Effectiveness of Platelet-Rich Plasma as an Adjunct to Core Decompression to Treatment Outcomes and Femoral Head Preservation in Avascular Necrosis of the Hip: A Meta-Analysis of Randomized Controlled Trials


Avascular necrosis (AVN) of the femoral head is a progressive, debilitating and multifactorial disease due to an intraosseous pathology leading to a decrease in functional outcomes hindering patients to perform daily activities.[1] It occurs in patients in the most productive age group of 25-50 years old in the male population with bilateral involvement in 59% of cases.[2,3] Patients with AVN reported symptoms of localized groin pain which can limit the range of motion (ROM), especially during passive internal rotation, and is associated with a decrease in the quality of life among patients usually presenting within two years from the onset of disease and in the absence of treatment.[2,4] Affected femoral heads would present with head distortion or collapse with arthritis. The pathology is based on multiple etiologies associated with a reduction in vascular supply to the subchondral bone of the femoral head leading to osteocyte death and eventual progressive collapse in the structure of the femoral head leading to arthritis of the hip joint.[2,3] When the articular surface has already collapsed, the disease usually does not regress which often leads to subsequent total hip arthroplasty (THA) at a young age.[2]


AVN can occur due to an atraumatic or traumatic cause. Majority of the patients present with a non-traumatic cause of AVN of the femoral head which is usually associated with the use of alcohol, glucocorticoids, hematologic disorders, pregnancy, chronic renal failure and other metabolic disorders. In 30% of non-traumatic AVN, the etiology of the disease is unknown and hence called idiopathic.[2,3] On the other hand, post-traumatic osteonecrosis of the femoral head is also a possibility with an incidence of around 20%-40% following femoral neck fractures and is typically related to the fracture pattern and involvement of the medial femoral circumflex artery (MFCA).[5]. Traumatic AVN occurs due to disruption of the vascularity to the femoral head, specifically the MFCA which is the major blood supply to the femoral head during adulthood.


The gold standard of treatment for late-stage Osteonecrosis of the Femoral Head (ONFH) is a total hip replacement which showed significant clinical success in this population. However, there are concerns with regard to its outcomes among young adults aged 25-50 years old undergoing joint arthroplasty. On the other hand, conservative methods such as physical therapy for the management of ONFH usually lead to poor outcomes with failure to provide long-lasting improvement. Early recognition and surgical treatment for patients with pre-collapse or early stages of ONFH are necessary for good clinical outcomes.[6,7] If diagnosed in the early stages of the disease, a more conservative surgical technique like core decompression may be done which is aimed to delay and prevent the need for future total hip arthroplasty. However, the efficacy of this procedure still remains to be controversial as failure of a single core decompression may not provide adequate bone healing in the necrotic area.[7]


Platelet rich plasma (PRP) is an autologous blood plasma that contains a concentrated and supraphysiological level amount of platelets and growth factors such as platelet-derived growth factors (PDGF), transforming growth factor beta 1 and 2 (TGF-b1, TGF-b2), IGFs and epidermal growth factors (EGF).[8-11] The use of PRP is considered safe and has shown positive effects on the stimulation of tissue healing with a rationale that additional platelets will exponentially increase the number of multiple growth factors mentioned above at the site of injury.[2,8,9] Aside from this, PRP used in conjunction with autologous bone graft is postulated to synergistically augment the growth and formation of bone.[5,10,11] In another study by Houdek, et al., they concluded that bone marrow-derived mesenchymal stem cells possess the capacity to transform into different mesenchymal cell types like osteoblasts, chondrocytes and adipocytes, thereby aiding tissue regeneration. When used alongside PRP in ONFH treatment, the growth factors in PRP enhance Multipotent Stem Cells' ability to differentiate into new bone and blood vessels. This collaboration between PRP and MSCs fosters osteogenesis and leads to improved healing process for early-stage ONFH leading to good clinical outcomes with 93% showing no progression of disease on MRI after 12 months.[11]


Core decompression is the most widely accepted hip-preserving treatment for early-stage osteonecrosis of the femoral head. Its main function is to reduce intramedullary pressure to allow adequate blood flow and promote new bone formation to reduce pathogenesis of the disease process.[2] It is known that core decompression relieves pain and helps in delaying or preventing disease progression by allowing creeping substitution of the necrotic area by bringing blood supply through drill tunnels.[2] The use of an autologous bone graft placed through the track of core decompression has evolved to be an appealing option for orthopedic surgeons.[7] In recent years, studies have been made regarding the use of biological adjuvants incorporated in autologous bone grafts to improve the outcomes of patients treated with Core Decompression (CD). However, findings were inconclusive and had heterogenous results with limited evidence.[2,10,11]


Materials and Methods

This study was performed in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items from Systematic Reviews and Meta-Analyses statement.


