Skip to main content
NCBI home page
Orthopedic Reviews logoLink to Orthopedic Reviews
. 2024 Jun 27;16:120053. doi: 10.52965/001c.120053

Femorotibial angle on short knee radiographs fails to accurately predict the lower limb mechanical alignment. A systematic review and meta-analysis on different femorotibial angle definitions and short knee radiograph types

Giancarlo Giurazza 1,2, Giovanni Perricone 1,2, Edoardo Franceschetti 1,2, Stefano Campi 1,2, Pietro Gregori 1,2, Biagio Zampogna 1,2,3, Umberto Gabriele Cardile 1,2, Giuseppe Francesco Papalia 1,2, Rocco Papalia 1,2
PMCID: PMC11213696  PMID: 38947178

Abstract

Introduction

Hip-knee-ankle angle (HKA) on Full Limb Radiographs (FLRs) is the gold standard for coronal knee alignment assessment. Despite the widespread utilization of the more convenient femorotibial angle (FTA) on either antero-posterior (AP) or postero-anterior (PA) short knee radiographs (SKRs), its definition and correlation with HKA remains controversial. This review is the first to systematically investigate FTA-HKA correlation and the effect of different FTA methods and SKRs.

Methods

Systematic literature search (Pubmed, Scopus, Cochrane Library) followed PRISMA guidelines, to evaluate studies examining the FTA-HKA correlation. Meta-analyses compared the 3 most common FTA methods, knee center determination method and SKR types.

Results

17 studies (2597 patients, 3234 knees) were included. The strongest correlation with HKA (r = 0.78) was found for FTA Method 1 (angle formed by lines drawn from the midpoint of tibial spines to points 10 cm above and below the joint line). No significant differences were observed when grouping the FTA methods by knee center assessment (Group I, r = 0.78; Group II, r = 0.77). AP SKRs showed a trend towards stronger FTA-HKA correlation compared to PA SKRs, in both Method 1 (r = 0.79 vs 0.75) and Method 3 (r = 0.80 vs 0.66).

Conclusion

Irrespective of its definition or type of SKR used, FTA lacks reliable accuracy in predicting the HKA in most knees. FLRs should be used whenever precise estimation of the patient’s alignment is necessary. Caution is warranted in interpreting studies investigating knee alignment or knee arthroplasty outcomes based on FTA.

Keywords: Femorotibial angle (FTA), Hip-Knee-Ankle angle (HKA), FTA-HKA correlation, Short Knee Radiograph, Full limb radiograph

1. Introduction

The emergence of patient-specific alignment techniques in Total Knee Arthroplasty (TKA) has renewed the longstanding interest of the orthopedic community on knee coronal alignment1–3 and has emphasized the importance of understanding and classifying its variations.4–6

Despite various methods for assessing coronal knee alignment, the gold standard remains the measurement of the mechanical axis using the hip-knee-ankle angle (HKA) from a weight-bearing full limb radiograph (FLR).7,8 Unlike standard short knee radiographs (SKRs), which capture approximately 10 cm of the distal femur and proximal tibia, FLRs provide a comprehensive view of the entire lower limb, aiding in the assessment of anatomical variations in the proximal femur and distal tibia. However, FLRs have drawbacks such as increased radiation exposure, specialized equipment requirements, and higher costs.

Consequently, many clinical studies have used the femorotibial angle (FTA) measured on SKRs to evaluate alignment, delineate “safety” boundaries and risk factors for failure in knee arthroplasty9 and still today many surgeons rely on SKR in their clinical practice, even for the analysis of post-operative outcomes.

The FTA definition varies greatly among different research papers, leading to various methods proposed for its measurement, both on fully-extended antero-posterior (AP) and semi-flexed postero-anterior (PA) SKRs. As a result, the FTA-HKA correlation is still a subject of ongoing debate, with a broad spectrum of values reported in the literature.10–13 To our knowledge, a systematic review on the subject has never been performed.

This study aimed to conduct a thorough review of existing literature to provide a comprehensive understanding of the relationship between FTA and HKA. Additionally, it aimed to perform a meta-analysis to quantify the FTA-HKA correlation comparing 1) the most commonly used methods for measuring FTA, 2) the method used to determine the knee’s center, and 3) AP vs PA SKRs. The hypothesis was that, irrespective of its definition and type of SKRs, FTA would not allow to predict the HKA in a substantial subset of patients.

2. Methods

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines have been utilized to carry out this systematic review.14

2.1. Eligibility Criteria

Only studies written in English and employing prospective or retrospective comparative designs were considered. All included studies examined the correlation between HKA measured on FLR and FTA measured on standard SKR. Exclusion criteria were applied to studies utilizing knee radiographs capturing more than 10 cm of the tibia and femur to assess the FTA.

2.2. Information Sources and Search

An electronic systematic search of PubMed (MEDLINE), Scopus and Cochrane databases was carried out by 2 reviewers, G.P and G.G., to identify eligible studies. The search was executed on November 12, 2023. The following search string was adopted: “hip-knee-ankle”[All Fields] AND (“angle”[All Fields] OR “angled”[All Fields] OR “angles”[All Fields]) AND (“diagnostic imaging”[MeSH Subheading] OR (“diagnostic”[All Fields] AND “imaging”[All Fields]) OR “diagnostic imaging”[All Fields] OR “x ray”[All Fields] OR “x rays”[MeSH Terms] OR “x rays”[All Fields]).

