Metacarpophalangeal (MCP) arthritis is more frequently associated with rheumatoid arthritis than with osteoarthritis. There are distinct presentations and treatment options for the two different etiologies. This chapter discusses both diagnoses, their individual presentations, the unique surgical options for treatment of both diseases, and the postoperative rehabilitation required for a successful outcome.
Rheumatoid disease leads to a predictable ulnar drift deformity with volar subluxation at the MCP joints.
Rheumatoid arthritis is a systemic disease that demands attention to the entire patient.
In patients with rheumatoid disease, wrist deformity should generally be addressed prior to MCP arthroplasty.
Osteoarthritis often affects only one digit, whereas rheumatoid arthritis usually affects all four.
Arthroscopy is a useful surgical tool in the treatment of osteoarthritis in the MCP joint.
Surface replacement arthroplasty is useful in patients with osteoarthritis and closely recreates the joint mechanics.
Treat wrist deformities before hand deformities.
Repair the radial collateral ligament and centralize the extensor tendon in patients with rheumatoid arthritis.
Supervised postoperative hand therapy is critical.
Arthroscopy can be a useful treatment modality and staging tool.
During surface replacement arthroplasty use intramedullary cutting guides and intraoperative fluoroscopy (proper alignment is critical to a good outcome).
Arthritic disease of the metacarpophalangeal (MCP) joints in the hand is a potentially debilitating malady. Within this diagnosis, there are two distinct presentations with markedly different treatments. Rheumatoid arthritis (RA) is the more common presentation and can be differentiated by a distinct pattern of associated destruction and deformity when compared to osteoarthritis of the MCP joint. Primary osteoarthritis of the MCP joint, excluding the thumb, is much rarer and often secondary to posttraumatic effects of fracture or ligamentous instability.
Like the distinctive arthritic etiologies that affect this particular joint, there is an emerging unique dichotomy of surgical treatment options. Traditional treatment options have evolved from soft tissue interposition to hinged metallic implants and now generally involve some sort of silicone arthroplasty. Silicone implants are designed to act as a temporary dynamic spacer while the body creates a more permanent stabilizing encapsulation. No specific fixation is used and the implant freely pistons within the medullary canal. More recently, surface replacement arthroplasty, which uses techniques similar to joint replacements in other parts of the body, has emerged as a new tool in the treatment of MCP arthritis. As with hip and knee replacements, the goals of MCP surface replacement arthroplasty are to remove a minimal amount of bone and provide a new nonbiologic bearing surface. Because of the differences in presentation, classification, and treatment, this chapter will discuss rheumatoid arthritis and osteoarthritis separately and the roles of these two treatment options for both.
The thumb metacarpophalangeal joint is unique in the presentation and treatment of both rheumatoid arthritis and osteoarthritis. Thus, the evaluation, classification, and treatment of thumb MCP arthritis will also be discussed separately.
More so than any joint in the body, the MCP joint of the patient with rheumatoid arthritis is subject to a predictable pattern of deformity, subluxation, and destruction. The evolution of this deformity is multifactorial but originates with the pathognomic synovitis and tenosynovitis of RA that allows supportive periarticular soft tissues to erode, stretch, and remodel under the influences of chronic deforming forces. The factors that lead to the characteristic volar and ulnar subluxation of the rheumatic MCP joint include the following:
The native anatomy of the metacarpal head predisposes to ulnar deviation in flexion as a result of the larger radial condyle.
The tendency of the hand to fall into a Z -pattern deformity—the wrist falls into radial deviation as the carpal bones slide down the radial slope. The metacarpals follow the carpus and also point in a radial direction. The digits now “zigzag” in the opposite direction with ulnar deviation at the MCP joints.
Synovitis within the MCP joint leads to attenuation of the radial sagittal hood, further exacerbating ulnar drift and ulnar subluxation of the extensor apparatus.
As the extensor mechanism falls into the ulnar intermetacarpal space volar to the center of rotation of metacarpal head, the extensor tendon becomes a relative flexor causing volar subluxation of the proximal phalanx in addition to ongoing ulnar deviation.
Volar subluxation is accentuated by attrition of the volar plate and collateral ligaments caused by the continuing synovitis.
Rheumatoid changes within the interosseous muscle complex lead to intrinsic tightness, which exacerbates volar subluxation ( Fig. 14-1 ).
