Role of Transarterial Angiography with Embolization in Deciding Surgical Approach to Juvenile Nasopharyngeal Angiofibroma: A Step-Ladder Approach

Abstract

Objective  This study was aimed to understand the usefulness of transarterial angiography and embolization in management of juvenile nasopharyngeal angiofibroma (JNA) and to apply the information obtained to stage the disease, select appropriate surgical approach, predict intraoperative bleeding, and prognosticate the disease.

Study Design  This study represents a retrospective review of the patients of JNA with major focus on transarterial angiography and embolization findings and its clinical and surgical implications.

Setting  The study conducted at a tertiary-care super-specialty referral center.

Participants  Forty-two patients who had undergone transarterial angiography and embolization followed by surgery over the period of 5 years from July 2015 till February 2019 were participated in this study.

Main Outcome Measures  Tumor vascular pattern based on transarterial angiography, staging of the disease based on extent and vascular supply, surgical approach selected, and prognosticating the disease were determined from this study.

Conclusion  Transarterial angiography with embolization becomes a prudent preoperative investigation for management of JNA. It provides a complete three-dimensional map of the tumor with stage-specific vascular pattern, reduces vascularity of the tumor, and predicts the sites from where bleeding can occur intra operatively. Thus, it helps in choosing the appropriate surgical approach aiding in complete tumor removal with minimal bleeding.

Introduction

Juvenile nasopharyngeal angiofibroma (JNA) is a benign tumor occurring exclusively in adolescent males. It usually arises from sphenopalatine foramen and the pterygoid body area presenting as unprovoked massive nasal bleed. The etiology of JNA is explained by many theories and their growth is influenced by various growth factors as specified by Nicolai et al. ¹ Though benign the tumor tends to expand the bones and grows through the foramens and fissures in the paranasal sinus (PNS) and related areas of the skull base. Surgical excision is the main stay of treatment, although there are alternative modalities of treatment like hormonal therapy, chemotherapy, and radiotherapy but, they fail to control the disease effectively as compared with the surgical excision (Nicolai et al). ¹

The tumor is staged based on its location and extent, among various staging systems described, the Fisch staging system is most widely used (Gleeson), ² but it does not consider the tumor vascularity. The University of Pittsburg Medical Center (UPMC) staging system incorporates this essential factor (Snyderman et al). ³ Stage of the tumor, appropriate surgical approach, expertise of the operating surgeon in excising all tumor, and intraoperative bleeding are the most important factors that determine the prognosis of the disease.

As the management of JNA mainly involves control of intraoperative bleeding and maximum tumor removal, understanding the vascular pattern of the tumor becomes essential (Chan et al). ⁴ This is where preoperative transarterial angiography and embolization plays a very vital and unreplaceable role (Moulin et al). ⁵ It not only maps the main arterial feeders but also helps in deciding the approach, the site where bleeding is anticipated intraoperatively, and weather there is a possibility of residual disease postsurgery

In the study, we aim at correlating the transarterial angiography findings with the stage of tumor and discuss how the aids in selecting surgical approach to minimize blood loss and maximize complete tumor removal and thus minimize residual disease and eventually reduce recurrence.

Materials and Methods

A retrospective systematic review of all cases of JNA from July 2015 to February 2019 was done in our tertiary-care referral hospital. A total of 68 patients were diagnosed as JNA based on the clinical and radiological evaluation. Of the 68 patients, 3 patients did not return to our hospital after the diagnosis, hence we had a study group of 65 patients. Of the 65 patients, 23 patients did not undergo preoperative transarterial angiography with embolization, 6 of them were having stage-I disease, while 17 of them had stage-II JNA Thus, we had a study group of 42 patients who underwent preoperative transarterial angiography with embolization ( Fig. 1 ). These cases were analyzed for digital subtraction angiography (DSA) finding, staging, stage-specific vascular pattern, and surgical approach.

Fig. 1.

Study group. DSA, digital subtraction angiography.

