was used as the effect measure. Survival rates were calculated and analyzed using R software. The overall implant survival rate was pooled across single-arm studies using a generalized linear mixed model with logit transformation to stabilize variances. For the single-arm meta-analysis of overall survival, between-study variance was estimated using restricted maximum likelihood and effect sizes pooled using the inverse variance method. To examine whether survival rates varied by follow-up duration, restricted cubic spline analysis was performed with 4 knots at the 5th, 35th, 65th, and 95th percentiles. The CI corresponding 95% was subsequently calculated and presented in a forest plot. The degree of heterogeneity among the studies was assessed using I² and Q statistics. Dichotomous outcomes were analyzed using both the fixed-effects and random-effects models. Heterogeneity was considered significant if I² > 50% and/or a Q-test P-value < 0.10. The model selection strictly followed the predefined threshold: fixed-effects model for I² ≤ 50% and Q-test P ≥ 0.10; random-effects model for I² > 50% and/or Q-test P < 0.10. This decision framework is reported consistently for each meta-analysis in the Results section. Based on the consistency of the results, the reliability of the combined results was assessed, and a funnel plot was created to examine potential publication bias. Given the analysis of 14 risk factors, we acknowledge the potential for Type I error inflation due to multiple comparisons. This limitation is discussed in detail in the Limitations section (Sect. 4). We examined 14 risk factors without applying formal correction methods for multiple testing (e.g., Bonferroni, false discovery rate). This approach was justified by the explicitly exploratory, hypothesis-generating nature of the study, the clinical interdependence of many examined factors (e.g., radiotherapy and hyperbaric oxygen therapy), and the conservative nature of Bonferroni correction which would substantially increase Type II error risk. All statistically significant findings are interpreted as requiring confirmation in prospective studies, and P-values are reported without adjustment. Ethical approval and study protocol Ethical approval was not required, as the investigation did not involve human or animal subjects and was based solely on published literature. The investigation was retrospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the application identifier: CRD42024604320. Results Study and identification and selection A total of 2613 documents retrieved from the electronic databases. No additional records were identified through other sources. Following duplicate removal, 1,990 unique records were screened based on titles and abstracts, resulting in the exclusion of 1,896 records that were not relevant to the review question. The full texts of the remaining 94 articles were retrieved and rigorously assessed for eligibility against the pre-defined inclusion and exclusion criteria. At this stage, 52 articles were excluded for the following reasons, categorized according to the inclusion and exclusion criteria defined in the Methods section: 21 studies were excluded for focusing on non-tumor populations; 15 were excluded as they were meeting abstracts, editorials, or case reports; 14 were excluded due to insufficient or un-extractable outcome data; and 2 were excluded as duplicate publications. Consequently, 42 studies fulfilled all criteria and were included in the qualitative synthesis. All 42 studies contributed data to the meta-analysis of the overall implant survival rate, while the number of studies available for meta-analyses of specific risk factors varied, as detailed in the Results. The complete study selection process is illustrated in the PRISMA flow diagram (Fig. 1). Fig. 1The PRISMA flow diagram of the search and screening process The PRISMA flow diagram of the search and screening process Quality assessment of the included studies The selected studies were evaluated using the NOS. Of the total, 22 studies (52.4%) scored 9, 14 