The evaluation of the rheumatoid hand can often be challenging and requires an appreciation of the differences among deformity, pain, and functional deficit. Patients with RA are often protective of the function that they do have and may be reluctant to “risk” surgery. This may in part result from a mismatch between the hand surgeon’s and the rheumatologist’s perceptions of the benefits of surgical intervention. Unfortunately, this often results in delayed referral and a “self-fulfilling” prophecy effect because advanced deformities, bone loss, and tissue changes limit the reconstructive possibilities and compromise the ultimate end result. Starting with a highly predictable surgery such as thumb MCP fusion is one strategy for gaining the patient’s trust.
Clinical assessment of the rheumatoid hand necessitates a more global evaluation of the patient. Because rheumatoid arthritis is a systemic disease, the involvement of multiple joints should be suspected. Lower extremity reconstruction typically should be performed before addressing the upper extremities. The reason for this is straightforward. The use of crutches associated with lower extremity surgery puts excessive and potential harmful strains on shoulder, elbow, and wrist arthroplasties. Bilaterality is common, as is proximal involvement of the wrist, elbow, shoulder, and cervical spine. The cervical spine, in particular, should be closely evaluated in any patient undergoing surgery. Subtle instability of the cervical spine, even subclinical, may become dangerous during intubation and positioning.
Before focusing on the hand, physical examination of the shoulder, elbow, and wrist must be performed. A reconstructed hand is not helpful if the patient cannot position the extremity in a plane of function. When focusing on the wrist, abnormal carpal alignment should be noted. As the supportive structures around the wrist become attenuated the carpal bones often slide down the radial/volar slope of the distal radial articular surface. This often results in radial deviation of the carpus and metacarpals. If present, this deviation must be addressed prior to the MCP joints (see Chapter 10 ). Additionally, the distal ulna should be evaluated, as should the presence of extensor tenosynovitis. A prominent distal ulna (and chronic tenosynovitis) can result in extensor tendon rupture, which must be addressed to restore active MCP motion.
As the examination finally moves to the hand and fingers, skin integrity, posturing, and function must all be assessed. Thin and fragile skin is not uncommon with patients with RA and creates the potential for wound healing problems. Poor nutrition may contribute to this potential problem and also should be considered and optimized. Next, general posturing, muscle atrophy, and passive range of motion of the digits need to be assessed. The mass effect of proliferative synovitis at either the elbow or wrist can cause a compressive neuropathy of the median, ulnar, or radial nerves resulting in weakness and atrophy that should be recognized and addressed before MCP reconstruction. Volar subluxation and ulnar deviation are the most common deformities of the MCP joints. Both swan neck and boutonnière deformities are common distal manifestations of the rheumatoid process. Often the treatment of the PIP deformity will either need to be staged or can precede any surgery at the MCP joint. Passive range of motion of the joints and the intrinsic tightness test ( Fig. 14-1 ) can identify contracted tissues (or muscles) that need to be released to achieve soft tissue balance. Finally, active range of motion including the functional range of motion of the MP joints and total composite finger flexion needs to be noted. A mismatch between passive and active ranges of motion requires further investigation. Possible etiologies include flexor tenosynovitis, extensor tendon ruptures, sagittal band attenuation with extensor tendon dislocation, or posterior interosseus nerve palsy. The Vaughn-Jackson syndrome is a condition of attritional rupture of the digital extensor tendons usually secondary to proliferative synovitis and prominent deformity of the ulna at the distal radioulnar joint. The ruptures typically begin ulnarly with the extensor digiti quinti and progress radially. If the fingers can be actively held in extension when passively extended but not actively extended from a flexed position, extensor tendon subluxation should be considered. Usually, the dislocated or subluxing extensor tendon slips off the metacarpal head in an ulnar-volar direction. This also adds a significant pathologic deforming force that contributes to digital ulnar drift and to MCP joint volar subluxation. This needs to be addressed usually at the same time as definitive treatment of the MCP joint arthritis. The rheumatoid thumb can develop several patterns of instability and deformity as will be described later in this chapter. Range of motion, posture, and instability of the thumb carpometacarpal (CMC) joint, MCP joint, and interphalangeal (IP) joint should all be specifically inspected.