Then, a manual search of the bibliography of each published study was performed, to find relevant articles that could potentially have been missed. The remaining articles were analyzed by 2 reviewers, to exclude studies not fulfilling the eligibility criteria. Studies included were categorized by study type, according to the Oxford Centre for Evidence-Based Medicine.

2.3. Selection Process

For each study, the following data were extracted: epidemiological characteristics of participants (age, sex, type of knee, grade of OA, lower limb alignment), type of SKR (standing AP, semiflexed PA, FLR cropped to the dimension of SKR), FTA measurement method and FTA-HKA correlation.

The statistical analysis was conducted using R software, version 4.3.2. Meta-analyses were performed to pool the correlations between the most common FTA measurements and HKA and to establish potential differences depending on the method used to determine the knee’s center and the type of SKR used (AP vs PA). The results were reported as correlation coefficients and 95% confidence intervals (95% Cl). A random-effects model was used for I2 > 70%. Correlation coefficients were categorized as follows15,16: perfect (r = 1), excellent (0.9 ≤ r ≤ 0.99), very good/very strong (0.8 ≤ r ≤ 0.89), good (0.6 ≤ r ≤ 0.79), fair/moderate (0.4 ≤ r ≤ 0.59), weak (0.20 ≤ r ≤ 0.39), and neglegible (r <0.20).

2.4. Quality of the Studies

The quality of included studies was assessed using the MINORS (Methodological Index for Nonrandomized Studies) score.17

3. Results

3.2. Study selection and characteristics

The literature search initially identified 712 papers, which were reduced to 589 after removing duplicates. Following the screening of titles and abstracts, 555 studies were excluded. Subsequently, the full text of 34 manuscripts was read, resulting in 17 studies that met the eligibility criteria for the systematic review. The PRISMA 2020 flow diagram outlining this process is depicted in Figure 1.

Figure 1. PRISMA 2020 flow diagram.

Figure 1.

Open in a new tab

Details of the included studies are summarized in Table 1, encompassing a total of 2597 patients and 3234 knees. Of the 17 studies, 3 were prospective while the remaining 14 were retrospective. The mean age of participants was 60.4 years (range: 14 to 81 years). Additionally, the average MINORS score across studies was 10.2/16 (Table 1).

Table 1. Study details.

Study Year of publication Study type Level of evidence N. of patients (knees) Sex (male:
female) 		Mean age (years SD) MINORS
Chang et al 2010 RS II 201 (201) 104:97 39.5 (±11.5) no OA group
69.5 (±5.8) OA and TKA group		 13/16
Colebatch et al 2009 RS III 40 (80) 0:40 53 (±8) 11/16
Felson et al 2009 RS III OAI:
143 (281)
BOKS:
183 (303)		 OAI:
48.9%:51.1%
BOKS:
56.8%:43.2%		 OAI:
60.5 (±9.7)
BOKS:
65.9 (±9.3)		 8/16
Gielis et al 2020 RS II 100 (110) 53:57 54 (±7.4) 11/16
Hinman et al 2006 RS III 40 (40) 16:24 64.7 (±9.4) 10/16
Iranpour-Boroujeni et al 2014 RS III 142 (266) 70:72 61.5 (±9.8) 8/16
Issa et al 2007 PS II 146 (146) 37:109 70 (±11) 11/16
Kraus et al 2005 RS III 57 (57) 17:40 65.7 (±11.9) female
68.8 (±8.1) male		 10/16
McDaniel et al 2010 RS III 50 (50) 7:43 62 (range 40–79) 8/16
Navali et al 2012 PS III 50 (100) 19:31 34.7 (±15.2) 11/16
Park et al 2015 RS III 262 (262) NR NR 8/16
Shang et al 2022 PS III 223 (223) NR NR 12/16
Sheehy et al 2011 RS III 120 (120) 47:73 63 (±8.4) 8/16
Skytta et al 2018 RS II 83 (103) 33:50 69.3 (range 49-81) 10/16
Stickley et al 2016 RS II 653 (788) 207:446 69.5 12/16
Van Raaij et al 2009 RS II 68 (68) 32:36 55 11/16
Zampogna et al 2015 RS III 36 (36) NR NR 11/16
Open in a new tab

KL = Kellgren-Lawrence

NR = Not Reported

OA = Osteoarthritis

PS = Prospective study

RS = Retrospective study

SD = Standard Deviation

TKA = Total Knee Arthroplasty

THA = Total Hip Arthroplasty

Demographic characteristics, FTA measurement methods and types of knee radiographs examined are reported in Table 2. The FTA-HKA correlation was analyzed on AP SKRs in 8 studies, on PA semi-flexed SKRs in 4 studies and on both in 1 study. Finally, 4 studies analyzed the correlation on AP FLRs cropped to the dimensions of a SKR. The alignment category was reported in 9/17 studies, comprising a total of of 816 varus, 350 valgus and 400 neutral knees.

