Is the Keller Bunionectomy Outdated?

By Mark Hofbauer, DPM, and Steven Kravitz, DPM

Yes, Mark Hofbauer, DPM, cites the limited indications of the procedure. He says the first MPJ fusion is a better option for hallux abductovalgus patients as it provides enhanced stability and biomechanical effects. For the past 10 to 15 years, there has been a great debate over whether the Keller bunionectomy has become an outdated procedure. Proponents on both sides have been able to make solid arguments. The Keller bunionectomy began to get a black eye years ago when surgeons stretched the indications for its use. This was due in part because it was an easy, quick procedure to perform. As time has passed, the indications have narrowed significantly, not necessarily to the point of extinction, but at least to the endangered list. When deciding on a procedure for hallux abductus valgus repair, one has to weigh many considerations including joint destruction versus joint preservation, the patient’s activity level and his or her chance for the best long-term results. The reason for failed hallux abductovalgus surgery is almost always poor procedure selection. Having said that, we need to focus on what procedures will provide the best long-term results for the patient. Notice that I did not say “what procedure works best in my hands” or make statements such as “People have been doing Kellers for years” or “All of my patients with Kellers do great.” These types of statements and thought processes only slow the progression of a more scientifically based process toward accurate procedure selection for patients with a hallux abductovalgus deformity. What’s the first question we should ask when evaluating a patient with hallux abductovalgus? Is a joint preservation procedure or joint destructive procedure indicated? The answer in this case is a joint destructive procedure. The second question is: What joint destructive procedure will provide the patient with the most predictable, functional, best long-term result? When answering this question for almost all patients, the answer is first metatarsal phalangeal joint (MPJ) fusion, not a Keller bunionectomy. Understanding The Limitations Of Keller Bunionectomies And The Strengths Of First MPJ Fusions If we take a look at the indications for a Keller bunionectomy, they are limited at best. Essentially, you would use this procedure for an elderly, apropulsive patient with bump pain. Period. How many of these types of patients walk into your office every day? Very few. Many studies have shown that the Keller bunionectomy has withstood the test of time. This may be true to some extent, but how many of these studies have looked at patients five years down the road? Very few. How many of these studies have looked at whether the patient may have done even better with a different procedure? None. Our job is to look at history with a critical eye, expand upon our knowledge base, learn new technologies and strive for the best procedure for the patient. When we look at the articles written over the years concerning first ray stabilization in the rheumatoid arthritis patient, this is where we begin to learn of the Keller bunionectomy’s lack of predictability and lack of stabilizing force as opposed to that which is provided by the first metatarsal phalangeal joint fusion. When examining this subgroup of patients, we see that patients score much higher following first metatarsal phalangeal joint fusion (in comparison to the Keller bunionectomy) with regard to stride length, push-off power, pain relief, cosmeses, lesser metatarsalgia and activity level. This is due in part to the stabilizing effect of the first ray with a fusion versus the instability of the intrinsics about the first metatarsal phalangeal joint created by a Keller bunionectomy. One needs only to review the exhaustive studies of first ray pathology and anatomy, precisely those of Dudley Morton, to understand the importance of stabilizing and balancing the forefoot. Many of you will argue that one can reestablish the intrinsics about the first MPJ following a Keller bunionectomy by suturing the intrinsics to the base of the remaining phalanx. However, the studies show that push-off power is lost after a Keller bunionectomy. It is the same concept as to why lesser digital fusions do better than arthroplasties. Fusions create more stability and convert the soft tissues and intrinsics into stabilizing forces. The biomechanical effects of the first MPJ fusion (including reduction of the intermetatarsal angle, conversion of the conjoined tendon to an active correcting force and the repositioning of the first ray in the sagittal plane) over time are ultimately what proves the procedure’s superiority over the Keller bunionectomy. Final Notes The versatility of the first metatarsal phalangeal joint fusion versus the Keller bunionectomy should in and of itself make any foot and ankle surgeon question the need to perform the Keller bunionectomy. When selecting a procedure for hallux abductovalgus repair, remember, the question should not be what would work for this patient, but what might work better than any other procedure for this patient. Sigvard Hansen, MD, probably said it best when he noted, “Reestablishment of the weightbearing axis, equal weightbearing distribution through the six weightbearing points of the forefoot, proper balancing of weight between the hindfoot and the forefoot, and participation by the toes in forefoot weightbearing in mid- to late stance and push-off should be the minimal goals of surgery for hallux valgus. This approach requires much more analysis and surgical expertise than cutting of a bump, but it is a logical approach that achieves consistently good results.”1 The goal should always be to restore normal forefoot balance and function. This can only be achieved by stabilizing the first ray. Indeed, history has proven that when it comes to choosing a joint destructive procedure, the first MPJ fusion works better than a Keller bunionectomy. Dr. Hofbauer is a Diplomate of the American Board of Podiatric Surgery and a Fellow of the American College of Foot And Ankle