Does the use of i-FACTOR bone graft affect bone healing in those undergoing periacetabular osteotomy (PAO) for developmental dysplasia of the hip (DDH)? A retrospective study

ABSTRACT

The aims of this study were to compare, in patients with and without the use of i-FACTOR bone graft during periacetabular osteotomy (PAO) surgery for developmental dysplasia of the hip (DDH), (i) bone healing at six-weeks post-operatively (ii) rate of complications. This was a retrospective review of case records. Participants were people aged 15-50 years undergoing rectus-sparing minimally invasive PAO surgery for DDH. Group 1: patients with i-FACTOR, Group 2: No i-FACTOR. The primary outcome was the rate of bone healing on radiographs at 6 weeks. The likelihood of bone healing was compared using logistic regression with Generalised Estimating Equations (GEE) and expressed as odds ratios (95% confidence intervals (CIs; P < 0.05)). The occurrence of complications was extracted from surgical records. The i-FACTOR group had 3-times greater odds of partial/full union than those without [adjusted odds ratio (95% CIs, P-value)]: [3.265 (1.032 to 10.330, P = 0.044)]. The i-FACTOR group had 89% partial/full union at 6-weeks, compared to 69% of the non-i-FACTOR group. Half of the patients had leaking of bone graft in the i-FACTOR group versus 10% in the non-i-FACTOR group, 26% of the i-FACTOR group and 12% of the non-i-FACTOR group had neuropraxia of the lateral femoral cutaneous nerve (LFCN). Complication rates were low, and similar between groups. However, the rate of LFCN neuropraxia and bone graft leakage was higher in the i-FACTOR. These findings should be confirmed in a future prospective randomised clinical trial and include outcomes such as pain and quality of life.

BACKGROUND

Developmental dysplasia of the hip (DDH) is a well-known cause of hip pain and dysfunction and is the leading cause of early-onset hip osteoarthritis eventually requiring total hip joint replacement [1–3]. Individuals are often affected at an age of peak educational, work or family commitments. The mean age for periacetabular osteotomy (PAO) to be undertaken is approximately 28 years [4], with the majority of patients (86%) being female [5]. Hip osteoarthritis is a leading cause of global disability [6], in the United States, demand for hip replacement surgery among younger age groups has been predicted to grow markedly, with over half of joint replacement recipients expected to be aged under 65 years by the year 2030 [7]. Hip replacement surgery rates in Australia will rise to 80 000 by 2030, reaching an unsustainable cost of $2 billion annually [8], leading to clear repercussions for individuals and for society as a whole. PAO is a hip preserving surgery undertaken with the aim of delaying progression of osteoarthritis in those with hip dysplasia. Studies have demonstrated improvements in pain, function, activity levels, overall health and quality of life following surgery [9–11]. As with any surgical procedure, there is a risk of complications. Complications associated with PAO have been shown to adversely affect long-term pain and activity [5]. The most common major surgical complication following PAO is non-union [5], with over half of patients demonstrating incomplete healing of osteotomy sites at six-months post-operatively [12].

The PAO procedure involves cutting and realigning the hip socket from the pelvis to provide increased acetabular coverage of the femoral head and improved stability of the hip joint [13]. Appropriate healing of these bony cuts is an important focus of early rehabilitation following surgery as full weight-bearing capacity is an important early goal for young active patients. Delayed-union and non-union of the pelvic osteotomy sites are associated with delayed recovery following the surgery [5]. i-FACTOR is a Peptide Enhanced Bone Graft that is designed to enhance the body’s natural bone healing process and has been shown to enhance bone grafts and provide higher rates of fusion when compared to other methods in spinal orthopaedic surgery [14]. The i-FACTOR Peptide Enhanced Bone Graft consists of a synthetic P-15 peptide absorbed onto the surface of hydroxyapatite particles. The P-15 peptide is a 15-amino acid polypeptide that mimics the cell-binding domain of Type I collagen. The P-15 peptide enhances cell adhesion [15–18], migration [15] and induces osteoblast cell proliferation and differentiation [16]. The mechanism of action for P-15 is via receptor-mediated anchorage of osteoblasts which initiates several signal transduction pathways that lead to the synthesis of growth factors, cytokines and bone morphogenetic proteins [16, 17] leading to new bone formation. Evidence of the bone-forming ability of i-FACTOR has been demonstrated in numerous spine indications [19–24] and long-bone non-unions and delayed unions [25]. If i-FACTOR can reduce the incidence of complications relating to bony union, this could be of particular benefit to the patient’s pain and activity levels. The use of i-FACTOR has not been investigated in patients undergoing PAO surgery.

