Management of Small, Minimally Invasive Sinonasal Hemangiopericytoma
Sinonasal hemangiopericytoma (SNHPC) formerly known as glomerulopericytoma prior to WHO reclassification in 2005 , was first described by Stout and Murray in 1942 . It is a rare mesenchymal tumour of perivascular cell origin , of unknown aetiology, with no gender predilection and presents from the 3rd to 5th decade of life with increasing incidence around the 6th-7th decade [3-5]. Hemangiopericytomas (HPC) usually occur in the lower extremities, pelvis and retroperitoneum, with 15 – 30% presenting in the head and neck and of these less than 5% are in the nose and paranasal sinuses [3,5], accounting for less than 1% of sinonasal tumours.
SNHPC commonly presents with epistaxis, nasal obstruction or as a painless mass. Less common symptoms include facial pain or pressure, headache, rhinorrhoea or because of pressure on adjacent structures [3,5,6]. On examination SNHPC commonly appears as a soft or rubbery mass which can vary in colour [6,7]. Tumour size is reported as ranging from 8 mm – 800 mm [5,8] with a mean size of 31mm, indicating that most are small. Between 40% and 44% of SHNPC present within the nasal cavity, the rest arising from within the paranasal sinuses or occupying multiple sites [5,6]. The diagnosis is usually made from the histology, the characteristic features being closely packed cells with a storiform, whorled or palisading pattern interspersed with thin-walled branching “staghorn” vessels , although a variety of patterns can be seen within each tumour . Immunohistochemical stains are positive for vimentin (98% of tumours), α-smooth muscle actin (92%) and muscle specific actin (77%) but commonly negative for CD34 and bcl2 . Histological markers of poor prognosis and recurrence (both local and metastatic) include prominent mitotic activity (>4 mitoses per 10 high powered fields), focal necrosis, haemorrhage, increased cellularity, cellular atypia and pleomorphism [3,6]. Duval and Thompson [5,8] both reported that tumour size did not correspond to recurrence for SNHPC; in contrast, Enzinger  found the 10 year actuarial survival rate for HPC was associated with size. Thompson  identified 3 features statistically significant for the development of recurrence:- duration of symptoms (>10months), bone nvasion and severe nuclear pleomorphism.
HPC predominately metastasises via a haematogenous route, most commonly to bone, lung and liver and in more than twothirds of patients is preceded by local recurrence . The metastatic rate for SNHPC (5-10%) is lower than for nonsinonasal HPC (12-60%) [3,7,9]; systematic reviews of SNHPC suggest the risk is lower (1.9%), but it should be kept in mind when assessing patients with SNHPC [5,6].
Dahodwala  recommended initial imaging to assess for deep invasion and involvement of surrounding structures; followed by surgical excision. With about 20% local recurrence despite complete excision  and documented cases of local recurrence after more than a decade, lifelong follow up is recommended . We present two cases of patients with small, minimally invasive SNHPC and pose the clinical question – should management of small SNHPC disease follow the same recommendations as for larger and invasive SNPHC?
Case 1:- A 76 year old woman was referred following an incidental MRI finding of a left nasal septal polypoid lesion, likely mucosal in origin, she was asymptomatic. On examination there was a 12mm, firm, sessile swelling of from the left side nasal septum. A biopsy confirmed SNHPC without poor prognostic features (figures 1) and an endoscopic complete resection was achieved. No recurrence has been noted at 18 months follow up.
Figure 1. This sinonasal glomangiopericytoma grows beneath theepithelium and compresses adjacent seromucinous glands. It exhibits a myxoid background. The cells are spindle in form, small, and blandappearing.
Case 2:- A 64 year old man presented with reoccurrence of symptoms of nasal obstruction. Five years previously he underwent trimming of the inferior turbinates; at that time nasal endoscopy found no signs of rhinosinusitis and the nasal septum appeared normal; there was no clinical indication for imaging. When he represented nasal endoscopy showed an erythematous posterior nasal polyp. A CT scan showed mucosal thickening within the paranasal sinuses, an atelectatic right maxillary sinus and opacification consistent with a left sided nasal polyp.
Figure 2. Sinonasal glomangiopericytoma showing more traditional features including dilated vascular spaces with perivascular hyaline (a), and those in “staghorn” shapes (b). The neoplastic cells are tightly packed and show uniform size and bland nuclei (b).
He underwent endoscopic sinus surgery and the polyp was removed and its origin on nasal septum cleared with a microdebrider. Histology showed a SHNPC with a positive excision margin, there were no adverse prognostic features (figure 2). Follow up at 8 months showed normal appearing nasal mucosa with no evidence of macroscopic recurrence.