Table 2. Types of knee radiographs, demographic characteristics, FTA methods and FTA-HKA correlations.

Study Knee Radiograph FTA Method Type of knee Lower Limb Alignment (knees) FTA-HKA
Correlation
Chang et al AP SKR Method 3 99 low grade OA, 102 end stage OA Varus: NR
Valgus: NR
Neutral: NR		 Control group:
M 0.678
F 0.664 e
OA group:
M 0.692
F 0.800
TKA group:
M 0.429
F 0.273
Colebatch et al AP SKR Method 1 KL 0-1 Varus: 47 (58.8%)
Valgus: 10 (12.5%)
Neutral: 23 (28.8%)		 0.81
Felson et al PA SKR Method 3 OAI:
KL 0: 29 (10.3%)
KL 1: 51 (18.2%)
KL 2: 96 (34.2%)
KL 3: 87 (31%)
KL 4: 0 (0%)
BOKS:
KL 0: 39 (13%)
KL 1: 71 (23.4%)
KL 2: 84 (27.6%)
KL 3: 99 (32.8%)
KL 4: 10 (3.2%)		 OAI:
Varus: 121
Valgus: 61
Neutral: 99
BOKS:
Varus: 174
Valgus: 60
Neutral: 69		 OAI:
0.66
BOKS:
0.68
Gielis et al AP SKR Method 1:
Fem3Tib3, Fem4Tib4
Method 2:
Fem2Tib2
Method 3:
Fem2Tib3, Fem2Tib4
Method 4:
Fem2Tib1
Method 5:
Fem1Tib1
Method 6:
Fem1Tib3, Fem1Tib4		 KL 0: 9 (8.2%)
KL 1: 23 (20.9%)
KL 2: 30 (27.3%)
KL 3: 36 (32.7%)
KL 4: 12 (10.9%)		 NR Fem3Tib3: 0.84
Fem4Tib4: 0.83
Fem2Tib2: 0.87
Fem2Tib3: 0.89
Fem2Tib4: 0.89
Fem2Tib1: 0.90
Fem1Tib1: 0.88
Fem1Tib3: 0.86
Fem1Tib4: 0.86
Hinman et al AP FLR (cropped to size of SKR) Method 1 KL 1: 3 (7.5%)
Kl 2: 10 (25%)
KL 3: 11 (27.5%)
KL 4: 16 (40%)		 Varus: 36 (90%)
Valgus: 3 (8%)
Neutral: 1 (2%)		 0.88
Iranpour-Boroujeni et al PA SKR Method 1:
Existing FTA Method
Method 8:
New FTA Method		 KL 0: 29 (10.9%)
KL 1: 39 (14.7%)
KL 2: 91 (34.2%)
KL 3: 87 (32.7%)
KL 4: 20 (7.5%)		 M:
Varus 75 (60%)
Valgus 13 (10%)
Neutral 38 (30%)
F:
Varus 38 (27%)
Valgus 60 (42%)
Neutral 42 (30%)		 Existing FTA Method: 0.68
New FTA Method: 0.72
Issa et al PA SKR Method 1 KL 0: 4 (2.7%)
KL 1: 18 (12.3%)
KL 2: 68 (46.6%)
KL 3: 46 (31.5%)
KL 4: 10 (6.8%)		 NR 0.86
Kraus et al AP SKR
PA SKR		 Method 1 KL 0: 3 (2.6%)
KL 1: 25 (21.9%)
KL 2: 30 (26.3%)
KL 3: 42 (36.9%)
KL 4: 14 (12.3%)		 M:
Varus 19 (55.9%)
Valgus 5 (14.7%)
Neutral 10 (29.4%)
F:
Varus 40 (50%)
Valgus 31 (38.7%)
Neutral 9 (11.3%)		 AP: 0.65
PA: 0.75
McDaniel et al PA SKR Method 1:
Method B, Method C, Method D
Method 2:
Method A
Method 3:
Method E		 KL 1, 2, 3 NR Method A: 0.50
Method B: 0.65 Method C: 0.55
Method D: 0.64
Method E: 0.59
Navali et al AP FLR (cropped to size of SKR) Method 1 NR Varus: NR
Valgus: NR
Neutral: NR
mean HKA 175.9 ± 11.3		 0.93
Park et al AP SKR Method 7 Pre-operative (KL NR)
Post-operative (TKA)		 Pre-operative:
- varus: NA
- valgus: NA
- neutral: 26/187
Post-operative:
- varus: NA
- valgus: NA
- neutral: 116/206		 Pre-operative: 0.90
Post-operative: 0.69
Shang et al AP SKR Method 7 Post-operative TKA Varus: 126 (56.5%)
Valgus: 14 (6.3%)
Neutral: 83 (37.2%)		 0.86
Sheehy et al AP FLR (cropped to size of SKR) Method 3 KL ≥2:
Complete dataset:
51%
Group 1:
70%
Group 2:
33%
Group 3:
20%
Group 4:
80%		 Group 1 (30): HKA ≥ 5.0° varus
Group 2 (30): HKA 0.0-4.9° varus
Group 3 (30): HKA 0.1-4.9° valgus
Group 4 (30): HKA ≥ 5.0° valgus		 Complete dataset: 0.88
Group 1:
0.41
Group 2:
0.45
Group 3:
0.50
Group 4:
0.66
Skytta et al AP SKR Method 5 Post-operative TKA Varus: 44 (43%)
Valgus: 33 (32%)
Neutral: 26 (25%)		 0.54
Stickley et al AP SKR Method 3 KL NR Varus: NR
Valgus: NR
Neutral: NR
Mean HKA: 174.7 ± 7.7°		 0.88
Van Raaij et al AP FLR (cropped to size of SKR) Method 3
FTb
Method 5:
FTa		 Medial compartment OA
KL 1: 27 (40%)
KL 2: 12 (18%)
KL 3: 28 (41%)
KL 4: 7 (1%)		 Varus: NR
Valgus: NR
Neutral: NR
Mean HKA: 173.1 ±3.7°		 FTa: 0.65
FTb: 0.34
Zampogna et al AP SKR Method 1: FTa1
Method 5: FTa2		 Pre-operative Medial compartment OA
(KL NR)
Post-operative (HTO)		 Pre-operative:
Varus 36
Valgus 0
Neutral 0
Mean HKA: 172.9°
Post-operative:
NR
Mean HKA: 181.3°		 FTa1
Pre-operative:
0.26
Post-operative:
0.53
FTa2
Pre-operative:
0.71
Post-operative:
0.79
Open in a new tab

