The Role of Tumor Extent and Post-Embolization Residual Vascularity in Endoscopic Endonasal Surgery for Juvenile Nasopharyngeal Angiofibroma

Article information

J Rhinol. 2025;32(3):150-154

Publication date (electronic) : 2025 November 27

doi : https://doi.org/10.18787/jr.2025.00037

Joo Hyun Park, MD, Donghyeok Kim, MD, Hyo Yeol Kim, MD, PhD, Yong Gi Jung, MD, PhD, Gwanghui Ryu, MD, PhD, Sang Duk Hong, MD, PhD

Department of Otorhinolaryngology-Head and Neck Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea

Address for correspondence: Sang Duk Hong, MD, PhD, Department of Otorhinolaryngology- Head and Neck Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Republic of Korea Tel: +82-2-3410-3579, Fax: +82-2-3410-6987, E-mail: kkam97@gmail.com

Received 2025 July 2; Revised 2025 July 8; Accepted 2025 July 27.

Abstract

INTRODUCTION

Juvenile nasopharyngeal angiofibroma (JNA) is a rare, benign, but locally aggressive vascular tumor that predominantly affects adolescent males. Management of JNA typically involves preoperative embolization aimed at minimizing intraoperative bleeding [1,2]. However, the impact of residual vascularity after embolization on surgical outcomes remains uncertain [2]. Various staging systems have been proposed over the years to classify JNA based on tumor extent and the involvement of surrounding structures [3]. More recently, there has been increasing interest in evaluating residual vascularity post-embolization as a possible prognostic factor, as reflected in the University of Pittsburgh Medical Center (UPMC) staging system [4].

The introduction of endoscopic surgical techniques has substantially improved the management of JNA by enabling more precise tumor resection and reducing morbidity [5-9]. Despite these advances, significant intraoperative challenges remain, such as considerable blood loss and the risk of residual tumor. Previous studies have indicated that the UPMC staging system—which incorporates both residual vascularity and the route of intracranial extension—offers superior prognostic value compared to traditional staging systems [10].

This study aims to evaluate the prognostic significance of residual vascularity relative to tumor extent in patients undergoing endoscopic endonasal surgery for JNA.

METHODS

Data collection

Collected data included patient demographics (age, sex), tumor extent, residual vascularity status following embolization, and clinical outcomes. Clinical outcomes assessed were estimated blood loss (EBL), transfusion requirement, total operation time, and the presence of residual tumor on the first postoperative imaging follow-up. Residual vascularity status was determined by a radiologist’s interpretation of post-embolization imaging, where the presence of post-embolization stain indicated RV+, and its absence indicated RV- (Fig. 1). Tumor extent was defined relative to the pterygomaxillary fissure (PMF): tumors involving only the pterygopalatine fossa were classified as limited, whereas those extending beyond the PMF were classified as extensive.

Fig. 1.

Two cases with similar anatomic extension, demonstrating the presence or absence of residual vascularity after embolization. A1 and B1: Enhanced computed tomography scans show extensive angiofibroma extending to the infratemporal fossa. A2 and B2: Angiography reveals extensive vascular staining surrounding the tumors. A3 and B3: Following embolization, A3 displays minimal vascularity, whereas B3 demonstrates residual vascularity. A4 and B4: Endoscopic views show minimal intraoperative bleeding in A4, but extensive bleeding in B4. A5 and B5: Postoperative computed tomography scans confirm no residual tumor in both A5 and B5.

RESULTS

Patient demographics

The mean age of the patients was 18 years, and all were male. The median follow-up period was 18 months, with a range of 1 to 88 months. Nine patients (31%) had feeding arteries originating from the ICA, and eight patients (27.6%) exhibited bilateral feeding arteries. Revision surgery was required in three patients (10.3%) (Table 1).

Table 1.

