Heterotopic ossification is a pathological process of abnormal bone formation in extraosseous soft tissues (Balboni et al., 2006). Although its pathogenesis is not fully understood, most scientists agree that it is a reactivation of the bone formation process (Lachiewicz et al., 2024). The process is initiated by an inflammatory response and local progenitor cells, such as chondrocytes and osteoblasts, proliferate and differentiate in response to the inflammatory response and are ultimately remodeled and mature to form ectopic bone tissue. The imbalance of inflammatory factors and the initiation of the inflammatory response play an important role in the process of heterotopic ossification (Hwang et al., 2022; Wang et al., 2022). As a common complication after hip surgery, HO has an incidence of 15%–90% after primary total hip arthroplasty (Herzberg et al., 2024; Łęgosz et al., 2019). Especially in patients with femoral neck fracture, ankylosing spondylitis, previous hip fracture and hypertrophic osteoarthritis (Peng et al., 2020; Neitzke et al., 2024). It is also worthy of note that the incidence of HO following hip arthroscopy is 8.3% (Bedi et al., 2012). The bony structure formed by HO may compress peripheral nerves and blood vessels, resulting in limited joint motion, pain, and delayed functional recovery, which significantly increases the difficulty of postoperative rehabilitation and the economic burden, and severely weakens the efficacy of surgery and patient satisfaction (Rama et al., 2009; Wang et al., 2023a).

At present, clinical strategies employed for the prevention of HO principally comprise pre-/postoperative irradiation and pharmacological interventions (Puga et al., 2025). Irradiation has been demonstrated to reduce the risk of HO by inhibiting the osteogenic differentiation ability of progenitor cells (Heyd et al., 1995; Marziano et al., 1996), Cai et al. found that the overall incidence of HO after THA was low in patients treated with irradiation, and that irradiation was the most effective method to prevent HO after THA (Cai et al., 2019). Despite the broad consensus regarding the efficacy of irradiation in preventing HO following THA, considerable controversy persists with regard to the optimal radiation dose, timing, and potential carcinogenic risk (Heyd et al., 1997; van Leeuwen et al., 1998). NSAIDs have been shown to exert their anti-inflammatory effects by inhibiting prostaglandin synthesis. Indomethacin has long been regarded as the “gold standard” for preventing HO, and diclofenac has been shown to have a unique advantage in the reduction of severe HO (Brooker grade III-IV) (Macfarlane et al., 2008). In recent years, selective cyclooxygenase-2 inhibitors (e.g., celecoxib) have gained prominence in clinical use due to their improved gastrointestinal safety and long-lasting anti-inflammatory efficacy (Barbato et al., 2012; Lavernia et al., 2014; Oberberg et al., 2021). Migliorini et al. found that celecoxib was the most effective in reducing the incidence of HO, followed by diclofenac and indomethacin by traditional mata analysis (Migliorini et al., 2021). However, Shapira et al. also found by meta-analysis that non-selective and selective NSAIDs had comparable efficacy in the prevention of HO, yet were more effective than irradiation (Shapira et al., 2021). However, the majority of extant studies have focused on a single drug class or included only a limited number of interventions, and have not systematically quantified the difference in efficacy between different NSAIDs.

Notwithstanding the body of evidence that supports the preventive value of non-steroidal anti-inflammatory drugs (NSAIDs) in comparison with radiotherapy, a number of key controversies remain unresolved (Anaspure et al., 2025). Firstly, the boundary between the efficacy of traditional non-selective NSAIDs (e.g., ibuprofen, naproxen) and newer selective inhibitors (e.g., The following aspects require further clarification: firstly, the results obtained from different studies have been inconsistent, or even opposite; secondly, there is a lack of multifactor corrected analysis of the relative comparative advantage of irradiation versus pharmacological interventions. To this end, this study is the first to undertake a simultaneous integration of randomised controlled trial data from 3,014 postoperative THA patients based on the Bayesian network meta-analysis (NMA) framework to systematically compare the preventive efficacy of ibuprofen, naproxen, celecoxib, indomethacin, diclofenac, erliximab and radiotherapy. The optimal intervention programme was elucidated by probability ranking and multi-node evidence network construction. The objective of this study is 3-fold: firstly, to overcome the limitations of conventional two-by-two comparisons; secondly, to furnish orthopaedic surgeons with a foundation for selecting individualised preventive regimens based on a high level of evidence; and thirdly, to provide theoretical support for the direction of future clinical trials.

