Hand Surgery
1st Edition

Osteoarthritis of the Hand and Digits: Distal and Proximal Interphalangeal Joints
Thomas W. Kiesler
The most common form of arthritis, osteoarthritis (OA), is a condition that is characterized by articular cartilage deterioration. Involvement in the hand is seen most commonly at the thumb carpometacarpal joint and at the interphalangeal (IP) joints. The cartilage changes are manifested by joint enlargement, pain, swelling, stiffness, contracture, and angular deformity. Nonoperative therapy is the cornerstone of treatment. In recalcitrant cases, however, surgical treatment is appropriate to provide pain relief and joint stability. IP joint débridement, mucous cyst excision, arthroplasty, and arthrodesis are the mainstays of surgical treatment.
The proximal interphalangeal joint (PIPJ) and the distal interphalangeal joint (DIPJ) are simple diarthroidal ginglymus (hinge) joints, which essentially allow motion in only one plane. Primary lateral stability is achieved by the strong radial and ulnar collateral ligaments. The bony architecture of the opposing proximal and distal bones provides secondary constraint (1). The biconvex distal surface articulates with the reciprocal proximal biconcave condylar surface. The central intercondylar ridge of the proximal bone interlocks with the distal central ridge. The bony and ligamentous anatomy together impart the necessary stability for pinch activities.
Besides stability, the PIPJ and DIPJ allow a wide range of motion for gripping activities. The normal sagittal plane arcs of range of motion at the PIPJ and the DIPJ are 0 to 105 degrees and 0 to 85 degrees, respectively. Functional arcs of motion, however, are only 36 to 86 degrees (with an average of 60 degrees) and 20 to 61 degrees (with an average of 39 degrees) at the PIPJ and the DIPJ, respectively (2). The value of the PIPJ to overall hand function is clearly evident by the fact that it contributes 85% of intrinsic digital flexion and 30% of the overall combined flexion of the finger (3).
Classic OA is a noninflammatory primary cartilage disease that is characterized by progressive articular cartilage deterioration and reactive new bone formation. A less common type of OA, erosive or inflammatory OA, involves the DIPJs and PIPJs of postmenopausal women. It has a more abrupt onset and more swelling and tenderness than classic OA (4). Posttraumatic OA is another subset of OA that has a clearly defined etiology: traumatic intraarticular cartilage disruption. Regardless of the type of OA, the pathogenesis is similar. Age, systemic factors, genetics, and trauma can predispose one to the development of OA. However, the specific joint that is affected and the severity of the disease are usually dictated by local biomechanical factors (5).
Articular cartilage is composed of water, collagen, proteoglycan, and chondrocytes. The biochemical changes that are seen in the cartilage of aging joints are increased water content, decreased protein content, increased collagen stiffness, and decreased proteoglycan mass, size, and proportion. Chondrocytes become larger, acquire increased lysosomal activity, and fail to reproduce. The combined overall effect of these changes is a loss of cartilage strength and elasticity (6). However, changes that are found in cartilage from symptomatic older osteoarthritic patients are different than those that are seen in the cartilage of older asymptomatic patients.
Pathophysiologically, the initiating event in cartilage degradation that leads to OA has been theorized to be mechanical stress that leads to altered chondrocyte metabolism, production of matrix metalloproteinases, and disruption of collagen matrix properties (7). As a process, therefore, OA is biochemically mediated. It is the local mechanical factors, however, that most likely initiate and perpetuate the process; the process begins with cartilage microfracture and fibrillation and ends in complete bone eburnation (5).
As the articular cartilage degenerates, the subchondral bone reacts to the abnormal load by cyst formation, sclerosis,

and marginal osteophyte formation (Fig. 1). Most commonly, the narrowing and wear of the joint space are asymmetric, thus resulting in an angular deformity. Joint stability and mechanics can be adversely affected, further altering the stresses across the joint and perpetuating the osteoarthritic process. Joint enlargement, joint incongruity, and pericapsular contracture result in loss of motion in the flexion-extension arc.
FIGURE 1. Interphalangeal joint osteoarthritis with osteophyte formation, subchondral sclerosis, and cyst formation.
The pathophysiology of mucous cysts, small ganglion cysts at the DIPJ, is much less well understood. Although not considered a clear etiologic factor in the production of the mucous cyst, OA has been reported to be an associated finding in nearly 80% to 100% of cases (8,9). Kleinert et al. (10) have suggested that the lesion arose from the joint capsule in the vicinity of a dorsal osteophyte, as they found a definite pedicle that led to the joint capsule of the DIPJ in all cases that were reported. As the cyst fills with fluid and enlarges, it can place pressure on the germinal matrix of the nail bed and can cause longitudinal grooving of the nail plate. Large cysts can cause thinning of the overlying skin.
FIGURE 2. Malaligned osteoarthritic interphalangeal joints.
OA is considered a universal problem of humans. There appears to be a heritable component in its distribution within the population (11). It is seen in at least one joint by the age of 75 years and has become the second most common cause of disability in adults in the United States (7). Evaluation of the prevalence of OA among 17 specific populations showed quite a wide variation, with Alaskan Eskimos who are older than 40 years of age having the lowest prevalence for both men (22%) and women (24%). The mean prevalence of the entire study group for persons 35 years of age and older was found to be 60% for men and women (11). The symptoms that are associated with classic OA often begin in the fifth to sixth decade of life. Women are more often affected than men of the same age.
History and Physical Examination
Clinically, the most common patient complaint is the insidious onset of joint pain and stiffness that interferes with function of the digit or hand, making pinch and grip activities difficult. The joint symptoms are initially activity related but may progress to the extent that they are noticed at rest and at night. In some cases, however, cosmetic concerns may be the sole reason that a patient presents to the surgeon’s office. Patients may complain of a “crooked,” malaligned joint (Fig. 2) or may give a history of having noticed painless periarticular joint enlargement; Heberdens’s nodes (Fig. 3) and Bouchard’s nodes are noted at the DIPJ and the PIPJ, respectively. Inspection of the osteoarthritic DIPJ usually shows some degree of Heberden’s nodes along the radial and ulnar aspects of the joint. Angular deformity can range from minimal to quite dramatic.

Large mucous cysts that are adjacent to the joint can cause longitudinal grooving of the nail plate (Fig. 4).
FIGURE 3. Large Heberden’s node around an osteoarthritic distal interphalangeal joint.
FIGURE 4. Mucous cyst that is adjacent to an osteoarthritic joint. Nail plate grooving may result if the cyst places pressure on the germinal matrix.
Similarly, the PIPJ usually shows some sign of enlargement. Cyst formation is less likely but is not uncommon (Fig. 5). More chronic PIPJ OA oftentimes shows some degree of angular deformity. Palpation of the joint line at the PIPJ or DIPJ elicits tenderness and may demonstrate an effusion. Radial and ulnar deviation stresses may demonstrate varying degrees of instability, depending on the severity of the disease. In chronic cases, range of motion is generally diminished compared to the unaffected joints and occasionally displays a coarse crepitus. Pinch and grip strength are often decreased due to discomfort or instability, or both.
Diagnostic Studies
Plain radiographs are generally all that is needed to confirm the diagnosis of an osteoarthritic joint. Dedicated, true, orthogonal posteroanterior, lateral, and oblique views are required to accurately evaluate the joint line. Magnified views are sometimes helpful to identify subtle abnormalities in small digits. Asymmetric joint space narrowing (Fig. 6) is the earliest radiographic finding, followed later by marginal osteophyte formation (Fig. 7). Subchondral cyst formation, broadening of the base of the phalanx, and subchondral sclerosis are the hallmark findings of advanced OA at the DIPJ and the PIPJ (Fig. 8) (12).
FIGURE 5. Large proximal interphalangeal joint periarticular cyst in an osteoarthritic joint. Note the angular deformity and joint space loss.
FIGURE 6. Asymmetric joint space narrowing at the proximal interphalangeal joint.
Radiographs that demonstrate more symmetric joint space narrowing, erosions, and considerable destruction may indicate an inflammatory disease process, such as rheumatoid arthritis, psoriatic arthritis, gout, or Reiter’s syndrome. In this case, laboratory tests, including erythrocyte sedimentation rate, rheumatoid factor, uric acid level, and complete blood count, should be obtained to further evaluate such a process.
FIGURE 7. Volar and dorsal osteophyte formation at the thumb interphalangeal joint.