Summary of Current Recommended Management for SNHPC
The investigation and management of SNHPC is largely consistent within the literature, however the use of additional investigations or interventions is dependent on the characteristics of the tumour and the patient. Computed Tomography and/or MRI allow tumour size, location, involvement of or invasion into surrounding structures and any major vascular involvement to be assessed. When the tumour is large, appears vascular or has a large feeding vessel identified then angiography is recommended and provides the opportunity for intravascular embolization to reduce bleeding risk during resection. [5,6,9]
Wide local excision with intraoperative frozen sections is recommended to ensure clear margins and reduce recurrence risk [4-6].
Figure 3. Management of Snhpc Limited to the Nasal Cavity Flow chart of suggested management plan for small SNHPC limited to the nasal cavity, either with or without prior biopsy
In a systematic review Duval  found the risk of recurrence following complete excision to be 20-22% compared to 71- 100% when excision was incomplete; 79% of recurrences occurring within the first 5 years.
For tumours with positive resection margins further surgery with intraoperative frozen section is recommended to ensure complete clearance.
If further surgery is contraindicated due to risk or patient preference then radiotherapy may delay and reduce the rate of recurrence . With recurrence reported more than a decade after surgery, lifelong follow up is recommended.
Management of Nasal Cavity SNHPC
In applying such recommendations from literature to our patients we found limited guidance, especially regarding management of disease limited to the nasal cavity, and in particular the role and timing of further management in a patient with positive resection margins reported after tumour removal. Given its benign appearance, it is reasonable to assume that many small, unreported SNHPC proceed to an excision biopsy without imaging and subsequently may or may not proceed to further surgery. With the aim to reduce unnecessary investigations and invasive interventions for patients with small minimally invasive disease we carried out a review of literature.
Imaging With Contrast +/- Angiography +/- Embolization
A systematic review by Duval  reported tumour sizes varying from 5-140 mm with, a mean of 31 mm; this suggests that, although large tumours are identified, SNHPC are predominately small tumours. Small SNHPC, such as our cases, are more often expansile masses rather than invasive lesions  and the role of imaging for small, fully visualised disease has not been well established. In both our patients the tumours were small and limited to the nasal septum, imaging did not provide additional relevant information, but in conjunction with the clinical/surgical findings, indicated that the disease was limited to the nasal cavity.
Wide Local Excision with Intraoperative Frozen Sections
The best predictor of recurrence is resection completeness and recurrence is not related to tumour size or location [4-6]. The preferred initial treatment is complete wide local excision with intraoperative frozen section to ensure clear margins; but is it always necessary? Some advocate for frozen sections in all cases of SNHPC , but many small lesions with either a preor post-operative diagnosis of SNHPC are completely excised by obtaining a 3-5mm cuff of macroscopically normal mucosa
Further Surgery +/- Radiotherapy (Case 2)
Our second case had positive margins on histology however that biopsy did not include the polyp base which was reduced with a microdebrider and at follow up revealed no macroscopic evidence of recurrence. So in our patient, reported with positive margins, do we re-operate to attain negative margins and risk creating a septal perforation?
A systematic review estimates risk of local recurrence for SNHPC with incomplete resection as 11.5 times higher compared to complete resection . However, recurrences often present as a slow growing mass similar to the original tumour and will be similarly managed to the primary tumour, usually without subsequent recurrence [9,12]. Radiotherapy confers little advantage except in the management of unresectable disease. [5,13]
It may be considered reasonable for our second patient with no poor prognostic histological features and no macroscopic findings to continue to be managed with regular follow up. Follow up should be more frequent during the first five years and should he present with signs of recurrence then imaging both local and for intra thoracic and intra-abdominal lesions should be considered.
SNHPC is a rare tumour, with low malignant potential, which predominantly presents as disease localised to the nasal cavity; while it is easy to focus on its propensity for local recurrence, it should not be forgotten to screen for systemic disease. There is limited data available in literature and this makes establishment of management guidelines difficult. While most recommendations for SNHPC appears to be fairly consistent these do not appear to differentiate between the limited sinonasal disease as present in our cases and more extensive disease. With localised disease removal is more readily achievable with a measurable margin, recurrence is less likely and should be readily detected at follow-up; this suggests that a less intensive management approach may be reasonable (Figure 3).
8.Thompson LDR, Miettinen M, Wenig BM. Sinonasaltype hemangiopericytoma. A clinicopathologic and immunophenotypic analysis of 104 cases showing perivascular myoid differentiation. Am J Surg Pathol. 2003, 27(6): 737-749.
13.Carew JF, Singh B, Kraus DH. Hemaniopericytoma of the head and neck. The laryngoscopy. 1999, 109(9):1409-1411.