Commentary & Perspective

Commentary & Perspective on
"Comparison of Robotic-Assisted and Manual Implantation of a Primary Total Hip Replacement. A Prospective Study"
by Matthias Honl, MD, et al.

Commentary & Perspective by
Anthony M. DiGioia III, MD*,
Institute for Computer Assisted Orthopaedic Surgery, Pittsburgh, Pennsylvania

Computer-assisted orthopaedic surgical (CAOS) based tools and technologies have come from the drawing board to clinical use. At some centers, CAOS tools are being used on a routine basis. However, technical and clinical challenges continue to hinder their widespread adoption. Most importantly, there is a need to demonstrate improvements in patient outcome if these tools are to be widely used in clinical practice.

This paper provides one of the first opportunities to examine patient outcome in a prospective, randomized study of patients who underwent primary total hip replacement with use of a robotic assistive tool to prepare the femoral canal for one design of an uncemented femoral implant. The authors are to be congratulated on the methodology employed. This is one of the first reports in the CAOS-related literature to be performed in a prospective and randomized manner.

It is important for readers to understand some of the basics of CAOS. This new paradigm, made possible by computer-assisted orthopaedic surgery, tightly integrates preoperative planning and imaging with the surgical intervention. The hope is that this new generation of surgical tools and sensors will permit surgeons to perform more accurate and minimally invasive techniques. Most importantly, it must be understood that the goals of CAOS technologies, even with active robotic systems as used in this study, are to assist and to complement, but not to replace, the surgeon. Ultimately, active robotic systems are only as good as a surgeon's preoperative plan and surgical technique.

These new capabilities are balanced by the potential negative aspects and "costs" of the introduction of new technologies into clinical practice. These costs include the additional time required for preoperative planning and intraoperative imaging and the expense of increased operating-room time. In addition, the systems themselves are expensive and their accuracy and validity must be determined.

Readers should also understand that CAOS technologies represent a spectrum of tools. At one end of the CAOS spectrum are the "active" robotics systems, such as the one used in this clinical trial. These robotic tools are capable of performing part of an operation autonomously. At the other end of the spectrum are the "passive" navigation systems, tools that are capable of providing additional information during a surgical procedure but that do not perform surgery. Rather, the surgeon controls the action and uses the system as a surgical information system.

This classification system is based on control and safety. The United States Food and Drug Administration has determined that passive navigation systems can be reviewed for the most part through the 510k processes because the surgeon is the end effector and controls the surgical tools. In contrast, active robotic tools currently require an investigational device exemption because these tools work under their own software and hardware control for at least a portion of the procedure.

In their report, Honl et al. present a minimum twenty-four-month follow-up that focuses on the learning curve in the early adoption of the technology and on the short-term outcome of patients. This study presents both the positive and negative aspects of an active robotic system. On the positive side, there was an improvement in the equalization of limb lengths and in the varus/valgus alignment of the implant, and a transient improvement in patient outcome at six and twelve months that subsequently was equal to that of the control group at twenty-four months. On the negative side, the robotic group experienced multiple system failures, longer operating-room time, a higher prevalence of heterotopic ossification, and increased dislocation rates and limp associated with abduction weakness. The authors also observed signs of a violation of the abductor mechanism in several of the revision cases, which necessitated surgical repair.

Several other authors have reported that there is an appreciable time commitment for both the surgeon and the staff in the early stages of application of active robotic systems and that it is necessary to perform between twenty and thirty procedures (and sometimes more) before the approach is established and they reach a level of comfort in using the system. In fact, in the largest clinical experience in Germany, there was a substantial reduction in operative time and an improved ease of use over time.

The selection of an appropriate femoral implant may be a very important factor in the safe use of the robotic system. After their initial experiences, several sites in Europe have switched to implants that are of a straight design. It is certainly not optimal to have to choose a particular implant design because of its compatibility with the limitations of the robotic system. Caution should prevail when extrapolating the use of the robotic system to implants that have anatomical or curved designs or designs that differ substantially from the type of implant that was used in this study.

Most importantly, the reader should understand that it is the surgeon who remains responsible for planning and performing surgery. The robotic system only implements the surgeon's plan. Active robotic systems require a greater understanding of the surgical variables that must be planned for prior to the actual surgery. For instance, many decisions that surgeons currently make during surgery (i.e., fit and fill, press fit, etc.) must be made preoperatively. This shifting of the decision process to preoperative planning presents new challenges for surgeons and should not be underestimated. For example, a very aggressive preoperative plan (maximize fit and fill to the endosteal bone surface) may lead to a "freeze" of the system as it moves to cut into the more sclerotic bone. There may be times when an aggressive preoperative plan, even when accurately implemented by the robotic tool, may produce such a tight fit that it is difficult to press fit the femoral implant and impossible to seat the component fully.

There are also other implant-related factors to consider. Regarding dislocation, the implant chosen permits the surgeon to "dial" in different version alignments of the femoral component even after the femoral component sleeve is implanted. The authors mention that the plan was to place the femoral prosthesis in 15° of anteversion. It is unclear whether this is the 15° of the sleeve itself, 15° of the anteversion of the femoral stem inside the sleeve, or 15° with respect to the femoral anatomical axis. It is also unclear how the authors measured acetabular alignment during the surgical procedure.

Complications involving the abductor muscles are of concern. These problems may also be related to the type of femoral implant chosen. If the implant has a straight stem, lateral burring is required to permit direct access to the femoral canal. Is it difficult to protect the abductor muscles when using the robot during an anterolateral approach? Could excessive retraction have caused the same phenomenon? These results suggest that certain implant designs may not be appropriate for use with robotic canal preparation. Several retrospective studies have used the same robotic system but with different implants and have reported fewer perioperative surgical problems.

The initial application of an external fixation device to immobilize the femoral shaft is not a surgeon or patient-friendly experience. The increased prevalence of nerve palsies was especially concerning, but I would not expect this fixation system to be solely responsible for the higher rate of nerve palsies. The authors also report difficulties with using the pins for the registration technique. The current updated robotic system has addressed both of these issues and does not require such intensive external fixation or the application of pins for registration prior to the surgical procedure. These trends are very important as we try to move toward less invasive (and not more invasive) techniques for joint arthroplasty.

Overall, this is an important paper in the field of computer-assisted surgery and specifically for active robotic systems. This study sets a standard as one of the first prospective and randomized studies in this area and provides useful information for surgeons who may be evaluating the technology. Surgeons should be aware that "passive" navigation systems are currently being studied intensively both in research and in active clinical use and that new developments in these systems may overcome many of the limitations associated with the active robotic systems. However, whichever robotic systems are used, the navigation tools for both total hip and total knee replacement will need to be validated with prospective and randomized studies that critically examine patient outcome.

*The author did not receive grants or outside funding in support of his research or preparation of this manuscript. He did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author is affiliated or associated.