AIMS

Primary aim: To compare bone healing at six-weeks post-operatively in those patients with and without the use of i-FACTOR (Flex FR) Peptide Enhanced Bone Graft during PAO surgery.

Secondary aim: To compare the rate of surgical complications between patients with and without the use of i-FACTOR bone graft during PAO surgery.

STUDY DESIGN AND PROCEDURES

This study was a retrospective review of records for patients undergoing PAO between January 2009 and March 2020 (inclusively). Surgery was performed across three different hospitals over this period. Because the study was a retrospective review of patient records, patients were not prospectively allocated to either group. The treatment (i-factor or demineralised bone matrix) was applied as part of routine clinical care. Which treatment the patient received was determined by the treating surgeon’s standard practice at the time of surgery. Ethical approval was gained from the La Trobe University Human Research Ethics Committee (HEC20201) and the study was prospectively registered with the Australia New Zealand Clinical Trials Registry (ACTRN12621000625875).

Participants

All consecutive patients undergoing PAO surgery, aged between 15 and 50 years, of a single fellowship-trained surgeon formed the cohort for this study. Due to the retrospective nature of the study, the sample size was one of convenience. Patients received i-FACTOR between November 2017 and October 2019, while all remaining patients did not. Only patients with a Kellgren–Lawrence score [26] of ≤2 were considered for PAO surgery and were thus included in this study.

Patients were excluded if they underwent femoral osteotomy at the time of PAO surgery; if they were diagnosed with another significant hip condition (e.g. trauma, rheumatoid arthritis, avascular necrosis, Perthes disease, slipped upper femoral epiphysis, osteochondritis dissecans, fracture, septic arthritis, bursitis or tendinitis) or if they suffered from neurological, other musculoskeletal or rheumatological diseases affecting hip function. Patients were also excluded if no X-rays were available for the appropriate time frame (X-rays need to be performed >4 and ≤10 weeks post-surgery).

Surgery

All surgeries were performed by a single fellowship-trained surgeon in the metropolitan region of Melbourne, Australia. The rectus-sparing minimally invasive Ganz PAO procedure consisted of four orthogonal cuts around the acetabulum to allow for multidirectional correction of the mobilised acetabulum. Acetabular reduction was performed under intraoperative fluoroscopy. Femoral head coverage, acetabular inclination, joint congruency, acetabular version and medial translation of the femoral head were then re-established. Preservation of the posterior column and fixation using 3–5 screws facilitated pelvic and acetabular fragment stability following correction. When i-FACTOR was used, following completion of the correction and securing fixation with ×4 long small fragment pelvic recon screws (60–110 mm), final biplanar fluoroscopy was taken to ensure that the surgical prescription was achieved and impingement was avoided. Then a washout was performed for cell saver to complete its autologous re-transfusion. Following this, local anaesthetic was injected into lumbosacral plexus through the psoas fascia. The ilial cut, posterior column cut and pubic ramus cut were then laid with i-FACTOR and care was taken to make sure this was contained and it could not expand through to tissue planes. Gelfoam was then used to minimise leakage. When i-FACTOR was not used, a demineralised bone matrix (allograft bone mixed in with a small amount of autograft) was used instead. No additional commercial product was applied to the demineralised bone matrix.

