Scaphoid Excision and Bicolumnar Carpal Fusion with Retrograde Headless Screws

Abstract

Background/Purpose  Scaphoid excision and partial wrist fusion is used for the treatment of scapholunate advanced collapse/scaphoid nonunion advanced collapse wrist arthritis. The purpose of this study was to report midterm functional and radiographic outcomes in a series of patients who underwent bicolumnar fusion of the lunocapitate and triquetrohamate joints using retrograde headless screws.

Methods  Twenty-three consecutive patients (25 wrists) underwent surgery with this technique from January 2014 to May 2017 with a minimum follow-up of 1 year. Assessment consisted of range of motion, grip, and pinch strength. Patient-reported outcome measures included disabilities of the arm, shoulder, and hand (DASH) and patient-rated wrist evaluation (PRWE) scores. Fusion rates and the radiolunate joint were evaluated radiographically. The relationship between wrist range of motion and midcarpal fusion angle (neutral position vs. extended capitolunate fusion angle > 20 degrees) was analyzed.

Results  Average follow-up was 18 months. Mean wrist extension was 41 degrees, flexion 36 degrees, and radial-ulnar deviation arc was 43 degrees. Grip strength was 39 kg and pinch 9 kg. Residual pain for activities of daily living was 1.6 (visual analog scale). The mean DASH and PRWE scores were 19 ± 16 and 28 ± 18, respectively. Patients with an extended capitolunate fusion angle trended toward more wrist extension but this did not reach statistical significance ( p  = 0.17).

Conclusions  With retrograde headless compression screws, the proximal articular surface of the lunate is not violated, preserving the residual load-bearing articulation. Patients maintained a functional flexion–extension arc of motion with grip-pinch strength close to normal. Capitolunate fusion angle greater than 20 degrees may provide more wrist extension but further studies are needed to demonstrate this.

Level of Evidence  This is a Level IV study.

Advanced degenerative osteoarthritis of the wrist secondary to scaphoid nonunion advanced collapse (SNAC) or scapholunate advanced collapse (SLAC) may become symptomatic and require surgical treatment. Proximal row carpectomy (PRC) and scaphoid excision with partial wrist fusion (SEPF) are two established procedures for the treatment of these pathologies. Even though PRC and SEPF share some indications, the latter is reserved as a motion-preserving surgery when the head of the capitate is arthritic but the radiolunate joint is spared. Motion preservation with SEPF occurs at the radiolunate and ulnotriquetral articulations. ¹ Different intercarpal arthrodeses have been described using the same biomechanical principles, including four- and three-corner fusion (4CF and 3CF), capitolunate, and bicolumnar arthrodeses. ^{2 3 4} In bicolumnar arthrodesis, the capitate is fused to the lunate and the triquetrum to the hamate, sparing the lunotriquetral and capitohamate joints from preparation. This fusion modification simplifies the surgical technique and prevents destabilization of the carpus. The fusion across columns is unnecessary because of competent intrinsic ligaments connecting the capitate/hamate and lunate/triquetrum. The main objective of this study was to retrospectively review the functional and radiological results of a series of patients with scaphoid excision and bicolumnar carpal fusion using retrograde headless screws.

Materials and Methods

We obtained institutional ethics review board approval for this study. This was a case series of 23 consecutive patients (25 wrists) who underwent a bicolumnar carpal fusion by the senior author (level of expertise: 3) from January 2014 to May 2017 due to a stage 2 or 3 SNAC/SLAC wrist. These patients returned for clinical and radiographic evaluation with a minimum follow-up of 1 year and were assessed by either the senior author or surgical fellow. After obtaining informed consent, each patient completed the disabilities of the arm, shoulder, and hand (DASH) and patient-rated wrist evaluation (PRWE) questionnaires and completed pain scores on a visual analog scale (VAS). The clinical assessment consisted of wrist range of motion measured with a goniometer, grip, and pinch strength measured using a JAMAR dynamometer (Sammons Preston Roylan, Bolingbrook, IL); the average of three measurements was used for data analysis. Motion and strength were compared with the unaffected contralateral side where possible.