studies (33.3%) scored 7–8, and 6 studies (14.3%) scored 6. There was a range of scores from 6 to 9, with a higher score representing a higher level of assessment quality [10]. The distribution of scores across the NOS is presented in Supplementary Table 2. All 42 included studies met the predefined threshold for high quality (NOS ≥ 6), with scores ranging from 6 to 9. As no studies were classified as low quality, sensitivity analysis stratified by study quality was not performed due to the lack of meaningful variability. Characteristics of the included studies In total, 42 clinical trials were included, encompassing 2,448 patients diagnosed with HNN and 10,164 dental implants. Table 2 summarizes the fundamental characteristics of each study. For the stratified survival analysis by follow-up duration, studies with clearly reported mean or median follow-up were classified accordingly. For studies reporting only a range of follow-up times, classification was based on the midpoint of the range. The pooled survival rate was 88% across all studies. Sub-group analysis by follow-up duration revealed similar survival rates (≤ 3 years: 90%; 3–5 years: 85%; >5 years: 89%) (Fig. 2A). Restricted cubic spline analysis revealed no significant association between follow-up duration and implant survival rate (Fig. 2B). Meta-regression showed no linear trend (P = 0.894) and ANOVA indicated no significant non-linearity (P = 0.102), supporting the validity of pooling studies with variable follow-up periods. Table 2Characteristics of the studies included in the meta-analysisAuthor, YearCountryAge (mean ± SD)Total patients/male/femaleTotal implantsOverall survival rate(%)Follow-up period (years)Aravind Buddula, 2011USA60.2 ± 11.348/29/1927187.823Michael Burgess, 2016Australia5159/35/2419993.475Cao Yingguang, 2003ChinaNR27/18/913164.895Matías Cuesta-Gil, 2009Spain52 ± NR111/80/3170692.639Kanchan P. Dholam, 2011IndiaNR12/NR/NR3562.862Nicole Ernst, 2016Germany65.8 ± NR36/29/719498.453F. Goker, 2022Italy63.33 ± 14.2819/9/108298.784.5Susanne Heberer, 2011Germany61.1 ± NR20/16/410298.041.2Sarah A. Hessling, 2015Germany55 ± NR59/39/2027297.065Wei Huang, 2012China45.2 ± NR24/18/68886.368.25Joannis Katsoulis, 2013SwitzerlandNR28/NR/NR10482.695D. Khadembaschi, 2021AustraliaNative 59 ± 1577/52/25221955Flap 50 ± 1877/46/31240885M.O. Klein, 2009Germany58.4 ± NR68/NR/NR19082.635Anke Korfage, 2010The Netherlands61.5 ± 11.550/35/1519592.825Kovács, A. F., 2001GermanyNR47/37/1016098.7510C. A. Landes, 2006Germany63 ± NR30/22/811499.123Sabine S. Linsen, 2012Germany55.7 ± 16.2566/43/2326287.0210 Mancha de la Plata, 2012Spain55.5 ± 8.730/22/822589.783.75Viivi Mattila, 2024Finland68 ± NR59/27/3219687.762.8C. NACK, 2015GermanyNR20/16/410275.495Giorgio Pompa, 2015Italy51 ± 1934/12/2216890.481.9R. H. Schepers, 2006The NetherlandsM:64.8 ± 10.2F:68.1 ± 10.448/29/1913998.562Ramona Schweyen, 2022GermanyM:63.2 ± 12.4F:63.0 ± 13.4118/83/3532889.905Richard J. Shaw, 2005USANR77/NR/NR36484.623.5Wolfgang Wagner, 1998Germany55 ± NR63/54/927572.8010Thomas Weischer, 1999GermanyNR40/NR/NR17590.863Thomas Weischer, 2001Germany57 ± NR24/19/511197.302.5Richard Werkmeister, 1999Germany55 ± NR29/23/610975.233Kaan C. Yerit, 2006Australia57.8 ± 14.271/56/1531675.008Duohong Zou, 2015USA42.4 ± NR32/18/1411091.8212Joongyo Lee, 2022Seoul South Korea63 ± NR27/15/129080.003J. N. Lodders, 2021the Netherlands60.3 ± 13.244/23/2120283.664.9Dominic P Laverty, 2019UK63.3 ± NR167/109/5877995.515Kanchan P. Dholam, 2013India46 ± NR30/18/128576.475Michael R. Fenlon, 2009UKNR41/NR/NR14587.593Brent Woods, 2019America56 ± NR20/13/710277.454G Granström, 1999America64.9 ± NR7833570.455M. C. Rana, 2016India60 ± 2546/NR/NR16267.285G. Sammartino, 2011Italy55.8 ± NR77/51/2617288.372L. L. Visch, 2002The Netherlands62 ± NR130/78/5244678.0310E. Schiegnitz, 2021Germany57.3 ± NR164/110/5471190.153.5F. Wolf, 2021Germany65.7 ± NR121/91/3075190.413.8SD Standard deviation, NR Not reported, M Male, F Female Table 2Characteristics of the studies included in the meta-analysisAuthor, YearCountryAge (mean ± SD)Total