Eligibility Criteria and Study Inclusion

This meta-analysis included all RCTs comparing the use of PRP as an adjunct to core decompression and bone grafting vs core decompression alone to patients with AVN of the hip with the following characteristics: The population in the study included adult patients of both sexes diagnosed with AVN or ONFH. The intervention assessed for the study was the use of PRP as an adjunct to standard core decompression with bone grafting compared to the control which is core decompression with bone grafting alone as treatment. The primary outcome of the study was functional and pain levels before and at the final follow-up measured using the Harris Hip Scores (HHS) and the Visual Analog Scale (VAS), respectively. The secondary outcomes were treatment failure assessed by the number of hips with survival from disease progression, which were patients with no signs of further worsening of the AVN based on serial radiographs until final follow-up, and the number of those patients needing further hip surgeries for the same problem. All outcomes assessed in the study included preoperative and post-treatment scores assessed with a minimum of 24 months follow-up. Only full-text and published RCTs from 2000 up to the year prior to the commencement of review (2021). The search was performed and not limited by language. Duplicates were removed and retrieved references were screened in two steps: the first step was to screen titles/abstracts for matching our inclusion criteria and the second step was to screen retrieved full-text articles for eligibility to meta-analysis.


Exclusion Criteria

The exclusion criteria included the following: (1) Unpublished studies; (2) Case reports, cohort, reviews, other study methodologies other than RCTs; (3) Animal experiments; (4) Full-text journals which are not available; (5) Application of other biologic agents other than PRP as an adjunct to CD for treatment of ONFH; (6) Diseases other than ONFH.


Search Methods for Identification of Studies

The study was based on a comprehensive literature search using PubMed, Cochrane Library, Science Direct, Google Scholar and Med Line including RCTs as well as previous meta-analyses published from 2000 up to prior to the commencement of the review (2021). Studies that compared the use of PRP as an adjunct to CD and bone grafting versus CD alone to patients with AVN of the hip in terms of function and pain scores as well as disease progression and the need for further hip surgery for the same condition were identified.


The main key search terms used were ((((Avascular Necrosis) OR (AVN) OR (Osteonecrosis)) AND (Femoral Head)) OR (((Platelet Rich Plasma) or (PRP)) OR (Core Decompression))) and (Randomized Controlled Trial). The articles gathered are not limited to the English language. Three independent reviewers first screened the search results from each of the databases by title and abstract alone. Studies that do not satisfy the inclusion criteria or include any of the exclusion as well were not included in the meta-analysis. This screening included removing duplicated studies. Studies should have the same population, intervention, control and outcome. Potentially relevant articles were then reviewed and subsequently screened by way of full-text eligibility. Any discrepancies were resolved between the authors as to whether the study will be excluded or removed.


Data Extraction

The researchers gathered all available data from the literature collected that have passed the initial screening. Information from the original studies was extracted based on their relevance to the topic regarding the “Effectiveness of Platelet Rich Plasma for Bone Graft Incorporation as an Adjunct to Core Decompression to Treatment Outcomes and Delay Progression of Avascular Necrosis of the Femoral Head”. Non-relevant studies or data such as those patients undergoing surgeries other than core decompression were disregarded. Parameters collected for comparison between the two techniques included assessment of functional score using HHS and pain score using VAS. Other outcomes evaluated included the presence of disease progression by radiography and the need for further hip surgery for the same pathology.


Risk of Bias Assessment

The researchers appraised the risk of bias for all studies collated based on guidelines in the Cochrane Handbook. This involved proper randomization and blinding of participants, surgeons and outcome evaluators. Data extracted and methods were reviewed by two main authors. The presence of bias was further subdivided into low-risk, unclear risk or high-risk. Any study with high-risk bias in even one category was categorized as high risk of having bias. Studies in the low-risk group are studies that had a low-risk bias for all categories. Otherwise, they were classified to the unclear risk group. Disagreements on the classification of bias scoring or data were discussed with a third author.


Statistical Analysis

Statistical analyses were performed using Review Manager Statistical Software, Version 5.4. A p -value ≤0.05 was considered statistically significant. A random effects model (REM), using the Mantel-Haenszel model, was employed in the analysis since the study did not assume one effect size among all the studies. This type of model in meta-analysis takes within-study and between-study variations into account. The means and standard deviations of the study’s outcome variables were utilized to compute the standardized mean difference (SMD). Statistical heterogeneity between studies was scrutinized using the Q statistics test, I 2 statistics and tau squared (τ 2 ) statistics (Higgins & Thompsons, 2002).



Results of the literature search: a total of 1041 potential records as seen in Figure 1 were identified from the databases. After the removal of duplicate studies, 948 studies were further screened. 928 studies were excluded for not meeting the PICOM requirements after screening the title and abstract. In all, seven full-text articles were assessed for eligibility, of which four trials were excluded due to the modified intervention group with the use of other ortho-biologic agents other than PRP instilled through the core decompression site. The remaining three studies were included in this meta-analysis.


Picture 1

Figure 1: Flow diagram for selecting studies.