The radiographic assessment parameters consisted of fusion rate and appearance of the radiolunate joint space on X-ray. The relationship between capitolunate fusion angle and wrist range of motion was also analyzed. We identified nine wrists fused with a capitolunate angle greater than 20 degrees of extension and nine wrists fused in neutral position and the difference in wrist motion was documented.

Statistical Methods

Routine descriptive statistics were calculated including means and standard deviation for wrist range of motion and grip/pinch strength measurements of both wrists where possible.

Difference in range of motion and grip/pinch strength between the operative wrist and contralateral control wrist was determined by using the Mann–Whitney nonparametric test. This was also used for the comparative analysis between partial arthrodesis in the neutral and extended positions. An α value of 0.05 was considered significant. All tests were considered statistically significant if the p -values were <0.05. Microsoft Excel 2008 for Mac (MicrosoftCorp, WA) and Stata 11 for Mac (StataCorp, TX) were used for statistical analysis.

Surgical Technique

The patient was positioned supine with the involved arm outstretched on a hand table. Under axillary block regional anesthesia, the arm was prepped and draped in a sterile fashion. A tourniquet was applied to the upper arm during the procedure. A dorsal midline longitudinal incision was made from the base third metacarpal to the level of the distal forearm. The extensor retinaculum was identified and full-thickness skin flaps were raised radially and ulnarly. The superficial radial nerve and dorsal sensory branch of the ulnar nerve were protected within the skin flaps. The extensor retinaculum was incised at the third extensor compartment and retinacular flaps were raised radially into the second compartment and ulnarly into the fifth compartment. The tendons were protected with Penrose drains and retracted. The sensory branch of the posterior interosseous nerve was isolated and a 1 cm section was excised en bloc to denervate the dorsal wrist capsule.

A ligament sparing radiocarpal capsulotomy was performed and the arthritic scaphoid was excised using K-wire joysticks, sharp dissection, and a small osteotome. The proximal surface of the lunate and radial lunate fossa was protected throughout. The articular surfaces of the capitolunate and triquetrohamate joints were denuded of cartilage and the subchondral bone was decorticated to create a bleeding cancellous surface. Care was taken to remove similar quantities of bone from the triquetrohamate and lunocapitate intervals, thus allowing for symmetric compression across the midcarpal interval. Autologous bone graft from the excised scaphoid was morcellized and interposed into the fusion sites ( Fig. 1 ). Great care was taken to adequately reduce the dorsal intercalated segment instability deformity. Depending on the patient́s needs and regional anatomy, the intercarpal fusion angle was left either neutral or slightly extended to theoretically achieve more wrist extension ( Fig. 2 ).

Fig. 1.

( A ) Degenerative scaphoid nonunion advanced collapse changes are noted. ( B ) The proximal row is subluxed dorsal to the distal row and the articular surfaces are prepared. ( C ) The midcarpal preparation completed prior to bone grafting and fixation. ( D ) Completed scaphoid excision and bicolumnar fusion with retrograde headless screws. Note the correction of the radial translation deformity of the distal carpal row.

Fig. 2.

( A ) Dorsal intercalated segment instability configuration of the carpus typical of scaphoid nonunion advanced collapse wrist. ( B ) Fusion with the capitolunate column in neutral alignment. ( C ) Intercarpal fusion with the capitolunate alignment in 20 degrees of extension.

Dorsal radiocarpal impingement was always ruled out before final fixation and in all cases two 3.5 mm cannulated headless compression screws (Arthrex, Naples, FL) were placed in a retrograde manner for intercarpal fixation. This almost universally was a 28 mm mini screw across the capitolunate articulation and a 24 mm mini screw across the triquetrohamate articulation. The entrance point was through the 3rd carpometacarpal (CMC) interval on axis with capitolunate column and 4th/5th CMC joint for the triquetrohamate column ( Fig. 3 ). In some cases, the entrance was placed more distal through the metacarpal base to optimize screw trajectory. In three cases, a third screw was used across the capitolunate interval when intercarpal compression and stability was not satisfactory. The position of the screws was assessed under live fluoroscopy. The dorsal capsule was closed and repaired without tension, skin flaps were closed, and a short-arm cast was applied for 6 weeks.

Fig. 3.

