Comparative Efficacy of Triamcinolone versus Dexamethasone in Local Injection for Stenosing Tenosynovitis of the Metacarpophalangeal Joint: A Retrospective Cohort Study Authors: Lin Huang, Jiabao Wei, Qinci Xie, Xiaofeng Chen, Jinying Wang, Junjie Zheng, Shiqiang Zhuo, Yongfeng Li, Zhixiao Yao, Gaofeng Huang Published: 2026 ABSTRACT AbstractBackgroundLocal glucocorticoid injection is first-line therapy for stenosing tenosynovitis. However, the comparative efficacy of lipophilic sustained-release triamcinolone versus hydrophilic short-acting dexamethasone remains underexplored. This study aimed to compare clinical outcomes of triamcinolone versus dexamethasone in metacarpophalangeal joint injection, with hypothesis-generating exploration of how differential pharmacokinetics may influence efficacy and recurrence.MethodsThis retrospective cohort study included adult patients receiving first-time local injection for stenosing tenosynovitis ranging from February 2021 to January 2025. Patients were divided into triamcinolone (10–20 mg + lidocaine) and dexamethasone (4 mg + lidocaine) groups. Propensity score matching (1:1) controlled for confounding. The primary outcome was 3-month overall efficacy (Quinnell grade I). Secondary outcomes included recurrence, second injection rate, surgical conversion, and local adverse events.ResultsAfter matching 97 patients per group, baseline characteristics were balanced. At 3 months, triamcinolone demonstrated superior efficacy (82.5% vs 63.9%, P=0.004). Triamcinolone also showed lower 6-month recurrence (12.5% vs 27.4%, P=0.009), second injection rate (8.2% vs 19.6%, P=0.021), and surgical conversion (4.1% vs 13.4%, P=0.024). Fat atrophy was numerically higher with triamcinolone (9.3% vs 3.1%, P=0.072). We hypothesize that the sustained-release depot effect of triamcinolone may enable prolonged modulation of the inflamed A1 pulley microenvironment, though this mechanistic interpretation requires prospective validation.ConclusionIn this retrospective matched cohort, triamcinolone was associated with superior 3-month efficacy and lower recurrence compared to dexamethasone. These findings are hypothesis-generating and warrant prospective dose-controlled confirmation.Keywords: triamcinolone, dexamethasone, stenosing tenosynovitis, metacarpophalangeal joint, sustained-release pharmacokinetics Introduction Stenosing tenosynovitis, commonly known as trigger finger, is a prevalent hand disorder characterized by progressive thickening and fibrocartilaginous metaplasia of the A1 pulley, leading to mechanical impingement of the flexor tendon. Affecting approximately 2–3% of the general population and over 10% of individuals with diabetes mellitus, the condition imposes a substantial burden on hand function and quality of life.1–3 The pathophysiological core of trigger finger extends beyond simple mechanical stenosis. Emerging evidence implicates chronic low-grade inflammation and metabolic dysregulation within the tenosynovial microenvironment. Histopathological studies have revealed infiltration of activated macrophages, lymphocytes, and mast cells in stenotic A1 pulley tissues, accompanied by elevated levels of pro-inflammatory cytokines including interleukin-1β (IL-1β), tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6).4,5 These inflammatory mediators drive fibroblast proliferation and extracellular matrix deposition and may induce metabolic reprogramming of tenocytes and synovial fibroblasts—a shift toward aerobic glycolysis similar to that observed in rheumatoid synovitis.6–8 This metabolic switch may sustain fibroblast activation, promotes resistance to apoptosis, and perpetuates a vicious cycle of inflammation and tissue remodeling.9–11 Local glucocorticoid injection remains the cornerstone of non-surgical management, offering simplicity, low cost, and a favorable safety profile. However, the selection of specific corticosteroid preparation remains largely empirical, with limited comparative evidence guiding clinical decisions. Glucocorticoids differ fundamentally in their physicochemical properties, which dictate pharmacokinetic behavior and, consequently, their duration of action. Lipophilic corticosteroids such as triamcinolone acetonide exhibit low water solubility. When injected locally, they precipitate as crystalline aggregates, forming a sustained-release depot that dissolves gradually over several weeks.12–14 This prolonged local residence time enables continuous suppression of inflammatory transcription factors and sustained modulation of target cell activity.15–17 The extended drug exposure may be advantageous for chronic inflammatory conditions like trigger finger, where complete resolution of pathological tissue remodeling may require weeks of uninterrupted anti-inflammatory signaling.7,18,19 In contrast, hydrophilic corticosteroids such as dexamethasone phosphate are highly water-soluble. They diffuse rapidly from the injection site, achieving high peak concentrations but extremely short