FIGURE 8. Findings of advanced osteoarthritic interphalangeal joint: angular deformity, broadening of the base of the phalanx, and cyst formation.
The natural history of OA is usually benign. Because there is currently no medical therapy for the underlying disease process in OA, treatment is symptomatic only. Consequently, patient education is crucial. Treatment can be appropriately administered only if the surgeon accurately comprehends the patient’s pain level and the disparity between the patient’s current level of activity and the desired level of activity. After confirmation that an inflammatory arthropathy is not present, the patient can generally be assured that, although pain control may at times be a formidable task, complete disability, as is sometimes seen in RA, is unusual. A thorough explanation of the status of the joint disease and the treatment options, along with their respective risks and benefits, educates the patient and allows him or her to actively participate in his or her care plan. Nonoperative means are nearly always used first.
The primary goal of nonoperative treatment is pain relief. The resulting increased function of the affected hand is usually a result of adequate pain control. Initially, symptomatic treatment includes controlling exposure to provocative activities that produce pain, swelling, and stiffness. This includes modification of work and leisure activities. Completely eliminating these activities is most important during acute flare-ups. Formal hand therapy can be used to instruct patients on joint protection techniques, especially in tip-to-tip pulp pinch and lateral key pinch (13). Therapy is also useful to teach techniques for edema control and the use of adaptive devices, as well as gentle range-of-motion and strengthening exercises. Preservation of motion may aid in the maintenance of hand function. Splinting can be used to immobilize a tender joint during a flare-up, to provide rest at night or after activities, or to provide support and protection during activities.
Medical treatment includes analgesic use, which is usually adequate if symptoms are mild and episodic. Acetaminophen use has been shown to be as effective as nonsteroidal antiinflammatory drug (NSAID) treatment (14). NSAID use is popular and can decrease symptoms. The side effects that are reported with prolonged use, especially in elderly patients, demand judicious use of these drugs. NSAID use is contraindicated in patients with recent or active peptic ulcer disease, bleeding diathesis, congestive heart failure, renal insufficiency, or current anticoagulation use.
If symptoms persist, intraarticular steroid injections may be administered. First introduced in 1954, intraarticular steroids can give temporary symptomatic relief during acute flare-ups but unfortunately cannot reverse the osteoarthritic process (15). Because of its relatively low solubility, triamcinolone preparations deliver the best long-term results (16). Because the steroid is administered locally, the systemic complications that are seen with oral steroids are rare. However, local complications can include subcutaneous fat atrophy and depigmentation. Joint sepsis can occur from improper technique. Some feel that repeated injections can lead to further joint destruction by inhibiting production of chondroitin sulfate (12) or could cause attenuation of the central tendon at the PIPJ, leading to a boutonnière deformity (16).
Mucous Cyst Excision
Mucous cyst excision is indicated in cysts that have created longitudinal grooving of the nail plate and in large cysts that have caused significant thinning of the overlying skin and risk rupture. Ruptured cysts create a direct path from the skin into the joint, setting the stage for bacterial contamination and the development of septic arthritis. Simple incision and drainage or aspiration have not been shown to effectively eradicate these lesions and may even promote infection of the cyst.
Early surgical techniques relied on radical cyst excision with skin grafting (8,17). Later techniques recommended cyst excision combined with special attention to osteophyte excision, joint débridement, and wound closure by local rotation flap or simple repair (9,10). Kleinert et al. (10) had no recurrences at 12- to 18-month follow-up, whereas Eaton et al. (9) reported only one recurrence in 50 cases, with follow-up ranging from 6 months to 10 years. All nail plate deformities resolved within 6 months. Both studies reported recovery of preoperative range of motion and no complications. Moreover, both studies clearly demonstrate that concomitant osteophyte excision and joint débridement are the critical adjuncts to cyst excision to minimize cyst recurrence.
Preoperatively, patients must thoroughly understand that mucous cyst excision and joint débridement do nothing to treat the underlying disease process of OA. Any expectations that this procedure alone can ameliorate the remainder of their joint symptoms or halt the osteoarthritic disease process must be dispelled.

FIGURE 9. A: Preferred rotation flap skin incision for mucous cyst excision in cases in which the skin is severely thinned or in which the cyst obliterates the eponychial fold. Note that the excised cyst has been triangulated. B: Elevated rotation flap that exposes the base of the nail plate, the terminal extensor tendon insertion, and the distal interphalangeal joint. C: The flap is sutured into place after rotation into the defect.
Author’s Preferred Technique
After obtaining adequate general or local anesthetic, the affected upper extremity is placed outstretched on an arm board and is sterilely prepped and draped. The extremity or digit is exsanguinated, and a pneumatic brachial or digital elastic drain tourniquet is set. The incision that the author chooses depends on whether primary wound closure is likely to be possible. Rotation flap coverage should be considered in situations in which the cyst is extremely large and the overlying skin is extremely thinned or in which the cyst involves the eponychium and its excision leaves a defect with exposed nail bed. If any or both of these situations are present, the author uses a small rotation flap technique that has been described by Atasoy (18). This incision allows for access to the radial and ulnar aspects of the DIPJ and allows adequate wound coverage (Fig. 9A). The cyst and overlying skin are triangulated and excised sharply down to the paratenon and germinal matrix. The full-thickness flap is developed and raised off the paratenon for later rotation and insetting (Fig. 9B).
If excision of the cyst does not likely require flap coverage, the author uses one of the standard approaches to the DIPJ (Fig. 10). The author’s preference is the Y-shaped incision. This incision usually allows placement of one of the distal arms directly over the cyst and again allows reliable access to the radial and ulnar aspects of the joint for débridement. Irrespective of the initial incision used, full-thickness skin flaps are raised by using an atraumatic technique to expose the terminal tendon and its margins. As the dissection continues distally, the cyst wall is encountered. The plane between the overlying skin and the cyst is developed, and the cyst is dissected free from the surrounding tissues. Care is taken to avoid damaging the germinal matrix at the distal extent of the incision. A longitudinal incision is made along the radial and ulnar margins of the terminal tendon, extending from the joint capsule proximally. The cyst and its stalk are traced to their origin in the joint capsule at the interval between the collateral ligaments and the terminal tendon. With the collateral ligaments protected, the cyst, its stalk, and the adjacent dorsal capsule are excised. A blunt elevator is used to retract the terminal tendon, and the joint is thoroughly inspected. All marginal osteophytes and remaining dorsal capsule around the joint