Radiographs are a routine and necessary component in the evaluation of the rheumatoid hand. In most cases a simple three-view plain x-ray examination is all that is needed. Advanced imaging such as magnetic resonance imaging (MRI) will add little to the management. Attention should be directed to the extent of arthrosis in the MCP joints and remaining joints in the hand and wrist. Additionally, the remaining bone stock must be critically evaluated. In severe, advanced RA the bony erosions and severe osteopenia may contraindicate arthroplasty at the MCP joint.
When evaluating the rheumatoid hand, it is critical to consider the patient as a whole. What are the patient’s current functions and limitations? What additional joints are affected by the disease process? What are the patient’s expectations? No surgery will cure the patient with rheumatoid arthritis of their rheumatoid disease process, but focused surgical treatments of involved joints in the hand can drastically improve their function and quality of life.
Nalebuff’s original classification system was modified slightly by Stirrat to include radiographic changes, clinical deformity, and pain as factors. The system stratifies disease into categories of early, moderate, late, and very late.
|Early||None, synovitis only||Minimal or none||Minimal|
|Moderate||Early erosions in metacarpal head||Moderate extension lag or ulnar drift||Minimal|
|Late||Articular destruction and/or subluxation||Increased extension lag and ulnar drift, passively correctable||Moderate|
|Very Late||Joint destruction, volar dislocation||Fixed ulnar deformity and volar extensor tendon displacement||Severe|
Although this type of classification system may be useful for staging of disease, it does not significantly help guide the treating physician when choosing treatment. Clearly, patients with early disease are best managed nonoperatively. Stirrat has described preventative surgical procedures such as synovectomy and soft tissue balancing in patients at the moderate stage. Patients in late and very late stages are best suited to salvage procedures such as arthroplasty.
RA is largely a disease that is managed nonoperatively. Early disease may be managed by nighttime splinting and hand therapy, but daytime splinting is often too cumbersome to be practical. All patients with an established diagnosis of RA should be involved with a rheumatologist familiar with the ever-changing landscape of pharmacologic management of the disease. There are several stages of medical management of rheumatoid disease, including use of nonsteroidal anti-inflammatory drugs (NSAIDs); oral and injectable corticosteroids; disease-modifying antirheumatoid agents; and the newest biologic drugs such as Enbrel, Remicade, and Humira. Local injection of corticosteroid can be effective in reducing pain and inflammation from the tenosynovitis. However, the risk of tendon rupture is not insignificant and all patients should be counseled on this risk.
The indication for surgery in the patient with RA is disease that is not amenable to modern medical nonoperative treatment. This may present as isolated pain or deformity or a combination of both. Some authors have advocated for earlier surgical intervention in the form of synovectomy and extensor tendon realignment. However, this has not been shown to significantly alter the disease course. In general, for patients who have failed nonoperative treatment of RA affecting the MCP joints (other than the thumb), the gold standard remains silicone arthroplasty. The surgical indication for MCP arthroplasty in the patient with RA is essentially unremitting pain, deformity, or both causing poor function of the hand. A decreased total arc of motion of less than 40 degrees or severe ulnar drift of greater than 30 degrees associated with focal MCP pain have been proposed as objective criteria for proper patient selection, but as the following cases illustrate, any deformity that inhibits function should be considered. Figure 14-2 illustrates an extraordinary example of a patient with fixed dislocations of each MP joint resulting in greater than 90 degrees fixed flexion contractures and severe ulnar drift. As a result of the severe contractures, the patient was unable to even perform adequate hygiene. Although this patient had very little pain, the deformity and poor function severely limited her activities of daily life. Figure 14-3 illustrates a case in which a patient presented with fixed, painful volar subluxation of the phalanges and very poor range of motion but very little of the characteristic ulnar drift. In this patient’s presentation, pain was far more debilitating than deformity. Both of these patients suffered from rheumatoid arthritis, but they had different clinical presentations. Both had excellent although different indications for surgical intervention. It should be noted that the first patient’s expected clinical outcome is not as optimistic as the second’s because of the extensive deformity. A frank, honest discussion with the patient must be held preoperatively in this regard. Surgical contraindications in these types of patients would include lack of adequate bone stock, inadequate skin coverage, or the presence of untreated proximal disease.
