Femoral Head Core Decompressions: Characterization of Subsequent Conversion to Total Hip Arthroplasty and Related Complications

Abstract

Background:

Core decompression is a minimally invasive joint-preserving approach for early-stage osteonecrosis. The rate at which core decompression patients require total hip arthroplasty (THA) and rates of perioperative adverse outcomes have not been well-characterized.

Methods:

Adult patients undergoing core decompression and/or THA with osteonecrosis of the femoral head were identified from the 2015 to 2021 Q3 PearlDiver M157 database. Those undergoing THA without or with antecedent core decompression were identified and matched 4:1 on age, sex, and Elixhauser Comorbidity Index. Postoperative 90-day adverse events were compared with multivariable analysis. Five-year rates of revision, dislocation, and periprosthetic fracture were compared by the Kaplan-Meier curve and log-rank tests.

Results:

Core decompressions were identified for 3,025 patients of whom 387 (12.8%) went on to THA within 5 years (64% within the first year). The median time from initial core decompression to THA was 252 days. For THA, 26,209 adults were identified and 387 had prior core decompression. After matching, there were 1,320 without core decompression and 339 with core decompression. No statistically significant differences were observed in 90-day postoperative adverse events or 5-year rates of revision, dislocation, or periprosthetic fracture.

Conclusion:

Core decompression may be an option for patients with osteonecrosis and does not seem to affect THA outcomes if required later.

Osteonecrosis of the femur head was thought to be the result of local ischemia due to compromised blood flow of the femoral head.¹ Core decompression is a minimally invasive joint-preserving approach for early-stage osteonecrosis that involves drilling one or more channels from the femoral neck to head through the subchondral necrosis area. This is theorized to relieve vascular pressure and generate new channels for angiogenesis.^2-4

The goal of core decompression is to limit/reverse the progression of osteonecrosis and prevent the requirement of total hip arthroplasty (THA). Owing to small numbers, the rate at which patients undergoing core decompressions go on to THA is poorly defined. A relatively small series of 24 hips by D'Ambrosi et al⁵ reported 50% survivorship of core decompression to subsequent THA at 75 months postoperatively, with a mean of 2.7 years from core decompression to THA for patients who did convert. However, the survival rate ranged from 29% to 80% depending on the severity of the osteonecrosis.⁵ Wei et al⁶ demonstrated similar results in a larger retrospective study of 1,537 osteonecrosis hips in which 52.4% converted to THA after an initial core decompression at 2 years postoperatively. Overall, core decompression has demonstrated moderately high rates of conversion to THA with a wide range of previously reported values.

Although core decompression has been established as a less invasive treatment option for those with osteonecrosis, the question has been raised if such an intervention affects outcomes of potential subsequent THA. The largest report to date totaled only 110 osteonecrosis hips from a meta-analysis by Deng et al⁷ and found no difference in blood loss or rate of revision at 2-year follow-up after THA. In a somewhat related consideration, femoral neck fractures stabilized with percutaneous screw fixation are at greater risk of adverse events after THA if conversion is later needed.^8-10

Given the wide range of reported rates of conversion from core decompression to THA, there is a need to investigate/study this special population and to better characterize the outcomes of potential subsequent THA after core decompression with a larger sample size, longer follow-up, and tracking of more specific THA-associated complications. This study aimed to address these points by leveraging a large, national, multi-insurance administrative database.

Methods

Data Set/Patient Cohort

This retrospective study used the 2015 to 2021 October PearlDiver M157 administrative database (Colorado Springs, CO), which is a large database that contains records from approximately 157 million orthopaedic patients and is well established for use in orthopaedic studies.^11-14 As the data are output in deidentified and aggregated nature, our Institutional Review Board deemed studies using this data set exempt from review.

Patients with a laterality-specified diagnosis of left (L) or right (R) hip osteonecrosis were identified using International Classification of Diseases, 10th Classification (ICD-10) coding. Patients who underwent a core decompression with an ICD-10 osteonecrosis diagnostic code on the same date as the procedure were identified using Current Procedural Terminology (CPT) coding. Similarly, patients who underwent THA with an ICD-10 diagnostic code of osteonecrosis on the same date as the procedure were identified using the CPT codes 27130 or 27132.

Exclusion criteria included patients aged younger than 18 years, diagnosis of infection, neoplasm, or trauma within 90 days before THA and patients who were not active for at least 90 days after THA.

Patients With Core Decompression Who Went on to Total Hip Arthroplasty

Patients who underwent core decompression were first analyzed. Those who went on to THA in the subsequent 5 years were identified. To ensure the ipsilateral hip was operated on for both the THA and prior core decompression, patients were considered to have subsequent THA only if the osteonecrosis diagnosis was ipsilateral between the two procedures. The code “Time Between” was then used within the PearlDiver interface to characterize the time between core decompression and THA for this study population.

A histogram categorizing the percentage of total core decompression patients who underwent THA was then created based on the provided percentile data. A box and whisker plot demonstrating the median, minimum, maximum, 25th percentile, and 75th percentile of time to THA from core decompression was also generated.

