Differential Pulley Release in Trigger Finger: A Prospective,Randomized Clinical Trial

Introduction

Trigger finger most commonly arises from a diameter mismatch between the first annular retinacular pulley (A1 pulley) and the flexor tendons.^1,2 Treatments involve conservative management, corticosteroid injections, and surgical release.^3-7 Although surgery is often curative, many patients experience persistence or recurrence of symptoms; recurrence rates of 7% to 23% have been reported, leading some patients to reoperation.^8,9

The transverse fascicular fibers of the palmar aponeurosis (PA) have been shown to encompass the flexor system immediately proximal to the A1 pulley, functioning similarly to a pulley (Figure 1). ¹⁰ Often, this accessory “pulley system” becomes involved in trigger pathology, and proximal release must accompany traditional A1 release to restore full flexor excursion.^11,12 Thus, these fibers have been previously termed the “PA pulley” or “A0 pulley” ¹³ ; we use the latter designation throughout this study. A preliminary cadaveric study simulated trigger finger pathogenesis by artificially constricting various pulleys. ¹⁴ Constriction of the A0 pulley was sufficient to induce triggering in all 4 fingers, often with less tension than needed for triggering of the A1 pulley. Both biomechanical and a priori literature suggest a larger role for A0 pulley than previously thought.

OPEN IN VIEWER

Figure 1.

The A0 pulley, made of the transverse fibers of the palmar aponeurosis, lies just proximal to the A1 pulley.

Thus, persistent symptoms after surgery may be, in part, due to uncorrected A0 trigger finger. Studies have yet to classify A0 constriction as a cause of trigger finger in the patient population. This work aimed to determine the clinical prevalence and patient factors involved in trigger finger.

Methods

Subject Recruitment

This single-surgeon, investigator-designed, and initiated, prospective, randomized clinical trial was conducted under the purview of the Yale University Human Investigations Committee (HIC #2000021438). Subjects aged ≥18 years were included if they were diagnosed with at least 1 trigger finger with symptoms of mechanical locking that required surgical release by the principal investigator and surgeon. This included trigger thumb, as previous studies have implicated a role for A0 release in patients with trigger thumb. ¹⁵ Patients who had received prior surgery for trigger finger were excluded. Upon consent, patients were presented with a preoperative survey consisting of 20 questions regarding patients’ clinical history (Table 1). Subjects were excluded if they refused or were unable to complete the preoperative survey.

OPEN IN VIEWER

Table 1.

Demographic and Clinical Variables of Patients Requiring Trigger Finger Release.

Demographics				History of current illness
	Total	A0	A1		Total	A0	A1
Number of fingers	30	17	13	Left or right hand
Sex				Left	11	5	6
Female	15	9	6	Right	19	12	7
Male	15	8	7	Finger
Average age	57.9	57.2	58.7	Thumb	7	7	0
Handedness				Index	4	1	3
Left	11	5	6	Middle	8	4	4
Right	19	12	7	Ring	5	3	2
Manual labor occupation	17	10	6	Small	6	2	4
Smoking history	5	4	1	Average pain level (1-10)	6.6	6.4	6.9
Medical diagnosis				Require active straightening	11	6	5
Rheumatoid arthritis	1	0	1	Nodule	21	12	9
Diabetes mellitus	10	6	4	Prior steroid injection	15	8	7
Amyloidosis	0	0	0
Past hand disease	25	12	13
Past hand procedures	20	10	10
Trigger finger release	7	3	4
Carpal tunnel release	8	4	4

All values are based on a total of 42 digits. There were no significant differences in any of the variables between patients who had A0 and A1 release first.

Intraoperative Technique

Trigger finger release was conducted by the senior author in following the institutional criterion standard, releasing both A1 and A0 pulleys. Local anesthetic injection was given in simple trigger releases, and preoperative sensory nerve blocks were administered in patients who had other procedures performed concurrently, requiring higher levels of anesthesia. Subjects were excluded if they required general anesthesia or were not able to demonstrate motor function intraoperatively. An upper arm tourniquet was used in all cases for hemostasis.

A simple, linear incision was designed over a resting distal palmar crease, and exposure of the flexor tendon sheath and pulley system was obtained.

Prior to release, the patient was asked to flex and extend the finger under careful examination and documentation. Any clicking, catching, locking, or triggering was noted. At this point, a coin toss was used to randomize initial release of either the A0 or A1 pulley (Figure 2). Following release of the chosen pulley, the patient was again asked to flex and extend the finger, and results were documented. The remaining pulley was then released, and, again, the clinical trigger status was recorded.

OPEN IN VIEWER

Figure 2.

Sequential intraoperative pathway taken following randomization of A0 or A1 pulley release. In this example, the pathway follows selection of A0 initial release.

The A0 pulley was deemed responsible if the initial release of A1 failed, but following release of A0 successfully resolved triggering. The A0 pulley was deemed at least partially involved if the initial A0 release completely or incompletely resolved the trigger status of the patient. This intraoperative triggering status was the primary end point of this study design.

Results

Comparison of Complete With Incomplete Resolution

There were no significant differences in demographics, clinical variables, or study end points between patients who had initial A1 or A0 release (Table 1). Similarly, neither initial A1 or initial A0 release was significantly associated with complete, incomplete, or failed resolution of symptoms (Figure 3). Significantly more patients with a history of diabetes (P < .05) demonstrated incomplete resolution following initial release.

OPEN IN VIEWER

Figure 3.

Results following initial pulley release.

