8. Treatment recommendations in the primary therapy of oropharyngeal and hypopharyngeal carcinoma, taking into account effectiveness, functionality and quality of life – Oropharyngeal and Hypopharyngeal Carcinoma

The indication for primary surgical treatment of oropharyngeal carcinoma is viewed heterogeneously worldwide. In German-speaking countries, there is a traditional preference for a primary surgical approach given reasonable R0 resectability (see also section 6.6 Surgical safety margin). Internationally, primary surgical therapies are considered comparable to primary radiotherapy or radiochemotherapy in terms of effectiveness, functionality and late toxicity. Especially in very early stages (cT1N0, p16-independent), surgery with radiotherapy alone is on the same level of recommendation, although large prospective randomized studies on this issue are still pending.

Some authors prefer monomodal therapies (i.e. surgery or radiotherapy alone) to a multimodal approach for reasons of principle. This means that if adjuvant radiotherapy is considered necessary after surgery, radiotherapy alone should be preferred (1 vs. 2 modalities) in order to reduce the accumulation of the various modality-dependent toxicities/comorbidities. Similarly, it is argued that in the case of a necessary postoperative risk constellation with necessary adjuvant radiochemotherapy, it is not better to give preference to primary radiochemotherapy (2 vs. 3 modalities). Cheraghlou et al. [246] conducted a registry study with 4443 patients with HPV-positive oropharyngeal carcinoma of all tumour stages based on the National Cancer Data Base, classified according to the current 8th TNM edition (Evident Table PICO-2, registry-based studies). In patients with stage I disease, treatment with definitive radiotherapy alone was associated with significantly reduced survival compared with radiochemotherapy (hazard ratio [HR], 1.798; P = 0.029) or surgery with adjuvant radiotherapy (HR 2.563; P = 0.002) or surgery with adjuvant radiochemotherapy (HR, 2.427; p = 0.001). There was no significant difference after primary radiochemotherapy compared to surgery with adjuvant radiotherapy. In patients with stage II disease, significantly poorer survival was observed in patients treated with a single modality (either surgery [HR, 2.539; P = 0.009] or radiotherapy [HR, 2.200; P = 0.030]) compared to treatment with radiochemotherapy. Primary radiochemotherapy and surgery followed by adjuvant radiochemotherapy were equivalent. In patients with stage III disease, triple-modality therapy (surgery + adjuvant radiochemotherapy) was associated with significantly improved survival (HR 0.518; P = 0.024) compared to treatment with definitive radiochemotherapy alone. Similar results in favour of the trimodal approach were also observed for advanced p16-negative oropharyngeal carcinoma (retrospective cohort with 131 patients; [247]). The data to date therefore show that the "mono-meets multimodal" approach is not easily transferable, at least for oropharyngeal carcinoma.

In general, a trend towards transoral resection (as opposed to classic open approaches) + neck dissection + adjuvant radio- (chemo-) therapy has been observed over the last 20 years. Transoral surgery (TOS) techniques, including transoral laser microsurgery (TLM) and transoral robotic surgery (TORS), have been promoted in retrospective comparisons with conventional surgery or primary radiochemotherapy as gentle, minimally invasive procedures with good late functional outcomes ([248]; [249]; [250]; [251]; [Weinstein, GS et al. 2007]; [253]; [254]; [255]).

In order to determine the current best evidence in the comparative consideration of transoral (TORS, TLM, TOS) versus open surgical procedures, PICO question 4 was asked(evidence table PICO-4: Primary transoral surgical treatment procedures vs. open classical surgery). Only retrospective, predominantly non-randomized comparative studies and systematic reviews/meta-analyses for different head and neck localizations (larynx, oro- and hypopharynx) of different tumour sizes (predominantly T1-2) with low or very low evidence were found. The endpoints were different (mortality [242][244][245], 2-year DFS [242], 3-year DFS [256][244], recurrences [256][244][245], tracheostomy postoperatively [244], swallowing function 1 week and 1 year [245], intraoperative blood loss [245][242][242][Biron, VL et al. 2017], complications[258], length of hospital stay[242][242][244], PEG after 1 month and 1 year [Biron, VL et al. 2017]). The meta-analyses indicate that TORS may have better disease-free survival (DFS) and a reduced risk of free flap reconstruction compared to open surgery. TORS was associated with fewer tumour-positive resection margins (R1), a lower number of recurrences, fewer intraoperative tracheostomies, a shorter hospital stay and a shorter duration of postoperative nasal tube feeding compared to open surgery [241][241][259]. Compared to the more invasive techniques of conventional surgery, TORS appears to be less time-consuming and associated with less access morbidity ([254]; [260]).

The only prospective randomized trials to date that have directly compared TOS (transoral techniques such as TORS, TLM, etc.) with radiochemotherapy are the ORATOR 1 (phase II; T1-2 N0-2 tumours; 34 patients per arm, 88% p16-positive) and ORATOR 2 trials (ORATOR 1; Nicholas AC et al. [237]; [232]; ORATOR 2, Palma DA et al. [Palma, DA et al. 2022]) (ORATOR: Oropharyngeal Radiotherapy versus Transoral Robotic Surgery, Evidence Table PICO-3). The primary endpoint of the ORATOR 1 study addressed quality of life related to swallowing function (MD Anderson Dysphagia Inventory; MDADI score) after one year, with a clinically meaningful difference defined as a difference of more than 10 points. In the ORATOR 2 phase 2 trial, TOS + neck dissection + reduced adjuvant radiation dose (de-escalation arm) was compared with definitive radiochemotherapy for T1-2 N0-2 oropharyngeal carcinoma (100% p16 positive) with the endpoint "overall survival". A 1:1 randomization and stratification according to smoking status (< 10 pack years, ≥ 10 pack years) was performed. Patients in the radiotherapy arm were irradiated with a reduced radiation dose of 60 Gy, based on the study protocol of the Phase II NRG-HN002 trial [Yom, SS et al. 2021]. ORATOR 2 had to be terminated prematurely due to unacceptably high grade V toxicities in the TOS group [Palma, DA et al. 2022].

The median follow-up time of the ORATOR 1 study in the most recent publication from 2022 [232] was 45 months. In the combined MDADI score (dysphagia), the functional outcome after radiotherapy was significantly better compared to TORS + ND, especially in the 1st year after therapy. With longer follow-up (maximum 5 years), the difference decreased significantly, but was still statistically significant in favour of radiotherapy in 3 out of 5 sub-assessments and in the overall score. Clinically, however, the differences are to be classified as less significant, so that in view of the small number of cases, a superiority of radiotherapy cannot be classified as certain. The results show that primary transoral surgery or the primary non-surgical approach are relevant treatment options for stage I-II oropharyngeal carcinomas.

The study is viewed critically from a surgical perspective: In particular, the surgical technique is held causally responsible for the high complication rate of the ORATOR studies. Zech et al. have provided a valuable interpretation of this from a German ENT specialist perspective with a long-standing background in transoral laser microsurgery (TLM), which is quoted in more detail below [262]. The high complication rate of the ORATOR2 study (as well as the previous study ORATOR) [263] [237] [Palma, DA et al. 2022] could be the result of the unusual operative technical approach of the surgeons from Canada and Australia, if one compares the surgical procedure with other studies. In particular, the safety margin of at least of at least 10 mm required in the study protocol does not correspond to the clinical standard and has already been criticised in several comments on the study [264] [265] [266]. Experiments on body donor specimens have shown that for radical tonsillectomies a safety margin of more than 2 mm is not possible due to the limited thickness of the upper pharyngeal constrictor muscle (Musculus constrictor pharyngis superior), which separates the tonsil bed from the carotid sheath [267]. MRI measurements of the muscle confirm a mean thickness of 2.4 mm (standard deviation = 0.8 mm) [268]. In the currently recruiting "best-of" study EORTC-1420 (NCT02984410), this is taken into account, and for the deep tonsillar margin it is not the safety margin, but rather  the integrity of the superior constrictor pharyngis muscle that is defined as a quality feature of an oncologically clean radical en bloc tonsillectomy in patients with low-stage oropharyngeal carcinoma [269]. The authors justify this with equivalent local control after resection with a safety margin of 3 mm compared to 5 mm in this localisation [270]. In the current NCCN (National Comprehensive Cancer Network) guidelines [270][136] oropharyngeal safety margins of even 1.5-2 mm (irrespective of HPV status) are sometimes considered acceptable, although the NCCN and this guideline generally recommend 5 mm as the current international consensus (see Chapter 6). In clinical practice, some ENT surgeons in the USA already opt for a wait-and-see approach [271] [272] after resection of small HPV-associated oropharyngeal carcinomas and a narrow resection margin just as frequently as for adjuvant/post-resection, as primarily recommended in the NCCN guidelines. However, prospective data on this type of de-escalation is still lacking, so that this discussion is only presented here, but cannot (yet) lead to a deviating recommendation in this guideline.

The safety margin of more than 1 cm required in the ORATOR2 study protocol requires resection with narrow resection margins, and the additional circumferential margin incisions therefore represent an exceptionally radical surgical procedure in comparison to standard clinical practice and in no way a common (consented) treatment concept. This impression is reinforced by the high rate of tracheotomies in the surgical arm (65%). In the analysis of transoral resections at the University Hospital of Cologne for T1-T4 HPV-positive and -negative oropharyngeal carcinomas, for example, a tracheostomy was only necessary in 12% of cases [273]. The combination of the supposedly carotid-approaching excisions and a ban on the use of regional and free flap plasty for defect coverage in the ORATOR studies therefore appears to be a very plausible explanation for the two surgical-related deaths (fulminant haemorrhage and cervical osteomyelitis) [274]. The German Society for Otorhinolaryngology, Head and Neck Surgery and the Working Group for Oncology therefore warn against uncritically transferring the results of the ORATOR2 study to clinical practice (Zech et al. LRO2023).

The unusually high rate of (post-)operative complications of transoral surgery in both ORATOR trials is verified by the results of the large Phase III trials "PATHOS" and EORTC-1420 (NCT02215265, NCT02984410), which are currently still recruiting and in which many German centres are also participating on the urgent recommendation of the IAG-KHT. In comparison to the ORATOR trials, operational quality standards and quality control measures were implemented in the study protocol of these trials [275]. In view of the planned number of over 1000 patients, more reliable data on the morbidity and mortality of transoral surgery can be expected here.

The efforts of the ORATOR authors to minimise surgical risks and in particular the probability of bleeding by recommending ligation of the external carotid artery or its branches in laterally located tumours are to be acknowledged, as they can reduce post-operative bleeding rates, but are not generally required (although recommended in this guideline if there is good access to the relevant vessels during neck dissection). Laccourreye et al. reported 3.6% post-operative bleeding in 514 transorally operated patients with laterally located oropharyngeal tumours, of which 1.5% occurred at home after discharge from the hospital. The bleeding occurred predominantly in the first week, but also up to the third week and no longer after the fourth week. Prophylactic ligation of the external carotid artery or its branches as part of the primary procedure was not performed and was not recommended by the authors [276]. Salassa and Hinni [277] came to similar conclusions after a comprehensive retrospective study of 701 patients after transoral tumour surgery in the years 1996 - 2006, which showed a total of 1.4% postoperative bleeding between the 0th and 17th day. Three patients (0.4%) in this group had a catastrophic life-threatening haemorrhage, two of whom died. The main cause of haemorrhage was identified as the lingual artery in four cases, the superior laryngeal artery in two cases and the facial artery in two cases. In order to prevent potential bleeding after transoral tumour surgery, the authors recommend that in the event of bleeding occurring transorally, electrocaustic or vascular clips should be used first for arterial diameters over 2 mm. Prophylactic ligation of the tonsillar inflow from the external carotid artery or the external carotid artery itself is not recommended here either. A German study from 2004 by Esriti and Mann from the ENT University Clinic in Mainz investigated bleeding in the context of laser-surgical transoral tumour removal in the ENT area [278]. In the collective of 223 patients, a total of 97 oropharyngeal carcinomas were treated. Postoperative haemorrhage was observed in a total of 6% of these patients, with oropharyngeal carcinomas accounting for 10%. Severe secondary haemorrhage occurred in a total of five patients, of whom two required ligation of the external carotid artery and one who required ligation of the lingual artery. The Mainz colleagues also did not call for prophylactic ligation of the external carotid artery. Pollei et al. conducted a comprehensive analysis of a total of 906 patients who underwent transoral surgery for oropharyngeal carcinoma [279]. A postoperative haemorrhage rate of 5.4% was described here. 76.3% of these postoperative haemorrhages required surgical revision. Serious bleeding episodes were rare and occurred in 1.1% of patients. Prophylactic ligation of the of the external carotid artery was performed in 15.6% of patients. This was considered particularly necessary for advanced tumour stages. The rebleeding rate correlated with increasing tumour stage. The authors found that prophylactic ligation of the external carotid artery tributaries showed no advantage over patients who were not prophylactically ligated, although the rate of post-operative haemorrhage did not differ. Even in the group of prophylactically ligated patients, there was one case of life-threatening secondary haemorrhage. The authors conclude that due to the high risk of secondary haemorrhage, particularly in advanced tumour stages with intraoperatively exposed vessels in the case of a simultaneous neck dissection performed at the same time, the external carotid artery could be ligated. Although prophylactic ligation of the supplying vessels prior to transoral resection is not currently regarded as an absolute "must", it is common practice in many hospitals in order to reduce the intraoperative and postoperative risk of haemorrhage.

Non-transoral conventional surgery with classic external approaches is currently being used less and less frequently for stage I-II tumours. Transmandibular approaches in particular are only very rarely required for the resection of oropharyngeal carcinoma due to the higher access morbidity in the context of the alternative access spectrum (transcervical lateral, suprahyoid). With regard to the individual surgical techniques of the primary tumour, reference is made to the existing surgical doctrines due to the high complexity of resection and reconstruction techniques [280] [281] [282] [283]. 

The question of "surgical therapy - de-escalation" for HPV-associated oropharyngeal carcinoma is currently unanswered.

Radiotherapy has undergone significant technical development in the last 20 years. Intensity-modulated radiotherapy (IMRT) is available throughout Germany. Most institutes use rotation procedures (Volumetric Arc Therapy: VMAT) for IMRT, which are even better than IMRT with fixed irradiation angles at limiting the high-dose irradiated area to the actual target volume. The benefits of this new technology for patients with head and neck tumours have been well documented in a randomized study [286] and case series [287][288] [289]. Compared to older 3D-conformal radiotherapy, IMRT reduces both acute (mucositis, xerostomia) and late side effects (fibrosis, xerostomia, swallowing function). The dose-response relationships for radiotherapy and their volume dependencies in relation to the long-term functional outcome, in particular swallowing function, are now well understood [290] [291] [Kierkels, RG et al. 2014], so that further optimization of radiotherapy in terms of radiotherapy optimized for swallowing function (Swallowing Sparing IMRT) is becoming increasingly popular, now that a randomized study has also demonstrated the benefit compared to standard IMRT [293]. To implement this technique, it is important to contour the organs at risk [294] [295] and the target volumes [Grégoire, V et al. 2018] [Grégoire, V et al. 2014] [Biau, J et al. 2019] in accordance with current international standards. The recommended doses in the target volumes should not be undercut in order to better protect the organs at risk. IMRT techniques should be regarded as standard for both primary radio- or radiochemotherapy and adjuvant radio- or radiochemotherapy. High-quality cross-sectional imaging (CT or MRI) before the start of primary therapy is a prerequisite for high-quality radiation planning.

Interruptions of radiotherapy during primary or adjuvant radiotherapy beyond the planned breaks at weekends lead to a reduced effectiveness of radiotherapy due to repopulation of remaining tumour stem cells and should therefore be avoided [299] [300] [300]. If there are nevertheless interruptions to radiotherapy, it is possible to make up for lost time by radiotherapy twice a day at intervals of >6 hours with an unchanged single dose. In contrast to increasing the total dose, this procedure does not lead to a higher probability of late side effects. Tumour stem cells are also repopulated in the interval between surgical resection and the start of radio- or radiochemotherapy. If the interval between tumour resection and the start of radiotherapy exceeds 6 weeks, a meta-analysis of the available data [301] shows a significant increase in relapse rates (odds ratio: 2.89; 95% VB 1.60 - 5.21). Suitable organizational measures must therefore be taken to ensure that this interval is not exceeded if possible.

Primary radiotherapy or radiochemotherapy for HPV/p16+ oropharyngeal carcinoma

After primary radiochemotherapy in patients with HPV/p16+ oropharyngeal carcinoma, randomized studies [303] [Mehanna, H et al. 2019] [304] [235] [236] and the large registry databases show 5-year survival rates of around 90% in stage I, around 80% in stage II and around 70% in stage III (8th UICC classification). The vast majority of patients received conventionally fractionated or slightly accelerated radiotherapy with 70-72 Gy within 6-7 weeks in combination with cisplatin-containing simultaneous chemotherapy.

The good results, particularly in stages I-II, have raised the question of whether it is possible to de-escalate the therapy without compromising survival in order to reduce the toxicity of the therapy. Various strategies have been tested: In 3 larger randomized studies [303] [Mehanna, H et al. 2019] [304] it was tested whether a weekly administration of cetuximab is as effective as 3 applications of 100 mg/m² cisplatin (2 studies) or 6-7 weekly applications of 40 mg/m² cisplatin for radiotherapy. The studies showed a significant and clinically relevant inferiority of cetuximab in survival and locoregional relapse rate. In addition, treatment with cetuximab was no less toxic overall than cisplatin. Thus, the simultaneous administration of cisplatin and radiotherapy remains the standard therapy for HPV/p16+ oropharyngeal carcinoma.

Yom et al. [Yom, SS et al. 2021] compared in a randomized phase II study in HPV/p16+ oropharyngeal carcinomas in study I-II (8th UICC classification) a single slightly accelerated intensity-modulated radiotherapy (60 Gy in 5 weeks) with a conventional fractionated intensity-modulated radiotherapy (60 Gy in 6 weeks) in combination with weekly 40 mg/m² cisplatin. A significantly higher locoregional relapse rate was found in the arm with radiotherapy alone. There were no differences in all other oncological endpoints, including toxicity. The registry data [235] [236] showed that in HPV/p16+ oropharyngeal carcinomas in stages I-II, survival was also significantly worse if concurrent chemotherapy was not administered during radiotherapy. Only in patients without lymph node involvement was radiotherapy alone as effective as radiochemotherapy in stages I-II [240] [235]. This opens up a certain scope for saving simultaneous chemotherapy in stages I cN0, although data from randomized studies are still pending.

In two small phase II studies, induction chemotherapy was initially applied to HPV/p16+ oropharyngeal carcinomas in stages II and III. If there was a good response to chemotherapy, the total doses of subsequent radio- or radiochemotherapy were reduced by 16-25 Gy. The PFS in patients treated with reduced doses was still above 80% after 3 years in both studies. However, due to the small case numbers of n=20 [284] and n = 62 [307], no treatment recommendations can be derived from this.

In summary, de-escalation of the total dose of radiotherapy or simultaneous therapy with cisplatin in stages I-III of HPV/p16+ oropharyngeal carcinoma remains a study question. According to the 3 larger randomized studies (1321 patients in total), which exclusively included HPV/p16+ patients with oropharyngeal carcinoma in stages I-III (8th UICC classification), conventionally fractionated radiotherapy with 5x2 Gy per week up to a total dose of 70 Gy in IMRT technique in combination with 3 x 100 mg/m² cisplatin or 40 mg/m² weekly cisplatin parallel to radiotherapy is the best investigated treatment concept. In other studies, slightly accelerated fractionations of radiotherapy with 70-72 Gy within 6 weeks were also used with good results [8]. However, the benefit of accelerated radiotherapy in HPV/p16+ patients with oropharyngeal carcinoma is uncertain.

For patients with oropharyngeal carcinoma who are not suitable for cisplatin, e.g. due to impaired renal function, there are results from a number of randomized studies that tested carboplatin in combination with 5-FU [308] [309][310] [311] [312] [309] [308], mitomycin C in combination with 5-FU [313], [Keane, TJ et al. 1993] or weekly docetaxel [315] simultaneously with radiotherapy compared to radiotherapy alone. In some of these studies, some patients with tumours of the oral cavity, larynx and hypopharynx were also included. For both combination chemotherapies applied simultaneously with radiotherapy, a significant advantage was shown in terms of survival and locoregional tumour control. Only studies with <=80 patients in the radiochemotherapy arm are available for monochemotherapy with carboplatin [316] [317] [310] [318], and there were no significant advantages for monotherapy with mitomycin C [319] [320] [321] in parallel with radiotherapy. The HPV/p16 status of the oropharyngeal carcinomas was not determined in any of these studies, so it is unknown whether the benefit shown also depends on the HPV/p16 status of the tumours. Weekly applications of carboplatin and paclitaxel, concurrently with radiotherapy, have been investigated in a number of non-randomized phase II trials and some retrospective cohort studies [322], [323], [324], [325], [326], [327], [328]. Oropharyngeal carcinomas represented the main group of treated squamous cell carcinomas of the head and neck, with no differentiation made according to HPV/p16 status. The non-randomized retrospective studies indicate a similar efficacy of carboplatin + paclitaxel compared to weekly therapy with 40mg/m² cisplatin in parallel with radiotherapy. In contrast, there is very little data available for the administration of paclitaxel alone [329]. Cetuximab in combination with radiotherapy alone was tested in a randomized study [330] with radiotherapy alone for squamous cell carcinoma of the oropharynx (63%), larynx and hypopharynx. A significant advantage in overall survival and in the locoregional recurrence rate was demonstrated. The p16 status was retrospectively determined in 182 oropharyngeal carcinomas (66%). This showed that the benefit of cetuximab was largely limited to the p16 positive tumours [306].

Primary radio- or radiochemotherapy for HPV/p16- oropharyngeal carcinoma

Results of randomized studies on the effect of chemotherapy applied in addition to radiotherapy for patients with exclusively HPV/p16- oropharyngeal carcinomas are not yet available, as simultaneous radiochemotherapy was already used as standard therapy before the different biology of HPV/p16+ oropharyngeal carcinomas was known. An evaluation of these studies in oropharyngeal carcinomas was therefore not carried out. Radiotherapy was compared with the combination of radiotherapy and chemotherapy in a total of 107 randomized studies (19,805 patients). The latest update of the MACH-NC meta-analysis [331] from 2021 shows that only simultaneous radiochemotherapy leads to an improvement in overall survival and PFS, whereas neither induction chemotherapy nor adjuvant chemotherapy lead to a benefit in this regard. In 58 randomized trials, radiotherapy was compared with simultaneous radiochemotherapy in 14,401 patients. The vast majority of these patients had received simultaneous chemotherapy containing cisplatin. There was an absolute survival benefit of 6.5% (95% CL 4.6-8.4%) after 5 years and 3.6% (95% CL 1.8-5.4%) after 10 years. 34.7% of these patients had oropharyngeal carcinomas. The benefit of simultaneous chemotherapy applied in addition to radiotherapy was the same for oropharyngeal carcinoma (HR 0.82) as for squamous cell carcinoma of the larynx (HR 0.81), hypopharynx (HR 0.88) and oral cavity (HR 0.82) ([331]. It is therefore assumed that the benefit of simultaneous chemotherapy is the same for HPV/p16- and HPV16+. The available registry data confirm this assessment [332]. This applies in particular to HPV/p16- oropharyngeal carcinomas in stages III-IVb, whereas the benefit of additional simultaneous chemotherapy in stages I-II is not well documented [332]. It should be noted that only 9.1% of patients in the MACH-NH meta-analysis were older than 70 years, only 5.4% had an ECOG >=2 and only 5.4% were treated in stages I-II (UICC 7th edition). While in the MACH-NH meta-analysis patients over 70 years of age did not benefit from simultaneous chemotherapy, the NCDB and SEER registry databases also documented survival benefits in patients over 70 years of age [333] [334]. Excluding this patient group in the absence of contraindications for simultaneous chemotherapy therefore does not appear justified.

For patients with a contraindication to cisplatin, randomized studies have also demonstrated a survival benefit for the simultaneous application of carboplatin and 5-FU [308] [309] [310] [311] [312] [309] [308] or mitomycin and 5- FU [313] [Keane, TJ et al. 1993] or weekly docetaxel [315]simultaneously with radiotherapy compared to radiotherapy alone. Only studies with <=80 patients in the radiochemotherapy arm are available for monotherapy with carboplatin [316] [317] or mitomycin C [319] [320] [310] [318], and there were no significant advantages for monotherapy with mitomycin C [319] [320] [321] in parallel with radiotherapy. Weekly applications of carboplatin and paclitaxel concurrently with radiotherapy have been investigated in a number of non-randomized phase II trials and some retrospective cohort studies [322], [323], [324], [325], [326], [327], [328]. Oropharyngeal carcinomas represented the main group of treated squamous cell carcinomas of the head and neck, with no differentiation made according to HPV/p16 status. The non-randomized, retrospective comparative studies indicate a similar efficacy of carboplatin + paclitaxel compared to weekly therapy with 40mg/m² cisplatin in parallel with radiotherapy. In contrast, there is very little data available for the administration of paclitaxel alone [329]. A benefit of cetuximab in combination with radiotherapy for HPV/p16 oropharyngeal carcinoma cannot be derived from the available data [330] [306] [335].

As with HPV/p16 negative oropharyngeal carcinomas, radiotherapy should be carried out using the IMRT technique with the best possible protection of the salivary glands, the swallowing tract and the oral cavity, without reducing the dose in the target volumes. The standard fractionation for simultaneous chemotherapy is conventional fractionation with 5x2 Gy per week up to a target volume dose of 70 Gy. Accelerated radiotherapy with a reduction of the total treatment time to 6 weeks or less has not shown any benefit in combination with simultaneous chemotherapy according to the results of 2 randomized studies [336] [309]. Hyperfractionated radiotherapy with 2 fractions per day 5 times per week with single doses of 1.2-1.25 Gy up to total doses of 74.4- 80.4 in combination with simultaneous chemotherapy were compared with hyperfractionated radiotherapy alone in several randomized trials [337] [338] [339]. The additional chemotherapy also improved survival and locoregional tumour control in combination with hyperfractionated radiotherapy. In the randomized studies on radiotherapy alone, which compared different fractionations of radiotherapy, only hyperfractionated radiotherapy proved to be significantly superior in terms of overall survival, as described in the above-mentioned studies, whereas accelerated radiotherapy regimens only improved locoregional tumour control but did not result in a survival benefit [Lacas, B et al. 2017]. To date, hyperfractionated radiotherapy in combination with simultaneous chemotherapy has not been compared with conventionally fractionated radiotherapy with simultaneous chemotherapy. Thus, it remains unclear whether hyperfractionated radiotherapy with simultaneous chemotherapy may be even more effective than conventionally fractionated radiotherapy in combination with simultaneous chemotherapy, as predicted by a network meta-analysis [Petit, C et al. 2021]. In the case of radiotherapy alone with contraindications to concurrent chemotherapy, hyperfractionated radiotherapy is a good option, although it is not certain whether patients with poor general health or very old patients who are not suitable for concurrent chemotherapy will benefit from hyperfractionated radiotherapy. It should be noted in all the studies on fractionated radiotherapy that only about one third of the patients had oropharyngeal carcinomas and their HPV/p16 status was unknown. However, the studies showed no evidence of different effects of fractionation between squamous cell carcinomas of the oropharynx, hypopharynx, larynx and oral cavity.

Postoperative radiotherapy

The effect of postoperative radiotherapy compared to no adjuvant radiotherapy has not been investigated in randomized studies. However, the results of cohort studies and prospective registry data showed a very clear superiority of adjuvant radiotherapy in patients with more than one tumour-involved lymph node, extracapsular tumour growth at the lymph nodes (ECE) or with tumours only barely resected in healthy tissue (<5mm), so that randomized studies were not conducted ([342], [343], [344], [345]). These studies included squamous cell carcinomas of the larynx, oropharynx, hypopharynx and oral cavity. The HPV/p16 status was not known in any of the collectives examined. The collectives studied were mainly treated in Western countries between 1970 and 1998 and in Asia in the period 1990-2008. Based on the available data on the low incidence of HPV/p16+ oropharyngeal carcinomas in these countries during the study period, it can be assumed that the oropharyngeal carcinomas in these studies were predominantly HPV/p16- [346], so that these data essentially apply to HPV/p16- oropharyngeal carcinomas. Soliman et al. 2022 [347] analyzed the data of 15,036 patients with HPV/p16+ oropharyngeal carcinoma who were documented in the National Cancer Database (USA) between 2010 and 2017 and who had received adjuvant radiotherapy or no adjuvant radiotherapy after primary surgical treatment, predominantly in stages I and II. Patients who had received adjuvant radiochemotherapy were excluded. In the "Propensity Score Matched" analysis, adjuvant radiotherapy led to a significant survival benefit of approximately 8% in absolute terms after 5 years in the overall collective and 10% for patients with risk factors (R1 resection, ECE, LVI, blood vessel invasion). Overall survival with postoperative radiotherapy was 90% and 87% after 5 years in the overall collective and in patients with risk factors, respectively, which is significantly higher than the values reported in the old data for patients with oropharyngeal carcinoma with unknown HPV/p16 status ([342], [343], [344], [345]). The available data suggest that adjuvant radiotherapy reduces the locoregional recurrence rate and improves survival in oropharyngeal carcinoma regardless of HPV/p16 status, provided that more than one regional lymph node is affected or the primary tumour was only barely resected in healthy tissue (<5mm). For patients with pT3/pT4 tumours, lymph node involvement with ECE or R1 resection, there are no reliable data on treatment outcomes without adjuvant radiotherapy for HPV/p16+ oropharyngeal carcinomas, as there is an international consensus that these patients, just like those with HPV/p16 tumours, require adjuvant radiotherapy [348].

Postoperative radiochemotherapy

The value of additional cisplatin-containing chemotherapy administered simultaneously with radiotherapy was investigated in 3 larger randomized studies, which included patients with squamous cell carcinoma of the oropharynx, hypopharynx, larynx and oral cavity [349], [350], [351]. HPV/p16 status was not determined in these studies. However, for the reasons mentioned above, it can also be assumed that the majority of oropharyngeal carcinomas in these studies were HPV/p16-. The results of the studies consistently showed that patients with evidence of ECE in the affected lymph nodes or R1 resection (defined in these studies as resection <5mm in healthy tissue) had a survival benefit from adjuvant radiochemotherapy compared to adjuvant radiotherapy alone. In an analysis based on the NCDB registry [352], the results of adjuvant radiochemotherapy were compared with those of adjuvant radiotherapy alone in 1127 patients with HPV/p16+ and 424 patients with HPV/p16- oropharyngeal carcinoma who had intermediate risk factors for recurrence (at least one of the following factors: pT3-T4, ≥2 tumour-involved lymph nodes, involved lymph nodes in level IV or V, lymph vessel invasion). In the "Propensity Score Matched" cohorts of HPV/p16+ and HPV/p16- tumours, no benefit in terms of survival was shown for the additional platinum-containing chemotherapy. Other endpoints were not investigated. Thus, in the presence of these intermediate risk factors, adjuvant radiotherapy alone is preferable.

Although lymph node involvement with ECE in HPV/p16+ oropharyngeal carcinomas is not taken into account in the N classification and staging, the results from various registry studies show that the detection of ECE in the lymph nodes of oropharyngeal carcinomas is associated with a significantly higher recurrence rate and poorer survival regardless of HV/p16 status ([353], [354], [355], [356]). However, if adjuvant radio- or radiochemotherapy has been applied to HPV/p16+ oropharyngeal carcinomas, the recurrence rate in the presence of ECE is only minimally increased and only detectable in larger collectives. The available data from the registry studies consistently show no benefit of adjuvant radiochemotherapy compared to adjuvant radiotherapy alone [353], [354], [355]. The results of randomized studies are not available. Data on the relapse pattern after adjuvant radiotherapy or radiochemotherapy for HPV/p16+ oropharyngeal carcinoma indicate that the poorer prognosis in patients with ECE to the lymph nodes is almost exclusively due to a higher rate of distant metastasis [357]. Apparently, the intensity of the platin-containing chemotherapy applied simultaneously with radiotherapy is not sufficient to reduce distant metastasis. The available registry data therefore indicate that adjuvant radiotherapy alone may be sufficient for HPV/p16+ oropharyngeal carcinoma, even in cases of lymph node involvement with evidence of ECE, and that additional chemotherapy may not be necessary.

In contrast, the available registry data show a clear advantage for platinum-containing chemotherapy applied simultaneously with adjuvant radiotherapy forHPV/p16- oropharyngeal carcinomas with evidence of ECE in the lymph nodes ([353], [354], [355]).

Less well documented is the influence of positive (<1 mm) or narrow resection margins (<5 mm) on the recurrence rate and survival of oropharyngeal carcinomas depending on HPV/p16 status. The available data from registry studies show a greater negative impact on the recurrence rate and survival for HPV/p16 oropharyngeal carcinomas than for HPV/p16+ tumours [356]. Regardless of HPV/p16 status, the recurrence rates do not differ significantly in the case of positive (<1mm) compared to narrow (<5 mm) resection margins, even though a trend in favour of larger resection margins can be seen [358], [271]. Only with tumour-free resection margins of >=5 mm do the differences become larger and in some cases statistically significant, provided that larger collectives were examined [358], [344]. Regardless of whether no adjuvant therapy or adjuvant radio- or radiochemotherapy has been carried out, the treatment results for HPV/p16+ oropharyngeal carcinomas are significantly better in terms of the recurrence rate and survival ([358], [356]). In a large collective of NCDB [359]patients with HPV/p16+ oropharyngeal carcinoma with a proven positive resection margin as well as when ECE was detected in the affected lymph nodes or when both risk factors were present, survival did not differ whether the patients had received adjuvant radiotherapy alone or radiochemotherapy. The results of randomized studies are not available. In summary, the benefit of adjuvant radiochemotherapy compared to adjuvant radiotherapy alone for positive or close resection margins in HPV/p16+ oropharyngeal carcinomas in contrast to HPV/p16- oropharyngeal carcinomas cannot be considered certain. However, patients with HPV/p16+ oropharyngeal carcinomas with locally advanced disease (pT3-4 or pN2-3) who also had ECE resected at the affected lymph nodes or only just in healthy tissue were significantly underrepresented in the previously described collectives, so that a relevant benefit of adjuvant radiochemotherapy cannot be ruled out in these cases. If postoperative radiochemotherapy is indicated, it should be noted that results of randomized studies are only available for cisplatin or cisplatin + 5-FU, which have shown a significant reduction in the relapse rate or an improvement in survival ([349], [350], [351]. In a randomized Japanese study (n=261), 3x 100 mg/m² cisplatin at 3-week intervals was compared with weekly 40 mg/m² cisplatin simultaneously with postoperative radiotherapy [360]. The non-inferiority of the weekly administration of cisplatin was demonstrated. However, the proportion of patients with oropharyngeal carcinoma (11%) was small.

For patients who cannot receive cisplatin, e.g. due to impaired renal function, there is little data available for postoperative radiochemotherapy. For mitomycin C, data from two non-randomized studies are available, which only showed a trend towards a survival benefit [321][361]. The weekly administration of docetaxel simultaneously with radiotherapy was compared with radiotherapy alone in a randomized study [315]. In a small subgroup of this study, the therapy was also given in the postoperative situation. There was a trend towards improved survival for the combined therapy. For carboplatin + 5-FU and carboplatin + paclitaxel, no results of randomized studies in the postoperative situation are available. However, it seems plausible to use these combinations in the postoperative situation by analogy with the efficacy of these combinations in primary radiochemotherapy (see 8.3.1). There are no data suggesting a benefit for the use of cetuximab in the postoperative situation.

For patients with an intermediate risk of recurrence (no ECE and resection >=5 mm in healthy tissue), 60 Gy (5x 2 Gy per week) was applied in the region of the former primary tumour and the affected lymph nodes in the vast majority of patients in the clinical trials and the registry data, regardless of HPV/p16 status. In the neighbouring unaffected lymph node regions, 45-50 Gy with 5x 1.8-2.0 Gy per week were administered electively. There is an international consensus for the selection of the lymph node levels to be electively irradiated, which can be regarded as a standard guideline [Biau, J et al. 2019]. For HPV/p16+ oropharyngeal carcinomas, lower doses of 50 Gy in conventional fractionation [Ferris, RL et al. 2022] or only 30 Gy with 2x1.5 Gy per day were also investigated in smaller phase II studies [362], provided there was an R0 resection and no ECE at the lymph nodes. In the phase II study mentioned [362] 208 patients were randomized between 60 Gy and 50 Gy. The PFS after 3 years was 90% with 50 Gy as well as with 60 Gy. In the single-arm study with only 30 Gy, the PFS was also just under 90%. However, the number of cases in the studies mentioned is still too small to be able to generally recommend a dose reduction. In patients with a high risk of relapse (ECE or resection <5 mm in healthy or R1), 36 Gy with 1.8 Gy twice daily in combination with docetaxel was applied in 22 patients in the study by Moore et al. 2020 [362], without a noticeably increased relapse rate being reported. In contrast, in the study by Ferris et al. [Ferris, RL et al. 2022], 66 Gy was given in conventional fractionation in combination with simultaneous cisplatin with a high risk of relapse. De-escalation of the radiation dose below the recommended dose of 66 Gy (5x2 Gy per week) remains the recommended standard in the high-risk situation regardless of HPV/p16 status. For elective adjuvant irradiation of neighbouring, unaffected lymph node sites, 45-50 Gy (5x1.8-2.0 Gy per week) is also recommended in the high-risk situation. IMRT techniques are also considered standard for postoperative radiotherapy.

Surgical treatment was described in more detail in chapter 8.1.1. With a particular focus on advanced oropharyngeal carcinomas (p16-pos. stage III, p16-neg. stage III and IVa), there is a complete lack of randomized studies that directly compare primary surgical treatment with radiochemotherapy. In this respect, registry studies are helpful for classification, although a lower level of evidence must be assumed. Cheraghlou et al. ([236]; Evidence Table PICO 2, registry-based studies) conducted a registry study with 4443 patients with HPV-positive oropharyngeal carcinoma of all tumour stages based on the National Cancer Data Base (NCDB, data from 2009-2013), classified according to the current 8th TNM edition. It should be noted that this study did not include data on smoking status and the surgical procedures used, but the stage-dependent modality comparison was one of the best results. In patients with stage III disease (n=450), triple-modality therapy (surgery + adj. radiochemotherapy) was associated with significantly improved survival (HR 0.52; P = 0.024) compared to treatment with definitive radiochemotherapy alone. The 3-year overall survival of stage III patients after radiochemotherapy was 71.9% and after surgery + adjuvant radiochemotherapy 84.9%. Similar results in favour of the trimodal approach were also observed for advanced p16-negative oropharyngeal carcinoma (retrospective cohort with 131 patients; [247]).

Amini et al. also conducted a registry study from the NCDB (2009 - 2011) with 3952 patients (2454 p16-pos.) and a somewhat less conclusive focus ([370]). The T3-T4 proportion was given as 20.8% and the reference group was surgical therapy alone (rarely relevant). All bi- and trimodal approaches were better than surgery alone, which was only comparable to radiotherapy alone. It was concluded that bimodal therapies appeared to be beneficial in HPV-positive oropharyngeal cancer. In HPV-negative patients, postoperative chemotherapy with radiotherapy was associated with an improvement in overall survival, while no significant benefit was observed in HPV-positive patients ([370]; conclusion PICO 3 evaluation). Kamran et al., 2018 ([366] Evidence Table PICO-3) conclude in a registry study (NCDB 2004-2013; 22,676 patients; subset 6872 with HPV status; 73.3% HPVpos.; 31.7% T3-T4) of a propensity-matched cohort for HPV-positive patients that patients treated primarily surgically have a better 3-year survival, especially in the subgroup of (HPV)-negative patients (p=0.06), than after primary radiochemotherapy. For HPV-positive status, this difference was not significant (p=0.38). Other registry studies presented in the PICO-3 Table ([371], [239], [372], [240]) came to a similar conclusion.

In summary, based on the registry studies for stage III and IVa, p16-negative oropharyngeal carcinomas, the majority of authors favour primary surgery followed by adjuvant radiotherapy or radiochemotherapy as the treatment of first choice. For p16-positive. The registry studies do not provide a uniform picture for p16pos. patients, although the largest registry study of 450 HPV-positive stage III patients [236] shows a significant superiority of primary surgery + adjuvant radiochemotherapy. Since all registry studies did not adjust for smoking status, the current data situation should be evaluated with the necessary caution. Overall, the primary surgical approach is currently to be favoured in the patient group addressed, provided that R0 resection is possible and a meaningful functional outcome can be expected. Information on the respective surgical techniques has already been provided in Section 8.1.1.

In contrast to laryngeal carcinoma, organ-preserving surgery is only rarely considered for carcinomas of the hypopharynx due to the advanced stage of the tumour at the time of diagnosis ([386]; [387]). In addition, patients with hypopharyngeal carcinomas are often in an unfavourable general clinical condition due to extensive comorbidities [388]. Frequently, there is a long-standing, severe nicotine and alcohol abuse, chronic bronchitis, COPD, malnutrition, liver dysfunction and a consecutive impairment of blood coagulation. The prognosis of hypopharyngeal carcinoma across all stages is the least favourable of all squamous cell carcinomas of the head and neck in Europe (5-year survival < 20%, [389]). A careful preoperative risk assessment is therefore necessary before therapeutic considerations, especially primary surgical measures, and intensive postoperative monitoring is therefore often planned.

Advances in open laryngeal surgery and TLM have now enabled oncologically safe R0 tumour resections with acceptable functional restrictions in some patients with hypopharyngeal carcinomas. One of the basic principles of oncological surgery is that a complete, histologically verifiable R0 resection (>5mm) must be guaranteed. If this is not reliably possible as part of an intended surgical procedure (such as a TLM or TORS), alternative treatment options (open extended partial resection and reconstruction, laryngectomy or non-surgical treatment options) must be chosen as a matter of priority. In the case of hypopharyngeal carcinomas, whether a complete tumour resection will be possible can only be assessed by reviewing the endoscopic findings (as part of diagnostic panendoscopy by an experienced tumour surgeon) and meaningful imaging (usually MRI, CT) and preoperative swallowing diagnostics by FEES (see also Chapter 7). In addition to TLM ([390], [391], [392], [393], , [387], [394], [395], [396], [397], [398]), TORS ([242]; [241]) has also found its way into the treatment of hypopharyngeal carcinoma worldwide.

Patients with smaller tumours of the medial wall of the piriform sinus and the postcricoid region, which are still movable on the cartilaginous base, as well as smaller and superficially growing tumours of the lateral wall of the piriform sinus and the posterior wall of the hypopharynx are particularly suitable for TLM. For tumours in the area of the apex of the piriform sinus and the oesophageal orifice, organ-preserving surgery is generally not feasible. In 228 consecutive patients with hypopharyngeal carcinomas treated at a TLM clinic with experienced surgeons, surgical treatment was only possible in 136 patients (60% of all patients). Of these 136 patients, 90 underwent total laryngectomy. Of the 46 patients who underwent organ-preserving surgery, 23 underwent open partial pharyngectomy and 23 underwent transoral laser surgery. This means that 34% of all patients operated on and 20% of all 228 patients in the entire series underwent organ-preserving surgery, half of which, i.e. only 10% of all patients, underwent transoral laser surgery [387]. Even though these data are older, the authors believe that the surgical approach to TLM in hypopharyngeal carcinoma has not changed to date.

Overall, it can therefore be assumed that only around 20% of all patients with hypopharyngeal carcinoma can undergo larynx-preserving surgery. Of these, around half can be operated on by laser surgery and the other half by open partial pharyngectomy.

If the tumour reaches the resection margins, a subsequent resection should be attempted in order to achieve an R0 result; as already explained in Section 6.6, a subsequent resection has no negative prognostic influence, particularly in the TLM of laryngeal/hypopharyngeal carcinoma [134].

There is extensive literature on the high validity of TLM. The indications for organ-preserving TLM in the hypopharynx range from T1 to T4a tumours with good results: Steiner et al. published local control rates over five years of 84% pT1; 70% pT2; 75% pT3; 57% pT4a; recurrence-free 5-year survival 73% for stages I and II, 59% for stage III, 47% for stage IVa [397]. Eckel et al. showed an (uncorrected) survival of 59.3% and a local control rate of 73.3% in large collectives after 5 years, with 76.5% of all patients showing a functioning larynx after this time [387]. Rudert et al. showed an uncorrected survival of 48% and a disease-specific survival of 58% (71% in stages I and II) [399]. In carefully selected patients with early hypopharyngeal carcinoma, transoral laser surgery (in combination with unilateral or bilateral neck dissection and, if necessary, postoperative radiotherapy) leads to very good oncological and functional results.