Total Hip Arthroplasty Outcomes for Those With Versus Without History of Core Decompression

Those with osteonecrosis undergoing THA with versus without history of core decompression were then defined. Those without a history of core decompression were matched to those with a history of core decompression 4:1 based on age, sex, and Elixhauser Comorbidity Index (ECI) using methods previously described.¹⁵

Ninety-day postoperative events were then identified based on ICD, ninth Classification (ICD-9) and ICD-10 coding using methods previously described. Serious adverse events (SAEs) were defined as at least one occurrence of surgical site infection, cardiac arrest, myocardial infarction, sepsis, prosthetic joint infection, venous thromboembolism (pulmonary embolism or deep vein thrombosis), or dislocation. Minor adverse events (MAEs) were defined as at least one occurrence of transfusion event, wound injury (wound dehiscence or hematoma), acute kidney injury, urinary tract infection, or pneumonia. Any adverse events represented the presence of at least one occurrence of either a SAE or MAE. Ninety-day readmissions were also identified using the “ADMISSIONS” code within the PearlDiver interface.

Five-year joint-specific adverse outcomes were additionally defined. These included revision-free survival, prosthetic joint dislocation, and periprosthetic fracture, each of which was identified using CPT, ICD-9, and/or ICD-10 coding.

Statistical Analysis

For univariate analysis, continuous variables (age and ECI) were compared using the Student t-test, and categorical variables (patient sex and incidence of all evaluated adverse events) were compared using the Pearson chi-squared test. Multivariable logistic regression, controlling for age, sex, and ECI, was used to determine the odds ratios (ORs), and 95% confidence intervals of 90-day adverse events for matched THA patients with versus without antecedent care decompression was then done.

Kaplan-Meier survival curves were constructed to compare 5-year survival with revision, periprosthetic fracture, and dislocation between the matched study groups using methods previously described.¹⁵ A log-rank test was conducted to evaluate any statistically significant difference between the matched study groups for each of these long-term adverse events.

All statistical analysis was done through PearlDiver Bellwether (PearlDiver) and GraphPad Prism v9.4.1 (GraphPad Software). Statistical significance was defined as P < 0.05 for all comparisons.

Results

Incidence and Time of Total Hip Arthroplasty After Care Decompression

A total of 3,025 patients who underwent core decompression with laterality-specified hip osteonecrosis were identified. Of these patients, 387 patients (12.8%) required subsequent THA (64% of which was completed within 1 year from core decompression) (Figure 1). Median time from core decompression to THA for these 387 patients was 252 days (25th percentile: 134 days; 75th percentile: 438 days) (Figure 2).

Figure 1.

Graph showing the survival of core decompression to total hip arthroplasty. Time-based histogram of core decompression patients who later required conversion to total hip arthroplasty at varying months after initial core decompression.

Figure 2.

Graph showing the time from core decompression to total hip arthroplasty in patients with osteonecrosis. Box and whisker displaying the time from initial core decompression to conversion to total hip arthroplasty in adult patients who underwent primary total hip arthroplasty with a history of prior core decompression for osteonecrosis.

Demographics of Patients With Osteonecrosis Undergoing Total Hip Arthroplasty

In total, 26,209 patients underwent THA with laterality-specified hip osteonecrosis (Supplementary Table 1, http://links.lww.com/JG9/A320). Of these 26,209 patients, 387 patients (1.5%) had undergone core decompression before THA. In the unmatched groups, patients with a prior core decompression tended to be younger (45.7 ± 12.4 versus 58.5 ± 13.7, P < 0.0001). No statistical difference was observed in sex distribution or ECI.

A successful 4:1 match based on age, sex, and ECI status was done. The matched cohorts consisted of 339 THA patients with a previous history of core decompression and 1,320 THA patients without a history of core decompression (Supplementary Table 1, http://links.lww.com/JG9/A320).

Short-term and Long-term Postoperative Outcomes

The matched cohorts of THA with versus without prior care decompression were not statistically significantly different in any of the 90-day aggregated or individual adverse events by univariable (Supplementary Table 2, http://links.lww.com/JG9/A321) or multivariable (Supplementary Table 3, http://links.lww.com/JG9/A322) analyses.

Regarding aggregated adverse events, no differences were found between the study groups for any adverse events (OR = 1.27, P = 0.1670), SAEs (OR = 1.22, P = 0.3988), or MAEs (OR = 1.15, P = 0.5100).

For severe adverse events, multivariable analysis demonstrated the following: surgical site infection (OR = 185, P = 0.3227), myocardial infarction (OR = 0.95, P = 0.9643), cardiac arrest (OR = 0.98, P = 0.9840), sepsis (OR = 0.95, P = 0.9190), and venous thromboembolism (OR = 1.13, P = 0.7200). In addition, THA-related complications such as dislocations (OR = 0.73, P = 0.4504) and prosthetic joint infection (OR = 1.21, P = 0.6852) were not statistically different between study groups.