are removed with a small rongeur. Intraoperative fluoroscopy may be advantageous to confirm adequate débridement. The tourniquet is released, and hemostasis is obtained. The wound is irrigated, and the skin is sutured with 4-0 nonabsorbable sutures. If a local flap was used, it is rotated into position to cover the defect (Fig. 9C). A bulky finger dressing is applied, maintaining full extension at the DIPJ with a gutter splint.
FIGURE 10. Standard approaches to the distal interphalangeal joint for mucus cyst excision, arthrodesis, or arthroplasty.
Postoperatively, the dressing and sutures are removed at 10 to 14 days. Edema control and scar massage are initiated, as are active, active assist, and gentle passive range-of-motion exercises. A gutter splint is fashioned for wear between exercise periods if any extensor lag is present. Strengthening exercises may be started at 3 to 4 weeks postoperatively.
Distal Interphalangeal Joint Arthrodesis
Arthrodesis is the most common form of surgical treatment for OA of the DIPJ once nonoperative management is no longer effective. The patient’s functional needs must be carefully assessed. DIPJ arthrodesis is indicated when pain, instability, or malalignment has become so severe that it interferes with function of the digit or hand, especially with pinch activities. Arthrodesis improves appearance, corrects deformity and instability, and, as a result of pain relief, increases strength and function. The trade-off, cessation of joint range of motion at the DIPJ, is generally not considered to be a severe functional limitation. An alternative to arthrodesis, DIPJ arthroplasty, is considered in persons whose vocation or avocation specifically requires maintenance of DIPJ flexion.
The goal of any arthrodesis is a stable bony union in a proper position within a reasonable period of time (19). Many techniques of varying complexity have been described for IP arthrodeses (DIPJ and PIPJ), each similar in that the remaining articular cartilage and subchondral bone are removed, and the joint is placed in the desired position and is held in this position until union is achieved. Moreover, common to each of the techniques is the tenet that careful preparation of the bone ends is critical. The bone ends are prepared in a “cup and cone” configuration (20) or with flat bony cuts (21) by using an oscillating saw or rongeur. The techniques differ in the shape of skin incision, the approach to the joint, the bone surface preparation, the bone fixation, and the use of bone graft. Although bone graft has been proposed as an internal structural support (19,22), supplemental graft is generally not required in the osteoarthritic patient. Severe deformity rarely, however, may mandate its use to correct malalignment.
Fixation options for DIPJ arthrodesis include a single Kirschner wire (K-wire) (23), crossed K-wires (24), external dynamic compression (25), an interosseous wire loop (21,26), a combination interosseous wire and intramedullary fixation (27), and compression screws (28,29 and 30).
Comparisons of the fixation techniques for DIPJ fusion have been performed (31,32). Engel et al. (31) found compression screw fixation to afford quicker return to work and significantly less lengthy immobilization time when compared to K-wires alone. Nonunion occurred in three cases in each treatment group. Time to union was equal. Wyrsch et al. (32) evaluated the biomechanical characteristics of the Herbert screw (Zimmer, Warsaw, IN) and tension-band wire fusion techniques in a cadaveric study. The Herbert screw was found to have superior anteroposterior bending strength, greater torsional rigidity, and similar lateral bending strength, all characteristics that may be clinically important in small joint arthrodesis.
The complications that are associated with DIPJ fusion have been reviewed by Stern and Fulton (33). They evaluated 185 DIPJ and thumb IP joint arthrodeses using crossed K-wires, the Herbert screw, or combination of K-wire and interosseous wire fixation. The highest overall complication rate occurred in the psoriatic arthritis cohort (44%). Patients who were surgically treated for an osteoarthritic joint had the second lowest overall rate of complications (21%), just slightly greater than chronic posttraumatic arthritis (17%). Of the 61 patients with OA, major complications included nonunions (eight), malunions (three), deep infection (one), and osteomyelitis (one). Nonunion rate among the three techniques was not statistically significant, 11% to 12%. Minor complications included dorsal skin necrosis (4%), cold intolerance (3%), paresthesias (3%), permanent PIPJ stiffness (3%), and superficial wound infection (2%).
Author’s Preferred Technique
A Y-shaped dorsal incision is made, centered at the DIPJ. The terminal extensor tendon is identified. Using an atraumatic technique, full-thickness skin flaps are raised and retracted with stay sutures. The terminal extensor tendon is transversely incised just proximal to the DIPJ level. Sharp subperiosteal dissection is used to elevate the margins of the capsule, while osteophytes are removed with a small rongeur. Care must be taken not to damage the proximal extent of the germinal matrix that is located at the extensor tendon insertion. The collateral ligaments are released from their proximal attachments radially and ulnarly. The joint is hyperflexed, allowing access and complete visualization of the bony surfaces. A small oscillating saw or rongeur is used to remove all remaining cartilage and subchondral bone from the opposing surfaces down to cancellous bone. Whether the fusion is to be in slight flexion or in a neutral position, meticulous mating of the bony surfaces is critical for the appropriate coronal, sagittal, and rotational alignment. Once alignment is finalized, fixation is applied.
The desired position of fusion dictates the type of fixation. For fusions in slight flexion (usually as much as 20 degrees), the author recommends K-wire fixation. Place

two to three 0.035-in. K-wires antegrade through the base of the distal phalanx, exiting at the tip pulp, just inferior to the nail plate. Place the first wire longitudinally, and confirm with fluoroscopy that it is centered within the distal phalanx. Place the remaining wires at an angle through an adjacent starting hole. Reduce the bony surfaces accurately, and drive the longitudinal wire retrograde into the middle phalanx (Fig. 11). Adjust rotational alignment, and, with compression maintained at the fusion site, drive the remaining wires across the fusion site into the middle phalanx. The wires should cross within the middle phalanx, not at the fusion site. Confirm accurate alignment and bony opposition visually and radiographically. Ensure that no distraction exists at the fusion site. The tourniquet is released, and hemostasis is obtained. The extensor mechanism and skin are closed separately. A well-padded finger dressing is placed. The wires may be cut beneath the skin or left above the skin level and capped.
FIGURE 11. Distal interphalangeal joint arthrodesis using Kirschner wires (k-wires).
Internal screw fixation, Herbert screw, or Acutak screw (Acumed, Beaverton, OR) is the implant of choice when fusion in neutral position is chosen. However, adequate bone stock and sufficient cross-sectional area to contain the screw must be present (33). Care must be taken to choose a screw with a distal thread diameter that is small enough not to violate or fracture the distal phalanx. Mini screws may be required in small phalanges.
The author has found it helpful to make a pilot hole in the distal and middle phalanges with a K-wire before using the Herbert hand drill. This helps prevent accidental cortical penetration and nailbed injury and assists in obtaining accurate provisional alignment. Under radiographic control, a 0.035-in. K-wire is placed just palmar to the nail and is advanced retrograde from the tip of the distal phalanx into the base. After ensuring that the wire is well centered within the distal phalanx in anteroposterior and lateral planes, the fusion site is reduced, and the wire is driven into the middle phalanx, creating the pilot hole. The K-wire is removed. A small stab incision is made in the fingertip at the site of K-wire insertion. The main Herbert drill (smaller) is placed retrograde through the pilot hole in the distal phalanx. Similarly, the drill is passed retrograde through the middle phalanx. The small Herbert drill is removed. The larger Herbert drill is then placed retrograde through the tip of the distal phalanx only.
A Herbert screw of appropriate size to extend from the distal phalanx tuft to the cortical isthmus of the middle phalanx is chosen. The screw is advanced retrograde to a depth that completely buries the trailing threads in the distal phalanx. The leading threads should stop at a point just proximal to the middle phalanx neck. Appropriate rotational alignment is confirmed by comparing the plane of the adjacent nail plates to that of the fused digit. Coaptation at the fusion site should be confirmed radiographically and by direct vision (Fig. 12). The tourniquet is released, and hemostasis is obtained. The extensor mechanism and skin are closed separately with nonabsorbable sutures. A bulky finger dressing is applied.
Postoperatively, sutures are removed at 10 to 14 days. Aggressive active, active assist, and passive range of motion of the PIPJ and the metacarpophalangeal (MCP) joint are started. The arthrodesis site is protected with an Orthoplast gutter splint. Appropriate edema control is initiated. Union is monitored radiographically. If K-wires are used, they can usually be removed by 6 to 10 weeks after surgery. Strengthening is then initiated.
FIGURE 12. Intraoperative radiograph of thumb interphalangeal joint fusion that shows accurate coaptation of the bony surfaces after placement of the internal screw fixation.

