Sage Journals: Discover world-class research

Abstract

Purpose: The purpose of this study was to evaluate associations between demographics and Coronal Plane Alignment of the Knee (CPAK) classification in pre-surgical TKA patients. Methods: This is a retrospective study of 1167 patients with knee osteoarthritis who underwent TKA. CPAK categories I-IX were determined by arithmetic mechanical hip-knee-ankle angle and joint line obliquity measurements from pre-operative bone length radiographs. Patient age, sex, body mass index (BMI) and race were collected. Chi-square test of independence and adjusted Pearson’s residuals evaluated associations between CPAK classification and demographics. Results: There was a significant association between CPAK phenotypes I-IX and patient sex (X² = 5.8, p < 0.01). A positive association was found between both men and CPAK phenotype I, and women and CPAK phenotype VII. A positive association was found between African American patients and CPAK phenotype III and a negative association was found between African American patients and CPAK phenotype I (X² =14.8, p-value = 0.01). There was no association between age and BMI with CPAK phenotypes (n.s.). Conclusion: These results indicate that there are unidentified sex and race differences that exist in the CPAK classification of native arthritic knees. Patient characteristics play a significant role in determining patient knee phenotypes. Further research should investigate whether these characteristics warrant inclusion in pre-operative preparations, aiming to enhance the personalization of arthroplasty procedures.

Keywords

coronal plane alignment of the knee demographics total knee arthroplasty

Introduction

Dissatisfaction after total knee arthroplasty (TKA) continues to be reported amongst patients, the question of how to better optimize total knee arthroplasty procedures persists. Historically, mechanical alignment (MA) techniques have been utilized with the goal of achieving a horizontal joint line perpendicular to the neutral mechanical axis.¹ Although this technique is highly successful with regards to implant longevity and low failure rates, the lack of consideration of a patient’s native alignment in MA has been hypothesized to contribute to decreased patient satisfaction.¹ In contrast, kinematic alignment (KA) technique aims to restore the arthritic deformity to the pre-arthritic alignment in 3 dimensions. Increased interest in KA alignment strategies have evolved as improved survival rates have been reported.²

The Coronal Plane Alignment of the Knee (CPAK) classification system utilizes arithmetic Hip-knee-ankle angle (aHKA) and joint line obliquity (JLO) to categorize patients into nine different knee phenotypes.³ This system has been proposed to better identify native knee phenotypes and guide intraoperative soft tissue balancing. Associations between certain CPAK phenotypes and intraoperative alignment techniques have been demonstrated to improve functional outcomes.⁴

The variability in CPAK phenotype frequency among different patient populations has been examined. Studies analyzing CPAK phenotypes in patients from Belgium and Australia have found similarities in distribution of phenotypes between arthritic and healthy knees with CPAK II, I and IV as the most common.³ Additionally, sex differences in CPAK distributions have been found with females exhibiting higher incidence of valgus phenotypes compared to men.⁵ Although differences in CPAK phenotypes between geographic locations and sex have been examined, the interaction of other patient demographics and CPAK classification have not been investigated.^3,5

To date, no studies have examined the interaction of sex, race, age, and BMI on CPAK phenotype. Therefore, the aim of this study was to examine the impact of patient demographics on CPAK classification for native arthritic knees. It was hypothesized there would be differences in frequency of CPAK phenotypes impacted by sex and race.

Methods

Institutional Review Board (IRB) approval was attained, and a retrospective review identified all primary total knee arthroplasty cases performed at a tertiary care health system located in the southeastern United States between January 2014 and December 2022. Inclusion criteria were patients between 30 and 91 years old who underwent a primary TKA and had pre-operative long leg standing radiographs. Patients who had prior knee arthroplasty were excluded.

A total of 1167 patients were included in the study. Demographic information including age at time of surgery, sex, race, and BMI were recorded for each patient. TraumaCad orthopaedic templating software (Brainlab, Munich, Germany) was utilized to complete radiographic measurements. The medial proximal tibial angle (MPTA) was defined as the medial angle formed between the tibial mechanical axis and joint line of the proximal tibia, and the lateral distal femoral angle (LDFA) was defined as the lateral angle formed between the femoral mechanical axis and the joint line of the distal femur. The MPTA and LDFA were measured on preoperative bone length films with TraumaCad.