For MAEs, blood transfusion (OR = 2.12, P = 0.0622), wound issues (OR = 1.23, P = 0.5809), acute kidney injury (OR = 0.89, P = 0.7560), urinary tract infection (OR = 1.11, P = 0.7730), and pneumonia (OR = 1.10, P = 0.8220) rates were not statistically distinguishable between the two cohorts.

Ninety-day readmission rates were also not different between osteonecrosis patients with or without a history of core decompression before THA (OR = 1.13, P = 0.5490). For 5-year outcomes, there was no difference in revision-free implant survival (P = 0.6000) or periprosthetic fracture (P = 0.7000) between the matched study groups (Figure 3). There was also no difference between matched study groups in 5-year prosthetic dislocation (Figure 4, P = 0.9000).

Figure 3.

Graph showing the 5-year implant survival and periprosthetic fracture rates after total hip arthroplasty. Kaplan-Meier curve comparing 5-year revision-free survival and periprosthetic fracture rates in adult patients who underwent primary total hip arthroplasty with a history of prior core decompression compared with patients without a prior core decompression. P-value resulting from a log-rank test is shown.

Figure 4.

Graph showing the 5-year dislocation rates. Kaplan-Meier curve comparing 5-year dislocation rates in adult patients who underwent primary total hip arthroplasty with a history of prior core decompression compared with patients without a prior core decompression. P-value resulting from a log-rank test is shown.

Discussion

Femoral head core decompression may be considered to address early-stage osteonecrosis.¹⁶ Previous studies assessing outcomes of this procedure have been limited by patient size and follow-up.^5,17,18 The rate of conversion of core decompression to THA and whether the core decompression influences the complication risk profile of potential subsequent THA remain poorly defined.

A total of 3,025 patients undergoing core decompression were identified in this study. Of these, 387 patients (12.8%) underwent subsequent THA for the ipsilateral hip within 5 years following the procedure (64% of these 387 patients were done within 1 year of the core decompression). Previous studies have estimated this rate to vary from 17% to as high as 80% depending on the stage of osteonecrosis and length of follow-up.^5,18 A systematic review reporting the outcomes of 2,540 osteonecrosis hips found a 38% conversion rate to subsequent THA at an average of 26 months after the initial core decompression.¹⁹ It is interesting that the conversion rate of core decompression to THA of this study was below these numbers, and the fact that most of these instances were within the first year after core decompression pointed to even lower failure rates in the following years.

When THA is done after core decompression, this study found no difference in evaluated 90-day perioperative outcomes relative to THA patients without prior core decompression. Matched multivariate analyses found no difference in individual, minor, serious, or any adverse events. This speaks to similar physiologic stresses related to the surgery. A case series from Lee et al²⁰ of eight hips with prior core decompression compared with 16 matched primary THA control subjects reported minimally increased surgical times and intraoperative blood loss without elevated risk of intraoperative femoral calcar crack or periprosthetic osteolysis.

Furthermore, when THA is done after core decompression, this study found no difference in 5-year survival to hip revision, periprosthetic fracture, or dislocation. These comparably hip-specific outcome metrics speak to similar outcomes of the joints being done. This aligns with the report of Lim et al²¹ of 36 hips treated with core decompression and subsequent THA matched with 39 osteonecrosis hips undergoing THA that found no difference in revision due to prosthesis mechanical failure at average follow-up of 4.6 years.

This study has limitations to consider. As with any study based on administrative databases, the study is limited by the quality of the coded data. Studies of osteonecrosis are limited by vagueness in coding, but the coding used within this study aligns with that used in other related studies.^22-24 Specific to this study, a number of patients needed to be excluded based on lack of sidedness data with their coding (required to ensure ipsilateral revision to the coded revision). Furthermore, this study was not able to assess the stage of osteonecrosis or hip-specific patient reported outcomes. Intraoperative data and surgical approach are also not encoded in the database. It would be ideal to compare and match cohorts of patients with similar early stage of osteonecrosis with or without core decompression and subsequent rate of THA to better delineate whether prior intervention affects THA outcomes. Owing to the heterogenous nature of osteonecrosis etiology and prognosis, the question could not be answered in the current administrative database study. We are also unable to avoid the potential selection bias of which patients were offered core decompression as compared with those being monitored or deemed too advanced on the stage of osteonecrosis and just waiting for arthroplasty procedures.

Conclusions

Overall, this study found that only 12.8% of core decompressions went on to THA within 5 years, most of which were in the first year after core decompression. In cases where THA was subsequently needed, no short or longer term postoperative differences were found relative to those undergoing THA without a history of core decompression. Together, these findings are supportive of considering core decompression in those with osteonecrosis when indicated.

Supplementary Material

jagrr-8-e24.00024-s001.docx ^{(14.6KB, docx)}

jagrr-8-e24.00024-s002.docx ^{(14.8KB, docx)}

jagrr-8-e24.00024-s003.docx ^{(14.5KB, docx)}