Patients presenting with history of nasal bleed were subjected to a diagnostic nasal endoscopy. If a vascular mass in the nasal cavity or nasopharynx were to be found in adolescent males, a working diagnosis of JNA was made. The patients were advised to undergo contrast-enhanced computed tomography (CECT) of PNS to evaluate the site and extent of the lesion and there by stage the JNA. In case of proptosis, cranial nerve deficits, and orbital or intracranial extension on CECT, the patients were subjected to an additional contrast-enhanced magnetic resonance imaging (CEMRI). Transarterial angiography was ordered in all patients diagnosed with JNA to study the vascular pattern of the tumor. With experience, we have stopped doing DSA for stages I and II as the vascular supply was more predictable. This procedure was done 24 to 48 hours prior to the surgical excision. A fresh coagulation profile, prothrombin time, and activated partial thromboplastin time with hemogram and platelet count was done prior to the transarterial angiography to confirm a normal coagulation status.

Transfemoral catheters were passed and advanced till the carotid bifurcation and dye was injected into the external carotid artery (ECA) and internal carotid artery (ICA) systems separately and a complete vascular mapping of the tumor was performed. The study was done bilateral. Superselective cannulations of each branch of the ECA was done, the subsites of the tumor supplied by each branch were studied in detail. Selective embolization of the feeding artery was performed using poly vinyl alcohol (PVA) particles (300–500 μ) and gel foam (150–300 μ) based on the caliber of the vessel. All the major branches of ECA can be embolized, whereas the ICA and its branches cannot be embolized as it can cause cerebral vascular occlusion and vision loss due to migration of particles to ophthalmic artery leading to acute permeant blindness.

Based on the CECT, CEMRI, and transarterial angiography with embolization findings, the tumor was staged based on the modified Fisch staging system (Shah et al). ⁶ Patients were operated within 24 to 72 hours of transarterial angiography with embolization. The surgical approach was decided based on the staging system, stages I and II were resected endoscopically because the tumor was completely accessible, and the bleeding, if at all occurs can be managed endoscopically, whereas stages IIIAa _e , IIIAa _i , IIIBa _e , IIIBa _i , and stage IV required an open approach not only to expose the entire tumor but also to handle the bleeders from branches of ICA that were not embolized during transarterial angiography.

The principle of surgery remains the same irrespective of the surgical approach planned, the tumor is fully exposed by a subperiosteal dissection, based on the staging that we follow the approach is planned. After adequate exposure “centripetal dissection” of the tumor is performed toward the pterygoids. The tumor attachments, medially to the nasal mucosa posteriorly to the nasopharynx, are released using electrocautery or coblation in a medial to lateral direction. The lateral limit of the tumor in the infratemporal fossa is exposed and separated from the fat in a lateral to medial direction; superior extension through superior orbital fissure into middle cranial fossa is dissected in a craniocaudal direction. The epicenter of the tumor, the probable site of origin of the tumor, is the site were the dissection is done at the end. After the tumor dissection and hemostasis are done, the probable site of origin, usually the pterygoid body or the Vidian canal region is drilled to remove the JNA tissue imbedded in the bony trabeculae. The drilling is done under endoscopic guidance for magnification and controlled targeted drilling. This step is mandatory to avoid residual disease which over time may grow, release vascular endothelial growth factor (VEGF) and other angiogenic factors and is capable to draw blood supply from unembolized vessels mainly from branches of ICA.

Results

A total of 42 patients were included in this study, their clinical parameters that were studied are summarized in Table 1 . The mean age of the study group was 15.9 years. The analysis of transarterial angiography showed internal maxillary artery (IMA) on the ipsilateral side supplying 90.4% (38 patients) of tumor, while the ipsilateral ascending pharyngeal artery supplied 64% (27 patients) of tumor. Contralateral IMA supplied 57% of tumors, whereas ascending pharyngeal artery supplied 40.4% of tumors.

Table 1. Stage specific study group.

