A novel glenoid classification for pre-operative planning in reverse shoulder arthroplasty: “Chandigarh Classification”

Patient recruitment, inclusion, and exclusion criteria

Institutional ethics approval was obtained from the institutional ethics review board before the initiation of the study. Data from a previously conducted CT-based study of 600 normal shoulders at our institute were used to assess average glenoid dimensions in our patient population (average glenoid max anteroposterior [AP] diameter 23.2 ± 0.6 mm). Based on this, we developed a new classification system that incorporates both glenoid size (as assessed by the beforementioned study) and glenoid wear pattern (Walch classification).

The new classification system was then studied retrospectively to assess frequency mapping in our patient population. A retrospective analysis of CT scans of patients undergoing RSA at our institute from January 2022 to March 2023 was carried out. Written informed consent was taken from all patients before using their CT scans in the analysis. Inclusion criteria were patients over the age of 60 years undergoing RSA at our institute during the retrospective study period, CT scans done at our institute (to eliminate cross-center bias), and those willing to participate in the study. Exclusion criteria were patients <60 years of age, patients who had CT scans done at outside centers, and patients unwilling to participate in the study.

Study power calculation

Using a correlation coefficient of 0.03 to study the null hypothesis, we found that the sample size should be 80 patients for a moderately high power and reduced error rate. Using 120 patients, the alpha score is <0.01.

CT scan-based measurements

Two measurements were taken for all patients on 2D CT and 3D reformatted CT scans: Degree of retroversion (Walch Classification) and maximum AP dimension of the glenoid face (circle concept). The best way to assess the glenoid interface is through the three-dimensional reconstruction of two-dimensional CT images. These aid in understanding and defining patient-specific glenoid anatomy and location of wear.^[14,15] Werner et al. showed that measurements of glenoid version and inclination were more accurate with 3D reconstruction images as compared to 2D reconstructions.^[16] Walch classified glenoid wear patterns as Types A, B, and C. Friedman et al. showed that CT scans were a good modality to assess these wear patterns.^[13] Since its inception, the classification has undergone refining by various authors, with the present classification as such [Figure 1]. For the present study to adapt to the new proposed glenoid model, we measured the retroversion according to the traditional and more simplistic Walch classification. Type A is centered or symmetric arthritis without posterior subluxation of the humeral head. Type B is characterized by asymmetric arthritis with posterior subluxation of the humeral head. Type C shows glenoid retroversion >25° (dysplastic in origin) regardless of glenoid erosion or the location of the humeral head with regard to the glenoid.^[17]

Close

Figure 1:

Modified Walch classification as it stands at present.^[17]

Export to PPT

The baseplate overhang was assessed using the circle concept to determine baseplate fit on glenoids [Figure 2]. A circle is defined as the maximum possible circular area placed over the patient’s inferior glenoid, which encloses it without any bone deficiency in any quadrant in a 3D CT scan-based enface view of the glenoid. The circle is used to represent the base plate of reverse arthroplasty (in this case, the smallest baseplate available, which is 25 mm but can be adapted per user), and the diameter of the circle relates to the size of the base plate. Thus, the circle uses the maximum AP dimension of the glenoid at its largest point to determine the ideal surface area available for baseplate fit. If the diameter of the circle (baseplate size) was larger than the maximum AP diameter of the glenoid, this was defined as O. If the diameter of the circle was smaller than the maximum AP diameter of the glenoid, this was defined as N.

Close

Figure 2:

Circle concept. The circle is overlaid on a 3D computed tomography scan-based enface view of the glenoid and sized for the baseplate dimensions (in the representative case for a 25 mm baseplate).

Export to PPT

Inter-observer reliability

To assess the inter-observer reliability, we selected 28 shoulders from our data set. The new classification model was sent to two previous fellows of the institute who were taught how to use the classification, and they were instructed that they would receive CT data for 18 patients and that they must classify each patient.

Data collection and statistical analysis

Retrospective analysis of pre-operative CT scans of enrolled subjects was carried out by a single author (SP). Frequency mapping of each CT scan was done according to the new proposed classification subtypes (AO, BO, CO, AN, BN, and CN). The age and gender of the patients were recorded at the time of index surgery. The statistical analysis was performed by a single-blinded author (JS) using IBM Statistical Package for the Social Sciences 20 software. Frequency mapping characteristics were studied using numerical charting. A simple t-test was used to assess for differences between groups with respect to age and gender. The significance level adopted was P < 0.05.