Distal Interphalangeal Joint Arthroplasty
An alternative to arthrodesis, DIPJ arthroplasty, can also be used to eliminate pain at the arthritic DIPJ. Its advantage over arthrodesis is the maintenance of some degree of motion at the DIPJ for pinch and grip activities. Postoperatively, range of motion has been reported to average 30 to 40 degrees, with an average extensor lag of 12 degrees (34,35 and 36). DIPJ arthrodesis sacrifices the fine fingertip control that is modulated by the action of the flexor digitorum profundus (37). DIPJ arthroplasty is usually, however, only rarely used. Most authors agree that DIPJ arthroplasty is indicated for patients whose occupations or avocations mandate the maintenance of some degree of DIPJ motion to perform fine manipulative activities (34,35,38). To be considered, however, sufficient bone stock must be present, and the flexor-extensor mechanism must be intact and functional. Relative indications include patients with severe OA involvement in multiple digits of the same hand or patients with an adjacent PIPJ fusion on the same digit (35,37).
The advantage of retained motion must be weighed against the potential complications of DIPJ arthroplasty. Postoperative lateral instability, long-term implant survival, pinch power, skin erosion, and implant breakage from repetitive stress are areas of concern. Wilgus (37) found lateral stability to be satisfactory in all 38 DIPJ arthroplasties at an average follow-up of 10 years. Forty-three percent were stable to lateral stress, 52% showed some instability with a definite end point, and one implant was judged to be grossly unstable. Seventy-one percent were found to have improved power. One implant eroded through the skin, one was removed for infection, and one was found to be broken postoperatively at 30 months. Similarly, Brown (34) reviewed his short-term results in 21 arthroplasties with an average follow-up of 26 months. He judged all implants to be stable. One implant required removal owing to skin erosion; however, no implant breakage was reported.
FIGURE 13. Arthroplasty approach to the distal interphalangeal joint requires takedown of the terminal extensor tendon or one of the collateral ligaments.
Operative Techniques
An approach to the DIPJ for implant arthroplasty requires adequate exposure for accurate preparation of the bony surfaces, osteophyte removal, and implant placement. Unlike the PIPJ, the anatomy of the DIPJ does not allow direct complete joint access without transection of the terminal extensor tendon (34,36,37) or detachment of one of the collateral ligaments (Fig. 13), (35,38). As a result, both approaches require a period of strict immobilization postoperatively to allow for healing of the periarticular tissues before starting range-of-motion exercises.
Author’s Preferred Technique
A Y-shaped dorsal skin incision is used. Full-thickness skin flaps are raised radially and ulnarly, using an atraumatic technique. Care is taken not to damage the germinal matrix distally. A longitudinal incision is made along the radial and ulnar margins of the terminal extensor tendon. The tendon is gently freed from the middle phalanx by using a small elevator. The author prefers to approach the DIPJ by incising the terminal extensor tendon 5 mm proximal to the joint (36,37). This approach not only spares the insertion of the terminal tendon, it preserves the stability of the joint by not violating the integrity of the collateral ligaments. The tendon is reflected distally, and marginal osteophytes are débrided. A small oscillating saw is then used to remove the head of the middle phalanx distal to the collateral ligament insertions (Fig. 14). This cut must be perpendicular to the long axis of the phalanx in the sagittal and coronal planes. Awls are used to identify the intramedullary canals proximally and distally. Radiographic confirmation that the awls are well centered within the canals is advised. The canals are enlarged with broaches to allow for the appropriate

implant. If necessary, a power bur may be used to allow acceptance of the implant. Care must be taken not to violate the volar or dorsal cortices. The trial prosthesis is inserted. The implant should fit flush with the bone ends, without buckling. Once accurate fit is confirmed, the permanent prosthesis is inserted by using a no-touch technique. With the joint in full extension, the terminal tendon is repaired by using 4-0 nonabsorbable suture (Fig. 15). While maintaining the joint in full extension, lateral stability of the implant is assessed. Under radiographic control, a 0.035-in. K-wire is passed retrograde from the tip of the distal phalanx into the volar portion of the flexor sheath to maintain extension at the joint. Care must be taken to avoid the implant. The tourniquet is released, and hemostasis is obtained. The skin is closed, and a bulky finger dressing with a volar gutter extension splint is applied.
FIGURE 14. After terminal tenotomy, the head of the middle phalanx is removed just distal to the collateral ligament origin.
FIGURE 15. With the prosthetic implant in position, the extensor tendon is repaired with nonabsorbable suture.
Postoperatively, the bulky dressing and sutures are removed at 2 weeks. Edema control is initiated. The DIPJ is supported continuously in full extension with a volar gutter splint, such as that used for treatment of a mallet injury (39). Most critical at this juncture is initiation of active, active assist, and passive range-of-motion exercises to the PIPJ and the MCP joint. The K-wire is removed at 4 weeks; however, immobilization at the DIPJ is maintained full time for another 4 weeks (8 weeks total) with the gutter splint (36,37). At 8 weeks, gradual active motion is initiated. If an extensor lag is noted during this time, a short period of full-time splinting is restarted. Resistive exercises with putty can usually be started at 10 to 12 weeks (40). Night splinting is continued for a total of 3 to 4 months.
Similar to DIPJ OA, surgical management of the end-stage osteoarthritic PIPJ is indicated when all nonoperative treatments have proven ineffective in providing pain relief in activities of daily living or occupational pursuits. If there is persistent unrelenting pain, gross malalignment, loss of motion, and complete cartilage loss, surgical treatment should be considered. The goal of surgery is to provide pain relief and to improve function and appearance. Definitive surgical treatment at the PIPJ, however, is the least well defined in the management of the osteoarthritic hand (41). Although arthrodesis and joint arthroplasty are the currently acceptable treatment options, determining which procedure is most appropriate is still an unsolved issue. The ideal surgical procedure for the osteoarthritic PIPJ would provide stability for lateral and key pinch in the radial digits and should maintain or attempt to restore the flexion arc in the ulnar digits for gripping activities.
Literature abounds that describes the numerous techniques and results of PIPJ arthrodesis that provide excellent stability, albeit at the cost of eliminating motion (19,20 and 21,28,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61 and 62). Similarly, the evolution and results of several motion-sparing PIPJ arthroplasty techniques have been well chronicled. These include perichondral resurfacing arthroplasty (63,64) and tendon or capsule interposition (65), silicone flexible implant (66,67), and prosthetic replacement (68,69,70,71,72,73,74,75 and 76). Studies that directly compare arthrodesis to arthroplasty and those that assess treatment solely in the osteoarthritic PIPJ are rare. As a result, interpretation of the individual series is difficult, and treatment recommendations are often based on extrapolations to the osteoarthritic patient from the rheumatoid and posttraumatic arthritic populations.
There is general agreement that lateral stability for pinch activities is of paramount importance at the index finger PIPJ. Although in most cases, arthrodesis of the index PIPJ is the most appropriate surgical procedure to accomplish this goal (77), Millender and Nalebuff (78) believed that arthroplasty was also appropriate in the index finger if the joint is well aligned with good collateral ligaments. In a 1987 review of their results, Stern and Ho (16) cautioned against the use of silicone implants in the PIPJs of the radial digits owing to concern that stresses on the radial collateral ligament during key pinch may lead to implant failure. As a result, they recommended against implant arthroplasty and favored PIPJ arthrodesis in all of the following circumstances of PIPJ OA: (a) index finger OA, (b) a young, active patient, (c) loss of bone stock with or without angulation in the coronal plane, (d) preoperative motion of less than 30 degrees, and (e) failed implant arthroplasty.
In an effort to clarify treatment recommendations, Pellagrini and Burton (41) reviewed their series of 43 procedures at the PIPJ, 83% of which were osteoarthritic, treated with arthrodesis, silicone interposition arthroplasty, or cemented arthroplasty. At an average follow-up of 4 years, silicone implants demonstrated a mean active range of motion of 56 degrees, an extensor lag of 3 degrees, and periarticular erosions in 27% of cases. At 2-year follow-up, the mean range of motion of the Biomeric arthroplasties was 66.4