CPAK classification was defined as outlined by MacDessi et al.³ Arithmetic HKA (aHKA) was calculated as MPTA minus LDFA, and JLO was calculated as MPTA plus LDFA. The established boundaries for aHKA are varus: less than −2°, neutral: 0 +/− 2°, and valgus: greater than 2°. The boundaries for JLO, as previously defined, are apex distal: less than 177°, neutral: 180 +/− 3° and apex proximal: greater than 183°. The CPAK matrix has the three columns of aHKA and three rows of JLO plotted against each other to create 9 different recognized phenotypes.

The interobserver reliability of the CPAK measurements were determined using the class 3 model intraclass correlation coefficient (ICC). A 2-factor analysis of variation without replication model was used to calculate the ICC with a 95% confidence interval. The benchmarking of ICC values utilized the following classification: less than 0.50 indicating poor reliability, 0.50 to 0.75 indicating moderate reliability, 0.76 to 0.90 indicating good reliability, and values exceeding 0.90 indicating excellent reliability.⁶ When examining the association between demographic categories and CPAK category, chi square test of independence and descriptive statistics were used. When chi square test of independence demonstrated a significant association, Pearson’s standardized residuals were utilized to assess which demographic factors and CPAK categories demonstrated the greatest discrepancy from expected values. A residual greater than 3 or less than −3 was considered a large contributor to the chi square. Statistical significance was set at p value <0.05. Statistical analysis was performed using R studio Posit team (2023, RStudio: Integrated Development Environment for R. Posit Software, PBC, Boston, MA).

Results

Of the 1167 patients included in the study, sex was identified in 1166 (99.9%) with 520 males (44.5%) and 646 females (55.4%). The interobserver reliability for CPAK measurements is demonstrated in Supplemental Figure 1 with all measurements demonstrating an ICC with “excellent” reliability except Right mLPFA which demonstrated an ICC of “good” reliability.

There was a significant association between patient sex and CPAK phenotype (X² = 5.8, p < 0.01) (Table 1, Figure 1). When examining all possible combinations of sex and CPAK phenotypes, Pearson’s residuals demonstrated that male sex and phenotype I was observed more than expected while female sex and phenotype VII was observed more than expected. When examining aHKA axis alignment category only, there is a significant association between sex and aHKA axis (X² = 15.6, p < 0.01). When examining all possible combinations of sex and aHKA alignments, Pearson’s residuals demonstrated that female sex and valgus deformity was observed more than expected (F = 13%, M = 9%) while female sex and varus deformity was observed less than expected (F = 55%, M = 66%).

Table 1.

CPAK frequency by patient sex.

CPAK type	Males (n = 520)	Females (n = 646)
I	46% (239)	29% (189)
II	13% (70)	19% (123)
III	6% (30)	8% (51)
IV	17% (89)	16% (103)
V	9% (48)	9% (58)
VI	6% (14)	4% (26)
VII	3% (17)	10% (64)
VIII	2% (10)	4% (23)
IX	<1% (3)	1% (9)

Figure 1.

Distribution of CPAK phenotypes by sex.

Race was documented in 1148 (98.4%) patients with 969 (84.4%) identifying as Caucasian/White, 141 (12.3%) identifying as Black or African American, 17 (1.5%) identifying as Asian and 21 (1.8%) identifying as American Indian or Alaskan native, native Hawaiian or other pacific islander, 2 or more races or did not disclose. There was a significant association between Caucasian/White and Black/African American race and CPAK phenotypes I-IV (X² =14.8, p-value = 0.01) (Table 2, Figure 2). When examining all possible combinations of white or black race and CPAK phenotypes, Pearson’s residuals demonstrated that black or African American race and CPAK phenotype III was observed more than expected (13%) as compared to white or Caucasian patients (6%) while black or African American race and CPAK phenotype 1 was observed less than expected (26%) as compared to white or Caucasian patients (37%). When examining aHKA axis alignment category only, there is a significant association between white or black race and aHKA axis (X² = 11.9, p-value <0.01). When examining all possible combinations of black or white race and aHKA alignments, Pearson’s residuals demonstrated that Black or African American race and valgus deformity was observed more than expected (20%) as compared to white or Caucasian patients (10%).