Demographic characteristics

Clinical outcomes

Patients without residual vascularity (RV-) had an EBL of 410.00±361.94 mL, while those with residual vascularity (RV+) had an EBL of 1,384.74±1,634.84 mL (p=0.115). Among patients with limited tumors, the EBL was 350.59±308.15 mL, whereas for those with extensive tumors, it was 2,037.50±1,757.47 mL (p<0.001) (Table 2).

Table 2.

Comparative analysis of surgical outcomes based on residual vascularity and tumor extent

In the RV- group, one patient (9.1%) required a transfusion, compared to seven patients (38.9%) in the RV+ group (p=0.110). Among patients with limited tumors, one patient (5.9%) required a transfusion, whereas seven patients (58.3%) with extensive tumors did so (p=0.003) (Table 2).

The total operation time for RV- patients was 162.50±115.70 minutes, compared to 242.21±187.45 minutes for RV+ patients (p=0.138). For patients with limited tumors, the operation time was 113.82±34.59 minutes, while for those with extensive tumors, it was 357.67±180.19 minutes (p=0.001) (Table 2) .

Residual tumors were observed in one patient (10%) in the RV- group and in four patients (21.1%) in the RV+ group (p=0.633). None of the patients with limited tumors had residual tumors, while five patients (41.7%) with extensive tumors did (p=0.007) (Table 2).

DISCUSSION

Our study aimed to evaluate the impact of residual vascularity and tumor extent on surgical outcomes in patients with JNA undergoing endoscopic endonasal surgery with preoperative embolization. Our findings demonstrate that extensive tumor involvement, particularly beyond the PMF, is a statistically significant predictor of increased EBL, higher transfusion requirements, longer operative times, and a greater incidence of residual tumor. While the presence of residual vascularity was associated with trends toward increased EBL, transfusion rates, and operative time, these differences did not reach statistical significance in our cohort. The patients in this study were exclusively male, with a mean age of 18 years and a median follow-up period of 18 months.

JNAs are characteristically highly vascular tumors, making PIAE a standard practice to reduce intraoperative blood loss and facilitate surgical resection [10]. Our observation that RV+ patients had higher mean EBL (1,384.74 mL vs. 410.00 mL for RV-) and a greater proportion requiring transfusion (38.9% vs. 9.1% for RV-) aligns with established knowledge that residual vascularity complicates surgery. Although these trends were not statistically significant in our dataset (p=0.115 for EBL, p=0.110 for transfusion), they are consistent with previous reports highlighting the challenges posed by persistent blood supply, especially from the ICA. For example, Snyderman et al. [4] noted that residual vascularity from ICA tributaries—which are generally not embolized due to the high risk of complications—strongly correlates with increased blood loss and the need for multiple procedures. Overdevest et al. [10] also demonstrated that JNA tumors supplied by the ICA had significantly higher mean EBL (1,738 mL) compared to those supplied solely by the external carotid artery (762 mL; p<0.01). Furthermore, tumors with bilateral vascular supply exhibited markedly increased mean intraoperative blood loss (2,146 mL) compared to those with unilateral supply (617 mL; p<0.01). In our cohort, 31% of patients had feeding arteries from the ICA, and 27.6% had bilateral feeding arteries. The absence of statistical significance for residual vascularity in our series, despite these observed trends, may be attributable to the relatively small sample size of 29 patients, in contrast to larger cohorts analyzed in systematic reviews or meta-analyses.