degrees. However, 71% required revision due to mechanical failure of the elastomer hinge. When compared with arthrodesis in the radial digits, grip and pinch strength in the arthroplasty patients was less likely to equal or to exceed the contralateral hand. They recommended against the use of implant arthroplasty in the radial digits, and proposed arthrodesis in 30 to 40 degrees of flexion for the index finger PIPJ (and occasionally in the long finger PIPJ if it is also used predominantly for pinch activities) as the procedure of choice. Silicone implant arthroplasty was recommended only for the ulnar digits of older patients.
Although arthrodesis is generally well tolerated in the radial digits, arthrodesis of the PIPJ in the ulnar digits severely impairs the hand’s ability to grip and can limit motion in adjacent digits owing to the quadraegia effect. Therefore, motion-preserving procedures should be given strong consideration, especially in the ulnar digits. However, before embarking on surgical treatment of the osteoarthritic PIPJ, one must be cautious and must consider the patient’s age and functional, vocational, and recreational needs, as well as which digit is involved. It is mandatory that all pertinent factors and their ramifications be discussed with the patient.
Proximal Interphalangeal Joint Arthrodesis
PIPJ arthrodesis is recommended for end-stage OA of the index PIPJ and in the ulnar digits when the need for stability outweighs that of range of motion. Early techniques relied on the use of bone pegs to provide stability until fusion occurred (19,22). Later, K-wires alone were used in conjunction with various bone preparation techniques: straight cuts (42), chevron bone cuts (43), cup and cone (20,44,45), convex-concave (46), or tenon (47). Concerns about distraction at the arthrodesis site with K-wires alone led to use of compression clamp K-wire fixation (25,48) or intraosseous wiring alone (21) or in combination with K-wires (26,49). Lister (26) reported a nonunion rate of 9% when using intraosseous wiring that was supplemented with K-wire fixation. In 1980, Allende and Engelem (50) reported on another internal fixation compression technique, tension band wiring, which was originally described by Segmuller (51). The tension band principle, conversion of the dorsal tension forces into palmar compressive forces at the bone ends, has proven effective in PIPJ arthrodesis. Given the stability of the construct, postoperative splinting is short-term, and early range of motion of adjacent digits is possible, theoretically preventing stiffness. Nonunion rates range from only 0% to 5% (50,52,53,54 and 55). Other minor complications of tension band wiring for PIPJ arthrodesis, which were reported by Stern et al. (55), include prominent hardware that requires removal (9%) and superficial infections (10).
Kovach et al. (79) performed an in vitro biomechanical analysis that compared crossed K-wires, intraosseous wiring, and tension band wiring at the PIPJ. They found tension band wiring to be superior in strength in anteroposterior bending and torsion. No significant difference was detected with regard to lateral bending. Ijsselstein et al. (56) retrospectively evaluated the clinical results of a series of patients who were treated with percutaneous K-wires or tension band wiring. Combining the results at the MCP joint, the PIPJ, and the DIPJ, they reported a 15% rearthrodesis rate and an infection rate of 18% in the K-wire group, compared to a 5% rearthrodesis rate and a 2% infection rate in the tension band wiring group. They believed that the disadvantages of longer operating time, more extensive dissection, and the possibility of hardware irritation were far outweighed by the stability and lower complication rates of tension band wiring.
Compression screw fixation for small joint arthrodesis was first introduced in 1984 by Faithfull and Herbert (28), who postulated that the compression that was afforded by the screw would decrease nonunion rates even further. They reported a 100% fusion rate at all joints in the digit. Similarly, a high fusion rate was reported by Ayres et al. (52) (98%) and Katzman et al. (57) (100%), with average time to union at the PIPJ being 6 and 8.1 weeks, respectively. Ayres et al. (52) reported that two (4%) of the Herbert screws were removed owing to painful prominence, and fracture of the dorsal cortical bridge occurred in four patients (7%). They believe that these complications (which occurred early in the series) are avoidable if one pays careful attention to detail in preparing the entrance hole at the proximal phalanx, not attempting to change rotational alignment as the distal threads engage, and ensuring that the screw is fully seated. Finally, Leibovic and Strickland (58) evaluated 224 PIPJ arthrodeses that were performed for various diagnoses and were secured with a number of different techniques. Average time to radiographic union, as judged by the presence of trabeculae crossing the fusion site, was most rapid with the Herbert screw (9 weeks) as compared to K-wire (10 weeks) and tension-band wire (11 weeks). Nonunion occurred in 21% of K-wire fusions and 4.5% of tension band fusions and was nonexistent in Herbert screw fusions. No nonunions were seen in the subset of osteoarthritic patients, and no cortical fractures occurred with placement of the screw. They emphasized making the dorsal starting hole well proximal from the arthrodesis site and then overdrilling the hole before screw insertion to avoid fracture.
Internal fixation with plates and screws (59,60) and external fixation devices (61,62) are reserved for special situations in which there has been extraordinary bone loss. Because this is generally not the case in the osteoarthritic PIPJ, these methods are rarely used in primary fusion.
Author’s Preferred Technique
The author prefers the tension band technique for PIPJ arthro-desis. A dorsal curvilinear incision is made, centered at the PIPJ. Full-thickness skin flaps are raised. The central slip is incised in the midline distally through its insertion. Alternatively, the joint may be approached via a Chamay incision,

