Pain Trajectories and Predictors Following Uterine Artery Embolization Using a Nitroglycerin–Lidocaine Protocol: A Prospective Study

Fatemeh Mahdavi Sabet: writing – original draft, writing – review and editing, visualization, formal analysis. Hamed Ghorani: project administration. Ali Mahdavi: project administration, supervision, data curation, methodology.

The authors affirm that this study was conducted in full accordance with the ethical principles approved by the Ethics Committee of Salamat Farda Hospital and the Shahid Beheshti University of Medical Sciences Ethics Committee (Ethical No. IR.SBMU.RETECH.REC.1404.1100). All procedures involving human participants complied with institutional ethical standards and with the Declaration of Helsinki (1964) and its subsequent amendments (most recently revised in 2013). Informed consent was obtained from all individual participants included in the study.

The authors have nothing to report.

Between April and August 2025, 37 patients underwent UAE. Three declined participation, resulting in a final cohort of 34 patients. The median age was 41.0 years (IQR: 6.25; range: 31–48 years). Baseline characteristics are summarized in Table  1 . Baseline characteristics of the study population ( n = 34). Note: Table  1 . Baseline demographic, clinical, and imaging characteristics of patients undergoing uterine artery embolization (UAE). Data are presented as numbers (percentages) or means ± standard deviations (SDs), as appropriate. Percentages are based on the total study population ( n  = 34). Cumulative percentages exceed 100% as patients frequently presented with overlapping clinical symptoms. Cumulative percentages exceed 100% because fibroid locations were assessed on a per‐patient basis, accounting for the presence of multiple fibroid subtypes within a single uterus. Calculated only for patients with a history of pregnancy ( n  = 12). The most common presenting symptoms were heavy menstrual bleeding (85.3%, 29/34) and back pain (85.3%, 29/34). The median number of fibroids was 2 (IQR: 2), and the median fibroid volume was 306.0 cm 3 (IQR: 235.0; range: 48.0–2000.0 cm 3 ). The median largest uterine dimension was 13.0 cm (IQR: 4.5; range: 3.3–19.4 cm). The median baseline hemoglobin was 12.1 g/dL (IQR: 1.35; range: 9.1–14.8 g/dL). All procedures were technically successful (100%, 34/34). The median number of PVA vials used was 4 (IQR: 2). No major adverse events or LAST were observed. Post‐procedural pain varied across time points (Figure  1 ). Mean VAS scores were 2.77 ± 2.03 at 0 h, 5.38 ± 2.81 at 1 h, 5.59 ± 2.28 at 2 h, 6.24 ± 1.67 at 3 h, 4.12 ± 1.68 at 6 h, 3.27 ± 1.48 at 12 h, and 2.69 ± 1.72 at 24 h. Pain peaked at 3 h post‐procedure and then gradually declined. A significant positive correlation was found between mean pain and fibroid volume ( ρ  = 0.51 [95% CI: 0.10–0.77], p  = 0.02). As summarized in Table  2 , following false discovery rate (FDR) adjustment, fibroid volume was positively correlated with post‐procedural pain at 1 h ( ρ  = 0.53 [95% CI: 0.12–0.78], adjusted p  = 0.05), and 2 h ( ρ  = 0.54 [95% CI: 0.14–0.79], adjusted p  = 0.05). Correlations at all other timepoints were non‐significant. Representative pain trajectories for patients at the upper and lower extremes of fibroid volume are illustrated in Figure  2 . Mean pain scores with standard deviation at different time points following uterine artery embolization (UAE). Pain was measured using a Visual Analog Scale (VAS, 0–10). The solid line indicates the mean pain score, while the error bars represent standard deviation ( n  = 34 patients). Individual data points for each patient are also plotted. Pain peaked at 3 h post‐procedure and gradually declined thereafter. Correlation between baseline fibroid volume and post‐procedural pain scores over 24 h with adjusted and non‐adjusted p ‐values. Note: Bold values represent statistically significant correlations ( p  < 0.05). Abbreviations: CI, confidence interval; FDR, false discovery rate; ρ , Spearman's rank correlation coefficient. Representative post‐procedural pain trajectories. Pain scores (VAS, 0–10) are shown for two extreme cases. (A) A 39‐year‐old woman with the largest fibroid volume (2000 cm 3 ) who required 13 vials of polyvinyl alcohol (PVA) particles. (B) A 37‐year‐old woman with the smallest fibroid volume (48 cm 3 ) who was treated via the radial artery using four vials of PVA particles. The volume of embolic agent (number of PVA vials) was not associated with pain at baseline ( ρ  = 0.35 [95% CI: −0.01–0.62], adjusted p  = 0.42), nor at any subsequent time point (adjusted p  = 0.51–0.89), or with overall mean pain ( ρ  = 0.17 [95% CI: −0.19–0.49], p  = 0.36). Furthermore, fibroid topography did not appear to influence patient pain, as there were no significant differences in mean pain scores based on the presence or absence of transmural (mean difference = 0.42 [95% CI = −0.90–1.73], p  = 0.50), intramural (mean difference = −0.73 [95% CI = −1.70–0.23], p  = 0.13), subserosal (mean difference = −0.23 [95% CI = −1.17–0.71], p  = 0.62), or submucosal (mean difference = 0.03 [95% CI = −1.07, 1.12], p  = 0.96) fibroids. Similarly, overall mean pain intensity did not significantly correlate with the largest baseline uterine dimension ( ρ  = −0.19 [95% CI: −0.51–0.18], p  = 0.31). Additionally, there was a non‐significant negative correlation between the number of fibroids and overall mean pain scores ( ρ  = −0.39 [95%CI: −0.69–0.01], p  = 0.06); no significant correlations were observed between the number of fibroids and pain scores at any individual time point (adjusted p ‐values ranging from 0.11 to 0.67). Stratifying the cohort by baseline clinical characteristics did not change the pain outcomes. Mean pain scores were similar regardless of pregnancy history (mean difference = 0.10 [95% CI = −0.89–1.09], p  = 0.83), concurrent adenomyosis (mean difference = −0.40 [95% CI = −1.66–0.86], p  = 0.48), or prior medical treatment for fibroids (mean difference = 0.93 [95% CI = −0.07–1.93], p  = 0.07). Ultrasound at 1‐month follow‐up demonstrated no evidence of treatment failure in any of the 34 patients. All patients were discharged within 24 h with continued NSAID therapy. Two patients required conservative management for post‐discharge pain without readmission. No readmissions were recorded during the study period.