Surgeons. He is on the staff of the Foot And Ankle Institute of Western Pennsylvania and has a private practice in McMurray, Pa. References 1. Hansen, S. “The First Ray,” Clinics In Podiatric Medicine and Surgery. July 1996. Editor’s Note: Dr. Hofbauer previously wrote the Counterpoint article for the feature, “Bunions: Are Proximal Osteotomies Necessary?,” which appeared in the August 2002 issue of Podiatry Today. No, Steven Kravitz, DPM, says the procedure is a proven treatment for those who have end-stage arthrosis. Using case study examples, he says the procedure is not limited to the geriatric population and also points out key modifications. The Keller procedure (hallucal proximal phalangeal base resection arthroplasty) works as well today as it has in the past and should be considered a staple in the treatment armamentarium for forefoot surgery. It is the alternative to the cosmetically improved result that one may attain with an implant arthroplasty. The Keller procedure can be used on any adult who has the appropriate pathology. While I agree it is most commonly used on geriatric patients, one should not assume that it is exclusively reserved for that age group. Selecting a joint destructive procedure for the first metatarsophalangeal joint (MPJ) is based more upon other factors than simply age. Considerations include the type of ambulatory activity the patient will participate in, the health of the patient, and the cosmetic result associated with a short first toe. The Keller stands as a tried and true, effective treatment for a patient with end-stage arthrosis. The health of the patient aside, there is no doubt that a joint salvage procedure is always preferred over any joint destructive procedure. That being said, there are circumstances in which joint salvage is not an option. When facing a situation that requires the selection of a joint destructive procedure, the Keller procedure should be one of several surgeries you consider. It is a simple, fast-healing, post-op ambulatory procedure that has a decreased risk of complications due to the absence of foreign body material that is inherent with implant arthroplasty. To help illustrate these concepts, let’s consider the following case studies. Case Study One: When A 33-Year-Old Construction Worker Presents With Traumatic Hallux Limitus The patient is a 33-year-old construction worker who lays down cinder block for basements and wall construction of commercial buildings, etc. He presents with a typical stage four traumatically induced hallux limitus of the right foot. In this classification system, stage four pertains to “end stage” arthrosis. In this case, he has less than 20 degrees total range of motion to the first MPJ along with inflammatory arthritis, which is a common finding in end stage arthrosis. Passively applying a minimal amount of motion to the hallux causes significant discomfort with the patient non-weightbearing. Radiographs show an absence of joint space, with typical medial and lateral flaring of the joint interface, which is often described as “trumpeting” of the phalangeal base and metatarsal head. A dorsal exostosis is noted on the first metatarsal head. This patient’s history indicates the pathology is the result of an accident in which a cinderblock had been dropped directly on the first MPJ of the right foot several years ago without medical follow-up. The apparent diagnosis is traumatic arthritis secondary to a suspected intraarticular fracture, a crush injury to the right first MPJ. Aside from smoking, the rest of this patient’s medical history is clear. When Job-Related Ambulatory Demand Rules Out Implant Use This presents a very challenging case. When inquiring about the patient’s employment, it became clear that when he lays the initial two or three levels of cinder block, he kneels on his right knee with his right foot out behind him. His lower leg is parallel to the ground and the first MPJ is maximally dorsiflexed to its end range of motion. There was concern that a digital fusion would have overstressed the hallux interphalangeal joint while the patient was in this position. It also would potentially cause significant pressure to the distal end of the hallux if dorsiflexion of the hallux was eliminated via fusion. Therefore, we did not consider arthrodesis to be an optimal option. We also questioned if an implant would be able to withstand the consistent dorsiflexionary end range of motion demand placed upon the prosthesis during the patient’s employment. We were concerned that a hinged prosthesis would have a high likelihood of ultimate failure and the non-constrained devices also were not designed to withstand that type of use. We explained to the patient that implant prosthesis was a viable option but because of the demand placed upon the foot during his employment, there was a higher risk of failure for him as opposed to other patients with a more average ambulatory demand. With a full explanation to the patient and two second opinions, the patient opted for the Keller arthroplasty. The patient has done well. Approximately 10 years have passed since the surgery and he has had no complications. The procedure was modified with an abductor tendon transfer and the patient is using functional orthoses, both of which are described in the following case study. Case Study Two: When Re-Implantation Is Not Viable A 54-year-old male with a silastic hemi-prosthesis presents with detritic synovitis and associated destruction of the remaining portion of the proximal phalanx to the degree that re-implantation is not viable. The procedure selected is a hemi-implant resection, essentially leaving what would otherwise be called a Keller procedure result. We added an abductor hallucis tendon transfer to the procedure in an attempt to decrease first ray hypermobility. During the surgery, we performed careful dissection proximally in order to identify the distal portions of the abductor hallucis muscle and associated tendon. The muscle tendon was moderately atrophied from lack of usage but we were still able to isolate and transfer the tendon to the first metatarsal head with an osseous anchor. After