patients/male/femaleTotal implantsOverall survival rate(%)Follow-up period (years)Aravind Buddula, 2011USA60.2 ± 11.348/29/1927187.823Michael Burgess, 2016Australia5159/35/2419993.475Cao Yingguang, 2003ChinaNR27/18/913164.895Matías Cuesta-Gil, 2009Spain52 ± NR111/80/3170692.639Kanchan P. Dholam, 2011IndiaNR12/NR/NR3562.862Nicole Ernst, 2016Germany65.8 ± NR36/29/719498.453F. Goker, 2022Italy63.33 ± 14.2819/9/108298.784.5Susanne Heberer, 2011Germany61.1 ± NR20/16/410298.041.2Sarah A. Hessling, 2015Germany55 ± NR59/39/2027297.065Wei Huang, 2012China45.2 ± NR24/18/68886.368.25Joannis Katsoulis, 2013SwitzerlandNR28/NR/NR10482.695D. Khadembaschi, 2021AustraliaNative 59 ± 1577/52/25221955Flap 50 ± 1877/46/31240885M.O. Klein, 2009Germany58.4 ± NR68/NR/NR19082.635Anke Korfage, 2010The Netherlands61.5 ± 11.550/35/1519592.825Kovács, A. F., 2001GermanyNR47/37/1016098.7510C. A. Landes, 2006Germany63 ± NR30/22/811499.123Sabine S. Linsen, 2012Germany55.7 ± 16.2566/43/2326287.0210 Mancha de la Plata, 2012Spain55.5 ± 8.730/22/822589.783.75Viivi Mattila, 2024Finland68 ± NR59/27/3219687.762.8C. NACK, 2015GermanyNR20/16/410275.495Giorgio Pompa, 2015Italy51 ± 1934/12/2216890.481.9R. H. Schepers, 2006The NetherlandsM:64.8 ± 10.2F:68.1 ± 10.448/29/1913998.562Ramona Schweyen, 2022GermanyM:63.2 ± 12.4F:63.0 ± 13.4118/83/3532889.905Richard J. Shaw, 2005USANR77/NR/NR36484.623.5Wolfgang Wagner, 1998Germany55 ± NR63/54/927572.8010Thomas Weischer, 1999GermanyNR40/NR/NR17590.863Thomas Weischer, 2001Germany57 ± NR24/19/511197.302.5Richard Werkmeister, 1999Germany55 ± NR29/23/610975.233Kaan C. Yerit, 2006Australia57.8 ± 14.271/56/1531675.008Duohong Zou, 2015USA42.4 ± NR32/18/1411091.8212Joongyo Lee, 2022Seoul South Korea63 ± NR27/15/129080.003J. N. Lodders, 2021the Netherlands60.3 ± 13.244/23/2120283.664.9Dominic P Laverty, 2019UK63.3 ± NR167/109/5877995.515Kanchan P. Dholam, 2013India46 ± NR30/18/128576.475Michael R. Fenlon, 2009UKNR41/NR/NR14587.593Brent Woods, 2019America56 ± NR20/13/710277.454G Granström, 1999America64.9 ± NR7833570.455M. C. Rana, 2016India60 ± 2546/NR/NR16267.285G. Sammartino, 2011Italy55.8 ± NR77/51/2617288.372L. L. Visch, 2002The Netherlands62 ± NR130/78/5244678.0310E. Schiegnitz, 2021Germany57.3 ± NR164/110/5471190.153.5F. Wolf, 2021Germany65.7 ± NR121/91/3075190.413.8SD Standard deviation, NR Not reported, M Male, F Female Characteristics of the studies included in the meta-analysis M:64.8 ± 10.2 F:68.1 ± 10.4 M:63.2 ± 12.4 F:63.0 ± 13.4 SD Standard deviation, NR Not reported, M Male, F Female Fig. 2Meta-analysis about the survival rate of dental implants in HNN patients (A) and Restricted cubic spline analysis showing the association between follow-up duration and implant survival rate B Meta-analysis about the survival rate of dental implants in HNN patients (A) and Restricted cubic spline analysis showing the association between follow-up duration and implant survival rate B Potential risk factors Extraction of potential risk factors Among the 42 studies included in this review, various factors potentially influencing implant survival rates were discussed. These 14 factors were selected for analysis based on their frequent mention in the clinical literature and hypothesized biological relevance to osseointegration in the compromised post-oncological environment. We extracted and analyzed the relevant data, identifying 14 factors for statistical analysis (Table 3). Table 3The summary of meta-analysis results about risk factorsVariableReferenceNumber of studiesModelI2(%)Risk Ratio95%CIP-ValueFemaleMale6Random510.980.931.030.42Smokingno Smoking9Random551.040.941.150.41T1-T2T3-T43Random611.030.961.110.43Radiotherapyno Radiotherapy18Random740.950.910.990.03Irradiation dose ≤ 50 Gy>50 Gy8Random621.080.981.200.11HBOno HBO5Random911.471.062.030.02Chemotherapyno Chemotherapy7Random580.990.931.040.64Grafted boneNative bone15Random580.920.880.970.0008Fibula flapIliac artery flap6Fixed00.970.921.030.30Grafted softno grafted soft3Random720.940.871.020.15Maxilla (implant site)Mandible16Random771.000.991.020.66Anterior (implant