S. No.	Age (y)	Stage	Side	I/L ECA	C/L ECA	ICA	Primary/recurrence	Modified stage	Surgical approach
Stage I
1	14	I	L	SPA	0	0	Primary	I	E
Stage II
1	24	II	L	IMA	0	0	Primary	II	E
2	20	II	R	IMA	IMA	0	Primary	II	E
3	16	II	L	IMA, APA	IMA	L ICA-C	Primary	II	E
4	22	II	L	IMA, APA	0	0	Primary	II	E
5	18	II	L	IMA, APA	IMA, APA	0	Primary	II	E
6	15	II	R	SPA, APA	SPA, APA	0	Primary	II	E
7	12	II	R	IMA, APA	IMA, APA	0	Primary	II	E + MFD
8	17	II	L	IMA	0	0	Primary	II	E
Stage IIIAa e
1	21	IIIa	R	IMA, FA,	IMA, FA, APA	0	Primary	IIIAa e	MFD + E
2	13	III	L	IMA, APA	IMA, APA	0	Primary	IIIAa e	MFD + E
3	17	III	R	IMA, APA	IMA	0	Primary	IIIAa e	MFD + E
4	14	IIIa	L	IMA, APA	IMA, APA	0	Primary	IIIAa e	MFD + E
5	16	IIIa	R	IMA, APA, AplA	0	0	Primary	IIIAa e	E(MD)
6	14	IIIa	R	IMA, APA	0	0	Recurrence	IIIAa e	MFD + E
7	18	IIIa	L	IMA, APA	IMA, APA	0	Primary	IIIAa e	MFD + E
8	14	IIIa	L	IMA, APA	IMA	0	Primary	IIIAa e	MFD + E
Stage IIIAa i
1	16	III	R	IMA, APA	IMA, APA	R	Primary	IIIAa i	MFD + E
2	15	IIIa	R	IMA, APA	SPA	R ICA-P	Primary	IIIAa i	E + MFD
3	17	IIIa	R	APA	APA	R ICA-C/O	Primary	IIIAa i	MFD + E
4	19	IIIa	L	IMA, AplA	0	L ICA	Primary	IIIAa i	MFD + E
5	13	IIIa	R	IMA, APA, AplA, MMA	0	R ICA	Primary	IIIAa i	MFD + E
6	15	IIIa	L	IMA, APA	IMA	L ICA-C	Primary	IIIAa i	MFD + E
7	16	IIIa	L	IMA	IMA, FA, APA	R ICA-O	Primary	IIIAa i	MFD + E
8	23	IIIa	L	IMA, APA	APA	L ICA-O	Primary	IIIAa i	MFD + E
9	13	IIIa	L	SPA, APlA, MMA, AcMA	0	L ICA-C	Recurrence	IIIAa i	MFD + E
10	17	IIIa	R	SPA, APA	0	R ICA	Primary	IIIAa i	E
11	25	IIIa	L	IMA, APA	IMA	R ICA-C	Primary	IIIAa i	MFD + E
12	14	IIIa	L	IMA, Apl A	IMA, APA	L ICA-P/C/O	Primary	IIIAa i	MFD + E
13	10	IIIa	R	IMA, MMA, APlA	0	B/L ICA-C, R-O	Recurrence	IIIAa i	MFD + E
14	15	IIIa	R	IMA, APA	IMA, APA	R ICA-C	Primary	IIIAa i	MFD + E
15	11	IIIa	L	IMA, APA	IMA, APA	R ICA-C	Primary	IIIAa i	MFD + E
Stage IVA
1	18	IVa	R	IMA, APA	IMA, APA	B/L	Primary	IVA	MFD + E
2	11	IVa	L	IMA	IMA, APA	B/L	Primary	IVA	E + MFD
3	12	IVa	R	IMA, APlA	0	R ICA	Primary	IVA	E + MFD
4	14	IVa	L	0	IMA	L ICA-O	Recurrence	IVA	MFD + E
5	11	Iva	L	IMA, MMA	0	L ICA	Primary	IVA	MFD + E
6	12	Iva	L	IMA, APA	IMA, APA	B/L ICA-O	Primary	IVA	MFD + E
7	15	IVa	R	IMA, APA	IMA, APA	B/L ICA-C, R-O	Recurrence (2 times)	IVA	MFD + E
Stage IVB
1	21	IVb	L	IMA, FA, STA	IMA	B/L	Recurrence	IVB	MFD + E
2	13	IVb	L	IMA, APA	IMA, APA	R ICA	Primary	IVB	E + MFD
3	16	IVb	R	IMA, MMA, FA, TFA	0	R ICA-O/C	Primary	IVB	MFD + E