The lead author Ali Mahdavi affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

In this prospective study, we investigated the predictors of post‐procedural pain following uterine artery embolization (UAE) utilizing a protocol incorporating prophylactic intra‐arterial nitroglycerin and low‐dose lidocaine. Our findings suggest that larger baseline fibroid volume is associated with increased post‐procedural pain. Furthermore, our observations suggest that while the low‐dose lidocaine regimen was safe, it may have been insufficient to blunt the peak of ischemic pain occurring at 3 h post‐procedure. In the absence of a control group, these findings should be interpreted as exploratory and hypothesis‐generating rather than indicative of analgesic efficacy. A principal finding of our study is the observed positive correlation between baseline fibroid volume and post‐procedural pain. Although a modest positive association between the volume of embolic material and immediate post‐procedural pain was observed, this did not reach statistical significance at any evaluated time point. Post‐UAE pain is primarily driven by ischemia and the subsequent inflammatory response within the uterus, particularly involving the normal myometrium [ 16 ]. Prior studies have demonstrated that pain correlates more closely with the degree of myometrial ischemia than with fibroid characteristics alone [ 6 ]. In this context, larger or more numerous fibroids often have a more complex and robust blood supply, necessitating a greater embolic load to achieve adequate devascularization, thereby increasing the extent of ischemic injury within adjacent uterine tissue [ 17 ]. An illustrative example consistent with this concept has been described by Leo et al. [ 18 ], who reported a patient with a large, highly vascular intramural fibroid requiring a substantial embolic load and experiencing severe post‐procedural pain with prolonged hospitalization of 14 days. These observations are consistent with the hypothesis that fibroid size, vascularity, and the extent of embolization collectively influence post‐embolization symptom severity. Given the limited sample size and exploratory design, these findings should be regarded as hypothesis‐generating and require confirmation in larger studies that also account for embolic load, vascularity, and infarction extent. While our observed peak timing differs slightly from previous reports, which range from an early 1‐h peak [ 10 ], to a delayed 6‐h peak depending on the intervention [ 12 ], it highlights a fundamental temporal mismatch. Lidocaine has a relatively short half‐life of approximately 90–120 min [ 9 ]; and its analgesic effect is therefore largely confined to the early post‐procedural period. Following embolization, arterial occlusion induces hypoxia and cellular necrosis, triggering a progressive release of inflammatory mediators that profoundly amplify visceral nociception [ 7 ]. Because this ischemia‐driven inflammatory response continues to evolve over several hours, the transient sodium channel blockade provided by lidocaine dissipates before the pain reaches its maximum intensity [ 19 ]. Consequently, the temporal overlap between peak drug efficacy and peak pain remains suboptimal. Dose may represent an additional factor. Previous studies demonstrating analgesic benefit have generally used higher total doses of intra‐arterial lidocaine (200 mg) [ 9 , 11 ], whereas lower or moderate doses have shown less consistent effects [ 10 , 20 ]. The 20 mg dose used in our protocol may therefore fall below the threshold required for sustained analgesic efficacy, particularly in the setting of evolving ischemic injury. To determine whether higher doses might be required for sustained analgesia, future studies should employ controlled, dose‐ranging designs with standardized administration protocols to better define the optimal lidocaine regimen that maximizes pain control while preserving embolization efficacy. A potential concern with intra‐arterial lidocaine administration is its possible detrimental effect on embolization efficacy [ 10 ]. Previous reports have suggested that intra‐arterial lidocaine administered prior to embolization or during the procedure may alter distal flow or induce vasospasm, potentially affecting embolic distribution and treatment outcomes [ 11 , 13 ]. This concern is supported by findings from Noel‐Lamy et al., who