The prevention of HO after THA has always been a significant concern (Wang et al., 2023b; Wang et al., 2023c). Consistent with prior studies, our findings indicate that among 926 cases in the control group, the incidence of HO was 55.2% (Herzberg et al., 2024), with 13.48% Of the 675 control cases evaluated using the Brooker scale developed grade III–IV HO. However, for all interventions, the incidence of Brooker grade III–IV HO was lower in the intervention group than in the control group. These findings suggest that both non-steroidal anti-inflammatory drugs and irradiation intervention have a beneficial effect in preventing HO after THA.

The heterogeneity in the direct comparison between diclofenac and control arises from variations in study sample sizes and the proportion of HO occurrence; however, both studies affirm diclofenac’s effectiveness in preventing HO after THA. This consistency bolsters our meta-analysis. Reis et al. reported that oral administration of diclofenac for 6 weeks reduced HO from 55% to 15% compared to placebo (Reis et al., 1992). Wahlstrom et al.’s study demonstrated that diclofenac completely prevented significant bone formation with minimal complications (Wahlström et al., 1991). A prospective study of 644 patients with THO reported similar results: 80% of patients in the diclofenac group did not develop ossification, and none developed Brooker grades Ⅳ HO (Jockheck et al., 1998). Sell et al. compared the incidence of HO following 2-week treatments with 150 mg and 75 mg of diclofenac per day, finding no significant difference between the two groups 6 months post-treatment, which supports a short-term, lower-dose treatment regimen (Sell et al., 1998). Haffer advocated a clear protocol, recommending the daily administration of at least 75 mg of diclofenac from the first day post-surgery for at least 9 days to effectively prevent HO (Haffer et al., 2022). Stable results from comparisons of diclofenac to irradiation further underscore diclofenac’s superiority (K̈olbl et al., 1998).

Etoricoxib, a selective cyclo-oxygenase-2 (COX-2) inhibitor, is likely ranked second in the comprehensive ranking. In the only study, etoricoxib demonstrated a similar HO incidence rate to diclofenac (17/45, 17/44), but etoricoxib had significantly fewer complications. As noted by Oberberg et al. in a study involving 194 participants, etoricoxib can effectively prevent HO after primary THA, offering a lower complication rate (Oberberg et al., 2021). A smaller study found Etoricoxib and indomethacin to have comparable preventive effects against HO. Among patients treated with etoricoxib, 31% developed Brooker grades I and II ossification, with 7% developing ossification, and no severe ossification was reported (Brunnekreef et al., 2013). Given the minor differences between the two drugs in our study and the scarcity of published studies, further research is warranted to provide a definitive comparison.

As a classic preventive measure against HO, local irradiation effectively reduces the occurrence of HO after THA (Heyd et al., 1995; Kruser et al., 2012). Both direct and indirect comparisons in our meta-analysis demonstrated the effectiveness of irradiation. We attribute the main cause of heterogeneity to sample bias across studies, but consistent preferences for irradiation were observed, and sensitivity analysis supports the robustness of our results. Controversy exists regarding the timing and dose of irradiation; some studies have employed postoperative fractionated irradiation. Kolbl et al. compared a single postoperative dose of 5 Gy with 7 Gy and found the latter to be an effective preventive regimen (K̈olbl et al., 1998). Conversely, numerous studies support the efficacy of low-dose and single-shot irradiation for preventing HO. Seegenschmiedt et al. found no difference in HO incidence between two doses of 5 × 2 Gy and 10 × 2 Gy (Seegenschmiedt et al., 1993). Lo et al. determined that a single postoperative dose of 700-cGy yielded results equivalent to fractionated irradiation (Lo et al., 1988), a finding corroborated by Hedley et al. (1989) and Healy et al. (1990). Heyd et al. noted good outcomes with preoperative irradiation, positing it as an alternative to postoperative irradiation (Heyd et al., 1997). A study involving 462 hips reported effective treatment with preoperative irradiation the night before surgery (Koelbl et al., 2003). Another study concurred, suggesting that patient management could be simplified and complications related to postoperative transportation reduced (Seegenschmiedt et al., 1994). Gregoritch et al. recommended irradiation within 4 h before surgery for satisfactory results (Gregoritch et al., 1994).

Concerns about radiation potentially causing loosening of joint prostheses or tumor occurrence have been addressed by several studies. Pakos et al. followed 97 patients (with three fatalities during follow-up) for 10 years and observed no tumors or prosthesis loosening in the irradiation and indomethacin group, similar to outcomes in the control group treated with indomethacin alone (Pakos et al., 2020). Significant heterogeneity was noted in comparisons between irradiation and diclofenac, attributable to disparate findings in the included studies. Kolbl et al. reported HO in 22 of 46 patients in the irradiation group compared to 6 of 54 in the diclofenac group (K̈olbl et al., 1998). In contrast, Sell et al. observed HO in 2 of 76 patients in the irradiation group and 18 of 77 in the diclofenac group (Sell et al., 1998), possibly due to differences in the timing and dosage of administered irradiation (the former study used a single dose before surgery, while the latter used three post-surgery doses). This calls for more robust studies to clarify the efficacy of diclofenac versus irradiation.

