Using Advanced Population Data to Fit Large Joint Prosthesis | Orthopedics This Week
Large Joints and Extremities

Using Advanced Population Data to Fit Large Joint Prosthesis

Courtesy of Stryker Corporation

As orthopedics moves into the med-intelligence era, one of the core promises is that new, large data sets will become available to surgeons to help inform individual patient treatment.

This new study can be viewed as an interim step along this road to a med-intelligence era.

Stryker Corporation, as literally every large integrated supplier of orthopedic and spine products is doing, has created an advance analytics and data base system to help surgeons use population level data to improve the specificity and customization of their treatment program. It is Stryker’s Orthopaedics Modeling and Analytics (SOMA) program.

SOMA is a proprietary database of 3D CTA scans and software. According to the most recent information available from Stryker, SOMA’s dataset includes 19,500 bones from all ethnicities. SOMA also offers proprietary software tools:

  • Bone Database Management Tool: Provides ability to research gender, ethnicity, and/or patient age.
  • Stryker Anatomy Analysis Tool: Allows the ability to analyze shape variation of bones, and to capture geometrical measurements such as bone density.
  • Stryker Implant Fitting Tool: Allows for automatic analysis of how well the implant design fits within or on the bone.

This last software tool, which provides information on the morphological aspects of osteoarthritic knees and the relationship of that data to implant positioning and balancing, is the subject of this new study.

Study co-author Nicolas S. Piuzzi, M.D., associate professor of Orthopaedic Surgery and co-director of the Musculoskeletal Research Center at Cleveland Clinic, explained to OTW the general purpose of this study. “Morphological characteristics of knees with OA [osteoarthritis] have been explored in the past, however only recently we have been able to develop and have access to large-scale biomorphometric computed tomography (CT) which may help outline morphological differences.”

Learning How to Incorporate Large Data Analysis Into a Working Practice

Dr. Piuzzi explained that collaboration with industry was foundational to organizing and completing this interesting large data set study. “A major critical step to making this large data analysis happen was the access and collaboration with industry which allowed us to use advanced analytics: the software Stryker Anatomy Analysis Tool that was used to segment bone surfaces on the CT images and create a virtual reconstruction of a corresponding bone using predefined landmarks and user-defined points, which were then mapped onto each individual knee for analysis.”

Dr. Piuzzi and his team collected morphological measurements for 10 different aspects of the knee.

  1. non-weight-bearing hip-knee-ankle angle,
  2. mechanical lateral distal femoral angle,
  3. medial proximal tibial angle,
  4. rotation of the posterior condylar axis relative to the surgical trans-epicondylar axis,
  5. ratio of medial to lateral posterior condylar offset,
  6. ratio of medial to lateral condylar radius,
  7. medial posterior slope,
  8. lateral posterior slope,
  9. medial coronal slope, and
  10. lateral coronal slope.

According to Dr. Piuzzi, this may well be the largest study of its kind. “We analyzed CT images of 965 normal knees, compared them with 193 OA knees (97 varus, 65 neutral, 31 valgus) and found significant differences in posterior condylar axis, condylar offset, and condylar radius as well as tibial slope in both the sagittal and coronal planes.”

“The OA group also had significantly lower medial posterior tibial slope and higher lateral posterior slope, with minor differences between the varus and valgus OA groups.”

“Compared with the normal group, the OA group was in overall varus (non-weight-bearing hip-knee-ankle angle, -2.2° ± 5.0° compared with -0.2° ± 2.4°) and had a significantly smaller medial posterior slope (8.4° ± 4.0° compared with 9.2° ± 4.0°), larger lateral posterior slope (9.2° ± 3.6° compared with 7.2° ± 3.3°), and smaller medial coronal slope (82.1° ± 4.3° compared with 83.9° ± 3.3°).”

The research team also observed differences among the OA subgroups for the medial coronal slope and lateral coronal slope.

“Compared with the normal group, the surgical trans-epicondylar axis of the OA group was less externally rotated relative to the posterior condylar axis (0.3° ± 1.5° compared with 1.2° ± 1.9°), and both the condylar offset ratio (1.01 ± 0.06 compared with 1.04 ± 0.07) and the condylar radius ratio (0.98 ± 0.07 compared with 1.03 ± 0.07) were smaller. Only the condylar radius ratio showed differences among the OA subgroups, with valgus deformity associated with a larger ratio.”

Of course, after conducting this study, the research team pointed out how access to this type of population level data can personalize large joint surgeries. “Enhanced understanding of various deformities can enable personalized implant positioning and balancing in total knee arthroplasty in an effort to continue improving clinical outcomes and optimizing procedural value.”

And, perhaps even more interestingly, this study may have exposed limitations of current implant positioning techniques. “Our results further question whether the use of conventional positioning methods, including those based on a fixed amount of rotation (e.g., posterior condylar axis +3° external rotation), soft tissue balancing, and a uniform posterior tibial slope, is sufficient to address the unique anatomy of each individual. With the advent and evolution of computer navigation and robotics and advancements in component design, there is a strong need to evolve toward more personalized treatment that utilizes implants and technology to help tailor the TKA to each individual’s anatomy,” said Dr. Piuzzi.


1 thought on “Using Advanced Population Data to Fit Large Joint Prosthesis

Share Your Thoughts

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.