Following PAO patients are usually in hospital four to five nights undertaking circulatory bed exercises and beginning mobilisation Day 1 following surgery, using a walking frame and progressing to crutches while under the supervision of an inpatient physiotherapist. Weight-bearing is restricted to 20% of body weight for the first 6 weeks and progressed if bone healing is sufficient on radiology. Outpatient physiotherapy typically occurs at 2 weeks post-surgery to check mobility and commence strengthening and range of motion exercises. Screws can remain place unless they cause irritation or break. In atraumatic cases, this occurs between 6–18 months post-operatively. Rehabilitation guidelines were identical for both treatment groups.

Medical records for all patients undergoing PAO surgery were deidentified by the surgeon and demographic, surgical and post-operative data extracted by two researchers.

X-rays

Each patient had undergone a standing weight-bearing anteroposterior (AP) pelvis radiograph using a standardised protocol at 6 weeks (>4, ≤10 weeks post-surgery), to assess bone healing. For this retrospective study assessment of X-rays was undertaken by two experienced, fellowship-trained, specialist musculoskeletal radiologists with at least 10 years’ experience. All X-rays were deidentified of patient information and dates. Preliminary screening was undertaken on 20 randomly selected X-rays to establish reliability and then consensus was established between the radiologists on each of the 20 X-rays prior to scoring the full cohort.

Outcomes

Primary outcome = Degree of bone healing (non-union versus partial/full union) identified on X-ray at 6 [4–10]  weeks following PAO surgery.

Secondary outcome = Complications identified in the medical notes related to PAO surgery. All complications were reported descriptively as present or absent. In addition, heterotrophic ossification was graded by the surgeon using the Brooker classification [27]. The Brooker classification grading was conducted at the time of screw removal, where heterotrophic ossification (HO) was graded 0–4. Grade 0 represents no HO, grade 1 represents small sections of bone within the soft tissues about the hip. Grade 2 represents bone spurs originating from the pelvis or proximal end of the femur, leaving at least 1 cm between femur and pelvis. Grade 3 is similar to grade 2 except the spurs of bone reduce the space between femur and pelvis to less than 1 cm. Grade 4 shows apparent bone ankylosis of the hip. Neuropraxia was recorded as either absent, present but not requiring additional surgical intervention, or present and requiring additional surgical treatment. No blood loss refers to no excessive blood loss reported or blood transfusions required.

Statistical analysis

Participant characteristics (age and sex) and complications were reported descriptively.

Inter-observer reliability for degree of bony healing (non-union versus partial/full union) was determined with the kappa statistic. Kappa scores were interpreted as ≤ 0 = no agreement, 0.01–0.20 = none to slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 = substantial and 0.81–1.00 = almost perfect agreement.

Hips were dichotomised into two groups: (i) PAO with the use of i-FACTOR and (ii) PAO without the use of i-FACTOR. For the primary outcome, dichotomous outcomes were evaluated (non-union versus partial/full union) using logistic regression with Generalised Estimating Equations (GEE) to account for paired limbs. Results were presented as odds ratios with 95% confidence intervals (CIs). The independent variable was intervention group (i-FACTOR: Yes/No) and covariates of age, sex, number of days from surgery to X-ray and year of surgery included in the analysis. We adjusted for number of days from surgery to X-ray to account for bony healing times, and for year of surgery to account for the learning curve of the treating surgeon. For the secondary outcome, rates of all surgical complications were descriptively reported. The relationship between i-FACTOR leaking and the complications of heterotopic ossification and neuropraxia (present or absent) were then explored logistic regression (adjusted for age, sex, days since surgery and year of surgery) with GEE.

Data was analysed using SPSS (v27.0; IBM Corp), Microsoft Excel, version 16 and statistical significance was assessed at P ≤ 0.05.

RESULTS

Between January 2009 and March 2020, 393 PAOs were undertaken on 295 patients [75% women; median age (±IQR); 22(±11)]. Numbers excluded from the study are described in Fig. 1 and group specific characteristics in Table I.

Fig. 1.

Study flow chart. (n = number of hips).