tissue residence times—typically hours to a few days.20 While this rapid kinetics may be ideal for acute inflammatory flares, it may be suboptimal for chronic stenosing tenosynovitis, where transient suppression may fail to durably alter the pathological microenvironment. Despite theoretical advantages favoring sustained-release preparations, direct comparative evidence from real-world clinical practice remains limited. Moreover, prior studies have not explicitly linked observed clinical outcomes to potential immunometabolic mechanisms underlying differential efficacy. This knowledge gap is clinically significant because therapeutic failure or early recurrence after hydrophilic corticosteroid injection may reflect inadequate duration of pharmacological effect rather than inherent drug potency. We hypothesize that the prolonged tissue residence of triamcinolone may translate into superior initial efficacy and lower recurrence through more durable suppression of inflammatory mediators and sustained modulation of pathological fibroblasts. This mechanistic framework, while grounded in established glucocorticoid pharmacology, remains speculative and requires direct biomarker validation in future studies. Therefore, this retrospective cohort study was designed to: (1) compare the real-world effectiveness of sustained-releasing triamcinolone versus short-acting dexamethasone for adult trigger finger; (2) evaluate recurrence rates and need for additional interventions as indirect measures of treatment durability; and (3) generate hypotheses regarding how differential pharmacokinetics may influence clinical outcomes. Patients and Methods Study Design and Ethics This was a retrospective cohort study between February 1, 2021, and January 31, 2025. Documented review and approval was obtained from review board of Jinjiang Municipal Hospital (Shanghai Sixth People’s Hospital Fujian). The study was conducted in accordance with the Declaration of Helsinki and local institutional guidelines. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements. Study Population Inclusion Criteria ①Adult patients (age ≥ 18 years) diagnosed with stenosing tenosynovitis of any digit; ②Receiving their first-ever local corticosteroid injection for the affected digit during the study period; ③Definite clinical diagnosis with baseline Quinnell grade II to V; ④Injected medication clearly documented as either triamcinolone acetonide or dexamethasone sodium phosphate, mixed with lidocaine; ⑤Minimum follow-up duration of 3 months with documented clinical outcomes via outpatient visit or structured telephone interview. Exclusion Criteria ①Uncontrolled diabetes mellitus, defined as HbA1c > 8% within 3 months prior to injection; ②Prior corticosteroid injection to the same site within the preceding 6 months; ③Co-existing inflammatory arthropathy, such as rheumatoid arthritis, psoriatic arthritis, gout; ④Local infection, skin breakdown, or surgical wound at the proposed injection site; ⑤Pregnancy or lactation documented in medical history; ⑥Incomplete follow-up data, including loss to follow-up before 3-month assessment without telephone contact. Treatment Protocol and Grouping All injections were performed by senior orthopaedic surgeons using a standardized clinical approach. The injection point was the site of maximal tenderness over the A1 pulley at the metacarpophalangeal joint crease. A volar approach was used in all patients. The needle was inserted perpendicular to the skin, advanced to contact the A1 pulley, and then withdrawn slightly before depositing the medication into the tendon sheath. Triamcinolone Group Triamcinolone acetonide injection (10 mg/mL) was mixed 1:1 by volume with 1% lidocaine without epinephrine. The total injection volume ranged from 1 to 2 mL, delivering 10 to 20 mg of triamcinolone. The exact dose was at the treating physician’s discretion based on the severity of triggering and estimated size of the A1 pulley. Treatment choice (triamcinolone vs dexamethasone) was determined by the attending physician’s preference, with no standardized protocol dictating agent selection. Triamcinolone dose was selected at the physician’s discretion based on the severity of triggering and estimated size of the A1 pulley. To examine whether dose selection introduced confounding, we performed a post-hoc analysis comparing baseline characteristics across triamcinolone dose subgroups (10 mg, 15 mg, and 20 mg). Dexamethasone Group Dexamethasone sodium phosphate injection (4 mg/mL) was mixed 1:1 by volume with 1% lidocaine without epinephrine. The total injection volume was fixed at 1 mL, delivering 4 mg of dexamethasone. All patients received a single initial injection. A second injection was permitted at the treating physician’s discretion if symptoms were only partially improved, which means reduction of Quinnell grade by 1 or 2 levels but not achieving grade I at the 6-week follow-up or initial improvement was followed by symptom recurrence before 12 weeks. Rescue surgical