Acute total hip arthroplasty for older patients with acetabular fractures: A meta-analysis

Abstract

Objective

Multiple treatment options for acetabular fractures in geriatric patients exist. However, no large-scale studies have reported the outcomes of acute total hip arthroplasty (THA) in this patient population. We systematically evaluated all available evidence to characterize clinical outcomes, complications, and revisions of acute THA for acetabular fractures in geriatric patients.

Methods

Meta-analysis of 21 studies of 430 acetabular fractures with mean follow-up of 44 months (range, 17−97 months). Two independent researchers searched and evaluated the databases of Ovid, Embase, and United States National Library of Medicine using a Boolean search string up to December 2019. Population demographics and complications, including presence of heterotopic ossification (HO), dislocation, infection, revision rate, neurological deficits, and venous thromboembolic event (VTE), were recorded and analyzed.

Results

Weighted mean Harris Hip Score was 83.3 points, and 20% of the patients had reported complications. The most common complication was HO, with a rate of 19.5%. Brooker grade III and IV HO rates were lower at 6.8%. Hip dislocation occurred at a rate of 6.1%, 4.1% of patients developed VTE, deep infection occurred in 3.8%, and neurological complications occurred in 1.9%. Although the revision rate was described in most studies, we were unable to perform a survival analysis because the time to each revision was described in only a few studies. The revision rate was 4.3%.

Conclusions

Acute THA is a viable option for treatment of acetabular fracture and can result in acceptable clinical outcomes and survivorship rates in older patients but with an associated complication rate of approximately 20%. Considering the limited treatment options, THA might be a viable alternative for appropriately selected patients.

1. Introduction

When managing acetabular fractures, open reduction and internal fixation (ORIF) is the treatment of choice for the younger population. However, ORIF is technically challenging in the elderly population in which it has a high risk of failure of fixation and incidence of posttraumatic osteoarthritis.^{1, 2, 3, 4} Moreover, the results of delayed total hip arthroplasty (THA) after initial ORIF often lead to suboptimal outcomes.^{5, 6, 7} Compared with historical results for ORIF alone or with delayed THA, ORIF with acute THA in elderly patients has shown better functional scores and lower complication rates.^{4,8, 9, 10, 11}

Although multiple studies have described treating acetabular fracture using primary THA, conclusions are difficult to draw because of the small numbers of patients and the variable methods of follow-up. In addition, no study has systematically evaluated the effectiveness and the complications of acute THA after acetabular fracture. Hence, we systematically evaluated all available evidence. The purposes of our study were to determine the overall demographic characteristics, characterize the complications associated with acute THA for acetabular fracture, and assess the revision rate for acute THA for acetabular fracture.

2. Methods

We performed a comprehensive literature search and evaluated all related published studies through December 2019. Using the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guideline,¹² we searched the databases of Ovid, Embase, and the United States National Library of Medicine using the search string ‘fract’ AND ‘acetab’ AND ‘arthropl’ OR ‘replace’, which returned 1930 studies. After excluding publications written in languages other than English, 1521 articles remained. After reviewing the abstracts of the 1521 studies, we identified 126 studies that examined the use of THA for the treatment of acetabular fracture. Cross-referencing the 126 studies yielded an additional 21 studies. Our inclusion criteria encompassed studies that evaluated the treatment of acetabular fractures with acute THA with a mean patient age of 60 years or older. Studies that examined THA as a delayed treatment of acetabular fracture, studies with a mean patient age younger than 60 years, case reports, systematic reviews, expert opinions, and follow-up reports of previously published studies were excluded. Each manuscript was assessed using the Methodological Index for Non-Randomized Studies (MINORS) criteria, which was modified to an all-or-nothing scale with which studies that adequately reported an index of the MINORS criteria received 1 point.¹³ Studies that did not report or inadequately reported a given criterion received 0 points. Studies with fewer than 6 points were excluded.¹⁴ Twenty-one studies were included in our meta-analysis.^{4,8,10,15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32} Two reviewers independently executed this method to prevent the exclusion of any relevant studies. Table 1 describes the indications for performing acute THA for each of the studies included in the analysis.

Table 1.

Indications for acute total hip arthroplasty.