Table I.

Participant characteristics

Participant characteristic	Group 1 (With i-FACTOR)	Group 2 (Without i-FACTOR)
Hips [n (% of cohort)]	72(18%)	321(82%)
Men [n (% within group)]	23(32%)	79(25%)
Women [n (% within group)]	49 (68%)	242 (75%)
Age at time of surgery [median (IQR)]	22 (9.5)	22 (11)
Number of days from surgery to follow-up X-ray [median (IQR)]	46 (7)	47 (9)
Excluded for reasons other than X-ray availability [n (% within group)]	14 (19%)	69 (21%)
All exclusions [n (% within group)]	21 (29%)	152 (47%)

RELIABILITY

The percent agreement for the degree of bone healing (non-union versus partial/full union) was 80%. The kappa score was 0.529 indicating moderate agreement.

PRIMARY OUTCOME

Logistic regression (with GEE) of use of i-FACTOR predicting likelihood of bone healing with adjusted analyses performed per hip, (covariates of age, gender, number of days from surgery to X-ray and year of surgery) was conducted. In patients undergoing PAO surgery, those in the i-FACTOR group had 3-times greater odds of partial/full union that those without [adjusted odds ratio (95% CIs, P-value): 3.265 (1.032 to 10.330, P = 0.044)]; (Table II). The included covariates did not significantly impact our primary outcome of bone healing. Eighty-nine percent of patients in the i-FACTOR group achieved partial or complete bony healing at 6 weeks compared to 69% of the non-i-FACTOR group.

Table II.

Results of logistic regression (with generalised estimating equations) predicting likelihood of bone healing with use of i-FACTOR

Variable	B	S.E.	Wald	df	P	Odds ratio	95% CIs
Group (i-FACTOR or no i-FACTOR)	1.183	0.5877	4.054	1	0.044	3.265	1.032 to 10.330
Age	−0.023	0.0249	0.820	1	0.365	0.978	0.931 to 1.027
Sex	0.473	0.5008	0.891	1	0.345	1.064	0.601 to 4.281
Days from surgery to X-ray	−0.016	0.0191	0.669	1	0.413	0.985	0.948 to 1.022
Year of surgery	−0.188	0.1192	2.477	1	0.115	0.829	0.656 to 1.047

S.E.= standard error; df = degrees of freedom; P = statistical significance <0.05.

SECONDARY OUTCOMES

The number of post-operative complications are reported in Table III for each group. Table III includes all participants, except those excluded for reasons other than X-ray availability. The most frequently reported complication was HO, with the same percentage reported in both groups (58%). Only 4% of reported nerve injuries required a subsequent procedure, the majority of reported cases (62%) relating to transient symptoms from the LFCN not requiring treatment. Fifty-two percent of patients in the I-FACTOR group had leaking of bone graft on X-ray at the six-week post-operative time point compared to 10% of the non-i-FACTOR group. The total number of complications for each group were 58 (i-FACTOR group) and 189 (non-i-FACTOR group). The results of logistic regression using GEEs showed that those with leaking of i-FACTOR did not have greater odds of either HO [adjusted odds ratio 1.405 (0.625–3.160, P = 0.411)] or LFCN neuropraxia [adjusted odds ratio 0.970 (0.900–1.045, P = 0.758)].

Table III.