New proposed classification – “Chandigarh classification”

The new classification combines the circle concept (to assess baseplate fit) with the traditional Walch Classification (to assess glenoid wear and version). Circle measurements are based on the smallest available baseplate diameter (25 mm). Circles are divided into two types: Type O (<25 mm maximum AP diameter) and Type N (≥25 mm maximum AP diameter). The Type O circles will have an overhang of the smallest baseplate on the native glenoid, whereas Type N circles will have a just or normal fit on the glenoid. The new classification is outlined below [Table 1 and Figure 3].

Close

Figure 3:

Diagrammatic representation of the new proposed classification system. O: Overhang of baseplate, N: Normal fit base plate. A/B/C: Walch classification.

Export to PPT

Frequency mapping data

A total of 120 patients were included in the retrospective analysis. A total of 68 females (57%) and 52 males (43%) were enrolled in the study. The average age of the patient population was 68.2 ± 6.4 years. 68% of patients (n = 82) were below 70 years of age and the remaining were more than 70 years of age at the time of index surgery.

According to the new classification, the frequency mapping data are presented in Table 2 and Figure 4. Most patients were either type AO (42%) or AN variants (36%). 55% of patients (n = 66) had overhang (Type O) of the baseplate on the glenoid. Less than 10% of patients (n = 8; 7%) had severely retroverted glenoids (Walch C).

Close

Figure 4:

Graphical representation of frequency mapping data of study population (n = 120 patients). O: Overhang of baseplate, N: Normal fit base plate. A/B/C: Walch classification.

Export to PPT

Subgroup analysis

Age and gender were assessed for variation between the types of new proposed classification. Neither age nor gender showed any significant correlation between the subgroups of the new classification.

Inter-observer reliability

Rater 1 and Rater 2 reliability data for the subset of 28 patients are presented in Table 3. A good to strong level of agreement was seen between both raters, with a kappa value of 0.878 with a statistical significance of <0.005.

Importance of baseplate and glenosphere positioning in shoulder arthroplasty

Reverse shoulder arthroplasty has seen increased uptake globally in recent years attributable to wider availability of the implant, better outcomes as compared to previous generations of prosthesis design, more surgical training and exposure, and expanding indications. With this increasing uptake comes a need to optimize outcomes, and one of the most important aspects of that is baseplate and glenosphere positioning in RSA. Glenosphere positioning has a significant impact on the clinical outcomes of RSA. Malpositioning increases the risk of complications, including dislocation, scapular notching, decreased range of motion, and component loosening.^[18-20]

Glenosphere positioning is determined by surgical technique, and reasons for its malpositioning include inaccurate assessment of pathological anatomy, incorrect choice of implant and/or positioning of the implant to correct pathology and inaccurate execution of preoperative plan at the time of surgery.^[21,22] All the above factors are influenced by both glenoid exposure and the degree of glenoid bone loss. In the setting of minimal glenoid bone loss, surgical decision-making is simpler with the use of intraoperative assessments such as the surface of the glenoid fossa, the “subchondral smile” method, or standard inferior tilt guides. However, some studies have shown that these references can also be unreliable.^[23,24] In the setting of moderate-to-severe glenoid bone loss, surgical decision-making is much more difficult, hence the role of pre-operative planning or patient-specific instrumentation techniques.

As glenosphere positioning is determined by baseplate positioning primarily and the baseplate by the glenoid face hence the role of pre-operative glenoid planning is paramount. To better understand glenoid morphology, a host of classifications has been developed by various authors, however, the mainstay remains the Walch classification.

Walch classification and its limitations

Walch et al. (1999) outlined a classification for glenoid wear in primary glenohumeral arthritis,^[10] with a higher Walch classification being a negative predictor for post-operative clinical function, as well as progression of posterior bone loss over time.^[11,12] In the present study, we have used the traditional Walch classification; however, the Walch classification has undergone modifications over time. Type A is centered or symmetric arthritis without posterior subluxation of the humeral head. Type A1 has minor central wear or erosion, whereas Type A2 has severe or major central wear or erosion. Type B is characterized by asymmetric arthritis with posterior subluxation of the humeral head. Type B1 has no obvious glenoid erosion with posterior joint space narrowing, subchondral sclerosis, and osteophytes. Type B2 has apparent or obvious erosion of the posterior glenoid, forming a biconcave appearance of the glenoid. Type C shows glenoid retroversion >25° (dysplastic in origin) regardless of glenoid erosion or the location of the humeral head with regard to the glenoid.^[17]