Table 2.

Frequency of CPAK phenotype by patient race.

CPAK type	Caucasian/White (n = 969)	Black or African American (n = 141)	Asian (n = 17)	Other (n = 22)
I	37% (362)	26% (36)	76% (13)	50% (11)
II	17% (163)	18% (26)	6% (1)	9% (2)
III	6% (59)	13% (19)	0	4.5% (1)
IV	17% (160)	16% (23)	0	32% (7)
V	1% (9)	10% (14)	0	0
VI	9% (91)	5% (7)	12% (2)	0
VII	3% (31)	9% (13)	0	4.5% (1)
VIII	7% (65)	<1% (1)	0	0
IX	3% (29)	1% (2)	6% (1)	0

Figure 2.

Distribution of CPAK phenotype by Patient Race (Other category includes the following: Did not disclose, 2 or more, American Indian or Alaskan native, native Hawaiian or other pacific islander).

BMI was identified in 1159 (99.3%) patients with 30 (2.6%) classified as healthy, 749 (64.6%) overweight and 380 (32.8%) obese. There was no association between patients with BMI classifications of overweight and obese and CPAK phenotype (X² = 8.1, p = 0.42) (Table 3).

Table 3.

CPAK frequency by patient BMI.

CPAK type	Healthy (n = 30)	Overweight (n = 749)	Obese (n = 380)
I	37% (11)	34% (257)	41% (158)
II	27% (8)	17% (130)	14% (55)
III	7% (2)	7% (56)	6% (21)
IV	7% (2)	17% (128)	16% (62)
V	13% (4)	9% (67)	9% (35)
VI	0% (0)	4% (30)	3% (10)
VII	7% (2)	7% (52)	7% (27)
VIII	0% (0)	3% (21)	2% (9)
IX	3% (1)	1% (8)	<1% (3)

Age was listed in 1154 (98.8%) patients with 65 (5.6%) patients aged 50 or less, 218 (18.9%) aged 51-60, 473 (41.0%) aged 61-70, 345 (29.9%) aged 71-80 and 53 (4.6%) aged 81 and above. There was no association between patient age and CPAK phenotype I-IV (X² = 7.5, p = 0.82) (Table 4).

Table 4.

CPAK frequency by patient age.

CPAK type	50 or less (n = 65)	51-60 (n = 218)	61-70 (n = 473)	71-80 (n = 345)	81 or greater (n = 53)
I	35 % (23)	35 % (76)	38% (181)	38% (130)	34% (18)
II	14 % (9)	17% (37)	17% (80)	18 % (61)	11% (6)
III	9 % (6)	6% (13)	7% (33)	7% (24)	9% (5)
IV	20 % (13)	12% (27)	20% (93)	14% (49)	19% (10)
V	3 % (2)	12% (26)	9% (44)	10% (33)	2% (1)
VI	5 % (3)	4% (9)	2.5% (12)	4% (13)	6% (3)
VII	3 % (2)	9% (19)	4% (19)	9% (31)	11% (6)
VIII	11 % (7)	3% (7)	2 % (7)	1% (3)	2% (1)
IX	0% (0)	2% (4)	1% (4)	<1% (1)	6% (3)

Discussion

The most important findings of this study were that patient demographics, such as race and sex demonstrated differences in CPAK phenotype frequency. However, no associations were found between age and BMI. With increased efforts to personalize TKA, phenotype classification systems such as CPAK offer surgeons the opportunity to further characterize patient’s pre-operative anatomy and can help guide operative planning. Previous work evaluating CPAK phenotypes has found geographical differences in phenotype frequency and associations between patient sex and aHKA alignment.^7,8 To date, there are no studies examining the association between demographic factors such as race, sex, age, and BMI impact on CPAK phenotype.