( A ) Preoperative scaphoid nonunion advanced collapse wrist. ( B ) A radiolunate wire is provisionally placed to correct the dorsal intercalated segment instability deformity, with subsequent placement of guidewires for the capitolunate and triquetrohamate headless screws. ( C ) Completed bicolumnar partial carpal fusion with retrograde headless screws.

Results

Twenty-three patients (25 wrists) were included in the study, including 7 women and 16 men; the average patient age at the time of surgery was 54 years, the average follow-up was 18 months (range: 13–32 months). Two patients were not considered for data analysis because their un-united partial carpal fusion had been converted to total wrist arthroplasty at the time of final follow-up.

The primary diagnosis was a SLAC wrist in 16 cases and SNAC in 9. Seven cases of SNAC wrists had at least two surgeries prior to the scaphoid excision and bicolumnar carpal fusion. Conversely, only four cases of SLAC wrists had one or two surgeries prior to the carpal fusion.

Range of Motion and Strength

Range of motion and grip/pinch strength measurements comparing the operative wrist with the contralateral control wrist are summarized in Table 1 . The extension, flexion, and ulnar deviation measurements were significantly less in the operative wrist, representing 69, 52, and 55% of the range of motion found on the contralateral control wrist respectively. Radial deviation angle, grip, and pinch strength measurements were not significantly different between operative and contralateral wrists.

Table 1. Range of motion and grip/pinch strength of operative wrist and contralateral control wrist.

	Bicolumnar arthrodesis	Contralateral control	p -Value
Extension, degrees	41 (±15)	59 (±10)	< 0.001
Flexion, degrees	36 (±14)	69 (±13)	< 0.001
Radial deviation, degrees	20 (±8)	26 (±12)	0.08
Ulnar deviation, degrees	23 (±9)	42 (±9)	< 0.001
Grip strength, kg	39 (±13)	43 (±12)	0.41
Pinch strength, kg	9 (±3)	9 (±3)	0.44

Patient Reported Outcomes

Mean PRWE score at final follow-up was 28 (± 18.3) and DASH score was 19 (± 16). Ten patients had jobs that demanded manual handling of loads, and seven of them returned to their prior occupation postoperatively. Five patients were retired prior to surgery and were able to resume their usual activities including several sports. All patients who had office or low physically demanding occupations were able to continue their jobs, although two patients decided to retire for reasons unrelated to their wrist. Mean current pain at rest on a 10-point VAS scale (0 = no pain; 10 = worst pain) was 1.6 (± 1.1); distribution of patients according to pain level is listed in Table 2 .

Table 2. Clinical score results.

	PRWE	DASH	Pain (VAS)/cases
Score at final follow-up	28 (± 18.3)	19 (±16)	0: four
			1: six
			2: nine
			3: three
			4 or more: one

Abbreviations: DASH, disabilities of the arm, shoulder, and hand; PRWE, patient-rated wrist evaluation; VAS, visual analog scale.

Relationship between Capitolunate Fusion Angle and Wrist Range of Motion

We were able to accurately identify radiographically nine wrists fused with capitolunate angle greater than 20 degrees of extension and nine wrists fused in a neutral position. Wrist extension was 38 ± 7 degrees in the neutral position group and 47 ± 13 degrees in the extended position group ( p  = 0.17). Wrist flexion was 35 ± 16 degrees and 35 ± 11 degrees in each group, respectively ( p   =  0.89).

Nonunion and Complications

Four patients underwent secondary procedures directly related to midcarpal arthrodesis. One patient required hardware removal and extensor tenosynovectomy due to an excessively long screw. One patient needed a revision for nonunion with autologous bone graft and repeat fixation that subsequently healed within 8 weeks of surgery. Two patients were converted to a total wrist arthroplasty secondary to nonunion of their intercarpal fusion with collapse and intra-articular screw penetration. Only one patient developed advanced radiocarpal osteoarthritis with moderate pain but refused conversion to total wrist fusion. Of note, no hardware removal was required for dorsal impingement symptoms ( Fig. 4 ).

Fig. 4.

( A ) Demonstrating a 62-year-old male with a Watson Stage 2 scaphoid nonunion advanced collapse wrist. ( B ) Three headless screws and additional bone graft from the distal radius were required because of osteopenia. ( C ) Unfortunately, the patient went on to symptomatic nonunion with intra-articular screw perforation. ( D ) He was revised successfully to total wrist arthroplasty.