Abbreviations: AcMA, accessory meningeal artery; APA, ascending pharyngeal artery; APlA, ascending palatine artery; B/L, bilateral; C/L, contralateral; E, endoscope-guided approach; ECA, external carotid artery; FA, facial artery; ICA, internal carotid artery; ICA-C, ICA-cavernous branches; ICA-O, ICA-ophthalmic artery; ICA-P, ICA-petrous branches; I/L, ipsilateral; IMA, internal maxillary artery; L, left; MFD, midfacial degloving; MMA, middle meningeal artery; R, right; SPA, sphenopalatine artery; TFA, transverse facial artery.

In our analysis of the arterial supply pattern of the JNA, we found that, in stage-II disease, ipsilateral IMA had supplied 87.5% of the tumors, the ipsilateral ascending pharyngeal artery (APA) was the second most common artery supplying 62.5%, while only one patient had supply from ipsilateral sphenopalatine artery (SPA) alone. Among the contralateral feeders, again IMA was the most common artery supplying 50% of stage-II tumor. Contralateral APA and SPA supplied 37.5 and 12.5% of tumors, respectively. One patient with stage-II disease had feeder from ipsilateral cavernous part of ICA (from the Vidian artery). Of the stage-III tumors, stage IIIAa _e was found in eight patients. Among these eight patients, all of them had vascular supply from the ipsilateral IMA, APA was supplying 85.7% of the patients, while ipsilateral facial artery (FA) was supplying 28.6%. Contralateral IMA supplied 71.4%, APA supplied 42.9%, and FA supplied 14.3%. Among the 15 patients with stage-IIIAa _i tumor, the ipsilateral IMA supplied 93.3% of the patients ipsilateral APA and FA supplied 66.7 and 26.7%, respectively. The majority of contralateral supply was from IMA, supplying 53.3% of patients. The contralateral APA supplied 46.7% and contralateral FA supplied 13.3% of the patients. The ipsilateral ICA supplied 80% of the patients with this stage, in addition to the ipsilateral ICA, the contralateral ICA supplied only 26.7% ( Fig. 2 ).

Fig. 2.

Stage-wise vascular pattern of JNA base on transarterial angiography. APA, ascending pharyngeal artery; C/L, contralateral; FA, facial artery; ICA, internal carotid artery; I/L, ipsilateral; IMA, internal maxillary artery; JNA, juvenile nasopharyngeal angiofibroma; SPA, sphenopalatine artery.

Ipsilateral IMA, APA, and FA supplied 85.7, 43.9, and 14.3%, respectively of stage IVA stage tumor, while contralateral IMA, APA, and FA supplied 71.4, 57.1, and none, respectively. All of the patients had supply from ipsilateral ICA while contralateral ICA supplied 57.1% of the patients ( Fig. 2 ).

The vascular mapping of the stage IVB tumors showed the ipsilateral IMA supplying tumor all the patients, 33.3% each by ipsilateral APA and FA. The contralateral IMA supplied 66.7% while APA supplied 33.3% of the patients. Both the ipsilateral and contralateral ICA supplied the tumors in 66.7% of the patients ( Fig. 2 ).