release (A1 pulley open or percutaneous release) was offered to patients who failed to achieve satisfactory improvement after two injections or who had persistent grade IV–V symptoms at 12 weeks. Patients were naturally divided into the triamcinolone group or dexamethasone group based on the medication documented in their electronic medical records at the time of injection. Data Collection of Variables and Sources Data were extracted from three sources: (1) hospital electronic medical records, (2) outpatient clinic notes, and (3) structured telephone interviews for patients who did not return for scheduled follow-up visits. Baseline Demographic and Clinical Variables The following variables were collected to characterize the study population and to serve as potential confounders for propensity score matching (Table 1). Table 1Baseline Demographic and Clinical VariablesVariable CategorySpecific Variables CollectedDemographicsAge (years), sex (male/female), body mass index (BMI, kg/m2)Clinical historySymptom duration (weeks from first symptom to injection), affected digit (thumb/index/middle/ring/little), affected side (right/left/bilateral), occupation (manual/non-manual), smoking history (current/former/never), diabetes mellitus (type 2, documented diagnosis)Disease severityBaseline Quinnell grade (II, III, IV, or V) assessed by the treating physicianPain assessmentBaseline Visual Analog Scale (VAS) pain score (0–10, with 10 = worst imaginable pain) Baseline Demographic and Clinical Variables Outcome Variables Primary Outcome The 3-month overall efficacy rate: Defined as achieving Quinnell grade I (normal movement, no snapping, catching, or pain) at the 3-month (±2 weeks) follow-up assessment, with patient-reported satisfaction and no need for further treatment. Secondary Outcomes 6-month recurrence rate: Among patients of Quinnell grade I at 3 months, the proportion who developed recurrent triggering or catching (Quinnell ≥ grade II) by the 6-month (±2 weeks) follow-up; Second injection rate: Proportion of patients receiving a second injection within 3 months of the first injection; Time to second injection: Number of days from first injection to second injection among those who required it; Surgical conversion rate: Proportion of patients who underwent open or percutaneous A1 pulley release surgery within 12 months of the first injection; Local adverse events: Subcutaneous fat atrophy, clinically visible skin dimpling at injection site, skin hypopigmentation of lighter skin color at injection site, persistent injection-site pain lasting >2 weeks, tendon rupture documented by ultrasound or surgical exploration, and deep infection. For patients who did not attend scheduled in-person visits, structured telephone interviews were conducted by trained research assistants using a standardized script to assess Quinnell grade, by asking the patient to actively flex/extend the affected finger and describe any snapping or catching, VAS pain score, and occurrence of adverse events. Outcome Measures of Operational Definitions Quinnell Grading System for trigger finger severity: Grade I: Normal movement. No pain, snapping, or catching during active flexion and extension. Grade II: Snapping or triggering present but no catching sensation. The finger moves smoothly without needing assistance. Grade III: Definite catching sensation present. The patient must overcome resistance to extend the finger, but extension is completed without external assistance. Grade IV: Catching or locking upon passive extension. The finger locks in flexion and requires passive manipulation by the other hand or an external force to extend. Grade V: Fixed flexion contracture. The finger is locked in a flexed position and cannot be passively extended. Visual Analog Scale (VAS) for Pain A 0–10 horizontal line, where 0 = no pain, and 10 = worst possible pain. Patients marked their current pain level during active flexion/extension of the affected finger. Definitions of Efficacy and Failure Complete success (cure): Quinnell grade I at 3 months. Partial success: Reduction in Quinnell grade by ≥1 level from baseline, but not achieving grade I. Treatment failure: No change in Quinnell grade from baseline, or worsening of grade, or need for second injection before 3 months due to intolerable symptoms. Definition of Recurrence For patients who achieved cure (Quinnell grade I) at 3 months, recurrence was defined as an increase to Quinnell grade ≥ II at any subsequent follow-up (6 months or 12 months). Sample Size Estimation Sample size was calculated a priori for the primary outcome (3-month efficacy rate): Triamcinolone group efficacy: 80%, Dexamethasone group efficacy: 60%. Using a two-sided chi-square test with α = 0.05 and power (1-β) = 0.80, the required sample size was calculated as 103 patients per group. Accounting for an anticipated 15–20% rate of incomplete data meaning loss to follow-up, or missing variables, we aimed to enroll approximately 240 patients. Propensity Score Matching (PSM) To minimize selection bias inherent to retrospective non-randomized studies, we performed 1:1 propensity score matching without replacement. Propensity score estimation: Propensity scores were estimated using a logistic regression model that included the following prespecified covariates: age, sex, body mass index, symptom duration, affected digit, affected side, manual occupation, smoking history, baseline Quinnell grade (categorized as II–III vs IV–V), baseline VAS pain score, and diabetes mellitus status. Matching algorithm: Nearest-neighbor 1:1 matching without replacement was performed using a caliper of 0.2 × standard deviation of the logit of the propensity score. The MatchIt package in R (version 4.3.1) was used for all matching procedures. Balance assessment: After matching, standardized mean differences (SMD) were calculated for all covariates. An SMD < 0.1 was considered indicative of good balance between groups. Additionally, chi-square tests for categorical variables and t-tests for continuous variables were performed to confirm no statistically significant differences remained (P > 0.05). Prespecified Subgroup Analysis To assess whether the treatment effect of triamcinolone vs dexamethasone differs across clinically important subgroups, we prespecified the following subgroup analyses (Table 2). Table 2Prespecified Subgroup AnalysisSubgroupDefinitionRationaleSeverityMild-to-moderate (Quinnell II–III) vs Severe (Quinnell IV–V)Severe disease has greater inflammatory and metabolic burden; may require sustained-release formulationDiabetes statusDiabetic (type 2) vs non-diabeticDiabetics have baseline immunometabolic dysregulation; may respond differentlySymptom duration<3 months vs ≥3 monthsLonger duration may indicate more established fibrotic pathology Prespecified Subgroup Analysis The severity-stratified analysis was considered the primary subgroup analysis and was powered for exploratory purposes, not for formal hypothesis testing. Results are presented with point estimates, 95% confidence intervals, and P-values for interaction when applicable. Statistical Analysis All statistical analyses were performed after propensity score matching, using the matched cohort (n=194, 97 per group) as the primary analysis set. Primary analysis: The primary outcome (3-month efficacy rate, binary: success/failure) was compared between triamcinolone and dexamethasone groups using the chi-square test. The number needed to treat (NNT) was calculated as 1 / (absolute risk reduction). Secondary analyses: (1) Recurrence rate: Compared using chi-square test for time-to-recurrence (censored at 6 months) with groups compared using the Log rank test. (2) Cox proportional hazards regression: Used to calculate the hazard ratio (HR) for recurrence, adjusted for baseline covariates (age, sex, diabetes, baseline Quinnell grade) to assess the independent effect of treatment assignment. (3) Second injection rate and surgical conversion rate: Compared using chi-square test. (4) Time-to-event outcomes (time to second injection, time to surgery): Compared using Mann–Whitney U-test (non-parametric). (5) Adverse events: Compared using Fisher’s exact test (due to low expected cell counts). Subgroup analysis: For the severity-stratified analysis (Quinnell II–III vs IV–V), the chi-square test was used to compare triamcinolone vs dexamethasone efficacy within each stratum. A test for interaction was performed using logistic regression with a treatment × severity interaction term. Sensitivity analyses: To test the robustness of our findings, we performed two sensitivity analyses: As-treated analysis: Including all patients before matching, with multivariable logistic regression adjusting for the same covariates used in PSM, plus triamcinolone dose and injection volume. Per-protocol analysis: Excluding patients who received a second injection before the 3-month primary outcome assessment to isolate the effect of a single injection. Statistical software: All analyses were performed with R version 4.3.1 using the MatchIt package for propensity score matching. Significance threshold: All tests were two-sided, and a P-value < 0.05 was considered statistically significant. No adjustment for multiple comparisons was performed for secondary outcomes. Results Patient Disposition and Baseline Characteristics After Matching After applying inclusion and exclusion criteria to the initial 512 screened patients, 242 patients were eligible for analysis. Propensity score matching successfully matched 97 patients from each group, with the remainder (45 triamcinolone, 3 dexamethasone) excluded due to lack of suitable matches. A patient flow diagram is presented in Figure 1. Patient Flow Diagram. Table 3 presents the baseline characteristics of the matched cohort. All covariates were well-balanced between groups, with no statistically significant differences (P > 0.05) and standardized mean differences <0.1. The mean age was 48.3±9.7 years in the triamcinolone group and 49.1±10.2 