AP = Antero-posterior

FTA = Femorotibial Angle

HTO = High Tibial Osteotomy

KL = Kellgren-Lawrence

NR = Not Reported

OA = Osteoarthritis

PA = Postero-anterior

SKR = Short Knee Radiograph

TKA = Total Knee Arthroplasty

3.2. META-analysis

In the 17 studies examined, 11 distinct methods of measuring the FTA were identified, as outlined in Table 2. For the purpose of result analysis, all methods that differed in the proximal point of the tibia (center of the base of the spines, center of the tip of the spines, and center of the tibial plateau) were grouped together, considering the differences negligible, thus resulting in a total of 8 methods (Table 2 and 3). Only preoperative data were included in the meta-analysis for studies that provided both preoperative and postoperative data.

Table 3. FTA methods and definitions.

FTA METHOD FTA DEFINITION STUDIES
Method 1 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the tibial spines and (b) the center of the tibial spines, and a tibial axis drawn between (a’) the center of the tibial spines and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines

  1. Hinman et al

  2. Zampogna et al

  3. Colebatch et al

  4. Navali et al

  5. Kraus et al

  6. Issa et al

  7. Gielis et al

  8. Iranpour-Boroujeni et al

  9. McDaniel et al
Method 2 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the center of the femoral notch, and a tibial axis drawn between (a’) the center of the femoral notch and (b’) the tibial mid-shaft 10 cm distal to the center of the femoral notch

  1. Gielis et al

  2. McDaniel et al
Method 3 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the center of the femoral notch, and a tibial axis drawn between (a’) the center of the tibial spines and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines

  1. Stickley et al

  2. Sheehy et al

  3. Van Raaij et al

  4. McDaniel et al

  5. Gielis et al

  6. Felson et al (database OAI)

  7. Felson et al (database BOKS)

  8. Chang et al
Method 4 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the center of the femoral notch, and a tibial axis drawn between (a’) the junction of the tibial meta-epiphysis and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines Gielis et al
Method 5 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the junction of the femoral meta-epiphysis, and a tibial axis drawn between (a’) the junction of the tibial meta-epiphysis and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines

  1. Zampogna et al

  2. Van Raaij et al

  3. Gielis et al

  4. Skytta et al
Method 6 Angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the junction of the femoral meta-epiphysis, and a tibial axis drawn between (a’) the center of the tibial spines and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines Gielis et al
Method 7 Angle formed by the intersection of a femoral axis drawn between (a) the most proximal part of the femoral mid-shaft included on the film and (b) the femoral mid-shaft 10 cm proximal to the joint line, and a tibial axis drawn between (a’) the tibial mid-shaft 10 cm distal to the joint line and (b’) the most distal part of the tibial mid-shaft included on the film

  1. Shang et al

  2. Park et al
Method 8 Angle formed by the intersection of a femoral axis represented by the perpendicular to a line tangent to the base of the femoral condyles and centered between the outer margins of the medial and lateral femoral epicondyles, and a tibial axis passing through the tibial mid-shaft between 1 and 10 cm below the tibial joint line Iranpour-Boroujeni et al
Open in a new tab

The 3 most commonly used FTA methods were Method 1, Method 3 and Method 5, evaluated in 9, 8 and 4 studies, respectively (Figure 2).

Figure 2. FTA-HKA correlation of the 3 most commonly used FTA methods.

Figure 2.

Open in a new tab

In Method 1 the FTA was defined as the angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the tibial spines and (b) the center of the tibial spines, and a tibial axis drawn between (a’) the center of the tibial spines and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines.