Discussion

In 1984, Watson and Ballet described the 4CF procedure to treat moderate-to-advanced wrist osteoarthritis with a SLAC or SNAC pattern, where the scaphoid is excised and the capitate, lunate, hamate, and triquetrum are fused. ⁵ This surgery was proposed as an alternative to PRC. The functional differences between these two methods are generally minimal, although there are reports of 4CF resulting in improved grip strength with PRC offering a greater arc of motion. ^{6 7} In addition, Wall et al ⁸ found that patients younger than 40 years of age at the time of PRC were at greater risk of needing conversion to total wrist fusion, which could be explained by the articular incongruity between the radiolunate fossa and the proximal pole of the capitate.

Several modifications have been proposed to the 4CF procedure, with the aim of simplifying the surgical technique, increasing the fusion rate, and minimizing hardware-related complications. Partial carpal arthrodesis can be achieved with a variety of fixation methods with or without bone grafting. In general, plates are associated with variable fusion rates and higher risk of hardware impingement on the radius in extension. ⁹ Wires and staples are also associated with high incidence of complications including nonunion. ^{10 11} While one patient in our series required a screw removal for excess length and tendonitis, no hardware removal was required for dorsal impingement. While our technique is reliable in situations with adequate bone stock, pathoanatomy such as large subchondral cysts in the lunate and capitate is a relative contraindication. In these scenarios, fixation such as multidirectional K-wires or alternative procedures such as total wrist arthroplasty may be more appropriate. This novel technique does not appear to have a higher rate of complications relative to traditional techniques. Our hardware removal rate (4%) and nonunion rate (12%) were in keeping with a systematic review of the literature on 4CF that reported rates of nonunion (5–14%) and need for hardware removal (3.3%). ^{12 13} Another theoretical limitation of this technique relates to the design of conical variable pitch headless screws. The thread density is highest on the trailing end of the screw. When inserted retrograde, the leading screw end has lowest thread density and may impair purchase in the distant bone. The lunate is mobile in this surgical scenario, which may create a construct at higher risk of mechanical failure.

To our knowledge there are no publications describing a series of patients treated with scaphoid excision and bicolumnar carpal fusion using retrograde headless screws with midterm clinical results. We believe the use of retrograde screws for midcarpal fusion has the advantage of not violating the articular surface of the proximal lunate, considering the radiolunate surface will be the main joint under load. With a similar assertion, Ball and Bergman ¹⁴ published the surgical technique for 4CF using retrograde headless compression screws. The main difference between their technique and ours was that both columns were fused together with either a hamatolunate or a hamatocapitate screw. Clinical outcomes were not published from their series.

Watson and Ballet et al ⁵ and Mulford et al ¹⁵ described two series of patients (22 and 11 wrists, respectively) where the fusion was performed using anterograde headless compression screws and compared with the available literature. In both series, the fusion rate was 100% and the range of motion and grip strength were comparable to those reported for traditional scaphoid excision and 4CF.

Another surgical adaptation from traditional 4CF technique was our modification of the fusion angle between the capitate and lunate to a more extended position. In theory, this modification should improve wrist extension at the expense of wrist flexion. If we consider most daily activities are made with the wrist in an extended position, ¹⁶ this gain in extension could improve clinical outcomes and patient satisfaction. In this study, patients with extended capitolunate fusion angle trended toward more wrist extension but this did not reach statistical significance ( p  = 0.17).

This study has several limitations, including a lack of preoperative clinical outcome measures to compare with postoperative scores. However, the use of contralateral wrist as a control group is widely used as an effective method to determine operative results in terms of motion, strength and pain control. The sample size of 25 wrists in this study is an appreciable number that allowed us to determine accurate clinical results. However, when we analyzed subgroups to find differences in wrist motion due to midcarpal fusion angle, the size in each group limited the power of the results. An average follow-up of 18 months with a minimum of 12 enables us to reliably describe midterm clinical outcomes and complication rates, though evaluation of long-term clinical results and survivorship of this procedure is still warranted.