Our study highlights the critical importance of tumor extent in predicting surgical outcomes. Tumors classified as “extensive” (i.e., extending beyond the PMF) were associated with significantly higher EBL (2,037.50 mL vs. 350.59 mL), increased transfusion rates (58.3% vs. 5.9%), longer operative times (357.67 min vs. 113.82 min), and a higher incidence of residual disease (41.7% vs. 0%) compared to “limited” tumors (all p<0.001, p=0.003, p=0.001, or p=0.007). This finding is closely related to the involvement of the infratemporal fossa (ITF), a complex anatomical space bounded by the posterior maxilla, lateral pterygoid plate, styloid process, and mandibular ramus. Endoscopic endonasal approaches to the ITF, as classified by Taylor et al. [5], involve progressive dissections—transpterygopalatine fossa, transmedial pterygoid plate, and translateral pterygoid plate approaches. These more extensive surgical approaches increase the risk of complications, such as V2/palatal numbness, Eustachian tube dysfunction, and trismus. Our findings underscore that greater anatomical involvement necessitates more complex surgical interventions. Bignami et al. [2] similarly identified ITF involvement as a “red flag” during surgical planning and preoperative counseling, directly correlating with an increased risk for transfusion.

The results of our analysis also inform the ongoing discussion regarding the utility of JNA staging systems. Contemporary staging systems, including the UPMC system, were developed to better accommodate current endoscopic surgical practices and to incorporate important prognostic factors such as the route of intracranial extension and residual vascularity following embolization. Snyderman et al. [4] demonstrated that the UPMC system more accurately predicted immediate morbidity—including blood loss, the need for multiple operations, and tumor recurrence—than earlier systems. Rowan et al. [3] further supported the predictive validity of the UPMC system for staged procedures, intraoperative transfusions, and postoperative residual disease, with higher area under the curve values than other conventional systems. However, our study did not clarify the precise prognostic role of residual vascularity, emphasizing the need for further research with larger patient cohorts.

Clinically, our results reinforce the importance of comprehensive preoperative assessment—including computed tomography, magnetic resonance imaging, and angiography—to achieve precise tumor localization and thorough characterization of vascular supply. Patients with extensive tumors, particularly those involving the ITF, should be carefully counseled regarding the increased risk of significant blood loss, the potential need for transfusion, longer operative times, and the possibility of residual disease. Although dural opening is uncommon in most ITF cases, skull base reconstruction with a nasoseptal flap is recommended if the dura is violated; otherwise, materials such as Surgicel, Floseal, and NASOPORE are used. The continued adoption of endoscopic endonasal surgery as the standard of care for JNA is justified by its minimally invasive nature and reduced tissue disruption compared to traditional open approaches. Nevertheless, successful outcomes depend heavily on meticulous surgical planning based on a detailed understanding of the tumor’s anatomical extent and vascular supply.

In light of the limitations of current staging systems, we explored a streamlined approach for categorizing patient cohorts and predicting clinical outcomes. Our method primarily relies on distinguishing tumor extent based on the PMF. We propose separating tumors into those of limited extent and those with significant intracranial extension. Additionally, for tumors that cross the PMF but do not extend intracranially, we specifically assessed for residual vascularity to anticipate the need for EBL management or transfusion (Fig. 2).

Fig. 2.

Proposed classification of JNA. JNA, juvenile nasopharyngeal angiofibroma; PMF, pterygomaxillary fissure.

Our study has several limitations. The retrospective design is inherent to its methodology. The relatively small cohort (29 patients) may have reduced statistical power to detect significant differences for certain variables, such as the effect of residual vascularity, which showed clear trends but did not reach statistical significance. Moreover, residual vascularity assessment relied on a qualitative evaluation (“presence” or “absence” of post-embolization stain) rather than a quantitative metric. The median follow-up period of 18 months is relatively short and may not adequately capture delayed recurrences; however, early postoperative magnetic resonance imaging is effective in detecting persistent disease. Future multicenter studies with larger sample sizes, longer follow-up, and standardized quantitative assessment of vascularity are needed to further elucidate these relationships.

In conclusion, our study demonstrates that tumor extent, especially when involving the ITF, is a significant determinant of surgical outcomes in JNA patients, with strong correlations to increased blood loss, transfusion requirement, operative time, and residual disease. While residual vascularity also tends to predict more challenging surgical parameters, its statistical significance should be further evaluated in larger patient populations.

Article information Continued