Indication	References
Displaced acetabular fracture with intraarticular comminution and/or protrusion	4,8,10,16, 17, 18,20, 21, 22, 23,25, 26, 27, 28, 29, 30,32
Acetabular impaction involving weight bearing zone	4,8,10,15,17,20,21,24,28, 29, 30, 31, 32
Femoral head cartilage loss, impaction, or fracture	4,8,10,17,19,20,21,24,27
Preexisting, severe hip osteoarthritis or avascular necrosis	4,10,16,17,19, 20, 21,28
Concomitant displaced femoral neck fracture	4,10,15,20,21,28
Radiographic evidence of osteopenia/osteoporosis	4,8,10,17,19, 20, 21,24,25,29
Anterior column fracture with posterior column involvement (intact or partially intact) or associated both column	18,25,31

We reviewed each study included in the analysis for population demographics and clinical outcomes, including the mean age of the population, types of acetabular fractures based on the classification by Letournel and Judet,³³ types of implants used in addition to THA, and type of prosthetic used (cemented or cementless). The complications that were specifically assessed included the presence of heterotopic ossification (HO), dislocation, infection, revision rate, neurological deficits, and venous thromboembolic event (VTE). For the purpose of our analysis, deep venous thrombosis and pulmonary embolism were grouped and defined as VTE. HO was recorded by Brooker grade, but if no grade was provided, it was assumed the HO was clinically significant (Brooker grade III or IV).³⁴ Only Brooker grades III and IV were included in the overall complication rate calculation.

All obtained results were recorded on an electronic spreadsheet (Microsoft Excel; Microsoft Office, Redmond, WA). Certain variables were not provided in some studies, such as participant demographics, Harris Hip Score, type of fracture, and surgical details. Therefore, each specific analysis was performed individually and studies were excluded from an analysis if information from that data point was not provided. Using statistical software (MedCalc version 15.2; MedCalc Software, Ostend, Belgium), we performed descriptive statistics for demographic characteristics. Complication rates and revision rates were analyzed using a random effects model of proportions when heterogeneous and a fixed effect model when not heterogeneous.

3. Results

Twenty-one studies evaluating 430 acetabular fractures managed with acute THA were included in our final analysis (Fig. 1). Thirty-nine percent of the participants were women, and 61% were men. The mean patient age was 72 years (range of the means, 60−81 years). The mean follow-up time was 44 months (range of the means, 17−97 months) (Table 2). In terms of fixation before THA, all patients in the included studies underwent some type of fixation with screws, cables, wires, plates, rings, or cages (Appendix).

Fig. 1.

Flowchart of study selection.

Table 2.

Study details and demographics.

Study	THA (n)	Female (n)	Male (n)	Age (y)		Follow-up (mo)
				Avg	Range	Avg	Range
Mears and Velyvis, 200210	57	–	–	69	26–89	97	24–144
Mouhsine et al., 200223	11	–	–	79	65–93	24	–
Tidermark et al., 200326	10	3	7	73	57–87	38	11–84
Beaulé et al., 200415	10	5	5	61	50–85	34	24–53
Mouhsine et al., 200422	18	–	–	76	65–93	36	12–46
Boraiah et al., 20098	18	8	10	72	55–86	47	12–120
Carroll et al., 20104	9	–	–	67	56–89	63	24–188
Herscovici et al., 201019	22	10	12	75	60–95	29	13–67
Chémaly et al., 201317	20	–	–	60	28–89	31	12–79
Malhotra et al., 201321	15	2	13	65	57–69	82	62–122
Chakravarty et al., 201416	19	6	13	77	57–90	22	2–80
Enocson and Blomfeldt, 201418	15	7	8	76	63–84	48	–
Rickman et al., 201424	24	8	16	77	63–90	24	8–38
Lin et al., 201520	33	18	15	66	47–92	67	12–170
Solomon et al., 201525	11	11	4	81	76–87	–	12–24
Boelch et al., 201727	9	4	5	80	63–90	–	–
Salama et al., 201728	18	6	12	66	35–81	22	12–36
Weaver et al., 201829	37	18	19	79	66–90	22	6–89
Borg et al., 201930	13	5	8	77	64–89	24	–
Giunta et al., 201931	27	4	23	69	57–84	48	12–84
Lont et al., 201932	34	10	24	71	56–92	17	0–72

THA, total hip arthroplasty; Avg, average.