creating a distally based flap of central tendon (80). The extensor tendon is reflected along with the joint capsule, protecting the lateral bands. The collateral ligaments are released from their attachments at the proximal phalanx, and the joint is hyperflexed. Osteophytes are removed with a rongeur. The articular surfaces are resected with an oscillating saw. The desired amount of flexion must be chosen before making the cuts. The index and middle fingers are best served with a fusion angle of 15 to 30 degrees to optimize their position for pinching activities. The ulnar digits, which are more involved in gripping activities, are more functional in 30 to 45 degrees of flexion (81). Both surfaces are cut, each at a slight angle, the sum of which produces the total amount of flexion required for the digit involved. Precise cuts are required to ensure that no angular or rotational malalignment exists.
Two parallel K-wires (0.035 or 0.045 in., depending on the size of the joint) are driven retrograde from the cut surface of the proximal phalanx to exit the dorsal shaft. The arthrodesis site is compressed with the surgeon’s fingers, and one of the K-wires is driven antegrade. The K-wires are advanced distally along the intramedullary canal or, for added fixation, are angled volarly to exit the volar cortex of the middle phalanx (81). After the first wire is advanced, rotational alignment and coaptation at the arthrodesis site are assessed. Small adjustments are made before fixing the final position with the second K-wire. Next, a 0.035-in. K-wire is driven transversely across the middle phalanx 1 cm distal to the arthrodesis site. This drill hole must be positioned dorsal to the axis of rotation. A strand of 25- or 26-gauge wire is passed through the hole distally and then around the base of the pins proximally in a figure-of-eight fashion. The wire is tightened with a needle driver in the standard fashion, compressing the arthrodesis site. The wires are bent slightly, cut close to the bone, and are rotated toward the bone surface to minimize irritation (Fig. 16). The wire is similarly cut and is positioned close to the surface of the bone. The extensor mechanism and skin are closed separately with nonabsorbable sutures. The digit is placed into a sterile bulky safe position dressing.
Postoperatively, the dressing can be changed within 5 to 7 days. The stability of the tension band construct allows adjacent joints to be mobilized immediately. Edema control measures are instituted. A custom-molded gutter splint is applied to the PIPJ. Sutures are removed at 10 to 14 days. The splint is continued until stress testing no longer elicits pain. Resistive exercises are started at approximately 8 to 12 weeks after surgery, after clinical union is achieved.
Proximal Interphalangeal Joint Arthroplasty
Many arthroplasty techniques have been proposed to preserve motion and function at the PIPJ as an alternative to arthrodesis. Resection arthroplasty and palmar plate interposition with or without flexor digitorum superficialis tenodesis and perichondral grafting are biologic arthroplasty techniques that use autogenous tissue (63,64 and 65). Carroll and Taber (65) reported on 30 patients who were treated with palmar plate interposition, of whom only 16 had satisfactory results. Lateral stability was problematic. Ostegaard and Weilby (63) combined palmar plate advancement with flexor digitorum superficialis tenodesis. They reported adequate stability and range of motion in three patients, one with OA. Seradge et al. (64) reported on 20 patients who underwent resurfacing arthroplasty of the PIPJ using the perichondrium of the sixth or seventh rib. They felt that their results were less than acceptable and recommended avoiding the procedure in arthropathies that resulted from healed pyarthrosis, in systemic diseases with joint involvement, in patients who require concomitant tendon reconstruction, or in patients who are older than 40 years of age.
FIGURE 16. Approaches to the proximal interphalangeal joint: longitudinal central tendon splitting and the Chamay approach, in which the triangular distally based flap of the central tendon is raised and reflected distally.
Prosthetic replacement has been extensively used at the PIPJ. Early prosthetic replacement with constrained hinged metal-metal or metal-plastic designs (82,83 and 84) met with considerable complications that were related to subsidence, cortical protrusion, loosening, implant failure, and metallosis. Swanson developed the silicone rubber flexible implant and began implanting the prosthesis in 1966. Since that time, it has been the most popular replacement material at the PIPJ and is considered by many to be the gold standard for PIPJ arthroplasty. In 1985, Swanson et al. (66) reviewed 424 joints, 153 of which were osteoarthritic, with an average follow-up of 5.14 years. Overall, 98.3% of patients had complete pain relief. Postoperatively, range of motion improved in all diagnoses, with 68% of the osteoarthritic

joints having greater than 40 degrees of motion. Complications included implant fracture (5.19%), ulnar deviation (lateral instability) greater than 5 to 10 degrees (3.7%), infection (0.36%), and implant dislocation (0.36%). The revision rate was 10.9%. Survivorship of the Swanson implant has been studied by Iselin and Conti (67) in posttraumatic arthritis. They found the arthroplasty to be quite durable, and it was successful in 91% of joints at 2 years, 87% at 5 and 7 years, and 81% at 9 years.
FIGURE 17. Proximal interphalangeal joint arthrodesis, using a tension band technique.
Concerns about clinical instability, particulate synovitis, bone remodeling, and implant fracture have prompted development of prostheses that are anchored to the bone itself: the osseointegrated titanium-silicone implant (68) and the unconstrained cemented surface replacement arthroplasty (69). Moller et al. (68) reported that osseous integration of the prosthesis occurred in 41 of 44 fixtures. The prosthesis was abandoned in its current design, however, owing to an unacceptably high (18%) rate of silastic spacer fracture. Linscheid et al. (69) recently reviewed their results of using a cemented surface replacement arthroplasty at the PIPJ in 47 patients, 24 of whom had OA. At an average follow-up of 4.5 years, average postoperative range of motion measured 14 to 61 degrees (an arc of 47 degrees), a gain of 12 degrees compared to preoperative measurements. The average extension deficit was 14 degrees. The best results occurred in those patients with OA. Instability was seen in only five joints, all with bone and capsular loss that was associated with rheumatoid or posttraumatic arthritis, none with OA. Specifically, prosthetic replacement of the index PIPJ finger was believed to tolerate well the lateral stresses that were associated with pinch. There was no evidence of radiographic loosening. Overall, complications were seen in 19 of 66 joints.
FIGURE 18. Volar approach to the proximal interphalangeal joint for replacement arthroplasty. The flexor sheath is reflected, allowing visualization of the flexor tendons. A3, third annular pulley.
FIGURE 19. After palmar plate release and collateral ligament recession from the proximal attachment, the joint is hyperextended into the “shotgun” position, exposing the joint surfaces.
Operative Techniques
Implant arthroplasty at the PIPJ can be performed from a lateral, dorsal, or volar approach (66,69,75,76). The lateral approach alleviates the need to violate the extensor mechanism; however, one collateral ligament and the

volar plate must be detached (69). Lindscheid et al. (69) believed the lateral approach posed difficulties in preparing the intramedullary canals and obtaining consistent alignment of the components and subsequently abandoned its use. The dorsal approach involves a longitudinal split of the central tendon (66), proximal central tendon release, central tendon insertion detachment (69), or development of a distally based flap of central tendon (80). Depending on the extent of violation of the extensor tendon mechanism, initiation of postoperative range-of-motion exercises may need to be delayed. Although the volar approach requires release of the volar plate and collateral ligaments, it spares the flexor and extensor tendon mechanisms. This allows for immediate active range-of-motion exercises (75,76). The need for collateral ligament repair is controversial (75,76,85).
FIGURE 20. The head of the proximal phalanx is removed by using an oscillating saw. The collateral ligaments are spared.
FIGURE 21. Permanent proximal interphalangeal joint implant insertion.
Author’s Preferred Technique
The author prefers the volar approach that was described by Schneider (75) for flexible silicone arthroplasty of the osteoarthritic PIPJ. The preferred anesthetic is an intermetacarpal nerve block and intravenous sedation. This allows for intraoperative assessment of range of motion. A wrist tourniquet is inflated after exsanguination. A volar Bruner incision is made, centered at the PIPJ. The radial and ulnar neurovascular bundles are identified, and a full-thickness skin flap is raised off the flexor sheath and is retracted with a suture. The flexor sheath is incised transversely just distal to the second annular pulley and just proximal to the fourth annular pulley and is reflected laterally (Fig. 17). A small elastic drain is used to retract the flexor tendons.
The palmar plate is released from the proximal phalanx, and the collateral ligaments are recessed from their proximal attachment (Fig. 18). The joint is “shotgunned” open with hyperextension, exposing the joint surfaces (Fig. 19). Osteophytes are removed with a rongeur. An oscillating saw is used to resect the proximal phalanx condyles perpendicular to long axis of the bone, removing only enough bone to allow for the thickness of the implant (Fig. 20). The base of the middle phalanx is not resected. The intramedullary canals are entered with the awl, and each is enlarged with broaches until the appropriate size is obtained. Care must be taken to ensure that the broaches are placed squarely within the canal to prevent rotational malalignment. The trial implant is placed, and active range of motion and stability are assessed. The permanent implant is inserted (Fig. 21). Radiographs are obtained to ensure proper alignment and placement of the implant (Fig. 22). The collateral ligaments are not repaired unless there is considerable