site)Posterior6Fixed451.071.021.120.008ImmediateDelayed4Random850.970.861.100.66Fixed prosthesisRemovable prosthesis3Random580.900.761.060.20HBO Hyperbaric oxygen, CI Confidence interval Table 3The summary of meta-analysis results about risk factorsVariableReferenceNumber of studiesModelI2(%)Risk Ratio95%CIP-ValueFemaleMale6Random510.980.931.030.42Smokingno Smoking9Random551.040.941.150.41T1-T2T3-T43Random611.030.961.110.43Radiotherapyno Radiotherapy18Random740.950.910.990.03Irradiation dose ≤ 50 Gy>50 Gy8Random621.080.981.200.11HBOno HBO5Random911.471.062.030.02Chemotherapyno Chemotherapy7Random580.990.931.040.64Grafted boneNative bone15Random580.920.880.970.0008Fibula flapIliac artery flap6Fixed00.970.921.030.30Grafted softno grafted soft3Random720.940.871.020.15Maxilla (implant site)Mandible16Random771.000.991.020.66Anterior (implant site)Posterior6Fixed451.071.021.120.008ImmediateDelayed4Random850.970.861.100.66Fixed prosthesisRemovable prosthesis3Random580.900.761.060.20HBO Hyperbaric oxygen, CI Confidence interval The summary of meta-analysis results about risk factors HBO Hyperbaric oxygen, CI Confidence interval First, we included three studies that investigated cancer stage as an influencing factor [11–13]. Second, tumor-related treatments considered as factors affecting implant survival included radiotherapy, chemotherapy, hyperbaric oxygen therapy (HBO), and bone and soft tissue grafting. Specifically, 18 studies focused on radiotherapy [11, 13–29], some of which further described the impact of radiotherapy dosage on implant survival. Using 50 Gy as a cutoff, we analyzed eight of these studies [11, 18, 26, 30–34]. Additionally, seven studies focused on chemotherapy [12, 15, 21, 24, 29, 35, 36]. Regarding bone grafting, 15 studies were identified [12, 13, 16–18, 21, 22, 24, 29, 30, 34, 36–39], including six that compared common grafted sources, such as the fibula versus the iliac bone [12, 28–30, 40, 41]. Three studies were included on the performance of soft tissue grafting [12, 21, 34]. Furthermore, five studies examined the use of HBO in post-radiotherapy patients [13, 29, 30, 37, 42]. Third, 17 studies focused on implant location as a factor influencing research outcomes. Among these, 16 studies distinguished between the maxilla and mandible [13, 15, 21–24, 29–34, 37, 41, 43, 44], and six compared anterior and posterior tooth positions [23, 25, 31, 33, 34, 37]. Four studies considered the timing of implant placement [22, 28, 39, 45]. Three studies investigated the impact of restoration type on implant survival rates [15, 30, 46]. Fourth, six studies included patient gender as an influencing factor [12, 21, 24, 37, 41, 46], and nine studies considered smoking status [12, 13, 15, 17, 22, 32, 34, 36, 39]. For factors supported by only 3–4 studies (e.g., cancer stage, soft tissue grafting, implant timing, prosthesis type), the meta-analytic results should be interpreted with caution due to limited statistical power. Meta-analyses of potential risk factors We extracted 14 factors that may influence implant survival rates. Given that the outcome of interest is implant survival, a RR < 1 indicates a lower survival rate (harmful effect) associated with the factor, while RR > 1 indicates a higher survival rate (protective effect). The following factors showed a statistically significant association with implant survival: Radiotherapy (18 studies) was associated with a significantly lower implant survival rate (RR: 0.95; 95%CI: 0.91–0.99; P = 0.03) (Fig. 3A). The effect size (RR = 0.95) indicates a 5% relative reduction in survival, corresponding to small absolute differences (e.g., 90% vs. 85.5% assuming high baseline survival). Whether this difference is clinically meaningful depends on the clinical context and patient risk profile. Bone grafting (15 studies) was associated with a significantly lower implant survival rate (RR: 0.92; 95%CI: 0.88–0.97; P = 0.0008) (Fig. 3B). Anterior (vs. Posterior) implant position (6 studies) was associated with a significantly higher implant survival rate (RR: 1.07; 95%CI: 1.02–1.12; P = 0.008) (Fig. 