In Method 3 the FTA was defined as the angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the center of the femoral notch, and a tibial axis drawn between (a’) the center of the tibial spines and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines

In Method 5 the FTA was defined as the angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the center of the femoral notch and (b) the junction of the femoral meta-epiphysis, and a tibial axis drawn between (a’) the junction of the tibial meta-epiphysis and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines

The FTA-HKA correlation coefficients were: 0.78 (95% CI: 0.66 to 0.86; I2 = 91%) for Method 1, 0.75 (95% CI: 0.61 to 0.84; I2 = 96%) for Method 3, and 0.72 (95% CI: 0.52 to 0.85; I2 = 91%) for Method 5.

The various methods were further classified into two groups based on their approach: those relying on a single reference point (Group I) and those utilizing two reference points—one for the tibia and one for the femur (Group II)—to establish the center of the knee (Figure 3). Group I, evaluated in 9 studies, showed a correlation coefficient of 0.78 (95% CI: 0.66 to 0.86; I2 = 91%), while, Group II, evaluated in 13 studies, showed a correlation coefficient of 0.77 (95% CI: 0.68 to 0.83; I2 = 95%).

Figure 3. FTA-HKA correlation based on the method used to determine the knee’s center: Group I, one reference point; Group II, one reference point for the proximal tibia e one for the distal femur.

Figure 3.

Open in a new tab

Finally, a trend towards stronger FTA-HKA correlation was found when using AP vs PA SKRs in both Method 1 and Method 3 (Figure 4). In Method 1, 6 studies used AP SKRs with a correlation coefficient of 0.79 (95% CI: 0.60 to 0.90; I2 = 93%), and 4 studies used PA SKRs with a correlation coefficient of 0.75 (95% CI: 0.62 to 0.84; I2 = 87%). Similarly, in Method 3, 5 studies used AP SKRs with a correlation coefficient of 0.80 (95% CI: 0.60 to 0.90; I2 = 96%), and 3 studies used PA SKRs with a correlation coefficient of 0.66 (95% CI: 0.62 to 0.71; I2 = 0%).

Figure 4. FTA-HKA correlation when using AP vs PA SKRs in Method 1 and Method 3.

Figure 4.

Open in a new tab

4. Discussion

The main finding of the current study, gathereing results from 17 articles involving a total of 2597 patients and 3234 knees, was that, irrespective of its definition and type of short knee radiographs, the femorotibial angle fails to predict the hip-knee-ankle angle in a significant subset of patients. To the best of our knowledge, this represents the first systematic review and meta-analysis on this topic.

Although the FTA11,12,18–23 is commonly defined as the angle formed by the intersection of two lines drawn from the midpoint of the tibial spines to a point situated 10 cm above and below the joint line and midway between the cortical bone surfaces (Method 1), alternative definitions have been explored in the literature.13,21–30 Notably, the chosen FTA definition seems to significantly impact its ability to predict the HKA, as observed by Gielis et al.21 In their study, they compared the FTA-HKA correlations with 6 of the 8 methods considered in the current systematic review and found the best correlation (r = 0.93) for Method 4, which defines FTA as the angle formed by the intersection of a femoral axis drawn between (a) the femoral mid-shaft 10 cm proximal to the femoral notch and (b) the center of the femoral notch, and a tibial axis drawn between (a’) the junction of the tibial meta-epiphysis and (b’) the tibial mid-shaft 10 cm distal to the center of the tibial spines.

Therefore, a meta-analysis was conducted on the three most commonly referenced methods to measure the FTA. Our results revealed Method 1 as the most valid, exhibiting a correlation coefficient of 0.78. Additionally, similar correlation coefficients were found for Methods 3 and 5, with values of 0.75 and 0.72, respectively. Notably, even within a specific FTA method, the FTA-HKA correlation can vary greatly, with values spanning from 0.2618 to as high as 0.93.19

An essential aspect in understanding the FTA-HKA correlation among various FTA measurement methods is the choice of reference point used to define the knee’s center. In advanced knee osteoarthritis, some degree of femur subluxation on the tibia is often observed. Therefore, it might seem logical to anticipate that methods utilizing a single reference point would be less reliable compared to those employing two separate reference points—one for the tibia and one for the femur. However, our meta-analysis did not support this assumption, as it revealed very similar results between Group I and Group II.

Several studies have investigated the FTA-HKA correlation using semi-flexed PA SKRs,10,20,22,23,30 which reflect the knee’s loading position, instead of fully-extended AP SKRs. The only study directly comparing the FTA-HKA correlation using both types of SKRs was conducted by Kraus et al,20 revealing correlation coefficients of 0.65 and 0.75 when employing AP and PA SKRs, respectively. In contrast, our meta-analysis yielded a stronger mean correlation coefficient for AP SKRs compared to PA SKRs, with values of 0.79 vs 0.75 and 0.80 vs 0.66 for Method I and III, respectively. These results suggest that AP SKRs may be a preferable option for assessing the FTA-HKA correlation.

It’s crucial to acknowledge substantial discrepancies across the included studies of the current systematic review. The reported differences between FTA and the HKA vary widely, ranging from less than 111,12 to more than 7°.10,13,24,25 The heterogeneity of these results may stem from factors other than variations in FTA measurement methods and type of SKR used, such as presence, direction and degree of femoral/tibial bowing or other extra-articular deformities,31 direction and degree of lower limb malalignment,25 sex,10,13,20 pre- vs post-operative state,18,28 and sample size across different studies.