instability. If needed, they can be reattached through drill holes in the proximal phalanx. The reflected portion of the flexor sheath is transposed deep to the flexor tendons and is reattached in such a manner as to resurface the volar portion of the joint. The tourniquet is released, and hemostasis is obtained. The skin is closed with 4-0 nonabsorbable suture. A bulky dressing with dorsal blocking splint is applied.
FIGURE 22. Posteroanterior (A) and lateral (B) postoperative radiographs of proximal interphalangeal joint arthroplasty.
FIGURE 23. Postoperative extension gutter splint. Note the edema control.
Postoperatively, the patient returns in 2 to 3 days for dressing change. A gutter splint is fashioned, holding the digit in full extension (Fig. 23). The lateral sides of the splint may be made higher to prevent medial and lateral stresses to the joint. Active and passive range-of-motion exercises are performed four times per day (Fig. 24). The splint is worn between exercise periods and at night. At 5 to 7 days postoperatively, dorsal taping or dynamic flexion splinting may be initiated if passive flexion is less than 70 degrees, provided that there is no extensor lag that is greater than 30 degrees. Sutures are removed at 2 weeks. At 3 weeks, exercises are increased to each hour, provided that there is no extensor lag that is greater than 25 degrees. Gentle strengthening is initiated at 8 weeks, beginning with a soft ball and progressing to putty. The patient should be weaned from the splint, and the splint should be discontinued by 12 weeks (86).
FIGURE 24. A,B: Postoperative active range-of-motion exercises for proximal interphalangeal joint arthroplasty.
1. Kiefahber TR, Stern PJ, Grood ES. Lateral stability of the proximal interphalangeal joint. J Hand Surg 1986;11:661–669.
2. Hume MC, Gellman H, McKellop H, et al. Functional range of motion of the joints of the hand. J Hand Surg 1990;15A(2):240–243.
3. Littler JW, Herndon JH, Thompson JS. Examination of the hand. In: Reconstructive plastic surgery, vol. 6, 2nd ed. Converse JM, ed. Philadelphia: WB Saunders, 1977:2971–2974.
4. Belhorn LR, Hess EV. Erosive osteoarthritis. Semin Arthritis Rheum 1993;22:298–306.
5. Dieppe P. Osteoarthritis and related disorders. In: Klippel JH, Dieppe P, ed. Rheumatology, 2nd ed. London: Mosby, 1988:8:11–12.
6. Miller MD. Review of orthopaedics. Philadelphia: Lippincott Williams & Wilkins, 1992.
7. Fife RS. Osteoarthritis; epidemiology, pathology, and pathogenesis. In: Klippel JH, ed. Primer on rheumatic diseases, 11th ed. Atlanta: Arthritis Foundation, 1997:216–218.
8. Constant E, Royer JR, Pollard RJ, et al. Mucous cysts of the fingers. Plast Reconstr Surg 1969;43:241–246.

9. Eaton RG, Dobranski AI, Littler JW. Marginal osteophyte excision in treatment of mucous cysts. J Bone Joint Surg 1973:55A:570–574.
10. Kleinert HE, Kutz JE, Fishman JH, et al. Etiology of the so-called mucous cyst of the finger. J Bone Joint Surg 1972;54:1455–1458.
11. Peyron JG. Epidemiologic and etiologic approach to osteoarthritis. Semin Arthritis Rheum 1979;8:288–306.
12. Swanson AB, Swanson G. Osteoarthritis of the hand. J Hand Surg 1983;8:669–675.
13. Linscheid RL, Dobyns J, Beckenbaugh R, et al. Mayo Clin Proc 1979;54:227–240.
14. Bradley JD, Brandt KD, Katz BP, et al. Comparison of an anti-inflammatory dose of ibuprofen, an analgesic dose of ibuprofen and acetaminophen in the treatment of patients with osteoarthritis of the knee. N Engl J Med 1991;325:87–91.
15. McGrath MH. Local steroid therapy of the hand. J Hand Surg 1984;9:915–921.
16. Stern PJ, Ho S. Osteoarthritis of the proximal interphalangeal joint. Hand Clin 1987;3:405–412.
17. Bourns HK, Sanerkin NG. Mucoid lesions of the fingers and toes. Clinical features and pathogenesis. Br J Surg 1963;50:860–866.
18. Atasoy E, O’Neill WL. Local flap coverage about the hand. Atlas Hand Clin 1998;3:179–234.
19. Moberg E. Arthrodesis of finger joints. Surg Clin North Am 1960;40:465–470.
20. Carroll RE, Hill NA. Small joint arthrodesis in hand reconstruction. J Bone Joint Surg 1969;51:19–12.
21. Robertson DC. The fusion of interphalangeal joints. Can J Surg 1964;7:433–437.
22. Potenza AD. Brief note: a technique for arthrodesis of finger joints. J Bone Joint Surg 1973;55:1534–1536.
23. Nemethi CE. Phalangeal fractures treated by open reduction and Kirschner wire fixation. Indust Med 1954;23:148.
24. Bunnell S. Joints. In: Boyes J, ed. Surgery of the hand, 4th ed. Philadelphia: JB Lippincott Co, 1948:320–324.
25. Braun RM, Rhoades CE. Dynamic compression for small bone arthrodesis. J Hand Surg 1985;10:340–343.
26. Lister G. Intraosseous wiring of the digital skeleton. J Hand Surg 1978;3:427–435.
27. Zavitsanos G, Watkins F, Britton E, et al. Distal interphalangeal arthrodesis using intramedullary and interosseous fixation. Hand Surg 1999;4:51–55.
28. Faithfull DK, Herbert TJ. Small joint fusions of the hand using the Herbert bone screw. J Hand Surg 1984;9:167–168.
29. Bednar MS. Distal interphalangeal joint fusion. Atlas Hand Clin 1988;3:1–16.
30. Teoh LC, Yeo SJ, Singh I. Interphalangeal joint arthrodesis with oblique placement of an AO lag screw. J Hand Surg 1994;19:208–211.
31. Engel J, Tsur H, Farin I. A comparison between K-wire and compression screw fixation after arthrodesis of the distal interphalangeal joint. Plast Reconstr Surg 1977;60:611–614.
32. Wyrsch B, Dawson J, Aufranc S, et al. Distal interphalangeal joint arthrodesis comparing tension band wire and Herbert screw: a biomechanical and dimensional analysis. J Hand Surg 1996;21:438–443.
33. Stern P, Fulton D. Distal interphalangeal joint arthrodesis: an analysis of complications. J Hand Surg 1992;17:1139–1145.
34. Brown LG. Distal interphalangeal joint flexible implant arthroplasty. J Hand Surg 1989;14:653–656.
35. Culver JE, Fleegler EJ. Osteoarthritis of the distal interphalangeal joint. Hand Clin 1987;3:385–401.
36. Zimmerman NB, Suhey PV, Clark GL, et al. Silicone interpositional arthroplasty of the distal interphalangeal joint. J Hand Surg 1989;14:882–887.
37. Wilgus EF. Distal interphalangeal joint silicone interpositional arthroplasty of the hand. Clin Orthop 1997;342:38–41.
38. Snow JW, Boyes JG, Greider JL. Implant arthroplasty of the distal interphalangeal joint of the finger for osteoarthritis. Plast Reconstr Surg 1977;60:558–560.
39. Swanson AB, Leonard JB, Swanson G. Implant resection arthroplasty of the finger joints. Hand Clin 1986;2:107–117.
40. Schwartz DA, Peimer CA. Distal interphalangeal joint implant arthroplasty in a musician. J Hand Ther 1998;11:49–52.
41. Pellagrini VD, Burton RI. Osteoarthritis of the proximal interphalangeal joint of the hand: arthroplasty or fusion? J Hand Surg 1990;15:194–208.
42. Burton RI, Margles SW, Lunseth PA. Small joint arthrodesis in the hand. J Hand Surg 1986;11:678–682.
43. Pribyl CR, Omer GE, McGinty L. Effectiveness of the chevron arthrodesis in small joints of the hand. J Hand Surg 1996;21:1052–1058.
44. Das GA, Belskey MR. Arthrodesis of the proximal interphalangeal joint with K-wire technique. In: Blair WF, ed. Hand surgery techniques. Baltimore: Williams & Wilkins, 1996:816–823.
45. McGlynn JT. Smith RA, Bogumill GP. Arthrodesis of the small joint of the hand: a rapid and effective technique. J Hand Surg 1988;13:595–599.
46. Watson HK, Shaffer, SR. Concave-convex arthrodesis in joints of the hand. Plast Reconstr Surg 1970;46:368–371.
47. Lewis RC, Nordyke MD, Tenny JR. The tenon method of small joint arthrodesis in the hand. J Hand Surg 1986;11:567–569.
48. Breitbart AS, Blat PM, Staffenberg DA, et al. An experimental study of small-joint compression arthrodesis. Ann Plastic Surg 1997;39:47–52.
49. McGrath JC, Vigil DV, Cohen MJ. A modified approach to cup-and-cone arthrodesis of the small joints of the hand. Contemp Orthop 1996;32:335–339.
50. Allende BT, Engelem JC. Tension-band arthrodesis in the finger joints. J Hand Surg 1980;5:269–271.
51. Segmuller G. Surgical stabilization of the skeleton of the hand. Bern: Hans Huber Publishers, 1977.
52. Ayres JR, Goldstrohm GL, Miller GJ, et al. Proximal interphalangeal joint arthrodesis with the Herbert screw. J Hand Surg 1988;13:600–603.
53. Khuri MS. Tension band arthrodesis in the hand. J Hand Surg 1986;11:41–45.
54. Uhl RL, Schneider LH. Tension band arthrodesis of the finger joints: a retrospective review of 76 consecutive cases. J Hand Surg 1992;17:518–522.