An association between patient sex and CPAK phenotype frequency was found in this study with 46% of males in CPAK phenotype I as compared to 29% of females. In contrast, females had a higher frequency of phenotype II at 19% and VII at 10% as compared to men who had frequencies of 13% and 3%, respectively. Additionally, when examining aHKA alignment, there was an association between sex and aHKA with females demonstrating a higher incidence of valgus deformity and less varus deformity as compared to men. These findings are similar to that of Steele et al., who analyzed 1946 knees using a deep learning neural network to render CPAK classifications and found that women had a significantly higher proportion of knees in valgus and significantly less knees in varus as compared to men.⁵ Similarly, Huber et al. who analyzed 8739 long leg radiographs found that CPAK Type I, II and III along with neutral and valgus alignment were more common in women.⁹ Overall, the findings of different CPAK phenotype frequencies between males and females is consistent with previous work demonstrating sex differences in lower extremity alignment. Differences have been demonstrated in static Q angle, anterior pelvic tilt, hip anteversion and thigh foot angles between healthy males and females.^10,11

This study identified a significant association with race and CPAK phenotypes. A positive correlation was observed between CPAK phenotype III and Black or African American patients. However, within this group, CPAK phenotype I was the most common, despite having a negative correlation, whereas CPAK phenotype VIII was the least common. Ethnic differences in knee phenotypes have previously been demonstrated with significantly lower tibial mechanical angle and higher femoral mechanical angle found in Asian patients than Caucasian patients.⁷ Additionally, CPAK frequency in Asian populations have been shown to have slight differences in observed classifications with CPAK II, I and III being the most common in healthy and arthritic knees.^12,13 A systematic review by Pagan et al. found geographical variations in CPAK phenotypes, CPAK phenotype I was the most common in India (58.8%), Japan (53.8%), and France (33.4%), whereas CPAK phenotype II was the most common in Australia (32.8%).⁸ Racial differences in alignment may be related to differences in the distribution of body weight and biomechanics among different racial and ethnic groups. Understanding the racial and ethnic differences in coronal plane alignment may have important implications for the diagnosis, prevention, and treatment of knee osteoarthritis and other knee-related conditions in diverse patient populations.

In addition, this study found no association of BMI or age with CPAK phenotypes. Our discoveries indicate that considerations beyond BMI and age might play a more crucial role in determining CPAK classifications. This observation is consistent with Nomoto et al.'s identification of similarities in CPAK phenotypes both before and after an eight-year period of osteoarthritic progression.¹⁴ Possible factors that could influence CPAK phenotypes include genetic factors, physical activity levels, and previous injury history.

One limitation of our study is the reliance on race data that is self-reported and recorded in the electronic medical record (EMR). While efforts are made to accurately capture this information, we acknowledge there is a possibility that patients may have entered their race incorrectly or declined to answer the question altogether. Additionally, we recognize that self-identified race is a complex and multifaceted construct that may not capture the full range of ethnic and ancestral backgrounds that contribute to a person’s identity. Additionally, there were few patients who identified as African American and Asian included in this study. Finally, the inclusion of patients with severe inflammatory arthritis and significant knee deformity may lead to inaccuracies in the measurement of anatomical parameters. Despite these limitations, we believe that our study provides valuable insights into the relationship between race and CPAK phenotypes. Additionally, the overall observed frequency of certain CPAK phenotypes is very low. A limitation encountered was observed CPAK phenotypes VI, VII, VIII, IX were so few for different demographic groups, which resulted in the inability to include these in statistical analysis. Future investigations of CPAK classifications could include examining their association with Patient-Reported Outcomes Measures (PROM), including generic and joint specific PROMs to evaluate the impact of CPAK on patient-reported outcomes. Additionally, assessing the inter-observer reliability of the CPAK classification or utilizing computed tomography images would enhance the validity and reliability of CPAK assessments.¹⁵

The high dissatisfaction rate in TKA in uncomplicated cases, could be attributed to the broad variability in alignment and the sex-specific distribution present in both classifications.¹⁶ Although this study demonstrates there is an association between sex, race, and alignment, this ultimately has no impact on TKA surgical technique because the goal of the surgeon is to maintain the joint line. However, the association between joint line and certain phenotypes based upon sex may help surgeons decide on prosthesis type such as asymmetric, symmetric, or custom designs.