Twenty of the 21 studies reported the type and subtype of the acetabular fracture (n = 416). In accordance with classification by Letournel and Judet,³³ elementary fractures accounted for 43% (n = 179) of the cases and 57% (n = 237) were associated fractures. Posterior wall fracture was the most common type of elementary fracture (19%, n = 77), and anterior column or wall and posterior hemi-transverse were the most common types of associated fractures (16%, n = 66). The rates of subtypes of acetabular fractures are presented in Table 3. In terms of clinical outcomes, the weighted mean Harris Hip Score at latest follow-up was 83.3 points (range of the means, 70−93 points).^{8,10,19,26,28,29,31}

Table 3.

Types of acetabular fractures.

Study	Elementary						Associated
	n	PW	PC	AW	AC	Tr	n	PCPW	T-s	AWCPHT	ABC	TPW
Mears and Velyvis, 200210	34	11	3	0	10	10	23	6	3	6	4	4
Mouhsine et al., 200223	0	0	0	0	0	0	11	3	8	0	0	0
Tidermark et al., 200326	5	0	0	0	0	5	5	0	0	5	0	0
Beaulé et al., 200415	8	0	0	4	4	0	2	0	0	2	0	0
Mouhsine et al., 200422	2	0	0	0	0	2	16	2	9	1	0	4
Boraiah et al., 20098	16	15	0	0	0	1	2	0	0	1	1	0
Carroll et al., 20104	–	–	–	–	–	–	–	–	–	–	–	–
Herscovici et al., 201019	0	0	0	0	0	0	22	0	0	7	6	9
Chémaly et al., 201317	10	6	2	0	1	1	10	1	4	0	3	2
Malhotra et al., 201321	8	3	3	0	0	2	7	3	0	2	0	2
Chakravarty et al., 201416	4	0	0	0	1	3	15	0	1	5	5	4
Enocson and Blomfeldt, 201418	11	0	0	0	10	1	4	0	0	4	0	0
Rickman et al., 201424	13	1	0	0	4	8	11	3	0	2	6	0
Lin et al., 201520	15	13	0	2	0	0	18	4	0	5	2	7
Solomon et al., 201525	1	0	0	0	1	0	10	0	0	7	3	0
Boelch et al., 201727	5	0	0	2	1	2	4	0	0	0	4	0
Salama et al., 201728	12	9	0	0	0	3	6	1	2	1	1	1
Weaver et al., 201829a	15	7	0	0	8	0	17	0	3	6	4	4
Borg et al., 201930	1	1	0	0	0	0	12	1	0	5	5	1
Giunta et al., 201931	15	9	2	0	0	4	12	0	2	5	5	0
Lont et al., 201932	4	2	0	0	1	1	30	2	19	2	5	2
Rate (%)	43	19	2	2	10	10	57	6	12	16	13	10

PW, posterior wall; PC, posterior column; AW, anterior wall; AC, anterior column; Tr, Transverse; PCPW, posterior column and posterior wall; T-s, T-shaped; AWCPHT, anterior wall or column with posterior hemi-transverse; ABC, associated both column; TPW, transverse and posterior wall.

^aFive patients had fractures classified as “other,” and were not included in the analysis.

Various techniques were used in these 21 studies in conjunction with THA for acute management of acetabular fracture. Most of the included studies used plates and screws for internal fixation. Four studies used ring devices,^18,19,27,32 one used a specific cage system (Octopus; DePuy Synthes, Johnson & Johnson, Raynham, MA),²¹ and four others implemented a cup-cage device (Appendix).^25,26,29,30 The estimated weighted mean blood loss was 925 mL (range of the means, 533−1163 mL),^{15, 16, 17, 18, 19, 20, 21,24,26,27,30, 31, 32} and the surgical procedures had a weighted mean length of 176 min (range of the means, 110−244 min) (Table 4).^{4,15, 16, 17, 18, 19, 20, 21, 22,24,26, 27, 28,30, 31, 32}

Table 4.

Estimated blood loss and operative time.