Two good general rules exist when approaching surgical reconstruction of the upper extremity in the patient with RA. First, reconstruction should begin proximally and proceed distally. Untreated shoulder, elbow, and particularly wrist disease should in general be addressed prior to consideration of MCP arthroplasty. Hand function should be considered after the ability to place the hand in space is addressed. Implants placed in hands with uncorrected wrist deformity have a poorer survivorship than implants in hands with no wrist deformity. Second, in consideration of contiguous joints, treatment should generally alternate between arthrodesis and arthroplasty. Most specifically, when both severe wrist and MCP disease are present, the most appropriate surgical treatment would be wrist arthrodesis and MCP arthroplasty. If deformity is present distally in the fingers, then proximal interphalangeal (PIP) fusion or soft tissue correction of the deformity would be warranted (as opposed to MCP and PIP arthroplasties in the same digit).
The first MCP arthroplasties were originally described in the 1950s using soft tissue interposition techniques. Many different materials were used for the interposition including the volar plate, extensor retinaculum, and perichondrium. However, these procedures were deemed unpredictable at best. Concurrently, Flatt’s metal hinge prosthesis and Swanson’s silicone spacer were introduced in the mid 1960s. The metal hinge prosthesis had early failures as a result of loosening and implant fracture. The Swanson silicone spacer, however, quickly became the most common surgical treatment for MCP rheumatoid arthritis because it proved to be reliable and the most successful. In Swanson’s original design of the silicone arthroplasty, titanium grommets were included to be placed over the implant both proximally and distally to protect the implant from potential fracture. The data on grommets remain equivocal. No paper definitively demonstrates a significant difference in survivorship with the use of grommets. Since Swanson’s original design of a straight, flexible silicone spacer, there have been several newer implants with unique design elements to arrive on the market. Weiss and colleagues compared the common Swanson implant (Wright Medical, Memphis TN) with the newer Avanta (Avanta Orthopaedics, San Diego, CA) and Neuflex implants (DePuy Orthopaedics, Warsaw, IN). In a cadaveric model the newer prebent implants demonstrated far less pistoning than the original Swanson implant. The authors also demonstrated that the mechanics, as measured by the instantaneous center of rotation and overall tendon moment arm, were superior in the Neuflex and Avanta implants. Some clinical studies have demonstrated some minor improvements in digital flexion with the Neuflex implant compared to the Swanson, but no study has revealed any improvement in long-term function or survivorship of any one implant. The proper choice of implant at this point remains a surgeon’s preference. Surface replacement arthroplasty should be considered only with extreme caution in RA because these procedures rely heavily on good bone sock and intact collateral ligaments, both of which are likely to be compromised in the rheumatoid hand.
Incision and Exposure
Swanson originally described his technique for MCP arthroplasty in 1972, and to this day the technique remains relatively unchanged. The incision is usually transverse and extends directly dorsal over each metacarpal head, taking care to preserve the dorsal venous drainage system. Placement of the perfect incision can be aided by the following trick: the pen marking for the incision can be barely seen at the top of the crest on top of the metacarpal heads when viewed from the proximal direction and when viewed from the distal direction. Alternatively, some surgeons advocate a two-incision technique with one incision placed longitudinally between the index and long metacarpal heads and one placed between the ring and small metacarpal heads. Following the choice of skin incision, the extensor mechanism is exposed, the radial sagittal hood will be attenuated, and the tendons will often be found subluxed ulnar to the metacarpal head. Preservation of the venous system and careful handling of the skin cannot be overemphasized whichever approach is chosen.
Soft Tissue Balancing
Swanson originally advocated incising the hood on the ulnar aspect to assist mobilization of the extensor tendon. Beckenbaugh and Linscheid modified this by suggesting incising the radial hood. This author’s preferred technique is to incise the attenuated radial portion of the hood, which will be imbricated during closure, to help centralize the extensor tendon over the metacarpal head, although it is sometimes necessary to release the ulnar side also to adequately centralize the tendon. Following a dorsal longitudinal capsulotomy a complete synovectomy is then performed. A rongeur can be used to “pluck” the diseased tissue away. Following synovectomy the collateral ligaments are sharply elevated off their insertion on the metacarpal head. Release of the ulnar collateral ligament is necessary to correct the ulnar deformity, while the radial collateral ligament of each digit is tagged for later repair. At this point the Bunnel intrinsic tightness test is compared to the preoperative test. If it remains significantly positive, then the ulnar intrinsic tendon is incised. Ulnar intrinsic release helps correct severe ulnar drift deformity and may also improve swan neck deformity in the digit. As an alternative to simple intrinsic release some authors advocate the crossed intrinsic transfer. Straub originally described the technique of transferring the ulnar intrinsic tendon to the radial side of the adjacent digit. Some authors have shown this to help reduce recurrence of ulnar drift, whereas others have indicated it may decrease MCP flexion or possibly contribute to swan neck deformity within the digit.