years in the dexamethasone group. Approximately 45% of patients in each group had severe disease (Quinnell grade IV–V), and 17–20% had diabetes mellitus. Table 3Baseline Characteristics After Propensity Score MatchingVariableTriamcinolone (n=97)Dexamethasone (n=97)P-valueAge (years, mean ± SD)48.3 ± 9.749.1 ± 10.20.571Female sex (n, %)71 (73.2%)68 (70.1%)0.631Body mass index (kg/m2, mean ± SD)24.2 ± 3.124.5 ± 3.30.508Symptom duration (weeks, median [IQR])8 [5–14]9 [6–15]0.392Thumb involvement (n, %)83 (85.6%)80 (82.5%)0.554Manual occupation (n, %)42 (43.3%)39 (40.2%)0.663Smoking history (n, %)19 (19.6%)21 (21.6%)0.725Baseline Quinnell Grade (n, %)0.813Grade II14 (14.4%)12 (12.4%)Grade III39 (40.2%)42 (43.3%)Grade IV31 (32.0%)28 (28.9%)Grade V13 (13.4%)15 (15.5%)Severe disease (Grade IV–V)44 (45.4%)43 (44.3%)0.884Baseline VAS pain score (0–10, mean ± SD)5.8 ± 1.56.0 ± 1.60.365Diabetes mellitus (n, %)17 (17.5%)19 (19.6%)0.712Abbreviations: SD, standard deviation; IQR, interquartile range; VAS, visual analog scale. Baseline Characteristics After Propensity Score Matching Abbreviations: SD, standard deviation; IQR, interquartile range; VAS, visual analog scale. Triamcinolone Dose Distribution The triamcinolone dose range (10–20 mg) and injection volume (1–2 mL) were at the treating physician’s discretion. To examine whether dose selection introduced bias, we performed a post-hoc analysis comparing baseline severity across triamcinolone dose subgroups. Among triamcinolone recipients, 58 patients (59.8%) received 10 mg (1 mL), 28 patients (28.9%) received 15 mg (1.5 mL), and 11 patients (11.3%) received 20 mg (2 mL). Importantly, higher doses were not preferentially administered to more severe cases: the proportion of severe disease (Quinnell IV–V) was 44.8% in the 10 mg subgroup, 46.4% in the 15 mg subgroup, and 45.5% in the 20 mg subgroup (P = 0.98), indicating no dose-severity confounding (Table 4). Table 4Triamcinolone Dose Distribution and Baseline Characteristics by Dose SubgroupCharacteristic10 mg (n=58)15 mg (n=28)20 mg (n=11)P-valueAge (years, mean ± SD)47.9 ± 9.548.7 ± 10.149.2 ± 9.80.89Female sex (n, %)42 (72.4%)21 (75.0%)8 (72.7%)0.96Severe disease (Grade IV–V)26 (44.8%)13 (46.4%)5 (45.5%)0.98Diabetes mellitus (n, %)10 (17.2%)5 (17.9%)2 (18.2%)0.99Symptom duration (weeks, median [IQR])8 [5–13]9 [5–15]8 [6–14]0.82 Triamcinolone Dose Distribution and Baseline Characteristics by Dose Subgroup Primary Outcome of 3-Month Overall Efficacy At 3 months post-injection, the overall efficacy rate (Quinnell grade I) was 82.5% (80/97) in the triamcinolone group, significantly higher than the 63.9% (62/97) observed in the dexamethasone group (χ2 = 8.42, P = 0.004). The absolute risk reduction was 18.6%, corresponding to a number needed to treat (NNT) of 5 (95% CI: 3 to 14). Subgroup Analysis of Severity Stratification Table 5 presents the primary outcome stratified by baseline disease severity. The treatment benefit of triamcinolone was most pronounced in patients with severe disease (Quinnell grades IV–V), where triamcinolone achieved 75.0% (33/44) efficacy compared to only 46.5% (20/43) for dexamethasone (P = 0.006). In contrast, among patients with mild-to-moderate disease (grades II–III), the efficacy rates were 88.7% (47/53) for triamcinolone and 77.8% (42/54) for dexamethasone, a difference that did not reach statistical significance (P = 0.131). The test for interaction was not formally significant (P for interaction = 0.091), likely due to limited power. Table 5Three-Month Efficacy Stratified by Baseline Quinnell GradeSeverity SubgroupTriamcinolone (n/N, %)Dexamethasone (n/N, %)Absolute DifferenceP-valueMild-to-Moderate (Grade II–III)47/53 (88.7%)42/54 (77.8%)+10.9%0.131Severe (Grade IV–V)33/44 (75.0%)20/43 (46.5%)+28.5%0.006 Three-Month Efficacy Stratified by Baseline Quinnell Grade Secondary Outcomes Table 6 summarizes the secondary outcomes. Among patients achieving cure at 3 months, the 6-month recurrence rate was 12.5% (10/80) in the triamcinolone group versus 27.4% (17/62) in the dexamethasone group (P = 0.009). The results confirmed significantly longer recurrence-free survival with triamcinolone (log-rank P = 0.008). Multivariable Cox regression demonstrated that triamcinolone use was associated with a 58% reduction in recurrence risk (HR = 0.42, 95% CI: 0.22–0.81, P = 0.009). Table 6Secondary Outcomes ComparisonOutcomeTriamcinolone (n=97)Dexamethasone (n=97)P-value6-month recurrence rate (among 3-month responders)12.5% (10/80)27.4% (17/62)0.009Second injection rate (within 3 months)8.2% (8/97)19.6% (19/97)0.021Surgical conversion rate (by 12 months)4.1% (4/97)13.4% (13/97)0.024Time to second injection (days, median [IQR])*56 [49–63]48 [42–56]0.090Notes: *Among patients who required second injection (triamcinolone n=8, dexamethasone n=19)*. Secondary Outcomes Comparison Notes: *Among patients who required second injection (triamcinolone n=8, dexamethasone n=19)*. The second injection rate was significantly lower in the triamcinolone