Rates of post-operative complications for both groups

	All (n = 310) n (%)		With i-Factor (n = 58) n (%)		Without i-Factor (n = 252) n (%)
Blood loss	0	(0)	0	(0)	0	(0)
Bone graft leaking	54	(17)	29	(50)	25	(10)
Intraoperative fracture	1	(<1)	0	(0)	1	(<1)
Infection	6	(2)	4	(7)	2	(1)
DVT	2	(1)	0	(0)	2	(1)
Nerve injury	47	(15)	16	(28)	31	(12)
LFCN altered sensation	31	(10)	9	(16)	22	(9)
LFCN, procedure required	11	(4)	6	(10)	5	(2)
Obturator	2	(1)
Femoral	0	(0)
Sciatic	0	(0)
Unspecified	3	(1)
Heterotrophic ossification*	179	(58)	32	(55)	145	(58)
Grade 1	3
Grade 2	165	(53)	32	(55)	133	(53)
Grade 3	3
Grade 4	1
Noted but not graded	7
Delayed union (no intervention)	1	(<1)	0	(0)	1	(<1)
Delayed union + procedure	11	(3)	2	(3)	9	(4)
Other	14	(5)	5	(9)	9	(4)
Allergy (stitches/skin reaction)	4
Lymphoedema	1
Broken screw	2
Delayed wound healing	1
Foreign body	1
Haematoma	2
Hypertrophic scar	2

*

reported at time of hardware removal.

LFCN = lateral femoral cutaneous nerve. Where number of total complications was less than n = 10, data for each sub-group was not reported. NR= not relevant.

DISCUSSION

People aged 15–50 years undergoing PAO surgery for hip DDH with i-FACTOR had three-times greater odds of partial/full bony union 6 weeks post-operatively compared to those without i-FACTOR. We showed that complication rates were low, and mostly similar between groups. However, the rate of LFCN neuropraxia and bone graft leakage was higher in the i-FACTOR group compared to the non-i-FACTOR group. There was no relationship between leaking of i-FACTOR bone graft and LFCN neuropraxia or HO. Importantly, these findings should be confirmed in a future prospective randomised clinical trial. In addition to bone healing and complications, outcomes that matter to patients such as pain, quality of life and return to sport should be evaluated over longer follow-up times.

Primary outcome: bony healing

We found that patients who had i-FACTOR during their PAO surgery were three-times more likely to have partial or full bony healing than those who did not have i-FACTOR. Previously, i-FACTOR has been shown to be safe, and result in similar outcomes for bone healing, pain, function and complications compared to local bone autograft 2 years following cervical fusion [14]. A RCT examining the effect of i-FACTOR compared to allograft in 100 elderly patients undergoing lumbar spinal fusion found that the bone healing rate was significantly higher in the i-FACTOR group (50% patients with healing) compared with 20% in the allograft group, measured at 1 year on CT scan [24]. Importantly however, there was no difference in patient-reported outcomes of pain or health-related quality of life. A previous large cohort study of patients undergoing PAO surgery for DDH, where no additional augmentation to improve bony healing was used, reported that only 45% (96 of 215) of the patients had complete healing of all osteotomy sites at the 6-month visit and 92% (225 of 245) demonstrated complete healing at 1 year [12]. While our findings suggest that the addition of a Peptide Enhanced Bone Graft might enhance bony healing, these need to be confirmed in a prospective RCT, and over longer follow-up times. We also need to understand whether i-FACTOR enhances outcomes that are important to young, active patients with DDH, such as pain, quality of life and return to sport following PAO surgery.

Secondary outcome: complications

Overall, we found that complication rates were low following PAO surgery for DDH in people aged 15–50 years in both groups. The exceptions included a high rate (approximately 50%) of heterotopic ossification in both groups, at the time of screw removal. This is favourable when compared to a study of 66 hips, where Wells et al. (2019) reported a higher incidence of heterotopic ossification (80%), and a higher rate (8%) of non-union requiring further open reduction and internal fixation surgery [5]. In our study in most cases, HO was seen at the time of screw removal just over the screw heads and was not present elsewhere. While this is classified as grade 2 according to the Brooker grading system, it could be considered a normal by-product of bone healing. It is unclear whether such heterotopic ossification is associated with worse outcomes such as pain, function and quality of life, and this should be explored in further studies. Importantly in our study, the rate of heterotopic ossification was not related to the use of i-FACTOR.

