A Prospective Evaluation of the Anatomy of the First Dorsal Compartment in Patients Requiring Surgery for De Quervain's Tenosynovitis

Abstract

Background  We prospectively evaluated the surgical anatomy during first dorsal compartment release for De Quervain's tenosynovitis, with special attention to the superficial branch of the radial nerve (SBRN). Additionally, the incidence of tendon instability during surgery was assessed.

Methods  This prospective cohort study consisted of 130 De Quervain's patients undergoing first dorsal compartment release. The treating surgeons recorded the type of incision used, the number of abductor pollicis longus (APL) and extensor pollicis brevis (EPB) tendon slips, the number of SBRN branches encountered, additional subcompartments created by any septations, and active/passive tendon stability.

Results  A singular first dorsal compartment was found in 37% of cases, whereas 55% of patients had two subcompartments and 8% had three. Multiple APL tendon slips (range: 1–4) were identified in 78% of patients. In contrast, a single EPB tendon was found in 92% of patients (range: 0–2). At least one SBRN was encountered in 61% of cases. Following surgery, instability was evident in 9% of patients, who had tendons perch with passive wrist flexion. In one of these patients (<1%), the tendons dislocated volarly out of the first dorsal compartment during active flexion.

Conclusions  The anatomical findings in our relatively large, prospective study of De Quervain's patients undergoing first dorsal compartment release are consistent with previous smaller and/or retrospective studies. Overall, we expect to encounter the SBRN during first dorsal compartment release in more than 50% of patients but are unconcerned if it is not visualized during a careful approach. Tendon instability has an incidence of 9%; however, dislocation is rare (<1%).

De Quervain's tenosynovitis is a common condition affecting the extensor pollicis brevis (EPB) and abductor pollicis longus (APL) tendons as they run through the first dorsal extensor compartment of the wrist. Several anatomical and surgical studies have suggested that a separate compartment for the EPB and/or multiple slips of the APL can predispose patients to this condition and can result in symptoms that remain refractory to nonsurgical treatment. ^{1 2 3 4} Surgical treatment of De Quervain's tenosynovitis involves release of the first dorsal compartment including the separate compartment for the EPB (if present) and all slips of the APL. A thorough exploration of the first compartment is necessary to ensure adequate release of all tendon slips, which otherwise may result in persistent postoperative symptoms if not completely decompressed.

Aside from the failure to identify and to release all tendon slips, two other complications involving anatomical variables can occur during De Quervain's surgery but have received far less attention in the literature. First, the superficial branch of the radial nerve (SBRN) is at a risk during surgical release. It emerges from between the brachioradialis and the extensor carpi radialis longus 7 to 9 cm proximal to the radial styloid and begins to branch approximately 5 cm proximal to the styloid. ^{5 6 7} Based on cadaveric studies, these branches are found overlying the first dorsal compartment directly in the De Quervain's surgical field in a large proportion of people. ^{6 7} Second, once the retinaculum has been released, the tether maintaining the EPB and APL tendons in the first dorsal compartment has been removed. This can result in postoperative volar subluxation of the tendons. ^{8 9 10} The recommended surgical technique is to release the first dorsal compartment along its dorsal margin, leaving a sling of retinaculum to restrain the tendons and prevent volar subluxation. ¹¹ The effectiveness of this has not been studied extensively.

Most of the existing data on the anatomical variability in patients treated surgically for De Quervain's tenosynovitis comes from retrospective or cadaveric studies. The purpose of this study was to prospectively evaluate the surgical anatomy during De Quervain's release with special attention to the SBRN and to assess the incidence of tendon instability during surgery.