3C). HBO (5 studies) showed a statistically significant but highly heterogeneous association with implant survival (RR: 1.47; 95%CI: 1.06–2.03; I² = 91%; P = 0.02). The extreme heterogeneity indicates substantial variability in effect estimates across studies, precluding definitive conclusions about HBO efficacy. This finding should be interpreted as hypothesis-generating only and requires validation in larger, methodologically rigorous trials(Fig. 3D).The following factors did not show a statistically significant association: gender (RR: 0.98; 95%CI: 0.93–1.03; P = 0.42), smoking (RR: 1.04; 95%CI: 0.94–1.15; P = 0.41), T stage (in HNN patients) (T1-T2 vs. T3-T4: RR: 1.03; 95%CI: 0.96–1.11; P = 0.43), irradiation dose ≥ 50 Gy (RR: 1.08; 95%CI: 0.98–1.20; P = 0.11), chemotherapy (RR: 0.99; 95%CI: 0.93–1.04; P = 0.64), fibula vs. iliac flap (RR: 0.97; 95%CI: 0.92–1.03; P = 0.30), soft tissue grafting (RR: 0.94; 95%CI: 0.87–1.02; P = 0.15), implant site (maxilla vs. mandible: RR: 1.00; 95%CI: 0.99–1.02; P = 0.66), timing of implant placement (RR: 0.97; 95%CI: 0.86–1.10; P = 0.66), and type of prosthesis (RR: 0.90; 95%CI: 0.76–1.06; P = 0.20). Forest plots for these analyses are provided in supplementary Fig. 1. Fig. 3The forest diagram of survival rates of dental implants in the radiotherapy group and the no radiotherapy group (A), the grafted bone group and the native bone group (B), the anterior group and the posterior group (C) and the HBO group and the no HBO group D The forest diagram of survival rates of dental implants in the radiotherapy group and the no radiotherapy group (A), the grafted bone group and the native bone group (B), the anterior group and the posterior group (C) and the HBO group and the no HBO group D Bias analysis Funnel plots were visually inspected for asymmetry suggestive of publication bias (Fig. 4). For the four primary comparisons (radiotherapy, bone grafting, implant site, HBO), visual inspection suggested relative symmetry for radiotherapy and bone grafting analyses, while the HBO funnel plot showed slight asymmetry potentially due to small-study effects. Fig. 4The funnel plots about radiotherapy (A), grafted or native bone (B), implant site (anterior or posterior) (C) and HBO D The funnel plots about radiotherapy (A), grafted or native bone (B), implant site (anterior or posterior) (C) and HBO D Discussion This meta-analysis retrieved relevant published literature and included 42 articles for further analysis to explore the survival rates of implants in HNN patients and their associated factors. Our pooled estimate of 88% survival aligns closely with the 87.8% reported by Eik Schiegnitz et al. [47] in their meta-analysis. The pooled survival rate of 88% was derived from studies with highly variable patient populations, treatment protocols, and follow-up durations (I² = 94.9%). While stratified analyses by follow-up period and restricted cubic spline modeling suggested temporal stability of survival estimates, we acknowledge that other sources of heterogeneity were not fully explored, including radiation exposure, grafting procedures, geographical region, and implant systems. The high residual heterogeneity indicates that study-level factors beyond follow-up time substantially influence outcomes. Consequently, the 88% survival rate should be viewed as a broad summary estimate across diverse clinical scenarios with considerable between-study variability, rather than a precise prediction for individual patients or specific treatment protocols. Clinicians should consider this uncertainty when counseling patients and planning treatment. Our finding that grafted bone was associated with significantly lower survival rates corroborates Fan et al.