Studies on the priority of naproxen in HO prevention after hip surgery support its effectiveness. In a randomized controlled trial, Gebuhr et al. administered 250 mg of naproxen three times daily for 4 weeks, significantly reducing the incidence of ectopic ossification after bone cement THA (Gebuhr et al., 1991). Vielpeau et al. found that the same dosage over 6 weeks was more effective than indomethacin and placebo (Vielpeau et al., 1999). A meta-analysis involving 269 patients indicated that naproxen was associated with significantly lower HO rates after hip surgery, including arthroscopic surgery (Ma et al., 2018). Another study corroborated these findings (Zhang et al., 2019). Macfarlane et al. reported that naproxen and diclofenac, which are equally effective as indomethacin, can be considered as first-line treatment for the prevention of HO after THA (Macfarlane et al., 2008).

Among the drugs studied, indomethacin was most frequently included and once regarded as the “gold standard” for the pharmacological prevention of HO (Macfarlane et al., 2008). A previous systematic review reported indomethacin as the most effective drug available at the time (Fijn et al., 2003). Both the effectiveness and safety of indomethacin have been confirmed (Tözün et al., 1992), and our meta-analysis corroborates these findings. Several studies have demonstrated that a 2-week regimen of indomethacin can effectively reduce the incidence of HO (Hofmann et al., 1999). Wuning et al. compared 100 mg of indomethacin daily over 7 days with a 14-day regimen, finding similar benefits between the two durations (Wurnig et al., 1997). However, shorter indomethacin courses were not more beneficial; Heide et al. reported that a 3-day regimen was ineffective in preventing HO post-THA (van der Heide et al., 2007). Dorn et al. found an 8-day course more effective than a 4-day course (Dorn et al., 1998). Another study indicated that a 5-day course failed to achieve the desired preventive effect (Koorevaar et al., 1999). Both short-term indomethacin treatment and single-dose irradiation of 6 Gy have been shown to reliably prevent severe HO, with no significant difference in effectiveness between the two (Bremen-Kuhne et al., 1997). Burd et al. reported no differential effect in preventing HO after acetabular fracture surgery between these approaches (Burd et al., 2001).

As an early selective COX-2 inhibitor, celecoxib has increasingly been preferred by physicians due to its low risk of gastrointestinal complications. One study involving 150 participants in celecoxib and 250 participants in indomethacin groups showed equivalent efficacy in preventing periarticular ossification post-hip replacement, with the celecoxib group experiencing significantly fewer adverse reactions (Romanò et al., 2004). Only one celecoxib study met the inclusion criteria: it included 250 patients treated with oral ibuprofen or celecoxib for 10 days, revealing that 50% of the patients in the ibuprofen group developed ectopic ossification, compared to 41% in the celecoxib group (Saudan et al., 2007). A multicenter prospective observational study following 480 patients for 12 months reported a 20% incidence of HO in those treated with celecoxib, with no cases of Brooker grade 3 or 4; the incidence of HO in untreated patients was 55%, and 8.9% were Brooker grade 3 or 4 (Barbato et al., 2012). Another retrospective study observed that 14 of 106 patients treated with celecoxib developed HO, 11 of whom were Brooker grades 3 or 4, compared with 35 of 188 who had HO, 26 of whom were Brooker grades 3 or 4 (Oni et al., 2014). Lavernia et al. also endorse celecoxib as an effective and safe preventive measure for HO post-THA (Lavernia et al., 2014). The results of this study show that the incidence of gastrointestinal side effects of selective COX-2 inhibitors is lower than that of non-selective COX-2 inhibitors. In addition, recent expert opinion also recommends selective COX-2 inhibitors, such as celecoxib and etoricoxib, as the first choice for preventing HO in patients with a first THA, but a comprehensive assessment of the patient’s condition and vigilance for cardiovascular events are needed. (Migliorini et al., 2022). Nevertheless, given the limited number of studies on celecoxib, we recommend conducting more randomized controlled trials to strengthen these findings.