hypothesized that distal vasospasm occurred when mixing lidocaine with embolic particles. In their study, this approach was associated with a trend toward fewer vials of embolic agent used and a significantly lower rate of complete fibroid infarction on follow‐up MRI. Importantly, they noted no such decrease in infarction rates when lidocaine was administered after embolization [ 11 ]. Similarly, Keyoung et al. reported that intra‐arterial administration of lidocaine resulted in pronounced vasospasm, leading to premature termination of the study, despite the use of a modified protocol incorporating a reduced dose of 100 mg [ 13 ]. Although evidence remains limited and heterogeneous, administering lidocaine after completion of embolization represents a pragmatic strategy to avoid this potential confounding effect while preserving procedural integrity. Furthermore, we deliberately chose to administer the lidocaine only after the true “near stasis” endpoint had been angiographically confirmed. The physiological rationale for this specific endpoint is that it allows for partial recanalization, thereby sparing the healthy myometrium from profound ischemia and inherently reducing post‐procedural pain [ 21 ]. To further optimize catheter flow and reduce the risk of vasospasm, intra‐arterial nitroglycerin was administered prior to embolization at a dose of 100 µg per uterine artery, to proactively prevent vasospasm and preserve arterial patency. Finally, we used a lower total dose of lidocaine than in previous studies to reduce potential pro‐thrombotic or vasoactive effects. This strategy was designed to mitigate the risk of incomplete embolization while providing effective pain control. Given the temporal limitations of lidocaine, there is increasing interest in the use of longer‐acting intra‐arterial anesthetics that may better align with the delayed peak of ischemic pain following UAE. Bupivacaine offers prolonged analgesia due to its higher hydrophobicity but carries concerns regarding cardiotoxicity with intravascular use [ 22 ]. Ropivacaine, a structurally similar agent with lower cardiac and CNS toxicity, has demonstrated effective analgesia for up to 9 h after UAE and has shown promising early results in emerging protocols, although further controlled evaluation of intra‐arterial use is needed [ 22 , 23 ]. While not yet widely adopted in routine practice, these agents may represent a potential direction for future investigation. Beyond intra‐arterial pharmacologic strategies, superior hypogastric nerve block (SHNB) is increasingly recognized as an effective modality for post‐UAE analgesia, significantly reducing opioid requirements and facilitating early discharge [ 4 , 24 ]. However, SHNB was intentionally excluded from the present study. Since our primary objective was to isolate the effects of intra‐arterial lidocaine within a standardized protocol, inclusion of a potent regional block would have substantially confounded pain assessment. Future studies should explore whether combining SHNB with optimized intra‐arterial strategies provides additive or synergistic benefit. Importantly, our experience demonstrated a favorable safety profile. We observed no intra‐procedural or post‐procedural complications attributable to lidocaine. This is consistent with other investigations of intra‐arterial lidocaine in the UAE. Similar to previous studies, no complications or adverse effects were seen in our cohort [ 9 , 13 ]. This study has several important limitations. Most importantly, the absence of a control group precludes any causal inference regarding the analgesic efficacy of the nitroglycerin–lidocaine protocol. Additionally, although systemic analgesia was standardized across patients, it may have attenuated observable differences in pain attributable to intra‐arterial interventions. The relatively small sample size further limits statistical power and may result in unstable estimates of association; accordingly, observed correlations should be considered exploratory. The intensive multimodal systemic analgesia protocol likely played a dominant role in pain control, potentially masking any modest effect of intra‐arterial lidocaine. Finally, follow‐up imaging was performed using ultrasound rather than MRI, limiting the ability to accurately assess the extent of fibroid infarction. Therefore, definitive conclusions regarding treatment completeness cannot be drawn.