Significant femoral and/or tibial bowing, present in a varying percentage of the population,32,33 can significantly impact the predictive accuracy of FTA for HKA, as highlighted by Oh et al.31 Their study revealed a twofold increase in the rate of misalignment categorization in knees with bowing compared to those without (26.2 vs. 13.1%). Incorporating a longer shaft length in the SKR could potentially capture the presence of extra-articular deformity, offering an explanation for the heightened FTA-HKA correlation observed by Chang et al13 when utilizing 15 cm shaft length, compared to the standard 10 cm. Similar results were also reported by Park et al28 and Sheehy et al.25

Additionally, research suggests that the FTA-HKA correlation may vary depending on the type of alignment and the severity of deformity, as demonstrated by Sheehy et al25 In their study, they found that using a shorter shaft length for assessing the FTA resulted in an increase in the FTA-HKA offset in varus limbs and a decrease in valgus limbs. Furthermore, inconsistent findings regarding offsets between HKA and FTA have also been reported between males and females in previous studies.10,13,20 While some authors10,20 reported a smaller offset in the female population, ranging from 3 to 3.5°, Chang et al13 reported opposite trends.

Various efforts have been undertaken to derive the HKA from the FTA through predictive equations.11,13,20 However, their reliability and external validity appear to be limited, likely as a consequence of the small sample sizes, as pointed out by Stickley et al.26 Indeed, their findings indicate that, when applying the gender-specific equation from Chang et al13 and the generalized equation from Kraus et al20 to 788 consecutive cases, they accurately predicted HKA within ±3° in only 61% of males and 63% of females, and 36% of males and 44% of females, respectively.

Finally, the predictive ability of the FTA for determining the HKA also appears significantly influenced by whether pre-operative or post-operative images are analyzed. In the TKA population, Park et al28 observed a more than twofold increase in the misclassification rate of patient alignment when using post-operative compared to pre-operative images (33.0 vs 13.9%), coupled with a lower FTA-HKA correlation (0.69 vs 0.90). Conversely, in a high tibial osteotomy population, Zampogna et al18 found that the FTA-HKA correlation was higher for post-operative compared to pre-operative images for both methods considered (FTa1: 0.53 vs. 0.26; FTa2: 0.79 vs. 0.71).

Our findings support the hypothesis that FTA lacks reliable accuracy in predicting the HKA in most knees. Moreover, inconsistent boundaries have been reported in the literature for categorizing alignment as neutral, varus or valgus based on the FTA values,9,34,35 thus limiting its utility also for broader analytical purposes, as shown by Shang et al.29 Therefore, we advocate for the use of FLRs whenever precise estimation of the patient’s alignment is necessary. Additionally, caution is advised in interpreting studies investigating knee alignment or outcomes after knee arthroplasty procedures based on FTA.

The findings of the present study need to be considered in light of its potential limitations. The considerable heterogeneity of the populations among the studies in terms of age, sex, type of alignment and presence vs absence of knee osteoarthritis, underscores the need for a cautious interpretation of the results. Additionally, a high level of heterogeneity was observed within the results of the meta-analysis, with the I2 exceeding 85% in nearly all reported correlations. Finally, the limited number and low evidence level of included studies, mostly categorized as level IV, as well as variations in FTA-HKA offsets, predictive equations, and FTA definitions, pose challenges in comparing results and may introduce bias in the absence of subgroup analysis.

5. Conclusions

Irrespective of the its definition or type of SKR used, FTA lacks reliable accuracy in predicting the HKA in most knees. The variability in offset between HKA and FTA is unpredictable and highly influenced by inter-individual factors. Full-length radiographs should be used whenever precise estimation of the patient’s alignment is necessary. Additionally, caution is advised in interpreting studies investigating knee alignment or outcomes after knee arthroplasty based on FTA.

Corresponding author

  • Name: Pietro Gregori

  • Institutional address: Research Unit of Orthopaedic and Trauma Surgery, Department of Medicine and Surgery, Università Campus Bio-Medico di Roma, Roma, Italy

  • Phone: +393400951122

  • Email: Pietro.gregori@unicampus.it

Authors’ contribution

RP had the idea for the article. GG was responsible for data acquisition, data analysis, and the writing of the manuscript. GP was responsible for data acquisition. GFP was responsible for data analysis. UGC and BZ were responsible for grouping the FTA methods with uniform definitions. EF, SC, and PG were responsible for reviewing and critically revise the manuscript. All authors have given final approval of the version to be published.

Conflict of interest

The authors declare that they have not conflict of interest.