group: 8.2% (8/97) vs 19.6% (19/97), P = 0.021. The median time to second injection among those who required it was 48 days (IQR: 42–56) in the dexamethasone group versus 56 days (IQR: 49–63) in the triamcinolone group (P = 0.09). Surgical conversion rate at 12 months was 4.1% (4/97) in the triamcinolone group versus 13.4% (13/97) in the dexamethasone group (P = 0.024). Safety Outcomes Local adverse events are summarized in Table 7. Subcutaneous fat atrophy occurred in 9.3% (9/97) of triamcinolone recipients versus 3.1% (3/97) of dexamethasone recipients (P = 0.072). Skin hypopigmentation rates were 5.2% (5/97) and 2.1% (2/97), respectively (P = 0.247). All cases of fat atrophy were mild (<1 cm in diameter), and no patient requested specific treatment. No tendon ruptures, deep infections, or systemic adverse events occurred in either group. Table 7Local Adverse EventsAdverse EventTriamcinolone (n=97)Dexamethasone (n=97)P-valueSubcutaneous fat atrophy9 (9.3%)3 (3.1%)0.072Skin hypopigmentation5 (5.2%)2 (2.1%)0.247Persistent injection pain (>2 weeks)2 (2.1%)4 (4.1%)0.683Tendon rupture0 (0%)0 (0%)—Deep infection0 (0%)0 (0%)— Local Adverse Events Sensitivity Analyses The as-treated analysis, including all patients before matching, with multivariable logistic regression yielded consistent results: triamcinolone was associated with significantly higher 3-month efficacy (adjusted OR = 2.59, 95% CI: 1.42–4.73, P = 0.002). The per-protocol analysis that excluded 27 patients who received a second injection before 3 months also confirmed the primary finding (84.6% vs 67.1%, P = 0.006). These sensitivity analyses support the robustness of our primary conclusion. Outcomes by triamcinolone dose subgroup: No significant differences in efficacy or recurrence were observed across triamcinolone dose subgroups, suggesting that the superior efficacy of triamcinolone was not driven by higher dosing within the studied range. Discussion This retrospective cohort study, strengthened by propensity score matching to control for confounding, provides compelling real-world evidence that a single local injection of the sustained-release lipophilic corticosteroid triamcinolone (10–20 mg) was associated with significantly superior clinical outcomes compared to the short-acting hydrophilic corticosteroid dexamethasone (4 mg) for adult stenosing tenosynovitis. The association of triamcinolone with better outcomes was consistent across multiple clinically meaningful endpoints: higher 3-month efficacy (82.5% vs 63.9%, P=0.004; NNT=5), lower 6-month recurrence (12.5% vs 27.4%, P=0.009; HR=0.42), reduced need for second injections (8.2% vs 19.6%, P=0.021), and fewer surgical conversions (4.1% vs 13.4%, P=0.024). These differences were most pronounced in patients with severe disease (Quinnell grades IV–V), where triamcinolone achieved 75.0% efficacy compared to only 46.5% with dexamethasone (P=0.006). Sensitivity analyses (as-treated and per-protocol) confirmed the robustness of these findings. The following discussion of potential mechanisms is speculative and hypothesis-generating. Our study did not measure tissue cytokines, metabolic enzymes, or epigenetic marks. The proposed mechanisms are extrapolated from established glucocorticoid pharmacology and prior histopathological studies of stenosing tenosynovitis. They are presented to generate hypotheses for future prospective mechanistic investigations rather than to assert demonstrated biological effects. Stenosing tenosynovitis is fundamentally a chronic inflammatory-fibroproliferative disorder. Histopathological analyses of stenotic A1 pulleys have consistently demonstrated: (1) inflammatory infiltrates including CD68+ macrophages, CD3+ T lymphocytes, and mast cells clustering around areas of fibrocartilaginous metaplasia;21,22 (2) elevated cytokine expression with 3- to 10-fold increases in IL-1β, TNF-α, and IL-6 mRNA and protein levels compared to normal pulleys;23,24 and (3) metabolic reprogramming of tenosynovial fibroblasts characterized by upregulation of glycolytic enzymes and lactate production—a shift that may support sustained fibroblast activation and resistance to apoptosis.25,26 This inflammatory-metabolic axis may be self-perpetuating. Inflammatory cytokines activate NF-κB signaling in fibroblasts, which not only drives further cytokine production but may also induce metabolic reprogramming toward aerobic glycolysis. Glycolytic metabolism produces lactate that acidifies the microenvironment, potentially activating matrix metalloproteinases and promoting tissue remodeling. Breaking this cycle may require sustained, not transient, suppression of inflammatory activity. Glucocorticoids suppress this axis through two complementary mechanisms: (1) transrepression—glucocorticoid receptor (GR) binding to NF-κB and AP-1 prevents their transcriptional activation of pro-inflammatory cytokine genes;27,28 and (2) metabolic modulation—GR activation promotes expression of regulators that may suppress glycolysis and promote