In conclusion, baseline fibroid volume may be associated with post‐UAE pain and may reflect the overall ischemic burden of the procedure. The use of prophylactic intra‐arterial nitroglycerin in combination with post‐embolization lidocaine was safe but did not appear to substantially alter the expected pain trajectory, potentially due to dose and temporal limitations. These findings should be considered hypothesis‐generating. Future adequately powered randomized controlled trials are warranted to investigate whether the nitroglycerin‐first approach can facilitate the safe use of higher, more effective doses of lidocaine, and to better define the multifactorial drivers of post‐UAE pain.

Uterine fibroids are the most common benign uterine tumors, predominantly affecting women in their reproductive years, with an increasing prevalence with age [ 1 , 2 ]. Uterine artery embolization (UAE), a minimally invasive procedure, was first introduced in 1995 as a successful alternative to traditional surgical treatments for symptomatic uterine fibroids [ 3 ]. Since then, it has become a well‐established minimally invasive treatment option, suitable for patients aiming to preserve their uterus [ 1 , 4 , 5 ]. Despite its advantages, a major limitation of the UAE is post‐procedural pain, which is a main feature of postembolization syndrome (PES) [ 6 , 7 ]. This is believed to be induced by transient ischemia of myometrium and subsequent tissue necrosis [ 4 , 6 ]. Typically, pelvic pain and cramping emerge within the first 2–3 h after the procedure, intensify over the next 8–12 h, and then gradually subside [ 8 ]. Several interventions have been proposed for pain management following UAE, including intra‐arterial lidocaine administration; however, existing literature demonstrates heterogeneity in both methodology and reported efficacy [ 9 , 10 , 11 , 12 ]. For instance, Keyoung et al. reported cases of moderate to severe vasospasm when 50 mg of 1% intra‐arterial lidocaine were administered into each uterine artery before embolization [ 13 ]. Conversely, Noel‐Lamy et al. reported no adverse events with lidocaine but found a significantly lower rate of complete leiomyoma infarction when lidocaine was mixed with embolization particles compared with administration after the embolization endpoint, with a nonsignificant trend toward the use of fewer PVA particle vials in the lidocaine–PVA group. They hypothesized that even at a diluted concentration of 0.5% lidocaine may induce distal vasospasm, thereby impairing embolization efficacy [ 11 ]. The present study aimed to characterize post‑procedural pain trajectories and identify predictors of pain in patients undergoing UAE within the context of a standardized analgesic protocol incorporating post‑embolization intra‑arterial lidocaine. Given that intra‐arterial lidocaine has been associated with vasospasm that may affect embolization efficacy, prophylactic intra‐arterial nitroglycerin was administered to mitigate vasospasm‐related confounding [ 11 , 13 ]. Nitroglycerin is a nitric oxide–mediated vasodilator widely used in endovascular procedures; it promotes smooth muscle relaxation and arterial dilation, thereby helping maintain luminal patency and facilitating distal embolic particle distribution [ 14 , 15 ]. Within this standardized framework, we sought to describe the temporal pattern of post‐embolization pain and evaluate its associations with baseline clinical and imaging variables.