Funding Statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References

  1. The influence of alignment on risk of knee osteoarthritis progression according to baseline stage of disease. Cerejo R., Dunlop D. D., Cahue S., Channin D., Song J., Sharma L. 2002Arthritis Rheum. 46(10):2632–6. doi: 10.1002/art.10530. [DOI] [PubMed] [Google Scholar]
  2. Alignment options for total knee arthroplasty: A systematic review. Rivière C., Iranpour F., Auvinet E., et al. 2017Orthop Traumatol Surg Res. 103(7):1047–1056. doi: 10.1016/j.otsr.2017.07.010. [DOI] [PubMed] [Google Scholar]
  3. Neutral mechanical alignment: a requirement for successful TKA: affirms. Lombardi A.V., Berend K.R., Ng V.Y. 2011Orthopedics. 34(9):e504–506. doi: 10.3928/01477447-20110714-40. [DOI] [PubMed] [Google Scholar]
  4. Functional knee phenotypes: a novel classification for phenotyping the coronal lower limb alignment based on the native alignment in young non-osteoarthritic patients. Hirschmann M. T., Moser L. B., Amsler F., Behrend H., Leclerq V., Hess S. 2019Knee Surg Sports Traumatol Arthrosc. 27(5):1394–1402. doi: 10.1007/s00167-019-05509-z. [DOI] [PubMed] [Google Scholar]
  5. Coronal Plane Alignment of the Knee (CPAK) classification. MacDessi S. J., Griffiths-Jones W., Harris I. A., Bellemans J., Chen D. B. 2021Bone Joint J. 103-B(2):329–37. doi: 10.1302/0301-620X.103B2.BJJ-2020-1050.R1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. The Chitranjan Ranawat award: is neutral mechanical alignment normal for all patients? The concept of constitutional varus. Bellemans J., Colyn W., Vandenneucker H., Victor J. 2012Clin Orthop Relat Res. 470(1):45–53. doi: 10.1007/s11999-011-1936-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Radiographic analysis of the axial alignment of the lower extremity. Moreland J. R., Bassett L. W., Hanker G. J. 1987J Bone Joint Surg Am. 69(5):745–9. doi: 10.2106/00004623-198769050-00016. [DOI] [PubMed] [Google Scholar]
  8. Axial lower-limb alignment: comparison of knee geometry in normal volunteers and osteoarthritis patients. Cooke D., Scudamore A., Li J., Wyss U., Bryant T., Costigan P. 1997Osteoarthritis Cartilage. 5(1):39–47. doi: 10.1016/S1063-4584(97)80030-1. [DOI] [PubMed] [Google Scholar]
  9. The effect of alignment and BMI on failure of total knee replacement. Ritter M. A., Davis K. E., Meding J. B., Pierson J. L., Berend M. E., Malinzak R. A. 2011J Bone Joint Surg Am. 93(17):1588–1596. doi: 10.2106/JBJS.J.00772. [DOI] [PubMed] [Google Scholar]
  10. Full-limb and knee radiography assessments of varus-valgus alignment and their relationship to osteoarthritis disease features by magnetic resonance imaging. Issa S. N., Dunlop D., Chang A.., et al. 2007Arthritis Rheum. 57(3):398–406. doi: 10.1002/art.22618. [DOI] [PubMed] [Google Scholar]
  11. Is there an alternative to the full-leg radiograph for determining knee joint alignment in osteoarthritis? Hinman R. S., May R. L., Crossley K. M. 2006Arthritis Rheum. 55(2):306–313. doi: 10.1002/art.21836. [DOI] [PubMed] [Google Scholar]
  12. Effective measurement of knee alignment using AP knee radiographs. Colebatch A. N., Hart D. J., Zhai G., Williams F. M., Spector T. D., Arden N. K. 2009Knee. 16(1):42–45. doi: 10.1016/j.knee.2008.07.007. [DOI] [PubMed] [Google Scholar]
  13. What should be considered in using standard knee radiographs to estimate mechanical alignment of the knee? Chang C. B., Choi J.-Y., Koh I. J., Seo E. S., Seong S. C., Kim T. K. 2010Osteoarthritis Cartilage. 18(4):530–8. doi: 10.1016/j.joca.2009.12.004. [DOI] [PubMed] [Google Scholar]
  14. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Page M. J., McKenzie J. E., Bossuyt P. M.., et al. 2021BMJ. 372:n71. doi: 10.1136/bmj.n71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Munro B. H. Statistical Methods for Health Care Research. Lippincott Williams & Wilkins; [Google Scholar]
  16. Evans J. D., editor. Straightforward statistics for the behavioral sciences. Brooks/Cole Pub. Co; Pacific Grove, CA: [Google Scholar]
  17. Methodological index for non-randomized studies (minors): development and validation of a new instrument. Slim K., Nini E., Forestier D., Kwiatkowski F., Panis Y., Chipponi J. 2003ANZ J Surg. 73(9):712–716. doi: 10.1046/j.1445-2197.2003.02748.x. [DOI] [PubMed] [Google Scholar]
  18. Assessing Lower Limb Alignment: Comparison of Standard Knee Xray vs Long Leg View. Zampogna B., Vasta S., Amendola A.., et al. 2015Iowa Orthop J. 35:49–54. [PMC free article] [PubMed] [Google Scholar]
  19. A comparative assessment of alternatives to the full-leg radiograph for determining knee joint alignment. Navali A. M., Bahari L. A. S., Nazari B. 2012Sports Med Arthrosc Rehabil Ther Technol. 4(1):40. doi: 10.1186/1758-2555-4-40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. A comparative assessment of alignment angle of the knee by radiographic and physical examination methods. Kraus V. B., Vail T. P., Worrell T., McDaniel G. 2005Arthritis Rheum. 52(6):1730–1735. doi: 10.1002/art.21100. [DOI] [PubMed] [Google Scholar]