Multivariable regression analysis revealed that diabetes status (P < .05) and female sex (P < .05) predicted incomplete resolution at first release (Table 2). Incomplete resolution was independent of A1 or A0 release, age, sex, specific finger affected, arthritis, or steroid injections.

OPEN IN VIEWER

Table 2.

Multiple Regression for Predictors of Incomplete Release After Release of Only 1 Pulley, A1 or A0.

Parameter	Odds Ratio	Confidence interval	P value
A0 or A1 release	1.06	0.96 to 1.00	.75
Age at surgery	0.98	0.71 to 1.58	.08
Sex	0.66	0.45 to 0.95	<.05
Digit involved	0.98	0.85 to 1.13	.75
Occupation	1.09	0.67 to 1.77	.70
Rheumatoid arthritis	2.64	0.72 to 9.78	.20
Diabetes mellitus	1.58	1.03 to 2.46	<.05
Past steroid injections	0.89	0.57 to 1.38	.59

Diabetes diagnosis and female sex were significantly associated with failed results upon first release.

Discussion

Trigger finger is a common and comprehensively described pathology with a defined surgical course, yet current treatments still often result in incomplete symptom relief or recurrence.^8,9 Most patients (70.6%) in this study demonstrated partial or complete symptomatic relief upon A0 release alone. Considering A1 division resulted in 31% of digits with no change in intraoperative symptoms requiring a secondary A0 release, the A0 pulley can be inferred to be primarily responsible for at least 31% of trigger fingers in this study. This ratio could be as high as 47% because half of the patients who received primary A0 release demonstrated complete resolution of symptoms. Diabetes mellitus diagnosis and female sex predicted incomplete symptom resolution, with neither initial A1 nor initial A0 release resulting in statistically better results. This association may offer an explanation for observed relative resistance of people with diabetes to success with steroid injection, as the pulley constriction may involve a wider area of the flexor tendon sheath. Further studies should seek to investigate why women might also experience poorer symptom resolution.

The A0 pulley in relation to trigger finger has been sparsely implicated in the hand surgery literature. Although the term “A0 pulley” was first used by Moutet in 2003, Manske and Lesker were the first to recognize the transverse fascicular fibers and paratendinous bands of the PA as a functional pulley, capable of biomechanical augmentation of flexor tendon tension.^10,13 Case reports previously identified 1 patient with persistent triggering intraoperatively following ring finger A1 division. ¹¹ On further dissection, the transverse fibers of the PA were stenosed, division of which resolved the patient’s symptoms. A second case reported a patient complaining of trigger finger in the middle finger, which presented with a “double-click.” A1 release abated finger locking, but triggering was persistent until the PA was divided in that report.

Such testimonials are identical to the experience of the principal investigator, prompting this study. Exclusive A1 release did not always resolve triggering during surgery, and proximal release was required, based on intraoperative testing. These findings led to our institutional protocol, including comprehensive A1 and A0 release for all patients with trigger fingers. Intraoperative time is minimally increased with swift division using tenotomy scissors just proximal to the A1 pulley. A0 pulley fibers, particularly if constricted, can be easily palpated with the edge of the scissors and do not require further exposure. According to cadaveric studies, mechanical excursion is preserved, as Manske and Lesker noted a total loss of 2.8% range of motion upon division of both pulleys compared with 1.0% with division of just the A1. ¹⁰

Prior reports lack any clinical mention of the A0 pulley as the focal etiology of trigger finger. Interestingly, the true recurrence rate reported in the hand surgery literature, particularly when percutaneous release is involved, is similar to the minimum 31% A0 pulley involvement seen in this study.^2,8,9,16 Percutaneous release likely isolates fewer pulley fibers and avoids any division of the A0 pulley, with up to 11% of patients requiring conversion to open surgery.^17,18 When open, overlap exists between the A1 and A0 pulley anatomy. Both pulleys are separated only by a few millimeters at most, and thus, many providers already release both routinely without discerning the A0 as a separate sheath. For example, release based on identification of the proximal A1 pulley edge by hand landmarks as described by Wilhelmi et al ¹⁹ would result in the division of both pulleys. Thus, these same landmarks might be used by providers who plan to release the A0 pulley. Nonetheless, further characterization of the A0 pulley remains important to ensure proper isolation and release.

This study is limited to results seen immediately intraoperatively and lacks long-term recurrence data. Although A0 release could theoretically have long-term biomechanical implications, these effects have been shown to be mostly negligible in previous studies (ie, decreases in maximal force of approximately ~1%). ¹⁰ Thus, while important to weigh these risks with potential benefits, we routinely elect to release the A0 pulley in all of our patients, regardless of the implicated digit and type of anesthesia. Because it is protocol to release both pulleys for patient benefit, we were unable to assess whether this A0 pulley release influenced long-term recurrence and/or biomechanical function in this study.

There were other limitations to this study. Surgeons were not blinded as to which pulley was released first, so inherent bias may have impacted recorded results. To limit this, observations were recorded objectively with preset variables (triggering, clicking, locking) and standardized against the preoperative examination. Although patients were randomized in this trial design, there may still be differences between cohorts that could have influenced study outcomes. For example, all patients with trigger thumb in our study were randomized to the A0-first intervention. Thus, future studies with increased recruitment may be necessary to validate these findings. Further studies might also aim to better define the clinical settings in which release of the A0 pulley may be most beneficial, including the role of A0 release for the treatment of mid trigger finger (pain rather than locking/catching), the differences in resolution related to the implicated digit (ie, fingers vs thumb), and the relative ease and efficacy of performing routine A0 release among patients receiving different forms of anesthesia (ie, local vs general).