The corresponding author (Ali Mahdavi) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained. All authors have read and approved the final version of the manuscript. Ali Mahdavi had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.

The authors declare no conflicts of interest.

We enrolled consecutive women aged 18 years or older with symptomatic, imaging‐confirmed uterine fibroids scheduled for UAE. Exclusion criteria included pregnancy, desire for future fertility, active pelvic infection, known hypersensitivity to lidocaine or iodinated contrast media, severe hepatic or renal dysfunction. This was a prospective single‐arm observational study approved by the Institutional Review Board (IR.SBMU.RETECH.REC.1404.1100). Consequently, all enrolled participants received the standardized nitroglycerin‐first lidocaine protocol during their UAE procedure. Informed consent was obtained from all individual participants included in the study. Baseline data, including patient demographics, obstetric history, and clinical symptoms, were prospectively collected. Pre‐procedural imaging (pelvic ultrasound or MRI) was used to document the number of fibroids, dominant fibroid volume, and the largest uterine dimension. Standard pre‐procedure laboratory tests were performed. All procedures were performed under conscious sedation by a single interventional radiologist (A.M.) with 10 years of experience. Vascular access was via the common femoral or radial artery. To prevent vasospasm, 100 µg of intra‐arterial nitroglycerin was administered into each uterine artery before embolization. Embolization was performed with polyvinyl alcohol (PVA) particles (250–350 µm, escalating to 300–500 or 500–700 µm) to the angiographic endpoint of near stasis, defined as cessation of forward flow for 10 cardiac cycles. Once this endpoint was angiographically confirmed, 2 mL of 1% lidocaine (20 mg per artery), diluted to 4 mL with normal saline, was infused intra‐arterially over 30–60 s. The sequence was repeated on the contralateral side. Following the procedure, patients received a standardized multimodal analgesia protocol. A continuous intravenous infusion was administered, containing morphine (20 mg) or fentanyl (500 µg) (depending on availability or operator preference), midazolam (5 mg), and ondansetron (6 mg) mixed in 100 mL of normal saline. This was delivered at a fixed rate of 4–8 mL per hour. Because this was a continuous fixed‐rate infusion rather than a patient‐controlled analgesia (PCA) device, individual opioid consumption metrics were not recorded. In addition to the continuous infusion, patients received scheduled intravenous ketorolac (a total of 60 mg) and intravenous acetaminophen (a total of 2000 mg) over the 24‐h monitoring period. Throughout this time, patients were monitored for sedation level, respiratory status, and signs of local anesthetic systemic toxicity (LAST). The primary outcome was the post‐procedural pain trajectory, measured with a 10‐point Visual Analog Scale (VAS) at baseline (0 h) and at 1, 2, 3, 6, 12, and 24 h; overall mean pain was calculated as the average of these time points. Secondary outcomes included technical success, adverse events, and 1‐month imaging follow‐up. All data were analyzed using R software (v4.3.2). The normality of continuous variables was assessed using the Shapiro–Wilk test. Normally distributed data were summarized as mean ± standard deviation (SD), whereas non‐normally distributed data were reported as median with interquartile range (IQR). Categorical variables were expressed as frequencies and percentages. Associations between continuous variables, including baseline fibroid volume and mean VAS pain scores, were analyzed using Pearson's or Spearman's correlation coefficients, as appropriate. To account for multiple testing across different time points, p ‐values for correlations were adjusted using the False Discovery Rate (FDR) method. Correlation estimates are presented alongside their 95% confidence intervals (CIs). A two‐tailed p ‐value < 0.05 (or FDR‐adjusted p ‐value, where applicable) was considered statistically significant. For comparisons of group means, differences between means (mean difference) were calculated along with their 95% CIs to provide an estimate of the effect size and its precision.