Conclusion

In examination of patient demographics and CPAK phenotype, we demonstrate that sex and racial differences exist in phenotype frequency. Therefore, findings of this study suggest a necessity for personalized alignment paradigms that account for variations in individual knee morphology such as personalized alignment methods and prosthesis design selection.

Supplemental Material

Supplemental Material - Patient sex and race are associated with differences in coronal plane alignment classification in native arthritic knees

Supplemental Material for Patient sex and race are associated with differences in coronal plane alignment classification in native arthritic knees by Caitlin Grant, Niall Cochrane, Mikhail Bethell, Christopher Holland, Jay Levin, Joshua Helmkamp and Thorsten Seyler in Journal of Orthopaedic Surgery.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical statement

ORCID iD

Mikhail Bethell

Supplemental Material

Supplemental material for this article is available online.

References

Jaffe

Dundon

Camus

. Alignment and balance methods in total knee arthroplasty. J Am Acad Orthop Surg 2018; 26(20): 709–716.

Howell

Shelton

Hull

. Implant survival and function ten years after kinematically aligned total knee arthroplasty. J Arthroplasty 2018; 33(12): 3678–3684.

MacDessi

Griffiths-Jones

Harris

, et al. Coronal plane alignment of the knee (CPAK) classification: a new system for describing knee phenotypes. Bone Joint Lett J 2021; 103-B(2): 329–337.

Clark

Steer

Khan

, et al. Maintaining joint line obliquity optimizes outcomes of functional alignment in total knee arthroplasty in patients with constitutionally varus knees. J Arthroplasty 2023; 38: S239.

Steele

Jang

Brilliant

, et al. Deep learning phenotype automation and cohort analyses of 1,946 knees using the coronal plane alignment of the knee classification. J Arthroplasty 2023; 38(6): S215–S221.

Koo

. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 2016; 15(2): 155–163.

Moser

Hess

de Villeneuve Bargemon

J-B

, et al. Ethnical differences in knee phenotypes indicate the need for a more individualized approach in knee arthroplasty: a comparison of 80 Asian knees with 308 caucasian knees. J Personalized Med 2022; 12(1): 121.

Pagan

Karasavvidis

Lebrun

, et al. Geographic variation in knee phenotypes based on the coronal plane alignment of the knee (CPAK) classification: a systematic review. J Arthroplasty 2023; 38: 1892.

Huber

Mitterer

Vallant

, et al. Gender-specific distribution of knee morphology according to CPAK and functional phenotype classification: analysis of 8739 osteoarthritic knees prior to total knee arthroplasty using artificial intelligence. Knee Surg Sports Traumatol Arthrosc 2023; 31: 4220–4230.

10.

Nguyen

A-D

Shultz

. Sex differences in clinical measures of lower extremity alignment. J Orthop Sports Phys Ther 2007; 37(7): 389–398.

11.

Tillman

Bauer

Cauraugh

, et al. Differences in lower extremity alignment between males and females: potential predisposing factors for knee injury. J Sports Med Phys Fit 2005; 45(3): 355–359.

12.

Hsu

C-E

Chen

C-P

Wang

S-P

, et al. Validation and modification of the coronal plane alignment of the knee classification in the asian population. Bone Jt Open 2022; 3(3): 211–217.

13.

Toyooka

Osaki

Masuda

, et al. Distribution of coronal plane alignment of the knee classification in patients with knee osteoarthritis in Japan. J Knee Surg 2023; 36(07): 738–743.

14.

Nomoto

Hanada

Hotta

, et al. Distribution of coronal plane alignment of the knee classification does not change as knee osteoarthritis progresses: a longitudinal study from the Toei study. Knee Surg Sports Traumatol Arthrosc 2023; 31(12): 5507–5513.

15.

Tarassoli

Corban

Wood

, et al. Long leg radiographs underestimate the degree of constitutional varus limb alignment and joint line obliquity in comparison with computed tomography: a radiographic study. Knee Surg Sports Traumatol Arthrosc 2023; 31: 4755–4765.

16.

Gunaratne

Pratt

Banda

, et al. Patient dissatisfaction following total knee arthroplasty: a systematic review of the literature. J Arthroplasty 2017; 32(12): 3854–3860.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB