Head and neck paragangliomas in Norway, importance of genetics, updated diagnostic workup and treatment

Abstract

Background

Head and neck paragangliomas (HNPG) are rare and predominantly benign tumours, originating from the neuroendocrine paraganglionic system. A considerable proportion of HNPGs are hereditary, depending on the population.

Aims/objectives

The purpose of this study was to estimate the rate of hereditary HNPGs in a Scandinavian (Norwegian) population, report long-term experience with HNPGs and offer all patients diagnosed an updated follow-up, with emphasis on identifying hereditary HNPGs through genetic screening and multifocality by 18 F-2-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT).

Material and methods

Our study was a partly retrospective and partly prospective cohort study. It included patients with HNPG diagnosed at Oslo University Hospital (OUH), Rikshospitalet between 1990 and 2017. The patients underwent genetic testing, ¹⁸F-FDG PET/CT and measurement of catecholamines and meta-nephrines in the plasma. All resection specimens and biopsies were subjected to histopathological review. The genetic testing protocol consisted of testing for mutations in the following genes; SDHD, SDHB, SDHC, VHL and RET.

Results

Sixty-three patients were included in the study with a median age of 49 years (range 12 − 80). Cranial nerve dysfunction was present upon diagnosis in 13%, and 14% had multifocal paraganglioma (PG) disease. Fifty-six patients (89% of all the patients) underwent genetic testing, and 29% of these had a PG related mutation. Seven of the eight patients (88%) with multifocal PGs who underwent genetic testing had a mutation. In two of the patients, the ¹⁸F-FDG PET/CT revealed unknown and subclinical multifocality.

Conclusions and significance

This is the first study with systematic genetic workup and PET/CT imaging in Scandinavia of HNPG patients. The mutation rate was within the lower range reported in the literature with respect to HNPGs. Combining genetic testing and PET/CT imaging in the diagnostic workup of HNPGs is valuable.

Chinese abstract

背景：头颈部神经节瘤（HNPG）很罕见, 主要是良性肿瘤, 源于神经内分泌旁神经节系统。相当一部分HNPG是遗传的, 取决于不同的人口。

目的：本研究的目的是评估遗传性HNPGs在一个斯堪的纳维亚人（挪威人口）中的比率, 报告患HNPG的经历, 并为所有患者提供最新的随访资料。工作重点是鉴定遗传性PG, 做法是通过18 F-2-氟脱氧葡萄糖（18F-FDG）正电子发射断层显像/计算机成像技术（18F-FDG PET / CT）进行遗传筛选和多焦点性研究。

材料和方法：我们的研究, 部分是回顾性研究, 部分是前瞻性研究, 包括了在1990年至2017年之间, 于Rikshospitalet奥斯陆大学医院（OUH）诊断为HNPG的患者。患者们接受了基因检测, 18F-FDG PET / CT和儿茶酚胺测定和血浆间肾上腺素。所有切除标本和活组织检查均经过组织病理学检查。遗传测试方案包括测试以下基因的突变：SDHD、SDHB、SDHC、VHL和RET。

结果：该研究纳入了63名患者, 中位年龄为49岁（范围12至80岁）。确诊时患有颅神经功能障碍的比例为13％, 多灶性副神经节瘤为14％（PG）。 56名患者（占所有患者的89％）接受了基因检测, 其中29％发生了与PG相关的突变。接受基因检测的八名多灶性PG患者中有七名（88％）发生突变。两名患者的18F-FDG PET / CT显示未知和亚临床多灶性。

结论和意义：通过系统的遗传检查和PET / CT进行的研究HNPG患者在斯堪的纳维亚半岛的影像学检查, 这样的研究是首次。在有关HNPG的文献中报告的突变率属较低范围内。结合基因检测和PET / CT成像进行诊断HNPG的检查很有价值。

Keywords:

• Head and neck paraganglioma
• multifocality
• genetics
• PET
• Scandinavia

Introduction

Tumours originating from the neuroendocrine paraganglionic system are defined as paragangliomas (PG). One important exception are tumours arising from the paraganglionic tissue of the adrenal medulla, they are named pheochromocytomas. There are two types of PGs, sympathetic and parasympathetic. Predominantly, PGs in the head and neck area are parasympathetic, hence normally they are non-functional and specifically not cathecolamine-secreting.

Head and neck paragangliomas (HNPG) are very rare, and represent only 0.6% of all head and neck tumours and 0.03% of all tumours [1]. Among HNPGs the carotid body PG (60%) is the most common subsite, followed by the jugular and/or tympanic (30%) and vagal PGs (10%) [2]. PGs may rarely be found in the larynx, sinuses and the thyroid gland.

PGs are predominantly clinically benign, and malignancy is usually defined as the presence of metastasis in non-neuroendocrine tissues. There are no well-defined histo-morphological criteria indicating malignancy [1].

PGs may occur sporadically or in hereditary forms. Hereditary forms count for approximately 1/3 (40%) of HNPGs and between 35 and 40% of all paraganglioma-pheocromocytomas [2,3]. This may depend on the population. In the head and neck area particularly, the hereditary PGs are frequently linked to mutations in the succinate dehydrogenase (SDHx) genes. The commonest mutations causing hereditary HNPGs are found in SDHD and SDHB, the SDHD mutation being the single most frequent [4]. The SDHx mutations are part of five hereditary PG syndromes which have different features; PGL1 (SDHD mutation, most frequent mutation, rarely malignant but often multifocal), PGL4 (SDHB mutation, second most frequent hereditary HNPG, with higher risk of malignant disease and functional PGs), PGL3 (SDHC mutation, rare), PGL2 (SDHAF2 mutation, rare and associated with multifocality) and PGL5 (SDHA mutation, rare) [5]. Additionally, there are other established familial syndromes that commonly include PGs; Neurofibromatosis 1 (NF1), Multiple endocrine neoplasia type 2 (MEN2, caused by RET mutation), Von-Hippel Lindau (VHL) and Carney-Stratakis dyad [6].

The commonly agreed diagnostic approach to HNPG consists of radiologic imaging and hormone analysis for ruling out secretory activity. During the last decade, whole-body positron emission tomography/computed tomography (PET/CT) was commonly used for the diagnosis of PGs. ¹⁸F-2-fluorodeoxyglucose (¹⁸F-FDG) PET/CT is particularly valuable rule out metastatic and/or multifocal disease, especially for PGs with underlying SDHx mutations [7,8]. Genetic screening has also become an integral part of the diagnostic algorithm in HNPGs.

Management of head and neck PGs ranges from expectancy, surgery, radiation therapy or a combination of these approaches. When choosing the type of treatment, age and general health of the patient, risk of damaging neurovascular structures, presence of secretory activity, and size of the PG are main factors taken into consideration.

Material and methods

Our study was partly retrospective and partly prospective cohort, and included patients with HNPGs diagnosed and treated at Oslo University Hospital (OUH), Rikshospitalet from 1990 through 2017. The patients diagnosed before 2012 were studied retrospectively, and the patients diagnosed after 2012 prospectively. OUH is a tertiary referral hospital and the primary location for the diagnosis and treatment of HNPGs in Norway. The study was approved by the Data Protection Office at OUH as a quality control project, and all included patients signed an informed consent. Only one patient declined to participate. The retrospective portion of the patients was identified by conducting searches for relevant ICD-9 and ICD-10 codes in the patient records and the pathology registry. All resection specimens and biopsies were subjected to histopathological re-examination by an experienced head and neck pathologist to reduce the risk of misclassification.

Hereditary PG was defined as positive family history as defined in the patient record by personal interview, a personal or family history suggestive of one of the known PG-syndromes (e.g. cerebellar hemangioblastoma raising the suspicion of von Hippel Lindau syndrome), and/or detection of a pathogenic or likely pathogenic mutation in a PG-gene on molecular analysis. Patients, who had not already been examined by genetic screening and ¹⁸F-FDG PET/CT, underwent these examinations if they consented. This count for patients diagnosed with head and neck PG before 2012, the year genetic screening and ¹⁸F-FDG PET/CT became routine parts of the diagnostic approach at our hospital in addition to ultrasound, CT scans and MRI. The genetic test panel consisted of sequencing and Multiplex Ligation-dependent Probe Amplification (MLPA) for mutations in the following genes; SDHD, SDHB, SDHC, VHL and RET. All hereditary cases were offered genetic counselling at the Department of Medical Genetics.

Prior to 2012, we also commonly used fine needle aspiration (FNA) examination to assess suspicious cervical masses, but later when genetic screening and ¹⁸F–FDG PET/CT became part of the clinical routine we usually did not perform FNA if the tumour has PG typical location and the imaging, particularly ¹⁸F-FDG PET/CT, supported the diagnosis. Furthermore, all patients underwent measurement of catecholamines and meta-nephrines in the plasma.

Treatment

Prior to surgery, most patients had a carotid (angiography) occlusion test to examine collateral cerebral circulation. Surgery was performed in collaboration between an Otolaryngology – Head and Neck surgeon and a vascular surgeon. The patients selected for surgery did not routinely undergo preoperative embolization of the tumours, as there exists no consensus regarding criteria for embolization and advantages with the procedure, furthermore it carries the potential risk of severe neurological complications [9].

Surgery was mostly performed according to common practice, yet in the later part of the series we adopted a ‘packing technique’ to reduce bleeding volume. For the most common procedure, carotid body PG, a small submandibular incision (6–8 cm) was sufficient, aided by intraoperative monitoring of the vagus nerve, and keeping close attention to the hypoglossal nerve and the cervical sympathetic chain. Blood loss was moderate, and no patients required blood transfusions. In cases of carotid body PGs that were operated, jugulodigastric lymph nodes were removed to facilitate access and examined in the routine histology, but formal neck dissections were not performed.

The patients selected for radiotherapy received radiation averaging 50 Gy.

Results

Population

The detailed demographics are presented in Table 1. Sixty-three patients were included in the study, with a total of 75 tumours, and in nine (14%) of the patients more than one tumour was identified. Thirty-two patients were diagnosed before 2012 and 31 after. The majority of the patients in our study were females, representing almost 75% of the patients. The gender difference was statistically significant (p < .001, Z-test). Median age upon diagnosis was 49 years (range 12−80). Median (mean) age was considerably lower if we only considered patients with hereditary PG: 37.5 (36.9) years versus 51.0 (53.3) among the patients with non-hereditary PG (p < .001, T-test). There were few malignant PGs (n = 4) in our study, all males. Only one patient (2%) had secretory activity from his HNPG.

Table 1. Population.

Gender	Number of patients	Median age at diagnosis	Patients with malignant PG	Patients with >1 tumour
Male	17 (27%)	41	4 (24%)	5 (29%)
Female	46 (73%)	52	0 (0%)	4 (9%)
Total	63 (100%)	49	4 (6%)	9 (14%)

Subsites, cranial nerve dysfunction and treatment

Table 2 depicts the anatomic distribution of the head and neck PGs, and a summary of the management of the PGs at our hospital. Half of the tumours were carotid body PGs, and we identified three PGs with uncommon subsites, one being a laryngeal PG and the two other thyroid PGs. Ten of the patients had cranial nerve dysfunction at the time of diagnosis. This was predominantly observed with jugular PGs, and most frequently affecting cranial nerves IX-XII. With respect to management of the head and neck PGs, half of them were treated with surgery, implying total excision for the carotid body PGs and the tympanic PGs, and in some cases subtotal surgery of the jugular PG, due to the proximity of vital structures. Hence, only 33% of the jugular PGs were operated, the rest were offered radiotherapy or expectant observation.

Table 2. PG types, cranial nerve dysfunction at diagnosis and absolute management.

PG type	Number of tumours	Cranial nerve dysfunction at diagnosis	Surgery	Radiation	Observation
Carotid body	41 (55%)	1 (2%)	22 (54%)	8 (20%)	10* (24%)
Jugular	18 (24%)	8 (44%)	6 (33%)	8 (44%)	4 (22%)
Tympanic	9 (12%)	0	6 (67%)	3 (33%)	0
Vagal	4 (5%)	0	0	2 (50%)	2 (50%)
Other (thyroid and laryngeal)	3 (4%)	1 (33%)	3 (100%)	0	0
Total	75 (100%)	10 (13%)	37 (49%)	20 (27%)	17 (23%)

*One patient was lost to follow-up, as he moved abroad.

The median follow-up time for our patients was 60 months (range 1–327 months). For the carotid body PGs which underwent surgery, only one recurred during follow-up. Among the surgically managed jugular PGs, we experienced re-growth in 33% of the cases. For the jugular PGs managed with radiotherapy, the rate was 29%.

Multifocality and genetics

As Table 3 shows, 56 (89%) of the patients underwent genetic testing in accordance with the protocol. We detected mutations linked to hereditary HNPG in 16 (29%) of the tested patients. Our cohort consisted of only one pair of siblings with SDHD mutations, the remaining patients with detected mutations were single cases, and three of them reported a positive family history. SDHD (eight patients) and SDHB (six patients) were the most prevalent mutations. Only four (25%) of the 16 patients with HNPG-linked mutation reported a positive family history. On the other hand, we found a mutation in all of the patients tested with a positive family history.

Table 3. Mutations and multifocality.

Mutation type among tested patients (n = 56)	Number of patients	Number of patients with malignant PG among HHNPGN	Number of patients with multifocal PG among HHNPGN
SDHB	6	0	0
SDHC	1	0	0
SDHD	8	1	7 (88%)
VHL	1	1	0
RET	0	0	0
Total	16 (29%)	2 (13%)	7 (44%)

HHNPGN: Hereditary head and neck paraganglioma.

Forty-four percent of the patients with mutations had multifocal PGs. Conversely, we found that seven of the eight patients (88%) with multifocal PGs who underwent genetic testing had a mutation. All of these seven patients had an SDHD mutation. There were a total of nine patients with multifocal PGs in our study, one of these patients declined to participate in the genetic testing.

Additionally, two of the four patients with malignant HNPG in our study had a mutation (SDHD/VHL). It is noteworthy that one of the patients with malignant PG did not undergo genetic screening as he died before the screening was offered.

¹⁸F-FDG PET/CT findings

Forty-one of the patients in the study had a ¹⁸F-FDG PET/CT examination. This examination supported the diagnosis in all our cases of head and neck PG, when performed preoperatively.

In two of these patients, the ¹⁸F-FDG PET/CT revealed unknown and subclinical multifocality. One patient presented with what was diagnosed as a unilateral carotid body PG, but the ¹⁸F-FDG PET/CT detected a small PG in his contralateral carotid body as well. Ultrasound and contrast-enhanced CT-scan were performed in advance without detecting the contralateral carotid body PG. In the other patient, with known bilateral carotid body PG, ¹⁸F-FDG PET/CT detected a pancreatic PG. Besides, in one of the four patients with malignant PG, ¹⁸F-FDG PET/CT identified an unknown lymph node metastasis.

Discussion

We present the first population-based study with regular genetic testing of patients with HNPGs in Scandinavia. In our study, a significant majority of the patients were females, the median age at the time of diagnosis was between the fourth and fifth decade, there were few malignant PGs, and a considerable proportion of the patients had hereditary PG. Consequently, most of our results resonate well with reports in the literature, and the significantly earlier onset of the hereditary forms of HNPGs in our study also agrees with institutions from other populations [3]. However, some differences were identified and outlined below.

Mutations and genetic screening

Among the patients who underwent genetic testing, only 29% had a mutation known to predispose for HNPG. This number is within the lower range reported in the literature with respect to HNPG. In contrast, some other European studies report higher rates: a Dutch study showed that 80% had a positive family history, and 99.5% of them had a mutation in the SDHx genes [10]. In a Spanish study, 40 patients were subjected to genetic testing, and 50% of them had a mutation [11].

Authoritative reports conclude that genetic testing for mutations must be an integral part of the work-up for all PG patients, for several reasons [12]. PGs that appear sporadic, with a negative family history, might in reality be hereditary. This may be due to a combination of factors including the occurrence of new mutations, reduced penetrance, imprinting in SDHD-related PG, and lack of appreciation among patients and doctors that certain combinations of tumours in different family members may signal heritability. Discovering a hereditary form of head and neck PG is deemed valuable, because it may help to estimate the risk of recurrence, malignancy, multifocality, pattern of inheritance/risk to relatives including risk of relevant syndromes. SDHB mutations have been reported to carry a higher risk of malignancy [13]; nevertheless, this is not reflected in our study. SDHD mutations are associated with multifocal PGs, and according to Boedeker et al., multifocal PGs are present in the vast majority of the SDHD patients [3]. In our study, we found that 88% of the patients with SDHD mutations had multifocal PG disease. Other benefits of discovering PG related mutations are the possibility of providing genetic counselling for particularly first-degree relatives.

Even though authoritative reports agree that genetic screening should be offered to all PG patients, there is a discussion about which genes should be included in the testing. Some argue that screening for SDHD, SDHB and SDHC is adequate [3], whereas others suggest including less frequent mutations in addition, as genetic testing has become more affordable and available [2]. Including genes very rarely involved in PG will allow detection of a low additional number of mutations, but will increase costs and, because of detection of Variants of Unknown Significance, complicate interpretation of variants and genetic counselling. We have so far opted to screen three common SDHx genes; SDHD, SDHB, SDHC, in addition to VHL and RET mutations in all patients, reserving analysis of the complete list of genes known to cause hereditary PG to special cases (familial, family history suggestive of a syndrome, malignant, multifocal, and early onset).

Imaging techniques

Ultrasound, contrast CT and MRI are the initial imaging modalities, which may suggest the diagnosis of a cervical PG. Although CT and MRI are useful and sensitive initial imaging methods, they may not be specific enough to identify that a neck mass is a PG. Therefore, reflecting trends in imaging of the head and neck of PGs in the literature and the advantages of PET/CT and, we started offering ¹⁸F-FDG PET/CT to all patients from 2012. In addition to the high sensitivity and specificity of ¹⁸F-FDG PET/CT, another major advantage is that it offers whole-body scan, which is helpful in order to detect metastatic or multifocal PG.

In the absence of histo-morphological criteria for defining a PG as malignant, whole body ¹⁸F-FDG PET/CT may be invaluable. With respect to head and neck PGs, less than 10% are considered malignant. Vagal PGs carry increased risk of being malignant [1]. In the case of multifocal PG with malignancy it may be impossible to define which PG caused the non-neuroendocrine metastasis, hence in our study we opted to count the number of patients with malignant PG disease instead of malignant PGs.

¹⁸F-FDG is the most frequently used tracer in PET/CT, however PET imaging with the somatostatin analogue ⁶⁸Ga-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-conjugated peptides (⁶⁸Ga DOTA-peptide) is believed to be the preferred PET examination when considering head and neck PG, as the majority of these express somatostatin receptors (SSTR) [14]. Although ⁶⁸Ga DOTA-peptide is the optimal PET tracer, both 6-l-¹⁸F-fluoro-dihydroxyphenylalanine (¹⁸F-FDOPA) and ¹⁸F-FDG PET are still considered excellent imaging techniques in PGs. The diagnostic performance for ¹⁸F-FDOPA is however lower in head and neck PG and metastatic disease related to SDH mutations [15]. ¹⁸F-FDG is particularly considered superior when it comes to PGs underlying SDHx mutations [7,8]. Several studies indicate that ¹⁸F-FDG PET/CT has a prognostic value, identifying patients with more aggressive disease and worse prognosis [16].

In our study, ¹⁸F –FDG PET/CT scans have been used, proving its significance, locating previously unknown multifocal PG in two patients, and malignant PG in one patient.

Furthermore, compared to modern imaging techniques, diagnostic fine-needle aspiration (FNA) has been deemed less relevant when suspecting a PG in the head and neck area. Some authors have warned against performing FNA if there is suspicion of PG because of the risk of catecholamine crisis. As most cervical PGs are non-secretory, this risk may be regarded as rather theoretical. Yet, the main reason to abstain from FNA is the limited diagnostic value [17]. According to our experience, if the tumour localization is suggestive of HNPG and PET/CT supports this, there is no requirement for cytological preoperative examination.

Treatment

Treatment of head and neck PG remains a contested subject in the literature. The three options available are surgery, radiotherapy, and expectant management, or a combination of these. Patient characteristics and preferences, primarily age and comorbidity, as well as tumour characteristics, size and proximity to vital neurovascular structures, determine which management option is chosen.

In general, younger patients without severe comorbidity or great risk of damaging adjacent neurovascular structures are consented for surgical resection. The presence of malignant disease or catecholamine secretion further accentuates the indication for surgical treatment. Radiotherapy may be the best option for PGs if surgery would require sacrifice of vital structures, patients with severe comorbidity or recurring PGs after surgery.

For carotid body PGs there is consensus that total surgical resection with a transcervical approach is the optimal treatment, given that the neurovascular risk is low if the surgeon is experienced [18]. More than half of our carotids’ PGs were subjected to surgery, and only one of the patients (5%) experienced iatrogenic cranial nerve injury. Twenty percent of the carotid body PG patients were offered radiotherapy. This was the chosen modality if the carotid body PGs were large, symptomatic, and surgical approach could not be justified due to either comorbidity or age. Twenty-four percent of the patients were observed and controlled regularly, predominantly if the carotid body PGs were small, asymptomatic, and age/comorbidity could not justify surgery. We advise that the patients who receive expectant management initially are followed-up annually, with MRI and with increasing intervals if stable.

Tympanic PGs are approached similar to carotid body PGs, as surgery carries a low risk for complications. In the presence of contraindications or multifocal PGs, radiotherapy is a reasonable alternative. Among our patients, only three of nine tympanic PG patients were given radiotherapy, and the rest offered surgery.

Contrary to tympanic PGs and carotid body PGs, many expert opinions suggest that the preferred treatment for jugular PGs is radiotherapy [19,20]. The anatomical localization of jugular PGs, surrounded by neurovascular structures, poses a significant morbidity risk, leading particularly to persisting dysfunction of cranial nerves, which may decrease the quality of life of the patients. Our experience in this study supports this, four of the six patients with jugular PGs who underwent surgery required cranial nerve sacrifice. Three of these patients with cranial nerve sacrifice were treated at hospitals abroad, specializing in skull-base PGs. Based on the present experience, for jugular PGs, we advise our patients radiotherapy aiming to achieve long-term tumour control. This approach requires a follow-up of minimum five years with imaging such as CT and MRI, and if there is tumour growth after radiotherapy, repeating radiotherapy or salvage surgery can be considered.

The treatment for rare vagal PGs is controversial. It is evident that the surgical approaches to vagal PGs often involve the sacrifice of the vagal nerve [18]. Only four patients presented with vagal PGs in our study, and given the risk of vagal nerve injury, we abstained from surgery in these patients. Two of the vagal PGs in our study were small, and the age of diagnosis for these patients suggested a limited life expectancy, consequently we opted for expectant management. The two other vagal PGs were subjected to radiotherapy. A systematic review shows that radiotherapy of vagal PGs may be better than surgery when it comes to tumour control and reducing the risk of major complications [20].

Treatment of multifocal PGs represents another topic of discussion in the literature. Our approach to these patients has been individualized for each patient and tumour. Hence, the treatment recommendations for the different subtypes of head and neck PGs, as outlined previously, principally apply for the different HNPGs in a patient with multifocal PG disease. Still, it must be emphasized that a step-by-step approach is recommended for these patients, and in general all treatment modalities can be utilized for the different PGs in one patient. In cases of multifocal PGs, it is vital to take into consideration parameters such as subtype of HNPG, presence of mutation and bilateral manifestations. For instance, if the multifocality consists of bilateral carotid body PGs, only one side may be operated first and the outcome closely evaluated particularly with regards to baroreceptor collapse and cranial nerve injury, and the other side possibly radiated, if required. Additionally, if the number of simultaneous HNPG does not justify a surgical approach, non-surgical treatment should be considered. In our study population we found a patient with five HNPGs, including bilateral vagal and carotid body PGs and unilateral jugular PG. For this patient we advised radiotherapy. Moreover, the presence of mutation among patients with multifocal PG is relevant when choosing treatment. For instance, SDHB mutations may carry a higher risk of malignancy; hence these patients may be advised surgery.

Conclusions

This is the first population-based study with routine genetic workup and PET/CT imaging in Scandinavia of HNPG patients. We found that 29% of the Norwegian patients with these tumours carried relevant mutations. The distribution of tumours located in the head and neck was congruent with other reports.

Genetic testing revealed PG related mutations in 16 patients (12 of them with a negative family history for PG), and 44% of the patients with mutations had multifocal disease, all with SDHD mutations. None of our patients with SDHB mutations had multifocal or malignant disease.

¹⁸F-FDG PET/CT supported the diagnosis in all our cases of HNPG, when performed preoperatively. Histology review was performed in all patients operated at our institution.

Combining genetic screening and PET/CT in the diagnostic approach to HNPGs is valuable when it comes to offering the PG patient early and individualized treatment and follow-up.

In terms of management, it should be individualized for all patients and radiotherapy may be a good alternative to surgery in older patients, and when the risk of nerve and/or critical vascular sacrifice is considerable.

Ethical approval

The study was approved by the Data Protection Office at Oslo University Hospital as a quality control project. The Norwegian Regional Committees for Medical and Health Research Ethics (REK) confirmed that the study is a quality control project.

Consent

Consent was obtained from all patients included.

Disclosure statement

The authors declare that they have no conflicts of interest. ICMJE Form for Disclosure of Potential Conflicts of Interest is filled out.

Data availability statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request, provided the request is accepted by the Data Protection Office.