Materials and Methods

Institutional Review Board approval was obtained prior to initiation of the study. Patients who underwent a primary surgical decompression of the first dorsal compartment of the wrist between September 2016 and February 2018 by a group of 14 board-certified, fellowship-trained hand surgeons were evaluated prospectively. Demographic data and intraoperative anatomical data from 130 patients formed the basis of this study. All cases were unilateral. The study cohort comprised 108 women and 22 men. This female predominance (83%) is comparable to the demographics of other De Quervain's studies. ^{1 2 3}

All surgeries took place under local anesthesia with or without sedation. Each surgeon performed the decompression using their preferred technique—either a longitudinal or a transverse/oblique incision along Langer's lines. Branches of the SBRN were identified and protected when encountered, and the retinaculum of the first dorsal compartment was incised at its dorsal-most aspect. The contents of the first dorsal compartment were explored, and all subsheaths were released. Once released, the stability of the tendons within the first dorsal compartment was tested passively in full flexion and extension. Then, the patient was asked to fully flex and extend the wrist actively, and the active stability was assessed. The tendons were classified as stable if they remained within the first dorsal compartment, perched if they rode up onto the volar ridge of the first dorsal compartment but did not dislocate, or unstable if they dislocated volarly from the first dorsal compartment in flexion. The treating surgeons completed a questionnaire detailing the type of incision used, the number of APL tendon slips, number of EPB tendon slips, the number of SBRN branches encountered, the number of additional subcompartments created by any septations, and active/passive tendon stability. Statistical analysis was performed using Fisher's exact test for comparison of nonparametric data.

Results

The mean age of the group was 57 years (range: 24–82). The average age of the female patients was 58 years compared with 52 years for the men ( p  = 0.03). A transverse/oblique incision was made along Langer's lines of the skin in 108 (83%) of cases, whereas the remaining 22 (17%) procedures were performed with a longitudinal incision. In 48 wrists (37% of cases), a singular first dorsal compartment without subsheaths was noted intraoperatively. Two total subcompartments were identified in 71 (55%) wrists, and three subcompartments were identified in 11 (8%) wrists.

There was substantial variation in the number of tendon slips discovered intraoperatively. Multiple APL tendon slips were noted 78% of the time (102 of 130 cases). A single APL tendon slip was found in 28 (21%) cases, two in 66 (51%), three in 30 (23%), and four in 6 (5%). A single EPB tendon was identified in 120 (92%) cases, two slips were found in 6 (5%) cases, and the EPB was absent in the remaining 4 (3%) cases. One or two SBRN were encountered in 50 and 11% of cases, respectively. Zero SBRN were encountered in 39% of the surgeries. There was no significant difference in the number of branches encountered based on the incision type ( p  = 0.8).

Following first dorsal compartment release, the APL and EPB tendons were found to be stable both actively and passively in 91% of cases. Tendon instability was noted in 9% of wrists after first dorsal compartment decompression: 12 patients were classified as perched with passive wrist flexion. In one of these patients (<1% of total cohort), the tendons dislocated volarly during active wrist flexion testing. This patient was immediately treated with a first dorsal compartment reconstruction using a brachioradialis tendon flap. Eight of the remaining 11 patients with perched tendons had the radial retinacular flap sutured to the subcutaneous tissues, as described by Bahm et al, and immediate postoperative thumb spica splinting, as advocated by McMahon et al. ^{1 9} The other three patients had no additional intraoperative treatment or postoperative splinting.

Discussion

Various studies have demonstrated a higher incidence of subsheaths dividing the APL and EPB in De Quervain's patients compared with normal cadavers. ^{1 2 3 12} Furthermore, De Quervain's patients tend to have multiple APL tendon slips but only a single EPB. ^{1 2 12} In a recent systematic review, Lee et al summarized the existing first dorsal compartment anatomical data from various De Quervain's case series. ¹² A septated first dorsal compartment was present in 59% of 470 patients (7 case series), multiple APL slips were noted in 73% of 320 patients (5 case series), and a single EPB was observed in 94% of 307 patients (4 case series). However, most of the available data come from retrospective studies. In our prospective study of 130 patients undergoing first dorsal compartment release, we found comparable rates of subsheaths (63%), multiple APL slips (78%), and a single EPB (92%). These findings in our prospective study with a relatively large cohort serve to substantiate the concordant findings from previous smaller and/or retrospective studies.

Unlike subsheath and tendon variability, however, less consideration has been given to other important anatomical variables in De Quervain's tenosynovitis. Specifically, the SBRN is vulnerable to injury given its close proximity to the first dorsal compartment and its susceptibility to form painful neuromas. ^{6 7 13} In a cadaveric study, Auerbach et al found that 75% (15/20) of specimens had an SBRN directly overlying the typical transverse incision for De Quervain's tenosynovitis. ⁶ Similarly, Abrams et al found that the nearest SBRN was on average only 4 mm away from the first dorsal compartment and laid directly on top of the first dorsal compartment in 35% (7/20) of cadaveric specimens. ⁷ Despite this anatomical data, the SBRN was commonly not encountered during our De Quervain's surgery.