‘s meta-analysis [48]. The osseointegration of implants in grafted bone vary due to differences in bone flap source, flap height and density, grafted skin flap, and the quality of the surrounding soft tissue [49, 50]. In this study, the most common sources of bone grafts for jaw reconstruction (fibula and iliac bone) were analyzed, and the results demonstrated comparable implant survival rates between fibula and iliac bone grafts. The comparable outcomes between fibula and iliac crest grafts suggest that the choice of donor site may be less critical than the fundamental difference between native and grafted bone microenvironment. No significant difference in implant survival rates was observed in areas with soft tissue grafts. Anatomic location also influences implant survival rates to some extent. Implants placed in the anterior region of the mouth generally exhibit higher survival rates than those placed in the posterior region. However, no statistical difference in implant survival rates was found between the maxilla and the mandible in this study. This contrasts with the conclusions of several other studies [51, 52], a discrepancy that may be attributed to advancements in implant technology and surgical techniques that have mitigated site-specific challenges. Radiotherapy is considered one of the risk factors for implant survival, whereas HBO may improve implant survival in irradiated tissues. This finding is consistent with a prior large meta-analysis reporting an 84.3% survival rate in irradiated tissues [53]. These findings align with those of multiple studies, which suggest that the sequence of radiotherapy and implantation does not impact implant survival rates. Some researchers recommend performing implant treatment 12 months post-radiotherapy to allow adequate time for tissue revascularization [54, 55]. Although higher radiation doses (e.g., > 50 Gy) are clinically associated with more severe tissue damage and potentially higher implant failure [33], The analysis of radiation dose > 50 Gy vs. ≤50 Gy was underpowered, based on only 8 studies with mixed designs (RR: 1.08; 95%CI: 0.98–1.20; P = 0.11; I² = 62%). The non-significant result should not be interpreted as evidence of no effect, but rather reflects uncertainty due to limited statistical power. This comparison requires further investigation in larger cohorts. In contrast, our analysis of five studies indicates that HBO is associated with a significantly higher implant survival rate (RR = 1.47), but this conclusion is tentative due to the small evidence base, extreme heterogeneity (I² = 91%), and lack of randomized controlled trials. HBO can enhance local blood circulation and promote the healing of damaged tissues [56]. However, there is currently no conclusive evidence supporting the improvement of implant survival rates in irradiated tissues following HBO-assisted treatment [42, 51, 57]. The optimal HBO protocol (pressure, duration, timing relative to implantation) remains undefined. Clinicians should weigh potential benefits against costs and patient burden on an individual basis. We caution against routine adoption of HBO based solely on this evidence. Furthermore, smoking is recognized as a risk factor for periodontitis and peri-implantitis, with long-term smokers exhibiting more severe bone loss [5, 58, 59]. However, this study did not find smoking to be a risk factor for implant survival rates in HNN patients. This may be attributed to the lack of detailed analysis on smoking quantity, frequency, and duration in the data relevant to this study. Therefore, continuous counseling and intervention for smoking patients remain essential, and smoking reduction and cessation should be key components of the long-term care plan for patients receiving dental implants. We included both benign and malignant head and neck neoplasms to reflect real-world clinical practice where implant rehabilitation is required regardless of tumor biology. While malignant tumors more frequently receive adjuvant radiotherapy and chemotherapy, potentially affecting implant survival, we addressed this heterogeneity by analyzing treatment-related factors (radiotherapy, chemotherapy) as separate variables rather than stratifying by malignancy status. This approach maximized statistical power while allowing examination of specific treatment effects. The T stage of cancer affects the complexity of surgery and the survival rates of tumor