Study	THA (n)	EBL (mL)		Op Time (min)
		Mean	Range	Mean	Range
Mears and Velyvis, 200210	57	–	–	–	–
Mouhsine et al., 200223	11	–	–	–	–
Tidermark et al., 200326	10	1100	700–1600	159	125–185
Beaulé et al., 200415	10	1060	100–3200	180	135–210
Mouhsine et al., 200422	18	–	–	165	120–180
Boraiah et al., 20098	18	–	–	–	–
Carroll et al., 20104	9	–	–	244	89–403
Herscovici et al., 201019	22	1163	300–4500	232	50–510
Chémaly et al., 201317	20	992	300–2000	171	105–315
Malhotra et al., 201321	15	835	450–1200	135	110–160
Chakravarty et al., 201416	19	700	220–1800	231	125–465
Enocson and Blomfeldt, 201418	15	665	250–1600	149	115–285
Rickman et al., 201424	24	1100	500–2500	193	130–280
Lin et al., 201520	33	852	200–5000	200	148–296
Solomon et al., 201525	11	–	–	–	–
Boelch et al., 201727	9	533	0–2000	189	136–266
Salama et al., 201728	18	–	–	122	65–180
Weaver et al., 201829	37	–	–	–	–
Borg et al., 201930	13	800	400–1700	188	175–321
Giunta et al., 201931	27	827	152–3019	110	65–170
Lont et al., 201932	34	1100	400–2700	169	97–310

THA, total hip arthroplasty; EBL, estimated blood loss; Op Time, operative time.

Using our random effects model, we determined that 20% of the patients had a reported complication (95% confidence interval [CI], 13.8%–27.6%) (Table 5 and Fig. 2). The most common complication was HO, with a rate of 19.5% (95% CI, 11.9%–28.5%) for any Brooker grade. The rate of clinically significant HO (Brooker grade III or IV) was lower at 6.8% (95% CI, 3.2%–11.7%). Hip dislocation occurred at a rate of 6.1% (95% CI, 4.0%–8.5%), 4.1% developed VTE (95% CI, 2.1%–6.8%), deep infection occurred in 3.8% (95% CI, 2.2%–5.9%), and neurological complications occurred in 1.9% (95% CI, 0.8%–3.4%) of patients who underwent acute THA for the management of acetabular fracture. The revision rate for the cohort was 4.3% (95% CI, 2.4%–6.8%).

Table 5.

Complications of acute total hip arthroplasty for acetabular fracture.

Complication Type	n	Rate (%)	95% Confidence Interval	References
HO (any grade)	386	19.5	11.9–28.5	8,10,15, 16, 17, 18, 19, 20, 21, 22, 23,26, 27, 28, 29, 30, 31, 32
HO (grade III or IV)	386	6.8	3.2–11.7	8,10,15, 16, 17, 18, 19, 20, 21, 22, 23,26, 27, 28, 29, 30, 31, 32
Dislocation	410	6.1	4.0–8.5	8,10,15, 16, 17, 18, 19, 20, 21, 22,24, 25, 26, 27, 28, 29, 30, 31, 32
VTE	410	4.1	2.1–6.8	8,10,15, 16, 17, 18, 19, 20, 21, 22,24, 25, 26, 27, 28, 29, 30, 31, 32
Infection	410	3.8	2.2–5.9	8,10,15, 16, 17, 18, 19, 20, 21, 22,24, 25, 26, 27, 28, 29, 30, 31, 32
Neurological deficit	410	1.9	0.8–3.4	8,10,15, 16, 17, 18, 19, 20, 21, 22,24, 25, 26, 27, 28, 29, 30, 31, 32
Revision	383	4.3	2.4–6.8	8,10,15, 16, 17, 18, 19, 20, 21, 22,24, 25, 26, 27, 28, 29, 30,32
Overalla	421	20.2	13.8–27.6	8,10,15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32

HO, heterotopic ossification; VTE, venous thromboembolic event, including deep vein thrombosis and pulmonary embolism.

^aOverall complication rate includes HO (grade III and IV), dislocation, VTA, infection, and neurologic deficit.

Fig. 2.

Forest plot of overall complications associated with acute THA for acetabular fractures.