What is surprising is that the analysis results do not support the effectiveness of ibuprofen in preventing HO. The heterogeneity of this analysis stems from the large disparities in sample sizes across three studies and their contradictory conclusions. Based on this research, the effectiveness of ibuprofen in preventing HO remains uncertain, and it ranked last in our meta-analysis. This outcome aligns with previous studies that also report inconsistencies regarding ibuprofen. Koorevaar observed 95 patients and noted that a 5-day ibuprofen regimen failed to prevent post-THA HO (Koorevaar et al., 1999). A comparative study concluded that daily oral administration of 100 mg of ibuprofen for 3 weeks was comparable to indomethacin for post-THA prophylaxis (Schneider et al., 2023). However, the indomethacin group was treated from 2017 to 2019, while the ibuprofen group was treated from 2019 to 2022, introducing potential bias. Although it appeared that fewer people in the ibuprofen group developed HO, our meta-analysis revealed no significant difference between ibuprofen and the control group. This contrasts with a previous meta-analysis that found ibuprofen generally effective in reducing HO incidence post-THA (Tariq et al., 2023). The discrepancy arises because one favorable study on ibuprofen was excluded based on our criteria, where one-third of patients in the ibuprofen group developed HO, compared to three-quarters in the placebo group. Fijn et al. also suggest that short-term ibuprofen use is ineffective (Elmstedt et al., 1985).

Since only one randomized controlled trial comparing ibuprofen with celecoxib met the inclusion criteria, the incidence of HO was lower in the celecoxib group than in the ibuprofen group (48/117 vs. 73/123). Celecoxib ranked only ahead of ibuprofen in our network analysis, which may be influenced by the potential ineffectiveness of ibuprofen.

In general, our research supports the effectiveness of diclofenac, irradiation, naproxen, and indomethacin in preventing HO after THA, while the efficacy of ibuprofen remains unconfirmed. Concerning the lower incidence of complications associated with celecoxib and etoricoxib, these drugs may hold promise, and we advocate for further studies to verify their benefits. Mild (Brooker grade I) to moderate (Brooker grade II) HO does not impair pain sensation or hip function (Slätis et al., 1978), hence our focus is primarily on interventions effective in reducing severe (Brooker grade III-IV) HO occurrences. Unfortunately, current data are insufficient for a NMA on this matter. Special attention must also be given to preventing gastrointestinal side effects when using NSAIDs. Despite failing to conduct a meta-analysis on this aspect, we are concerned that gastrointestinal side effects were the most common reason participants withdrawal from studies, even though some studies used gastric protective drugs (Knelles et al., 1997; Kienapfel et al., 1999).

The occurrence of HO is multidimensional, and its clinical course is significantly regulated by surgical technique, gender and comorbidities, in addition to pharmacological interventions and radiation therapy (Wang et al., 2024). A series of studies have demonstrated a significant correlation between the choice of surgical approach for THA and the occurrence of HO. The incidence of HO after modified anterolateral approach and STD-Bauer approach was found to be significantly higher than that of minimally invasive surgical routes such as direct anterior approach (DA) (Hürlimann et al., 2017). Furthermore, a retrospective study confirmed that the incidence of HO in the group of DA approach (19.4%) was significantly lower than that in the group of direct posterolateral approach (36.1%) (Alijanipour et al., 2017). This phenomenon may be closely related to minimally invasive surgery for the extent of soft tissue stripping and minimisation of trauma. Furthermore, comprehensive measures such as meticulous intraoperative removal of bone debris, avoidance of cement fixation and reduction of local haematoma formation have been shown to reduce the risk of HO (Kantak and Shah, 2017; Johnson et al., 2025). Gender as an independent variable of HO occurrence: although the incidence of postoperative HO is slightly higher in male patients than in females, the available evidence is not yet sufficient to support a gender-stratified management strategy (Aprato et al., 2023). A review of the extant literature on comorbidity-related studies indicated that patients with a history of previous HO, bilateral proliferative osteoarthritis (men), ankylosing spondylitis, diffuse idiopathic osteomalacia, and Paget’s disease constituted a high-risk group for postoperative HO after THA (Iorio and Healy, 2002). Nevertheless, the study was unable to analyse the subgroups in depth due to the incompleteness of the sample data. It is important to note that HO prophylaxis must adhere to the principle of multimodal management, encompassing preoperative risk assessment, intraoperative minimally invasive procedures, and postoperative dynamic management. The present study employed a net meta-analysis to compare the different effects of six commonly used NSAIDs as well as irradiation in HO prophylaxis after THA, with the aim of providing some references for the development of clinical strategies.

In terms of preventive efficacy, diclofenac and etoricoxib demonstrated the most favorable performance in preventing HO after THA within this network meta-analysis. Irradiation, naproxen, and indomethacin are also satisfactory options, while ibuprofen is ineffective. Given the advantages shown by etoricoxib and celecoxib, further randomized controlled trials are recommended to clarify their effects. Our conclusions require confirmation through additional high-quality studies.