It is unclear why we identified fewer SBRNs than expected given the existing data from cadaveric studies. Potentially, this relates to the inherent differences between surgical approaches and anatomical studies; we were performing our standard surgical approach with routine retraction of soft tissues without trying to discretely dissect out the SBRN. Surgeons should be aware that the SBRN is at a risk given its proximity to the incision and the first dorsal compartment; however, it may not need to be directly visualized. Alternatively, as seen with tendon and subsheath variability, it is possible that the branching pattern of the SBRN may differ in patients with De Quervain's tenosynovitis versus normal cadavers and that therefore the incidence of overlying branches is less in surgical patients. ^{1 2 3 12} While the true incidence of SBRN injury following first dorsal compartment release is unknown, our findings may help to explain why it is a relatively uncommon complication.

Similar to SBRN injury, tendon instability following first dorsal compartment release is a commonly discussed potential complication and an appropriate theoretic concern, but few reports documenting its occurrence or treatment exist. ^{8 9 10} To prevent subluxation of the tendons, Burton and Littler recommended incision of the first dorsal compartment over the EPB at its dorsal margin, Bahm et al described transcutaneous fixation of the retinacular flap to prevent anterior displacement of the APL tendon, and McMahon et al emphasized the importance of thumb immobilization for 7 to 10 days in order to maintain proper alignment of the tendons over the radius as the soft tissues heal. ^{1 9 11} While it is possible that these techniques minimize the risk of postoperative subluxation, the incidence of this complication remains unknown.

In our study, we found that 9% (12/130) of patients demonstrated some tendon instability following first dorsal compartment release. All of these patients perched with passive wrist flexion but only one (<1%) dislocated with active wrist flexion. The single patient with dislocating tendons had an immediate first dorsal compartment reconstruction using a brachioradialis tendon flap, as described by McMahon et al. ⁹ Eight of the 11 patients with perched tendons had the radial retinacular flap sutured to the subcutaneous tissues, as described by Bahm et al, and immediate postoperative thumb spica splinting, as advocated by McMahon et al. ^{1 9} The other three patients had no additional intraoperative treatment or postoperative splinting. In all 12 instances, the decision regarding whether to augment first dorsal compartment release was made intraoperatively by the individual surgeon. Since our study is isolated to intraoperative findings, it is unknown if any of our patients developed postoperative instability and/or if augmenting first dorsal compartment release made any clinical difference. However, in patients with stable tendons during intraoperative passive and active testing, it seems less likely that late instability would develop.

The strengths of this study include its relatively large number of patients and prospective methodology. Also, we evaluated anatomy relevant to potential complications that have been frequently discussed but inadequately assessed in the literature. We feel that these data provide surgeons with information regarding the SBRN and instability factors that may be encountered during the actual De Quervain's surgery.

There are also several limitations. Since this was an observational study, we do not have outcome data. Therefore, we do not know the percentage of patients who developed postoperative SBRN symptoms or whether any patients developed late, symptomatic tendon instability. Moreover, while the 14 participating surgeons defined tendon instability in an identical fashion, the assessment of instability inherently remains somewhat subjective. However, we feel that this is reflective of clinical practice and generalizes our results.

Based on our data, we expect to encounter the SBRN during first dorsal compartment release in more than 50% of patients but are unconcerned if it is not visualized during a careful approach. Provided that the soft tissues are gently retracted and that the first dorsal compartment is directly visualized during release, we would expect the likelihood of SBRN injury to be low. Furthermore, we counsel our patients that tendon instability following first dorsal compartment release is rare. However, we assess the stability of the tendons intraoperatively to identify patients in whom instability may be an issue. If perching is encountered, we consider retinacular flap augmentation and postoperative splinting. If the tendons were fully dislocated, we contemplate addressing this immediately with brachioradialis reconstruction in order to avoid potentially symptomatic, postoperative instability.

Funding Statement

Funding None.