As part of the PICO 4 analysis (Evidence Table PICO 4), we became aware of the only comparative study worldwide to date on TORS versus open surgery, in which only hypopharyngeal carcinomas were treated [242]. In n=56 patients, the tumour stages were distributed as follows: T1-T2 80%, T3 16%, T4 14% (male proportion 98%, mean age 64.7 years). The evidence is very low, as this is a retrospective study without matching and increased T1 tumours in the intervention group (TORS) and thus there is an imbalance between the intervention and control groups. The 2-year DSF showed no difference between the groups (relative risk: 1.08; CI 95% 0.83 - 1.41), with lower intraoperative blood loss and shorter hospital stay in the TORS group.

Transoral resection (preferably TLM) of hypopharyngeal carcinomas should be avoided and the legitimacy of converting the operation to laryngectomy should always be discussed and obtained. It is therefore advisable to plan and determine the therapeutic procedure in detail for each individual patient in advance as part of the interdisciplinary tumour board, whereby the individual situation and special wishes of the patient concerned must of course be taken into account. Subsequent radiotherapy cannot be considered an appropriate completion of an incomplete tumour resection.

With regard to functional impairment, the extent of the expected postoperative swallowing impairment in particular must be assessed in advance. In particular, the following questions must be clarified:

• in view of the planned extent of resection, whether temporary or prolonged aspiration must be expected,
• whether the patient's overall clinical situation suggests that temporary aspiration can be tolerated, and
• whether a temporary tracheostomy or PEG may be required.

On the other hand, more serious vocal impairments are generally not to be expected with transoral resection of hypopharyngeal carcinomas, and serious airway obstructions are also generally not to be expected with atraumatic surgical technique and perioperative shielding with antibiotics and steroids.

Some unfavourably located T2 hypopharyngeal carcinomas are reserved for the classic transcervical, i.e. open, techniques. Please refer to section 8.5.1.

The registry data from the USA [385], Canada [400] and the Netherlands [12] unanimously show that the treatment results in terms of overall and progression-free survival for squamous cell carcinoma of the hypopharynx in stages T1-2 N0 do not differ significantly depending on whether primary surgical treatment, primary radiotherapy or primary radiochemotherapy was carried out. Accordingly, the current recommendations of the ESMO and NCCN guidelines also list all of these treatment options without preference [401]. Particularly in the case of larger T2 tumours (e.g. postcricoid), where a primary laryngectomy would be necessary surgically, organ preservation by means of primary radiochemotherapy or induction chemotherapy followed by radiotherapy in the event of a good response to induction therapy or laryngectomy in the event of a poor response should be discussed on an interdisciplinary basis, as described in more detail under 8.5.2.

The same principles apply to the dosage and fractionation of radiotherapy as for HPV/p16-positive oropharyngeal carcinomas (see 8.1.3), i.e. radiotherapy was carried out in most studies on radio- or radiochemotherapy with 5 x 2 Gy per week up to target volume doses of 70 Gy in the area of the primary tumour and the affected lymph nodes. Slightly accelerated and also hyperfractionated radiation regimens with biologically equivalent doses give approximately equivalent results (described in detail in 8.1.3).

An advantage in terms of relapse-free survival or overall survival with simultaneous radiochemotherapy compared to radiochemotherapy alone has not been established for squamous cell carcinoma of the pharynx, larynx and oral cavity in stages I+II. In the patient population of the MACH-NC meta-analysis, only 5.5% of patients were in stages T1N0 and T2N0 and only 15.5% of patients had hypopharyngeal carcinoma [331]. In an earlier evaluation of the MACH-NC meta-analysis, stages I-II were evaluated separately for all tumour locations. There was no benefit of additional chemotherapy compared to radiotherapy alone for these early stages [368]. For locally advanced squamous cell carcinomas, however, a survival benefit from chemotherapy applied simultaneously with radiotherapy is well documented [368][331]. Whether patients with larger T2N0 hypopharyngeal carcinomas benefit from radiochemotherapy cannot be ruled out. If radiochemotherapy is indicated for these cases after interdisciplinary discussion, the same criteria apply with regard to the selection of chemotherapy as for locally advanced hypopharyngeal carcinomas (see 8.5.3).

Results of randomized studies on the value of radio- or radiochemotherapy compared to no adjuvant therapy are not available for squamous cell carcinomas of the hypopharynx or for oropharyngeal and laryngeal carcinomas or tumours of the oral cavity. This applies both to stages T1-2N0 and to locally advanced tumours. However, the results of systematic reviews and registry studies [342], [343], [344] indicate an improvement in relapse-free and overall survival if risk factors are present, although the subgroups of hypopharyngeal carcinomas in stages T1-2N0 were very small. In the registry studies, the narrow resection margin (<5 mm, N+ and ECE) was identified as a risk factor, whereas other factors such as G3, L1 and V1 were not investigated. For T1-2N0 carcinomas resected <=5 mm in healthy tissue, the available data suggest that postoperative radio- or radiochemotherapy is likely to be beneficial. For patients with other risk factors (G3, L1, V1), a benefit can neither be excluded nor confirmed.

For locally advanced squamous cell carcinoma of the pharynx, larynx and oral cavity, several randomized studies have shown an improvement in relapse-free survival [349], [350], [351]); in a meta-analysis [402] improved overall survival for patients in whom resection <=5mm was performed in healthy tissue was also shown. However, no patients with T1-T2 N0 tumours were treated in these studies, so that the benefit of additional simultaneous chemotherapy to postoperative radiotherapy is unknown.

In principle, some T3 hypopharyngeal carcinomas can also be safely resected using transoral procedures, preferably TLM. Please refer to Section 8.4.1. The vast majority of stage III and IV hypopharyngeal carcinomas will be reserved for open transcervical techniques in the event of a surgical procedure:

Transcervical laryngeal and partial hypopharyngeal resection

These resections are exceptional situations with strict indications because they can result in a lengthy rehabilitation phase of months to years for the patients concerned. In this case, a hemilaryngectomy is combined with a partial resection of the hypopharynx (medial and/or lateral piriform sinus wall) for laryngeal tumours growing transglottically on one side. Depending on the extent of the resection, the laryngeal skeleton can be built up with a rib cartilage graft and the hypopharynx reconstructed with a forearm graft. The creation of a wide piriform sinus with a thin forearm graft maintains the flexibility of the reconstructed larynx and thus enables elevation during the act of swallowing. This forward and upward movement of the larynx with simultaneous opening of the oesophageal inlet is the prerequisite for aspiration-free food intake [[5]; [410]; [411]; [412]; [413]; [281]; S3-Larynx chap. 7.5 [2]]. The laryngotracheal flap (LTF) as a local tissue transfer should also be mentioned here [414]. By using a laryngotracheal approach for hypopharyngeal tumour excision, the contralateral LTF can be used to reconstruct the hypopharyngeal defect. Although the contralateral uninvolved laryngotracheal tissue remains intact, this does not appear to increase the tumour recurrence rate. This may also reduce the use of complicated regional or free flaps. According to Chu et al. from Teipei, 75% of hypopharyngeal defects could be reconstructed with this flap without other flaps. Postoperative complications are rare, with only 2% of patients having a pharyngoesophageal stenosis and 5% a pharyngocutaneous fistula [415].

Pharyngolaryngectomy

As primary therapy, pharyngolaryngectomy is still an important part of the therapeutic options for advanced-stage carcinomas. In principle, direct closure of the pharynx is attempted if there is still sufficient residual tissue in the hypopahrynx. The use of a voice prosthesis is also recommended for voice rehabilitation using the primary technique (detailed description in S3-LL Laryngeal carcinoma [2]). The ideal method for hypopharyngeal reconstruction should preferably have the following characteristics: single-stage procedure, high success rate of tissue transfer, low removal morbidity, low fistula and stenosis rates, restoration of the ability to speak and swallow, successful reconstruction achievable in a heavily irradiated area and tolerance of postoperative radiotherapy (overview [415]).

Complete, i.e. circular hypopharyngeal resection generally requires reconstruction of the alimentary canal. The forearm graft is the method of choice for this. It is important to plan the graft sufficiently wide in order to be able to model the widest possible "tube". To prevent postoperative stenosis, the lower anastomosis with the oesophageal inlet should be dissolved by triangular exchange [411].

Another variant of stenosis prophylaxis with simultaneous formation of a very wide neohypopharynx was described by Bootz et al. [416]. In this case, the forearm graft is sutured to the prevertebral fascia with the longitudinal edges in a U-shape. This reduces shrinkage and stenosis to a minimum. Alternatively, the reconstruction can also be performed with a pedicled flap. This is suitable for poor vascularization after extensive resection and in the context of salvage resections after primary radiochemotherapy. Voice rehabilitation is performed with voice prostheses that are placed in the anterior esophageal wall below the caudal anastomosis.