55. Stern PJ, Gates NT, Jones TB. Tension band arthrodesis of small joints in the hand. J Hand Surg 1993;18:194–197
56. Ijsselstein CB, van Egmond, DB, Kovius SE, et al. Results of small-joint arthrodesis: comparison of Kirschner wire fixation with tension band wire technique. J Hand Surg 1992;17:952–956.
57. Katzman SS, Gibeault JD, Dickson K, et al. Use of a Herbert screw for interphalangeal arthrodesis. Clinical Orthop 1993;296:127–132.
58. Leibovic SJ, Strickland JW. Arthrodesis of the proximal interphalangeal joint of the finger: comparison of the use of the Herbert screw with other fixation methods. J Hand Surg 1994;19:181–188.
59. Wright CS, McMurtry RY. AO arthrodesis in the hand. J Hand Surg 1983;9:932–935.
60. Kleinert JM, Gateley D. Proximal interphalangeal joint fusion: special situations. Atlas Hand Clin 1998;3:31–39.
61. Bishop AT. Small joint arthrodesis. Hand Clin 1993;9:683–689.
62. Seitz WH, Sellman DC, Scarcella JB, et al. Compression arthrodesis of the small joints of the hand. Clinical Orthop 1994;304:116–121.
63. Ostgaard SE, Weilby A. Resection arthroplasty of the proximal interphalangeal joint. J Hand Surg 1993;18:613–615.
64. Seradge H, Kutz JA, Kleinert HA. Perichondral resurfacing arthroplasty in the hand. J Hand Surg 1984;9:880–886.
65. Carroll RE, Taber TH. Digital arthroplasty of the proximal interphalangeal joint. J Bone Joint Surg 1954;36:912–920.
66. Swanson AB, Maupin BK, Gajjar NV, et al. Flexible implant arthroplasty in the proximal interphalangeal joint of the hand. J Hand Surg 1985;10:796–805.
67. Iselin F, Conti E. Long-term results of proximal interphalangeal joint resection arthroplasties with a silicone implant. J Hand Surg 1995;20:S95–S97.
68. Moller K, Sollerman C, Geijer M, et al. Early results with osseointegrated proximal interphalangeal joint prostheses. J Hand Surg 1999;24:267–274.
69. Lindscheid RL, Murray PM, Vidal MA, et el. Development of a surface replacement arthroplasty for proximal interphalangeal joints. J Hand Surg 1997;22:286–298.
70. Neibauer JJ, Landry RM. Dacron-silicone prosthesis for the metacarpophalangeal and interphalangeal joints. Hand 1971;3:55–61.
71. Hage JJ, Yoe EPD, Zevering JP, et al. Proximal interphalangeal joint silicone arthroplasty for posttraumatic arthritis. J Hand Surg 1999;24:73–77
72. Doi K, Kuwata N, Kawai S. Alumina ceramic finger implants: a preliminary biomaterial and clinical evaluation. J Hand Surg 1984;9:740–749.
73. Ashworth CR, Hansraij KK, Todd AO, et al. Swanson proximal interphalangeal joint arthroplasty in patients with rheumatoid arthritis. Clinical Orthop 1997;342:34–37.
74. Dryer RF, Blair WF, Shurr DG, et al. Proximal interphalangeal joint arthroplasty. Clinical Orthop 1984;185:187–194.
75. Schneider LW. Proximal interphalangeal joint arthroplasty: the volar approach. Semin Arthroplast 1999;2:139–147.
76. Lin HH, Wyrick JD, Stern PJ. Proximal interphalangeal silicone replacement arthroplasty: clinical results using the anterior approach. J Hand Surg 1995;20:123–132.
77. Amadio PC, Wood MB. Alternative reconstructive procedures of the proximal interphalangeal joint. In: Morrey BF, ed. Reconstructive surgery of the joints, 2nd ed. New York: Churchill Livingstone, 1996:279–286.
78. Millender LH, Nalebuff EA. Commentary on osteoarthritis of the proximal interphalangeal joint. Hand Clin 1987;3:405–412.
79. Kovach JC, Werner FW. Palmer AK, et al. Biomechanical analysis of internal fixation techniques for proximal interphalangeal joint arthrodesis. J Hand Surg 1986;11:562–566.
80. Chamay A. A distally based dorsal and triangular tendinous flap for direct access to the proximal interphalangeal joint. Ann Chir Main 1988;7:179–183.
81. Leibovic S. Arthrodesis of the proximal interphalangeal joint of the finger using tension band wiring or Herbert screws. Atlas Hand Clin 1998;3:17–30.
82. Brannon EW, Klein G. Experiences with a finger-joint prosthesis. J Bone Joint Surg 1959;41:87–102.
83. Flatt AE. Restoration of rheumatoid finger-joint function: interim report on trial of prosthetic replacement. J Bone Joint Surg 1961;43:753–774.
84. Condamine JL, Benoit JY, Comtet JJ, et al. Proposed digital arthroplasty critical study of the preliminary results. Ann Chir Main 1988;7:282–232.
85. KirkpatrickWH, Kozin SH, Uhl RL. Early motion after arthroplasty. Hand Clin 1996;12:73–86.
86. Cannon N, ed. Diagnosis and treatment manual for physicians and therapists, 3rd ed. Indianapolis: The Hand Rehabilitation Center of Indiana, 1991.