4. Discussion

ORIF is well accepted as the treatment of choice for acetabular fracture in the younger population; however, no guidelines have been definitively established for the management of acetabular fracture in the elderly. Maintaining anatomic fracture reduction and avoiding posttraumatic osteoarthritis in this group of patients are challenging.^2,35 Acute THA combined with ORIF can be an option for properly selected patients and can allow for earlier rehabilitation and weight bearing, lower need for reoperation, and better functional scores compared with ORIF alone or delayed THA.¹⁵ Acute THA is a viable option and can be used by the treating surgeon in appropriately selected patients, including older patients with osteopenia and/or osteoporosis, comminuted acetabular fractures not amenable to ORIF, acetabular protrusion, marginal impaction, femoral head injuries, and femoral neck fractures (Table 1). To our knowledge, this study is the first meta-analysis of complications after acute THA for the treatment of acetabular fracture.

THA has been used in acute management of elementary and associated acetabular fractures. Various acetabular fracture subtypes are included in this study, with posterior wall, anterior column or wall, and posterior hemi-transverse being the most common fracture subtypes. These subtypes were also most common in a 16-year prospective epidemical study based in the United Kingdom.³⁶ The overall complication rate of 20% in our selected patient population was higher compared with all patients undergoing elective THA.³⁷ The complication rate associated with acute THA, however, is lower than that reported (38.8%) for older patients undergoing ORIF for acetabular fractures.³⁸

HO frequently occurs after acetabular fracture as a result of the substantial inflammation, extensive muscle dissection, blood loss, length of procedure, and possible coexisting head trauma.¹⁷ In the series presented herein, the most common complication was HO (19.5%). The rate is lower than what others have described for the acute operative management for acetabular fractures. In a 2005 meta-analysis, the incidence of HO was 25.6% for 613 cases of displaced acetabular fracture that received operative treatment within the first 4 weeks after trauma.³⁹ The rate of clinically significant HO (Brooker grade III or IV) was much lower in the present study, reported to be 6.8%. This indicates that most HO cases are not clinically relevant in patients undergoing acute THA for acetabular fractures.

The prevalence of dislocation after primary THA has been reported to be 2%–4% by various studies of various follow-up lengths.^{40, 41, 42} With the use of larger diameter femoral heads in recent years, dislocation rate in primary THA has plateaued at approximately 2% depending on the approach.⁴³ However, the dislocation rate in the cohort evaluated in this study was higher at 6.1%, likely due to the traumatic insult to the surrounding tissues that would otherwise provide stability of the THA.

The revision rate for this series was 4.3% with a mean follow-up time of 44 months. With variation in follow-up length, it is difficult to predict the long-term success of this procedure considering that most THA revisions occur after the first decade. In addition, the infection rates after primary THA for a diagnosis of osteoarthritis have been reported to be <1%,¹⁴ which is lower than our reported infection rate of 3.8% for acute THA after acetabular fracture. The difference in the rate of infection can be attributed to several factors, such as longer operative time required for stabilizing the fractures and lack of optimization of the patients before non-elective surgery. Reports have shown a high incidence of failure (50%) with cemented acetabular components in delayed THA after acetabular fractures,⁴⁴ but failure rates seem to be much lower (3%–19%) for cementless acetabular components.^45,46 The revision rate described herein is at the low end of the delayed cementless THA, but acute THA has the added benefit of avoiding a second surgical insult and the challenges associated with revision surgery after ORIF (subclinical infection, residual nonunion and pelvic deformity, bone loss, osteonecrosis, and retained instrumentation).⁴⁷

This study had several limitations. The 21 studies included for analysis were retrospective case series with heterogeneous designs, interventions, and follow-up periods. The internal fixation methods varied in individual studies as necessitated by the fracture types.⁵ Although this added heterogeneity to the cohort, the primary goal of internal fixation is to obtain sufficient reduction and fixation of the fracture to stabilize the cup when performing arthroplasty.¹¹ The studies were unified, however, by the use of acute THA for definitive treatment. Our analysis also focused on the overall and specific complications after THA, which were variably reported. Although the revision rate was described in most studies, we were unable to perform a survival analysis because the exact time to each revision was described in only a few studies. Furthermore, the studies varied by fracture type, surgical technique, and approach. We were unable to perform subgroup analyses for complication and revision rates based on these factors because of the lack of granular data inherent in meta-analyses. Future studies should examine the effects of these clinical and surgical factors on the complication and revision rates. The rates of complications and revisions could be affected by the variable follow-up, which should be addressed by future studies. Although this study was restricted to include articles written in the English language, the findings are representative of the current literature considering that most studies on acute THA for acetabular fractures are in English. In addition, our inclusion criteria allowed studies with patients younger than 60 years as long as the mean age was 60 years or older. Those studies were included to reduce heterogeneity and provide more meaningful results. Although these limitations are present, this is the largest study describing acute THA for the treatment of acetabular fractures in an elderly population.