the surgery, we emphasized orthoses with a first MPJ (Morton’s) extension to increase load to this area of the ball of the foot. The patient did not have any post-op complications and was actually able to do aggressive walking four to five miles a day in a “power walking” regimen. He had done so for years and this may have been one of the factors that accelerated the failure of the implant in this case by increasing the weightbearing load to the joint. I could argue that another procedure may have had a better long-term prognosis due to the patient’s active lifestyle. That aside, the fact that the patient was able to carry out the “power walking” activity postoperatively following implant removal and abductor hallucis tendon transfer suggests that the implant served solely as a joint spacer and did not provide any biomechanical advantage to enhance joint function and propulsive gait. That said, it should be noted that we cautioned the patient preoperatively about potential complications with implant removal, especially related to his power walking activity. However, he experienced no deficit in function but rather an increased ability to perform power walking due to resolution of the pain associated with the preoperative detritic arthritis. He had already accommodated to power walking with a hemi-prosthesis for the previous six or seven years and had no negative sequelae transitioning to the foot without the prosthetic device. This emphasizes the point that MPJ prostheses primarily serve as joint spacers and have minimal, if any, ability to improve the functional deficit that occurs when one removes the proximal phalangeal base. How Key Modifications To The Keller Can Reduce The Risk Of Post-Op Sequelae The Keller procedure has had many modifications described by various authors. I have found all of these to be helpful. For hallux abductus surgery, Ganley described transplanting the resected extensor hallucis brevis to the medial side of the joint capsule as an “autogenous graft” in order to reinforce the medial capsular fibers and decrease the possibility of developing postoperative hallux abductus positioning. Guido LaPorta, DPM, was the first one to describe the abductor hallucis tendon transfer to me. This procedure involves attaching the tendon with an osseous anchor to the medial aspect of the first metatarsal head where the medial epicondyle is located or would have been located. Keep in mind that it is commonly removed during most first MPJ arthroplasties as a consequence of medial eminence resection. If one cannot identify the abductor hallucis tendon, one may harvest the tibial sesamoid ligament for attachment as it is an extension of the abductor hallucis muscle tendon. A biomechanical analysis of this small strap muscle demonstrates that it crosses the first ray axis at 45 degrees and distant to it so that the muscle has a plantarflexory lever arm on the ray. I believe this has been very effective in decreasing first hypermobility that has been described as a sequelae to the Keller arthroplasty. Last but not least, when you are treating active people with the Keller arthroplasty, you can assist them postoperatively via the use of orthoses with Morton’s extensions. By increasing weightbearing to the first MPJ, these orthoses decrease transfer lesser metatarsalgia and add some degree of propulsive function to the first ray. I also attach the long flexor hallucis tendon to the under surface of the remaining proximal portion to assist toe purchase and decrease the possibility of post-op, cock-up hallux. Employing a simple drill hole and the suture of your preference works well. Final Notes These simple modifications can enhance the positive postoperative yield of the Keller arthroplasty. Postoperative orthotic usage is also a key adjunct to the overall management of the patient. In conclusion, first MPJ implant prostheses simply provide the benefits of increased toe length and a better cosmetic result. However, the simple Keller arthroplasty offers a potential for less complications in that it is less affected by osteoporosis and other considerations that can impact upon the viability of a prosthetic implant. It also allows for immediate postoperative weightbearing. While I agree the Keller procedure is most commonly used for the geriatric population, it need not be exclusively used for that age group. The overall health of the patient, the quality of osseous tissues, and the postoperative ambulatory expectation for the patient are all factors that one should take into consideration. Dr. Kravitz is a Fellow of the American College of Foot and Ankle Surgeons and is the Executive Director of the American Professional Wound Care Association. References 1. Kravitz SR, LaPorta GA, Lawton, JH: KLL progressive staging classiification of hallux limitus and hallux rigiidus. The Lower Extremity 1:55, 1994. 2. Root ML, Orien WP, Weed JH: Normal and Abnormal Function of the Foot Vol. II, Clinical Biomechanics Corp, Los Angeles, 1977. 3. Drago JJ, Oloff L, Lacobs AM: Conprehensive review of hallux limitus. J Foot Surg 23:213-220, 1984. 4.Vanore JA, Christensen CC, Kravitz SR, et al. Clinical Practice Guideline Diagnosis and Treatment of First Metatarsophalangeal Joint Disorders. J Foot Ankle Surg 42:3, 2003. 5. Shereff MJ, Bejjani FJ, Kummer FJ: Kinematics of the first metatarsophalangeal joint. J Bone S Surg Am 68:392,1986. 6. Vanore IV, Corey SV. Hallux limitus, rigidus, and metatarso-phalangeal joint arthrosis. In: Marcinko DE, ed. Comprehensive Textbook of Hallux Abducto Rigidus Reconstruction. St. Louis: Mosby Year Book; 1992: chap 10. 7. Banks AS, McGlamry ED: Hallux limitus and rigidus. Pp. 600-616. In McGlamry ED (ed). Comprehensive textbook of foot surgery 2nd edition. Williams and Wilkins, Baltimore, MD, 1992. 8. Root ML, Orien WP, Weed JH: Normal and Abnormal Function of the Foot Vol. II, Clinical Biomechanics Corp, Los Angeles, 1977. 9. Brunet, J.A. Pathomechanics of complex dislocations of the first metatarsophalangeal joint. Clin Orthop. 332:126-131, 1996. 10. Guido LaPorta, DPM, verbal communication. 11. James Ganley, verbal communication.

Add new comment