Analysis of the stage-specific surgical approach showed that stage-I tumor was endoscopically resected, seven out of eight patients (87.5%) of the stage-II tumors were endoscopically resected, while only one patient (12.5%) underwent midfacial degloving (MFD). In patients with stage IIIAa _e, , seven out of eight patients (87.5%) had endoscope assisted MFD approach, while one (12.5%) had endoscopic modified Denker's approach. Fourteen out of fifteen (93.3%) patients with stage IIIAa _i, had endoscope-assisted MFD approach, while one patient (6.6%) had endoscopic excision. All the patients with stage-IV tumor underwent surgery by endoscope-assisted MFD approach ( Table 1 ).

In the study, we had six patients with recurrent tumor. Of the six patients, one was our own follow-up patient, the other five were operated in other centers and were referred for the management of recurrent tumor. One patient had stage IIIAa _e with vascular supply from only ipsilateral IMA and APA and underwent excision by MFD approach. Two patients had stage IIIAa _i . _. One of them had tumor supply from middle meningeal artery and accessory middle meningeal artery on the ipsilateral side and ipsilateral cavernous part of ICA while the other patient had feeders from ipsilateral IMA, middle meningeal artery (MMA), FA and bilateral ICA, bilateral cavernous part, and the contralateral ophthalmic part. Two of the recurrences were of stage IVA, one had feeder from contralateral IMA and ipsilateral ICA; this patient had undergone the primary surgery in our center, had developed recurrence after 3 years, the primary tumor was stage IIIAa _e , had vascular supply from bilateral IMA and APA, and was operated by MFD assistance with endoscope. The other patient with stage-IVA recurrence had blood supply from bilateral IMA and APA and bilateral ICA. Only one patient with recurrence had stage IVB tumor with vascular feeders from bilateral IMA and bilateral ICA. All the recurrent tumors were surgically resected by a combined endoscopic and MFD approach.

Discussion

Angiofibroma, due to their benign nature and relatively slow growing course, expands bony walls and enters into foramen and fissures of the PNS, ITF, orbit, and skull base. The major concern about surgical resection are its deep-seated difficult anatomical location and extensive vascularity which together makes it challenging to control intraoperative bleeding (Tyagi et al). ⁷ As the patients are adolescent males, the maximal allowable blood loss during surgery is also less and the possibility of rapid hemodynamic instability thus, minimizing intraoperative blood loss becomes more important in these patients (Gross), ⁸ DSA and embolization can help in this regard. Though we have various methods to reduce bleeding like direct puncture embolization (DPE), coblation, or laser (Shah et al). ⁶ These all aim at working extravascular and all these have their own limitation.

Transarterial angiography with embolization aims at working at the intravascular level. The tumor vascularity is studied in detail, each and every branch of the ECA is superselectively cannulated, and dye is passed into specific area of tumor that each particular branch supplies. Thus, transarterial angiography provides a detailed analysis of the subsites of the tumor and its feeding vessels. Transarterial angiography provides a stage-specific vascular pattern, this gives the surgeon a thorough knowledge of the vessels suppling the specific part of the tumor ( Fig. 3 ). Superselective embolization of the feeding vessels using PVA particle and gel foam knock-off the blood supply which can be evident from the disappearance of tumor blush in the post embolization dye injection. Embolization makes the angiofibroma less vascular, thus aiding in drastically reducing intraoperative blood loss. The “angio” component of the angiofibroma is reduced preoperatively as much as possible and gives the surgeon the confidence to operate on a “less angio” and “more fibrous” benign tumor.

Fig. 3.

Transarterial angiography patterns of JNA. ( A ) Vascular supply by ipsilateral IMA and APA, ( B ) disappearance of tumor blush post embolization, ( C ) vascular supply by ipsilateral APA, ( D ) vascular supply from contralateral IMA and APA, ( E ) vascularity from ipsilateral MMA in recurrent tumor, ( F ) vascular supply from ipsilateral ICA-cavernous part, and ( G ) vascular supply from ipsilateral ICA-petrous part. APA, ascending pharyngeal artery; ICA, internal carotid artery; IMA, internal maxillary artery; JNA, juvenile nasopharyngeal angiofibroma.

Transarterial angiography with embolization gives vascular mapping of the tumor and reduces the vascularity as much as possible. The operating surgeon now knows which part of the tumor is supplied by which branch of ECA or ICA and also knows which part still has its vascularity undisturbed (areas supplied by branches of ICA as they are not embolized), the potential areas of intraoperative bleed. Based on this, the surgeon can plan the surgical approach as to which approach can give the best possible access for complete tumor removal and space for instrumentation to control bleeding. Studies have shown few drawbacks of performing preoperative transarterial angiography with embolization and have attributed it to be the cause of residual tumor. Mishra and Verma ⁹ have discussed that how the transarterial angiography and embolization can lead to increase in the friability of the lesion, especially the finger like projections that go into the cancellous bone and are left behind during resection. These are the reason for residual disease and subsequent recurrence. This can be overcome by drilling the specific areas where there is the possibility of cancellous bone involvement, like the Vidian canal and the pterygoid base (Howard et al, Thakar et al, and Tyagi et al ). ^{7 10 11}

In our study, all stage-I tumors were adequately excised by endoscopic approach. Of the stage-II tumors, all except one was excised by endoscopic approach. One patient had the tumor resection done by endoscope-assisted MFD approach, as the patient was only 12 years old with lesser body weight, and hence it allowed less blood loss, and since the tumor had its vascularity from bilateral IMA and APA, we decided to do an open approach to reduce blood loss. As our experience with stage-II tumors increased, we could understand that the vascularity is more predictable and bleeding is controllable. In our study group, we had 23 patients who did not undergo DSA with embolization preoperatively, they were stages I (6 patients) and II (17 patients) tumors and they were in the later period of our study. We subsequently stopped doing DSA with embolization for stages I and II, as we are confident in complete tumor removal by endoscopic approach with minimal blood loss.

Among the stage IIIAa _e tumors, one patient had blood supply from only ipsilateral IMA, APA, and AplA (ascending palatine artery), we attempted modified Denker's approach and we were able to completely excise the tumor and control the bleeding. All the other patients underwent resection by endoscope-assisted MFD. We examine the cavity with endoscope and perform controlled drilling of suspicious sites of residual tumor like the Vidian canal, pterygoid base and its marrow. Among the 15 patients with stage IIIAa _i , one underwent excision by endoscopic approach as the tumor vascularity was only from the ipsilateral SPA, APA, and a small feeder from ipsilateral ICA. As the ICA feeder was very small, we were confident about control of bleeding and tumor removal, so we decided to go with endoscopic resection and, in any point, if we had felt difficulty in resection or bleeding control, we always had an option of going ahead with an open approach. Thus, a surgeon must be confident about performing an open approach in early phases of their learning period and as they gain experience, endoscopic approach should be attempted. The stage IVA had vascular supply from ipsilateral IMA, AplA, and ICA, we had attempted endoscopic approach but could not handle the bleeding and had to convert into MFD approach, later we realized that feeder from AplA, and a branch of FA is difficult to control. In all the cases of the stage-IVB MFD, endoscopic approach was done. In all the recurrence cases, we were very clear to adopt an open approach due to its unpredictable vascular feeder pattern. In the cases with recurrence vascularity was from MMA, accessory meningeal artery (AcMA), and from ICA-cavernous part, as well as ophthalmic arteries. In these cases, the bleeding was more as these ICA branches cannot be embolized. Tumors with gross skull base involvement and intracranial extradural component were resected with a team, consisting of a neurosurgeon and wherever required, consultation of maxillofacial surgeon was taken. Interventional neuroradiologist was part of the team and performed the transarterial angiography, and embolization and was on standby for any vascular intervention if required during surgery.