Alternatively, jejunum transfer is recommended as a replacement after circular laryngopharyngectomy. Due to the segmental blood supply of the jejunum, up to 20 cm of jejunum can be harvested on the basis of a single vascular arcade. The transfer of the vascular mesentery with the jejunum is a further advantage of this flap, as it allows all dead spaces to be filled and all important vascular structures to be covered. In patients with limited oesophageal dilatation ("small" T4b), free jejunal flaps have an overall success rate of 90 - 100%. In addition, fistula and stricture rates are acceptably low [417][418]. Clinical experience with the jejunal flap has shown a high rate of successful recovery of swallowing function, with earlier rehabilitation and restoration of swallowing compared to other reconstructive methods. However, some patients may suffer from intermittent dysphagia due to uncoordinated peristalsis during swallowing. Voice rehabilitation is also a major problem with this visceral transfer [419]. A tracheoesophageal puncture with voice prosthesis insertion results on average in a less satisfactory voice than is achieved with skin flaps. Excessive mucus production tends to clog the prosthesis and results in a typically wet voice that lacks volume [419]. The free jejunal flap also requires a laparotomy to harvest the graft. The need for intra-abdominal surgery exposes the patient to additional abdominal morbidity and even mortality [420][419][420]. Adhesions, intestinal bleeding, intestinal obstruction and anastomotic or abdominal wound dehiscence are among the possible complications that have led to a reluctance to perform jejunum transfer in recent years.

A very rarely used alternative for voice rehabilitation is the construction of a speech siphon with simultaneous reconstruction of the alimentary canal from overlong jejunal segments and was also described by Remmert et al. [[421], [422]]. The procedure is primarily suitable for younger patients due to the need for abdominal surgery and has the advantage of a very short rehabilitation period of a few weeks to regain swallowing and vocal function ([2] Chapter 7.5).

Pre-therapeutic tracheostomy

When a hypopharyngeal carcinoma with relevant laryngeal involvement is first diagnosed, there may already be considerable dyspnoea, which often necessitates a pretherapeutic tracheostomy (analogy Ch. 7.1 [2]. Patients with extensive laryngeal carcinomas are often admitted to hospital for emergency treatment. A primary tracheostomy is often performed as an emergency measure. It is known from retrospective observations that tracheotomy, which is necessary before a laryngectomy or partial laryngectomy, can have a negative effect on the prognosis [423][424][425][426]. In particular, recurrences in the tracheostoma area are described more frequently in such cases, which are difficult to treat with both surgery and radiation. If a recurrence occurs in the area of the tracheostoma, the prognosis with regard to survival deteriorates significantly [427][427][428]. If a laryngectomy is considered as the primary treatment option, the procedure should be performed promptly in the event of dyspnoea in order to prevent a pre-therapeutic tracheostomy, or the patient should remain intubated until the laryngectomy (exceptional cases). Furthermore, attempts can be made to avoid tracheotomy, e.g. by endoscopic tumour debulking during the first examination under anaesthesia. If the decision is made to undergo primary radiotherapy, a pre-therapeutic tracheostomy is unavoidable, whereby this should be included in the radiation field to prevent recurrence in the tracheostoma area [429].

Pharyngolaryngectomy + oesophageal resection + gastric pull-up

The very radical and extensive surgery for T4b hypopharyngeal carcinomas with oesophageal infiltration mentioned at the beginning is only possible and sensible in very few cases. Surgery-related mortality is particularly high in this patient group due to the high level of comorbidities. The literature reports a mortality rate of between 5% and 25% together with an overall incidence of complications of between 26% and 55% [430][430][431][431][432] [433]. In this respect, gastric pull-up should therefore be reserved for exceptional situations (alternative: colonic interposition). In any case, a surgical partner team that is highly experienced in oesophageal surgery is essential for the success of the operation [381].

Primary radiochemotherapy

For locally advanced squamous cell carcinoma of the pharynx, larynx and oral cavity, the MACH-NC meta-analysis [331], based on 34 randomized studies with 6788 patients, showed an absolute survival benefit of 6.5% after 5 years for simultaneous radiochemotherapy compared to radiotherapy alone (p<0.001). The tumour location was known in 4650 patients. 756 patients in this group had locally advanced hypopharyngeal carcinoma in stages III-IVb (M0). In this subgroup, the benefit of simultaneous radiochemotherapy in terms of survival compared to radiotherapy alone was slightly lower (hazard ratio 0.88, 95% confidence interval 0.75-1.04) than for oropharyngeal carcinoma (hazard ratio 0.81, 95% confidence interval 0.71-0.95) and not formally statistically significant. However, independent evaluations of registry data [385] [400] [12] show a significant survival advantage of 10-15% in favour of simultaneous radiochemotherapy for locally advanced hypopharyngeal carcinomas. The advantage was highest in patients under 60 years of age and hardly detectable in patients over 70 years of age. In contrast, there was no advantage for induction chemotherapy or adjuvant chemotherapy in the MACH-NC meta-analysis [331]. Irrespective of this, induction chemotherapy can be used if a primary laryngectomy would be necessary surgically in order to achieve laryngeal organ preservation depending on the response to induction chemotherapy, as described in more detail under 8.5.2.

The survival benefit for simultaneous radiochemotherapy compared to radiotherapy alone is best confirmed for the simultaneous application of cisplatin-containing chemotherapy and radiotherapy [331]. Most experience exists for 3x100 mg/m² cisplatin and weekly cisplatin (40 mg/m²) as well as for cisplatin in combination with 5-FU.

For patients in whom cisplatin is not an option, e.g. due to poor renal function, randomized studies in which patients with hypopharyngeal carcinoma were also treated demonstrated a survival advantage for the simultaneous application of carboplatin in combination with 5-FU [308] [309] [311] [312], mitomycin C in combination with 5-FU [313] [Keane, TJ et al. 1993] or weekly docetaxel [315] simultaneously with radiotherapy compared to radiotherapy alone. Only studies with <=80 patients in the radiochemotherapy arm are available for monochemotherapy with carboplatin [316] [317] [310] [318] and there were no significant advantages for monotherapy with mitomycin C [319] [320] [321] in parallel with radiotherapy. Weekly applications of carboplatin and paclitaxel concurrently with radiotherapy have been investigated in a number of non-randomized phase II trials and some retrospective cohort studies [322], [323], [324], [325], [326], [327], [328]. Hypopharyngeal carcinomas represented only a small proportion of the treated squamous cell carcinomas of the head and neck. The non-randomized, retrospective comparisons indicate a similar efficacy of carboplatin + paclitaxel compared to weekly therapy with 40mg/m² cisplatin in parallel with radiotherapy. In contrast, there is very little data available for the administration of paclitaxel alone [329]. A benefit of cetuximab in combination with radiotherapy for hypopharyngeal carcinoma cannot be derived from the available data [330][306][335].

In comparison to radiotherapy alone, simultaneous radiochemotherapy with cisplatin and other chemotherapies not only increases hematotoxicity, which is usually well manageable, but also the severity of acute mucositis with the associated swallowing difficulties. Other typical acute side effects of radiotherapy, such as skin reaction, xerostomia and loss of taste, are less severely increased by simultaneous chemotherapy, but the individual effects add up so that radiochemotherapy is significantly more stressful for patients overall [434]. An increase in the rate of late effects of radiotherapy due to additional chemotherapy is less well documented and has mostly been described as non-significant or borderline significant in individual studies [338] [435] [436] [437]. Due to the different ways in which late effects are documented, a valid meta-analysis of these data cannot be meaningfully carried out. Nevertheless, there is little doubt that the late side effects are also increased by the additional chemotherapy [434]. The working group led by J.A Langendijk [438] was able to prove in prospective, systematic analyses [439], taking into account the individual dose distribution [440] of radiotherapy, that the rate of long-term swallowing disorders after combined radiochemotherapy is significantly higher than after radiotherapy alone [441] [442]. However, the same working group was also able to show that these negative effects can be at least partially reduced by using modern radiotherapy techniques such as swallowing sparing IMRT [290]. However, the effects described above are poorly documented for the subgroup of hypopharyngeal carcinomas, although swallowing disorders are to be expected particularly frequently due to the location of the tumours.

Laryngeal organ preservation

The topic of laryngeal organ preservation programs was initiated in the 1990s by the first large randomized studies in the USA and Europe with the introduction of definitive non-surgical protocols instead of laryngectomy, which in the event of failure resulted in salvage laryngectomy. For the history and the main treatment options established for advanced laryngeal carcinoma, please refer to the S3 Guideline on Laryngeal Carcinoma [2].