Characteristics of Included Studies

Table 1: Characteristics of the Included Studies

Author (Year)


Study design




Functional Score (HHS)

Pain Score (VAS)

Disease Progression

Need for further Surgery

Aggarwal (2020)

Total [n=43]









Early stage (stage I and II) of ANFH as diagnosed by magnetic resonance imaging (MRI) and staged by Ficat and Arlet staging


Pre-operative, 63-65 months




Xian (2019)

Total [n=46]







RCT; single-blinded

Post-traumatic ONFH of Association of Research Circulation Osseous (ARCO) stages II to III


Pre-operative, 36 months



Yang (2019)

Total [n=90]


PRP+CD + Oral Alendronate sodium








Pre-operative, 24 months





Of the three studies, all were RCTs. Two studies were conducted in China while one study was conducted in India. Two articles (Aggarwal and Xian) [2,7] were published in English, while one study (Yang) [12] was in Chinese language with publication time from 2019-2020. The study done by Yang, et al. [12] was translated into the English language to gain full assessment of the study methodology and data.


Evaluation of Treatment Outcomes

The study’s primary outcomes include an assessment of the function and pain scores of patients using the HHS and VAS system, respectively, for both the treatment and control groups. Data gathered for assessment included preoperative baseline scores and postoperative scores at the final follow-up (at least 24 months) to assess the effectiveness of treatment. HHS is a widely known scoring system that is considered to be reproducible and objective for hip pathologies with a maximum score of 100 as the best outcome.[13] VAS score is an essentially quantitative method of pain assessment with a horizontal 0-10 point scale with 10 being the most intense pain.[14]


On the other hand, secondary outcomes included treatment failure until the date of final follow-up (at least 24 months). This was assessed with the presence or absence of disease progression with serial radiography and the number of patients that need further hip surgery due to the same condition.[15]


Risk of Bias Assessment

Figure 2 and Figure 3 show the risk of bias graph and summary, respectively. It is noted that the study of Yang, et al. 2019,[12] showed a high risk of bias for selective reporting due to a shorter time of final follow-up with only 24 months compared to 36 months by Xian, et al. 2019,[7] and Aggarwal, et al. 2020 [2] with 64 months. It was also seen that the study of Yang, et al. 2019 [12] showed the presence of a high degree of bias as it also used oral alendronate as an adjunct to the PRP+CD group.

Picture 2

Figure 2: Risk of bias graph presented as percentages across all included studies.


Picture 4

Figure 3: Risk of bias summary for each risk of bias items in all included studies.


I. HHS Preoperative (Baseline) and Postoperative (at Final Follow-up)

It can be gleaned from Figure 4a-b, that analysis of the pooled data of preoperative baseline HHS showed a non-significant standardized mean difference (SMD) between PRP+CD versus CD alone using the random effects model (SMD = 0.32, z = 1.62, p = 0.11, 95% CI = –0.71 to 0.07). This finding denotes that the preoperative baseline HHS were not statistically different between the two groups. On the other hand, the postoperative HHS (at final follow-up) showed a statistically significant difference favoring higher HHS for the PRP+CD group compared to the CD alone group (SMD = 2.04, z = 3.80, p = 0.0001, 95% CI = 0.99 to 3.09). This denotes that patients who received PRP+CD had significantly higher functional scores compared to the control group as assessed with a higher HHS score upon final follow-up (at least 24 months).


Analysis of the functional outcome included all the three studies as seen in Figure 4a-b, using the HHS system which included pain, function, range of motion and presence or absence of deformity with 100 having the best score. All patients for both groups underwent this scoring system. However, HHS results in three studies between the preoperative baseline scores prior to the HHS score at the final outcome were not specifically mentioned in the study.


Picture 5

Figure 4a: Preoperative (Baseline) Harris Hip Score comparison between Platelet Rich Plasma + Core Decompression vs Core Decompression alone



Picture 6

Figure 4b: Postoperative (at Final Follow Up) Harris Hip Score comparison between Platelet Rich Plasma + Core Decompression vs Core Decompression alone


II. VAS score Preoperative (Baseline) and Postoperative (At Final Follow up)

The analysis of pooled data of preoperative baseline VAS score, as seen in Figure 5a-b, showed a non-significant standardized mean difference (SMD) between PRP+CD vs CD alone using the random effects model (SMD = -0.06, z = 0.33, p = 0.74, 95% CI = –0.39 to 0.28). This finding denotes that the preoperative baseline VAS scores were not statistically different between the two groups. On the other hand, the postoperative VAS scores (at final follow-up) showed a trend towards higher VAS scores for the CD alone group compared to the PRP+CD group (SMD = -2.25, z = 1.90, p = 0.0001, 95% CI = –4.56 to 0.07), however, this finding was not statistically significant. This denotes that patients who received PRP+CD had similar pain scores compared to the control group upon final follow-up (at least 24 months).


Analysis of the pain outcome included the two studies as seen in Figure 5a-b, using the VAS score with 10 having the worst pain to 0 with no pain . All patients in these two studies for both groups underwent this scoring system. VAS scores at final follow-up were measured at 24 months for the study of Yang, et al.,[12] while the study of Xian, et al.[7] was measured at 36 months.