patients, but no relationship was found between implant survival rates and cancer stage. Therefore, comprehensive surgical treatment for HNN appears to achieve osseointegration rates comparable to non-oncological populations, though this may reflect patient selection and surgical expertise rather than an inherently optimal biological environment. Additionally, no statistical differences were observed among other implant and restoration protocols. This study employed a systematic literature retrieval strategy to ensure comprehensive and accurate information retrieval. This process encompassed both recent research findings and influential studies from the past, thereby incorporating a large number of high-quality and diverse research samples. This approach enhanced the generalizability of findings to diverse clinical settings, though we acknowledge that observational pooled data cannot ensure statistical representativeness due to potential selection biases in primary studies. This study has several limitations that should be considered when interpreting the findings. First, the analysis of 14 factors raises the possibility of Type I error inflation due to multiple comparisons. Therefore, the statistically significant findings, while reported, should be interpreted as exploratory, and the p-values viewed in light of this multiplicity. Second, there was significant clinical and methodological heterogeneity across the included studies, including variations in implant systems, and surgical protocols. While statistical models were employed to address heterogeneity, this variation limits the precision of pooled estimates. Third, we included both benign and malignant HNN to reflect clinical reality; however, most primary studies did not stratify outcomes by tumor biology (benign vs. malignant), precluding formal subgroup analyses to evaluate whether implant survival differs between these groups independent of treatment exposure. This represents an unmeasured source of heterogeneity that may confound the observed associations between adjuvant therapies and implant outcomes. Fourth, a major constraint is the absence of adjustment for potential confounders (e.g., oral hygiene, systemic health) in the original studies. Fifth, our analysis was limited to aggregate proportion data; a more nuanced time-dependent survival analysis (e.g., using time-to-event data) was not feasible due to the lack of individual patient data or Kaplan-Meier curves in most reports. Finally, we assessed publication bias primarily through visual inspection of funnel plots. Formal statistical tests for funnel plot asymmetry (e.g., Egger’s test) were not performed because Egger’s test has low statistical power with fewer than 10 studies and can produce misleading results in such scenarios. This represents a methodological limitation that should be considered when interpreting the robustness of our findings. Conclusion This systematic review and meta-analysis indicate that dental implant therapy in HNN patients can achieve a favorable aggregate survival rate of approximately 88%. The synthesis suggests that radiotherapy (RR = 0.95) and bone grafting (RR = 0.92) are associated with statistically significant but modest reductions in implant survival rates. Conversely, data from limited studies suggest potential benefits from anterior implant placement (RR = 1.07) and hypothesis-generating evidence for adjunctive HBO (RR = 1.47, based on 5 studies with high heterogeneity). Given the inherent limitations of the included studies—primarily their retrospective design, significant clinical heterogeneity, and the inability to control for confounding factors—these conclusions should be interpreted with caution. They underscore the viability of implant rehabilitation in this complex patient population while highlighting the need for careful, individualized risk assessment. Future prospective studies with standardized reporting are essential to strengthen the evidence base and refine clinical guidelines. Supplementary Information Supplementary Material 1. Supplementary Material 1.