The classification has several limitations, especially the inconsistent reproducibility among orthopedic surgeons. As the Walch classification has expanded to include subtypes, intraobserver and interobsever reliability has decreased.^[17,25] The distinction between type B and C glenoids due to dysplastic wear versus glenoid wear has also drawn controversy. The Walch classification also addresses only posterior wear, with no descriptions of anterior or superior wear patterns, and relies on 2D scans for assessment rather than 3D techniques, which are more adequate in the modern era of shoulder planning.^[26]

Although the Walch Classification or its modification helps with pre-operative planning for the purpose of determining posterior asymmetry of the glenoid and the need for potential bone grafting or building the glenoid intra-operatively, it offers little in terms of the fit of the baseplate on the glenoid. In our practice, one of the common problems we have been facing is the ability to determine baseplate fit or overhang of the baseplate in our region, where the glenoid sizes are comparatively smaller as compared to Western populations. To this effect, we had previously performed a CT-based study to assess glenoid morphology (under publication), where we found that the average glenoid dimensions in our population (mean: 23.2 ± 0.6 mm) are smaller than the 25 mm baseplate. This means that some amount of baseplate overhang will exist in the majority of our patients, even with these small baseplates [Figure 5]. The morphology of the glenohumeral joint varies geographically.^[27,28] Most of the literature on the morphology of the glenoid refers to the Western population. There are few such data for the Indian population.^[29,30] Most Indian studies are anatomical studies on a cadaver with a small sample size, and a few are based on CT scans.^[29,31,32] The CT scan gives us the advantage of assessing the morphology of the glenoid and humeral head preoperatively.

Close

Figure 5:

Baseplate overhang seen in (a) South Asia due to smaller glenoid dimensions as compared to (b) western population.

Export to PPT

This concept is important as it allows us to determine how much of the glenoid surface area of the native glenoid, irrespective of version, is available for the baseplate. Thus, baseplates can be sized accordingly and pre-operatively without the need for specific or advanced pre-operative planning software.

New classification addresses both fit and version of glenoid

With the new classification, both baseplate fit and version are factored into the pre-operative planning in a single classification. Technically, the AN fit is the best fit as it provides normal baseplate fitting along with a concentric wear pattern, whereas the CO fit is the worst fit as it provides an overhang of the baseplate with dysplastic wear of the glenoid. Frequency mapping data showed that the vast majority of patients were in the AO or AN subtypes. What is surprising to see is the large number of patients with an AO subtype, suggesting an overhang of the small baseplate on the native glenoid. This is an important assessment as it highlights that a significant proportion of patients in our population may not have the ideal baseplate coverage despite a decent version.

The new classification is simple and reproducible as it has six types and is based on readily available 2D CT scans and reformatted 3D CT scans, which are available at most hospitals. The reproducibility of the new classification system was tested using two raters with a high degree of inter-observer reliability. The new proposed classification should help surgeons in clinical decision-making by not only factor in the fit of the glenoid baseplate but also the version of the baseplate and the need for potential augmentation. Although our model has been developed with a 25 mm baseplate (the smallest baseplate size available to us in our region), surgeons may adapt the classification according to the smallest baseplate size available in their region.

Frequency mapping data versus available literature

The frequency mapping data of the new classification system yields some salient findings. First, 78% of patients had a Walch A glenoid, with <10% having a Walch C glenoid. Importantly, 55% of patients were in overhang with a circle diameter of <25 mm. The circle diameters method, established by Sugaya et al.,^[33] allows us to better understand baseplate fit. Circle diameter measurements from our study show significantly lower measurements as compared to the max glenoid width of previous studies,^[34-38] with 55% of patients having a circle diameter <25 mm. This bears biomechanical significance because although using smaller baseplates has not been shown to lead to higher strain at the baseplate-glenoid interface,^[39] it may actually increase the impingement-free range of motion of the shoulder.^[40]

Limitations of the new classification

The new proposed classification has its own limitations. First, the baseplate size used is 25 mm (the smallest available baseplate in our region), however, this may be adapted as per the user. Second, the traditional Walch classification rather than the modified Walch classification has been used. The rationale behind this was to keep the proposed new classification as simple to reproduce as possible, especially for younger surgeons.