oxidative phosphorylation in inflammatory cells and fibroblasts.29–31 However, both mechanisms are reversible upon drug withdrawal. Therefore, the duration of glucocorticoid exposure may be as critical as the potency for achieving durable disease modification. Triamcinolone acetonide has low water solubility (approximately 0.02 mg/mL). When injected as an aqueous suspension, it forms a crystalline depot within the A1 pulley region. Dissolution occurs slowly, releasing active drug over 3–6 weeks.32 We hypothesize that this sustained release may achieve several advantages. The prolonged presence of triamcinolone may maintain transrepression of NF-κB target genes for weeks, potentially preventing the rebound inflammation that may occur after short-acting corticosteroid washout. This hypothesis is indirectly supported by our time-to-second-injection data: among patients who required repeat injection, the triamcinolone group had a median of 56 days before needing rescue therapy, compared to 48 days in the dexamethasone group (P=0.09). Sustained GR signaling may promote chromatin remodeling at inflammatory gene loci—recruitment of histone deacetylases and repression of histone acetyltransferases—resulting in prolonged gene silencing that could outlast the actual drug presence.33,34 This hypothesis could explain why triamcinolone was associated with a 58% reduction in recurrence risk (HR=0.42), a durable effect that persisted for months after the drug had been fully cleared. Prolonged GR activation may allow sufficient time for metabolic reversion in tenosynovial fibroblasts, shifting from glycolysis back toward oxidative phosphorylation, potentially reducing lactate production and collagen synthesis.35,36 In contrast, dexamethasone phosphate has high water solubility (>10 mg/mL), leading to rapid diffusion and clearance with a tissue residence half-life of only 24–48 hours.37,38 This brief exposure provides transient inflammatory suppression-sufficient to reduce symptoms but potentially insufficient to durably alter established pathology. The 27.4% recurrence rate observed with dexamethasone in our study may reflect this limitation. Our findings are consistent with and extend the findings of prior prospective studies. A previous study directly compared triamcinolone and dexamethasone for trigger finger injection in a randomized controlled trial, reporting a significantly higher cure rate for triamcinolone (73.2% vs 54.2%) and a similar trend toward higher fat atrophy (11.6% vs 3.8%).39 The slightly higher absolute efficacy rates in our study (82.5% vs 63.9%) may be attributed to differences in study design: our real-world protocol permitted a second injection at 6 weeks for partial responders, whereas the RCT limited treatment to a single injection. Notably, both studies demonstrate the same relative advantage for triamcinolone reinforcing the robustness of this treatment effect across study designs. Another study compared triamcinolone with another lipophilic agent, methylprednisolone, and found no significant difference between them, but importantly, both outperformed historical data on placebo or less potent preparations.40 A prospective study established that higher corticosteroid doses were associated with better outcomes, particularly in patients with diabetes.41 This dose-response relationship further supports the depot hypothesis. Our study adds to this literature by: (1) providing real-world effectiveness data from a larger matched cohort; (2) explicitly generating hypotheses regarding potential immunometabolic mechanisms; (3) conducting prespecified severity-stratified analysis to identify the subgroup most likely to benefit; and (4) reporting time-to-event outcomes that capture treatment durability. Our findings have several implications for clinical practice, though these should be interpreted cautiously given the retrospective design. First, for initial therapy in adult trigger finger: Triamcinolone (10–20 mg) may be a reasonable first-line choice over dexamethasone. Choosing triamcinolone was associated with an 18.6 percentage-point higher probability of achieving complete cure at 3 months (NNT=5) in this cohort. Furthermore, triamcinolone was associated with a lower risk of needing a second injection (8.2% vs 19.6%) and a lower long-term risk of requiring surgery (4.1% vs 13.4%). Second, for severe disease (Quinnell grades IV–V): Dexamethasone showed efficacy of less than 50% in this subgroup. In these advanced cases, a long-acting agent like triamcinolone may offer a higher chance of success. Third, for mild-to-moderate disease (Quinnell grades II–III): The choice is less critical. Both agents achieved reasonable efficacy (88.7% for triamcinolone, 77.8% for dexamethasone), and the difference was not statistically significant (P=0.131). In this subgroup, shared decision-making could consider patient preferences regarding the trade-off between potentially higher efficacy (triamcinolone) and lower cosmetic side effect risk (dexamethasone). Fourth, regarding safety and informed consent: The numerically higher rate of local fat atrophy with triamcinolone (9.3% vs 3.1%) represents a potential trade-off. Prolonged local glucocorticoid exposure suppresses adipocyte differentiation via GR-mediated downregulation of PPARγ.42,43 Importantly, all observed cases were mild and self-limited. Clinicians should discuss this risk with patients. Several limitations warrant consideration when interpreting our results. First, retrospective design and residual confounding. Despite propensity score matching on key covariates, we cannot exclude residual confounding from unmeasured variables, such as detailed hand activity levels, psychological factors, socioeconomic status, injection technique nuances. Only a randomized controlled trial can eliminate confounding by indication. Second, lack of direct immunometabolic biomarkers. We inferred potential mechanisms from clinical surrogate indicators like time to resolution, recurrence patterns, and time to second injection. Direct measurement of tissue cytokines, metabolic enzymes, or epigenetic marks would provide mechanistic confirmation. Such invasive sampling was not feasible in this retrospective study but should be considered for future prospective trials. Third, single-center and generalizability. This study was conducted at a single tertiary hospital in China. Patient demographics, physician practice patterns, and injection protocols may differ from primary care settings or other geographic regions. Multi-center validation is needed. Fourth, sample size limitations for secondary outcomes. While our post-matching sample of 97 per group was adequately powered for the primary efficacy outcome, it was underpowered for certain secondary analyses, eg, fat atrophy comparison, diabetes subgroup analysis, and test for interaction between treatment and severity. Fifth, follow-up and data completeness. We relied on a combination of outpatient records and telephone follow-up. Telephone assessment of Quinnell grade, while validated in some contexts, may be less accurate than in-person examination. Additionally, mild fat atrophy and skin hypopigmentation may have been underreported. Sixth, unequal corticosteroid dosing: The triamcinolone dose was not fixed (10–20 mg), whereas dexamethasone was administered as a fixed 4 mg dose. While our post-hoc analysis did not reveal dose-severity confounding or significant outcome differences across triamcinolone dose subgroups, we cannot exclude the possibility that higher triamcinolone doses contributed to the observed efficacy difference. This represents a major limitation of our retrospective design. Dose-response studies are needed. Conclusion In this retrospective propensity score-matched cohort, local injection of triamcinolone was associated with superior 3-month efficacy, lower 6-month recurrence, and reduced need for additional interventions compared to dexamethasone in adult stenosing tenosynovitis. These clinical advantages may be attributable to triamcinolone’s sustained-release pharmacokinetics, which we hypothesize enable more durable modulation of the inflamed A1 pulley microenvironment. However, this mechanistic interpretation remains speculative and requires direct biomarker validation. The therapeutic benefit appeared most pronounced in severe disease (Quinnell grades IV–V). In mild-to-moderate disease, both agents were effective, and shared decision-making may consider the trade-off between slightly higher efficacy with triamcinolone and lower cosmetic side-effect risk with dexamethasone. Given the retrospective design and dose imbalance between groups, these findings should be considered hypothesis-generating. Prospective randomized controlled trials with dose-matched protocols and mechanistic biomarkers are needed to confirm these observations and elucidate underlying mechanisms. Pending such confirmation, triamcinolone may be a reasonable first-line choice for local injection therapy in adult trigger finger. Funding Statement The author(s) declared that financial support was received for this work and/or its publication. The authors declare funding from Jinjiang Municipal Hospital Research Project (2023LC03), Quanzhou Science and Technology Plan Project (2023NS100; 2025QZNYZ043), Scientific Research Fund of Shanghai Sixth People’s Hospital (ynhglg202404), and Shanghai University of Medicine & Health Sciences University-level Research Fund (Grant No. XJJJ2502-2). Data Sharing Statement The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author Gaofeng Huang. Ethics Statement Documented review and approval was obtained from review board of Jinjiang Municipal Hospital (Shanghai Sixth People’s Hospital Fujian). Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements. Disclosure The authors declare that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. References Associated Data Data Availability Statement The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author Gaofeng Huang.