In older patients with poor bone quality, the optimal management of acetabular fractures remains poorly defined. Acute THA for acetabular fractures can result in higher revision and complication rates compared with primary THA but has complication and revision rates similar to those associated with ORIF and delayed THA for acetabular fractures. However, because of the limited alternatives in this selected group of patients, THA might be the best alternative. Surgeons should be aware of this possibility, and patients and their families should be educated regarding all the possible risks of this procedure. Further prospective studies are needed to diminish the limitations and define the best management for acute acetabular fractures.

Funding

No external funding was received for this work.

Author contributions

Julio Jauregui: conceptualization and design of the study; data collection; data curation and statistical analysis; formal analysis; writing, original draft; writing, review and editing.

Tristan Weir: data collection; data curation and statistical analysis; formal analysis; writing, original draft; writing, review and editing.

Jin Chen: data collection; writing, original draft.

Aaron Johnson: conceptualization and design of the study; resources; writing, original draft.

Neil Sardesai: data collection; writing, original draft.

Aditya Maheshwari: formal analysis; resources; writing, original draft.

Theodore Manson: conceptualization and design of the study; formal analysis; project administration; resources; supervision; writing, original draft; writing, review and editing.

Declaration of competing interest

TM and his institution receive money from various law firms for providing expert testimony; TM’s institution received a grant from the Orthopaedic Research and Education Foundation; TM receives money from AO North America and the Maine Review Course as payment for lectures. No other relationships, conditions, or circumstances present potential conflicts of interest.

Appendix A. Surgical Techniques by Study

Study	Internal Fixation	Total Hip Arthroplasty Technique
		Femoral	Acetabular	Approach
Mouhsine et al., 200223	Cables	C	NC	NR
Mears and Velyvis, 200210	Screws and braided cables	C (70%), NC (30%)	NC	AL (67%), EL (5%), O (28%)
Tidermark et al., 200326	Cage	C	C	P (40%), AL (60%)
Mouhsine et al., 200422	Cables	C	NC	P (56%), O (44%)
Beaulé et al., 200415	Plates and screws	NR	NC	A (100%)
Boraiah et al., 20098	Plates and screws	C (older), NC (younger)	NC	P (100%)
Herscovici et al., 201019	Plates and screws, Ganz ring	C (91%), NC (9%)	C (50%), NC(50%)	A (14%), P (86%)
Carroll et al., 20104	NR	NR	NR	NR
Chémaly et al., 201317	Plates and screws	NC	NC	EP (100%)
Malhotra et al., 201321	Cage	NC	NC	P (100%)
Rickman et al., 201424	Plates and screws	C	NC (cup), C (liner)	NR
Enocson and Blomfeldt, 201418	Burch-Schneider ring	C	C	AL (87%), PL (13%)
Chakravarty et al., 201416	Percutaneous screws	NC (89%), C (11%)	NC	P (100%)
Solomon et al., 201525	Cage	NR	NC (cup), C (liner)	EP (100%)
Lin et al., 201520	Plates and screws	NC (79%), C (21%)	NC (97%), C (3%)	NR
Boelch et al., 201727	Burch-Schneider ring, plates, screws	NC (89%), C (11%)	C (100%)	P (100%)
Salama et al., 201728	Plates and screws	NC (100%)	NC (100%)	P (100%)
Weaver et al., 201829	Plates and screws, cage	NC (100%)	NC (100%)	A (NR%), P (NR%)
Borg et al., 201930	Cage	C (100%)	C (100%)	P (85%), AL (15%)
Giunta et al., 201931	Plates and screws	C (100%)	C (100%)	P (78%), A&P (22%)
Lont et al., 201932	Plates and screws, ring	NC (38%), C (62%)	C (100%)	P (100%)

C, Cemented; NC, Not cemented; NR, not reported; A, Anterior; P, Posterior; AL, Anterolateral; EL, Extended lateral; PL, Posterolateral; EP, Extended posterior; O, Other.