  21. Predicting the mechanical hip-knee-ankle angle accurately from standard knee radiographs: a cross-validation experiment in 100 patients. Gielis W. P., Rayegan H., Arbabi V.., et al. 2020Acta Orthop. 91(6):732–7. doi: 10.1080/17453674.2020.1779516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. A new method to measure anatomic knee alignment for large studies of OA: data from the osteoarthritis initiative. Iranpour-Boroujeni T., Li J., Lynch J. A., Nevitt M., Duryea J., OAI Investigators 2014Osteoarthritis Cartilage. 22(10):1668–1674. doi: 10.1016/j.joca.2014.06.011. [DOI] [PubMed] [Google Scholar]
  23. A comparison of five approaches to measurement of anatomic knee alignment from radiographs. McDaniel G., Mitchell K. L., Charles C., Kraus V. B. 2010Osteoarthritis Cartilage. 18(2):273–7. doi: 10.1016/j.joca.2009.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Conventional knee films hamper accurate knee alignment determination in patients with varus osteoarthritis of the knee. van Raaij T. M., Brouwer R. W., Reijman M., Bierma-Zeinstra S. M. A., Verhaar J. A. N. 2009Knee. 16(2):109–111. doi: 10.1016/j.knee.2008.10.003. [DOI] [PubMed] [Google Scholar]
  25. Does measurement of the anatomic axis consistently predict hip-knee-ankle angle (HKA) for knee alignment studies in osteoarthritis? Analysis of long limb radiographs from the multicenter osteoarthritis (MOST) study. Sheehy L., Felson D., Zhang Y.., et al. 2011Osteoarthritis and Cartilage. 19(1):58–64. doi: 10.1016/j.joca.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Standard Radiographs Are Not Sufficient for Assessing Knee Mechanical Axis in Patients With Advanced Osteoarthritis. Stickley C.D., Wages J.J., Hetzler R.K., Andrews S.N., Nakasone C.K. 2017J Arthroplasty. 32(3):1013–7. doi: 10.1016/j.arth.2016.09.024. [DOI] [PubMed] [Google Scholar]
  27. Comparison of Standard Anteroposterior Knee and Hip-to-Ankle Radiographs in Determining the Lower Limb and Implant Alignment after Total Knee Arthroplasty. Skyttä E. T., Lohman M., Tallroth K., Remes V. 2009Scand J Surg. 98(4):250–253. doi: 10.1177/145749690909800411. [DOI] [PubMed] [Google Scholar]
  28. The Inadequacy of Short Knee Radiographs in Evaluating Coronal Alignment After Total Knee Arthroplasty. Park A., Stambough J. B., Nunley R. M., Barrack R. L., Nam D. 2016J Arthroplasty. 31(4):878–82. doi: 10.1016/j.arth.2015.08.015. [DOI] [PubMed] [Google Scholar]
  29. Using short knee radiographs to predict the coronal alignment after TKA: Is it an accurate proxy for HKA on full-length images? Shang G., Hu M., Guo J., Hao X., Xiang S. 2022J Orthop Surg Res. 17(1):340. doi: 10.1186/s13018-022-03235-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Can Anatomic Alignment Measured from a Knee Radiograph Substitute for Mechanical Alignment from Full Limb Films? Felson D.T., Cooke T.D.V., Niu J.., et al. 2009Osteoarthritis Cartilage. 17(11):1448–52. doi: 10.1016/j.joca.2009.05.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Short knee radiographs can be inadequate for estimating TKA alignment in knees with bowing. Oh S.-M., Bin S.-I., Kim J.-Y., Lee B.-S., Kim J.-M. 2020Knee Surg & Relat Res. 32(1):9. doi: 10.1186/s43019-019-0020-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Tibial bowing and tibial component placement in primary total knee arthroplasty in valgus knees: Are we overlooking? Palanisami D., George M. J., Hussain A. M., Md C., Natesan R., Shanmuganathan R. 2019J Orthop Surg (Hong Kong) 27(3):2309499019867006. doi: 10.1177/2309499019867006. [DOI] [PubMed] [Google Scholar]
  33. Longitudinal Shapes of the Tibia and Femur are Unrelated and Variable. Howell S.M., Kuznik K., Hull M.L., Siston R.A. 2010Clinical Orthopaedics & Related Research. 468(4):1142–8. doi: 10.1007/s11999-009-0984-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. The influence of postoperative coronal alignment on revision surgery in total knee arthroplasty. Morgan S. S., Bonshahi A., Pradhan N., Gregory A., Gambhir A., Porter M. L. 2008Int Orthop. 32(5):639–642. doi: 10.1007/s00264-007-0391-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. The relationship between the survival of total knee arthroplasty and postoperative coronal, sagittal and rotational alignment of knee prosthesis. Kim Y. H., Park J. W., Kim J. S., Park S. D. 2014Int Orthop. 38(2):379–85. doi: 10.1007/s00264-013-2097-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Orthopedic Reviews are provided here courtesy of Open Medical Publishing

RESOURCES