Following exposure and contracture release the bony surfaces are prepared to receive the implant. It is easiest to begin on the metacarpal side because the metacarpal osteotomy will aid in exposure of the phalanx. Small Hohmann retractors are placed on either side of the metacarpal neck to protect the volar structures. The osteotomy is planned at the most distal extent of the metaphyseal flare of the metacarpal head, just distal to the insertion of the collateral ligaments. The osteotomy should be perpendicular to the long axis of the bone in both posteroanterior and lateral planes with a tendency toward a flexed cut to promote flexion. The cut surface is then smoothed with a bone rasp to remove any sharp edges that may compromise integrity of the implant. The metacarpal canal is then identified with a sharp awl under manual control. Sequential broaching with the rectangular broaches is then performed, keeping the broach and metacarpal both parallel to the floor. On occasion it is necessary to use a power burr to expand the medullary canal. This should only be performed with a Swanson tip burr with a rounded leading edge to prevent perforation of the cortex. Any other type burr tip must be used with extreme caution because perforation of the fragile cortex can easily occur. Following preparation of the metacarpal, the phalanx is exposed. If the joint space at this point is still limited and contracted, the volar plate may be released form the volar metacarpal neck. It is usually unnecessary to perform a formal osteotomy of the proximal phalanx, but any osteophytes should be removed to contour the phalanx into a flat surface. The phalangeal canal is again located with a sharp awl. As with the metacarpal, cautious use of the Swanson tip burr may be helpful to expand the proximal phalanx canal. Sequential broaching is then performed to allow placement of the largest possible trial implant without overstuffing the joint. Some authors recommend broaching the index phalanx in slight supination to promote a supinated position of the finger to facilitate pinch to the thumb tip after final implant position. This is performed by holding the digit slightly supinated while positioning the broach perpendicular to the floor.
Implant Selection and Placement
Following broaching the largest possible trial implant should be selected based on the largest broach used. The trial should easily fit into the medullary canals without force. With the trial in place the surgeon should then check for passive motion and passive correction of deformity. Overstuffing of the joint with an inappropriately large implant will lead to decreased motion and possibly increase the likelihood of implant fracture. If the stems of the implant fit well in both medullary canals but the passive range of motion is limited, then further resection of the metacarpal neck should increase the joint space and improve range of motion. When the final implant is selected it is placed with great care using smooth forceps.
Prior to final component implantation two holes are created in the radial metacarpal neck cortex of each digit with a 0.045-inch K-wire to facilitate repair of the radial collateral ligament. Double-armed nonabsorbable 4-0 suture is placed through these drill holes prior to final implant placement. This suture is then used to formally repair the radial collateral ligament as an additional step to prevent recurrent ulnar deformity. At this step the capsule is not always repairable secondary to attenuation, but if possible it is repaired with 4-0 suture reefing toward the radial side. The radial sagittal band is then imbricated (using a “pants-over-vest” technique) on closure again with 4-0 nonabsorbable suture. The ulnar sagittal band, if incised during exposure, is not closed. The joint should then be taken through a full range of motion to determine if the extensor tendon is adequately centralized. If the tendon still has a tendency to sublux, it can be sutured loosely to the underlying capsule. Alternatively, a strip of juncturae or a small strip of the radial tendon may be mobilized and sutured to the radial capsule. Following this repair, the passive range of motion should again be evaluated to confirm that the tendon has not been overly corrected. Skin closure is performed over a drain. Again, as during exposure, a meticulous “no touch” technique should be used because the rheumatoid skin is extremely fragile. Last, a bulky dressing and a short arm splint with the MCP joints supported in full extension should be placed.
Preserve dorsal venous drainage and sensory nerves
Ulnar intrinsic release when necessary
Repair of radial collateral ligament
Centralize extensor tendon
Early supervised splinting and therapy program
Meticulous “no touch” surgical technique to protect skin
Failure to correct wrist deformity prior to MCP surgery
Improper handling of fragile rheumatoid skin
Unrecognized breach of medullary canal leading to implant malposition