We also reported 26% of patients in the i-FACTOR group and 11% in the non-i-FACTOR group had transient symptoms associated with LFCN neuropraxia. Neuropraxia of the LFCN has been previously identified as a minor post-operative complication [4], while Ganz et al. (1988) describe relative frequency of LFCN dysesthesias in PAO surgery [13]. However, LFCN neuropraxia has also been associated with worse self-reported function and general health [28]. While we found that neuropraxia of the LFCN was higher in patients with i-FACTOR, it is not clear what the longer-term impacts of this finding is on patient-reported outcomes. In addition, our findings of LFCN neuropraxia compare favourably to those reported previously for PAO and total hip arthroplasty, where 60% of patients had LFCN neuropraxia at 3 months post-PAO [29], and as many as 80% of patients after total hip arthroplasty [30].

Leaking of bone graft was reported in half of all patients in the i-FACTOR group on the six-week follow-up X-rays, compared to 10% in the non-i-FACTOR group, but this was not related to HO or LFCN neuropraxia. It is possible that the addition of Gelfoam to the i-FACTOR graft increased the volume in a limited space and thus resulted in increased leakage. It is also unclear whether leakage impacted patients’ symptoms, given the relationship between hip imaging findings and symptoms is known to be poor [31, 32]. To date however, no research has investigated the mechanisms or impacts of leakage, and this needs further investigation in future studies.

Stress fractures of the inferior pubic ramus following PAO are a commonly reported complication following PAO [33], however we did not report any incidence of this in either group.

Clinical implications

Our findings have important implications for clinicians and patients, who need to consider the potential benefits of faster bony healing and thus earlier mobility, versus the potential for side effects. This should be considered during the shared decision-making process, especially for patients who might benefit from earlier mobility.

Strengths and limitations

This study has several strengths and limitations that should be acknowledged. This was the largest study to explore the effects of a Peptide Enhanced Bone Graft such as i-FACTOR on healing in pelvic surgery in young active adults. Data was de-identified, and the data extractors and analyst were blinded to group allocation. We conducted a reliability analysis with moderate inter-rater reliability between radiologists. Our study was prospectively registered on a public clinical trials registry.

However, we also acknowledge that the retrospective nature of the study meant that we were limited to extracting data that was collected as part of routine clinical care and could not collect patient-reported outcomes such as pain, quality of life and return to sport as these were not consistently measured. Factors that may affect bone healing (such as smoking history and severity of dysplasia) were not recorded for either group but should be investigated in future prospective studies. In addition, despite both groups following the same post-operative protocol, data on when precisely patients were cleared to fully weight bear was not recorded and should be collected in future studies.

While our sample size was strong, there was a disparity in the sample size for the two groups, where the numbers in the i-FACTOR group were lower than the non-i-FACTOR group sizes. We only measured bone healing at approximately 6 weeks. This time frame was chosen as some bony union would be expected by this time, patients are usually encouraged to begin weight-bearing at this time, and all patients routinely were X-rayed at this point post-operatively. It was not known whether the volume of i-FACTOR bone graft was greater than that of the demineralised bone graft, which may have contributed to increased leakage in the i-FACTOR group. Finally, it was not possible to blind the radiologists to the use of i-FACTOR, as it was visible on the X-rays. This may have created a differential misclassification bias, potentially inflating the effect estimate in favour of i-FACTOR in scoring the extent of bone healing.

CONCLUSIONS

Our retrospective cohort study showed that people aged 15–50 years undergoing PAO surgery for hip dysplasia with i-FACTOR had three-times greater odds of partial/full bony union six-weeks post-operatively than those without i-FACTOR. Complication rates were low, and similar between groups, apart from LFCN neuropraxia (26% of the I-FACTOR group versus 11% of the non-I-FACTOR group) and bone graft leakage (50% versus 10%). These findings need to be confirmed in a prospective randomised clinical trial, where outcomes that are important to patients (pain, quality of life and return to sport) are also evaluated.

DATA AVAILABILITY

Deidentified aggregated data will be provided on reasonable request.

ACKNOWLEDGEMENT

None declared.

FUNDING

This work was supported by Cerapedics.

CONFLICT OF INTEREST STATEMENT

None declared.

Author notes

†

Joint first authors.