We advise for a staged strategy in training and decision making on the surgical approach in the management of JNA. To start with, surgeons should train themselves to be confident with open approaches, gradually understand the tumor anatomy and as they become more and more experienced, a staged switch to endoscopic approaches would be the best way of training ( Fig. 4 ). Once the confident with open approach is gained, a more conservative endoscopic approach should be opted, but the golden rule is not to compromise with tumor exposure and its removal, as well as to control bleeding while removal, irrespective of the approach. Open approaches like lateral rhinotomy, MFD, maxillary swing, and transpalatal approaches are popularly used for exposing JNA. The choice of the open approach should be based on the surgeons training and confidence, which again depends on the experience of the surgeon. The best way is to get trained under a mentor and practice any one approach for best results.

Fig. 4.

Protocol for decision making in surgical approach to JNA. JNA, juvenile nasopharyngeal angiofibroma.

As far as deciding between endoscopic and open approach is concerned, the decision should be based on the vascular pattern of the tumor and experience of the surgeon. In stages I and II JNA, the vascular supply is mainly by the branches of maxillary artery and APA which are predictable and bleeding from them are controllable for surgeons who are trained in performing functional endoscopic sinus surgery (FESS). When it comes to managing a stage-III JNA, the decision making between an open approach and endoscopic approach should be guided by the experience of the surgeon. In a stage IIIAa _e and IIIBa _e tumors, a beginner or a less-experienced surgeon should get well-trained open approach. After gaining adequate experience with open approach, one should opt to start the surgery by an endoscopic approach and try to perform the entire surgery with endoscope, but if the surgeon is not able to continue the surgery due to difficulty in exposing the tumor of is not able to control bleeding, then the surgeon should switch over to an open approach. As the vascularity of this stage is only by the ECA system and a surgeon who has adequate training in handling the ECA feeders by an open approach, trying to control it by an endoscope in the guidance of a mentor is the next level of training. This strategy will help the young surgeons to understand their own limitations in performing the surgery using the endoscope. Again, this sort of a training and experiences have to be gained under an experienced mentor. A less-experienced surgeon who is not able to perform a stage IIIAa _e or IIIBa _e by endoscopic approach, should operate up on a stage IIIAa _i and IIIBa _i surgery by an open approach, as the feeders are from both ECA and ICA systems. After gaining adequate experience of performing stages IIIAa _e and IIIBa _e by endoscopic approach, one can start a stage-IIIAa _i and stage-IIIBa _i with endoscope and follow the same training strategy as in IIIAa _e and IIIBa _e . For an experienced surgeon, approaching a tumor supplied by ICA system (IIIAa _i and IIIBa _i ) by an endoscopic approach may not be difficult. As far as stage IV is concerned, majority of the tumor vascularity is unpredictable and feeders from ICA system will be higher, a less-experienced surgeon can be foe and adequate tumor exposure and handling the bleeding from ICA system may not be a comfortable task, hence an open approach is advisable. An endoscopic approach in case of a stage-IV tumor is not contraindicated, but requires a greater level of experience to handle the ICA system. In case of a recurrent JNA, deciding upon the surgical approach should be governed by the location of the tumor, the vascularity it derives, and of course the experience of the surgeon. A staged training under an experienced mentor is mandatory.

Conclusion

Preoperative transarterial angiography with embolization provides a complete vascular map of the tumor, helps in understanding the relationship of vascular supply with the stage of the tumor, predict the part of the tumor that can bleed intraoperatively, and essential to decide on the appropriate surgical approach for complete tumor resection and to minimize the bleeding, thus it must be considered as an essential investigation in the management of JNA stage III. Stages I and II can be managed by endoscopic approach without transarterial angiography as the vascular supply is predictable and bleeding can be controlled. Operating surgeon has to adopt a staged-training strategy, starting with open approaches in the initial phases of the training for smaller tumors (stages IIIAa _e and IIIBa _e ) and as the experience increases, gradually switching over to endoscopic approach, even for tumors of advanced stage and unpredictable vascular pattern.

Funding Statement

Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.