There are only a few randomized studies (Phase III: EORTC 1996 [234]; EORTC 2009 [443]), (Phase II: TREMPLIN [444]; DELOS II: [445]) for organ preservation of advanced hypopharyngeal carcinoma as an alternative to pharyngolaryngectomy. The first large randomized study and the only one that directly compared a non-surgical with a surgical treatment procedure was initiated by the EORTC and recruited 202 patients with untreated, operable hypopharyngeal (78%) and laryngeal (22%) carcinomas [234]. Patients were randomized into an experimental arm with up to three cycles of induction chemotherapy with cisplatin/5-FU followed by standard fractionated radiotherapy up to a total dose of 70 Gy and a standard arm with partial pharyngolaryngectomy and postoperative radiotherapy. Patients in the experimental arm who developed a partial remission after the first cycle of induction chemotherapy received a second and third cycle. Patients who developed a complete remission at any point during induction chemotherapy were randomized to radiotherapy. Patients who developed less than a partial remission after the first cycle and less than a complete remission after the third cycle were assigned to pharyngolaryngectomy without prior radiotherapy. After induction chemotherapy, 54% of patients developed a complete remission in the area of the primary tumour and 43% in the area of the primary tumour and the locoregional lymphatic drainage pathways. After five years, there was no significant difference in survival between the two study arms (30% versus 35% in the surgical arm). The 3- and 5-year survival rates with functional larynx were 42% and 35%, respectively. In patients who achieved complete remission after induction chemotherapy and subsequently received only radiotherapy, a functional larynx was still present in 58% of patients after 5 years. It should be noted that only 6% of the included patients had a T4 tumour (T3: 75%, T2: 19%), so that no valid statement can be made from this study as to whether the induction concept can also be offered to patients with T4 tumours without a survival disadvantage compared to primary surgical therapy +/- radiotherapy or radiochemotherapy.

The risk of distant metastasis was significantly lower in the induction chemotherapy arm (25% versus 36%). In the second EORTC study mentioned, an alternating chemotherapy protocol was compared with a sequential chemotherapy protocol with a standard arm of the protocol described above. In 450 patients, there was no difference between the two arms [443]. A study directly comparing primary simultaneous radiochemotherapy and primary laryngectomy has not yet been conducted.

If the aim is to achieve functional laryngeal preservation using induction chemotherapy, a taxane-containing chemotherapy protocol has proven to be advantageous (taxane, platinum, 5FU: TPF) [446], [447]. With high response rates and good selection options, only a few salvage operations are usually necessary. The additional application of docetaxel during induction therapy also increased the laryngeal preservation rate 3 years after therapy from 57.5% to 70.3% (p=0.03) in a randomized study [446].

This effect was also confirmed in the DeLOS II study, albeit without the additive value of cetuximab for induction of organ preservation and overall survival. Early laryngectomy prior to radiotherapy was performed in 30% of patients after lack of response to short induction (1 cycle TPF/TP and subsequent endoscopic response assessment), which led to comparatively low toxicities/complications with very good 2-year survival [445]. In the DeLOS-II study, 50% of 170 patients had hypopharyngeal carcinomas with 69% stage IV tumours. The 2-year survival was 70% with 49% laryngectomy free survival.

Furthermore, induction therapies are being modified in the direction of less toxic substances, and 5-FU is being partially replaced. The randomized phase II TREMPLIN trial showed that cetuximab can achieve equally good results as simultaneous administration of cisplatin with a different toxicity profile (greater cutaneous toxicity, less haematotoxicity and nephrotoxicity), which may be advantageous for some patients [444].

Critics of organ preservation programmes cite the high function-limiting late toxicity, the higher non-tumour-related mortality and the comparatively high complication rates if salvage surgery is necessary. The supporters, on the other hand, cite the high rate of organ preservation with adequate patient selection, the possibility of salvage surgery as a curative option and the lack of a survival disadvantage. Careful individual counselling of the patient is therefore of particular importance in this situation. In the S3 guideline on laryngeal cancer, the option of primary non-surgical laryngeal organ preservation was included as an alternative to laryngectomy and should not be withheld from patients as a realistic treatment option in the consultation. This recommendation is adopted for hypopharyngeal carcinoma in the same way as for laryngeal carcinoma.

Postoperative radiotherapy

The indication and implementation of adjuvant radiotherapy and radiochemotherapy do not differ in principle from those of HPV/p16-negative oropharyngeal carcinoma (see 8.1.4).

The effect of postoperative radiotherapy compared to no adjuvant radiotherapy has not been investigated in randomized studies. Results of cohort studies and prospective registry data indicate that in patients with more than one tumour-involved lymph node, with extracapsular tumour growth at the lymph nodes (ECE) or with tumours only barely resected in healthy tissue (<5mm), adjuvant radiotherapy is clearly superior [342], [343], [344], [345]. These studies included squamous cell carcinomas of the larynx, oropharynx, hypopharynx and oral cavity. The proportion of hypopharyngeal carcinomas was between 12% and 43%. The available data suggest that adjuvant radiotherapy reduces the locoregional recurrence rate and improves survival in hypopharyngeal carcinomas if more than one regional lymph node is affected or if the primary tumour was resected just inside the healthy area (<5mm). For patients with pT3/pT4 tumours, lymph node involvement with ECE or R1 resection, there are no reliable data on treatment outcomes without adjuvant radiotherapy for hypopharyngeal carcinoma, as there is an international consensus that these patients require adjuvant radiotherapy [348].

For patients with an intermediate risk of recurrence (no ECE and resection >=5 mm in healthy tissue), 60 Gy (5x 2 Gy per week) was applied independently in the region of the former primary tumour and the affected lymph nodes in the vast majority of patients in the clinical trials and the registry data. In the adjacent, unaffected lymph node regions, 45 - 50 Gy with 5 x (1.8 - 2.0) Gy per week were administered electively. There is an international consensus for the selection of the lymph node levels to be electively irradiated, which can be regarded as a standard guideline [Biau, J et al. 2019]. IMRT techniques are also considered standard for postoperative radiotherapy.

Postoperative radiochemotherapy

The value of additional cisplatin-containing chemotherapy administered simultaneously with radiotherapy has been investigated in 3 large randomized trials for locally advanced squamous cell carcinoma of the oropharynx, hypopharynx, larynx and oral cavity [349], [350], [351]). The proportion of hypopharyngeal carcinoma in these studies was between 10% and 20%. Subgroup analyses by tumour location are not available from these studies. In the registry studies on hypopharyngeal carcinoma, the effect of postoperative radiochemotherapy was not compared with the effect of postoperative radiotherapy alone. Nevertheless, it seems plausible to assume that the results of the 3 randomized studies also apply to hypopharyngeal carcinomas, even if it cannot be ruled out that the effect sizes may differ slightly from those of the other tumour localizations. The results of the studies consistently showed that patients with evidence of ECE in the affected lymph nodes or an R1 resection (defined in these studies as resection <5mm in healthy tissue) have a survival advantage after adjuvant radiochemotherapy compared to adjuvant radiotherapy alone, so that postoperative radiochemotherapy is considered standard for these patients.

If postoperative radiochemotherapy is indicated, it should be noted that results of randomized studies are only available for cisplatin or cisplatin + 5-FU, which have shown a significant reduction in the relapse rate or an improvement in survival [349] [350], [351]. In a randomized Japanese study (n=261), 3 x 100 mg/m² cisplatin at intervals of 3 weeks were compared with weekly 40 mg/m² cisplatin simultaneously with postoperative radiotherapy [360]. The non-inferiority of the weekly administration of cisplatin was demonstrated. The proportion of patients with hypopharyngeal carcinoma in this study was 34%.

For patients who cannot receive cisplatin, e.g. due to impaired renal function, there are little data available for postoperative radiochemotherapy. For mitomycin C, data are available from two small randomized studies that only showed a trend towards a survival benefit [321], [361]. The weekly administration of docetaxel simultaneously with radiotherapy was compared with radiotherapy alone in a randomized study [315]. In a small subgroup of this study, the therapy was also given in the postoperative situation. There was a trend towards improved survival for the combined therapy. No results of randomized studies are available for carboplatin + 5-FU and carboplatin + paclitaxel in the postoperative situation. However, it seems plausible to use these combinations in the postoperative situation by analogy with the efficacy of these combinations in primary radiochemotherapy (see 8.5.3). There are no data suggesting a benefit for the use of cetuximab in the postoperative situation.

In postoperative radiochemotherapy in the studies described, the regions with the highest risk of recurrence (lymph nodes with ECE and areas of scarce resection) were irradiated with 5 x 2 Gy per week up to 66 Gy. The adjacent regions and electively irradiated lymph node sites also received between 54 and 60 Gy in conventional fractionation [349] [350]. The frequency of grade III/IV late side effects in the US study [349] after radiochemotherapy was between 1 - 6% higher than after radiotherapy alone in various categories, although no statistical significance was achieved. To reduce side effects, postoperative radiotherapy or radiochemotherapy is now routinely carried out using IMRT techniques.

The indication for elective neck dissection in patients with oropharyngeal and hypopharyngeal carcinomas in tumour stages T1 and T2 with a clinically negative lymph node status depends largely on the expected occult cervical metastasis rate. Cervical lymph node sonography, computed tomography and/or magnetic resonance imaging are currently standardly used to determine the clinical N-status. The detection accuracy can be increased by advanced procedures such as positron emission tomography (comprehensive description of imaging in section 7.3.) and/or sentinel lymph node biopsy [453], [454]. Studies on sentinel lymph node biopsy indicate that it could be equivalent to elective neck dissection in terms of patient survival in patients with early-stage head and neck tumours with clinically negative neck lymph node status ([225], [455]; see also Section 7.6.). Elective neck dissection is currently the gold standard for determining nodal status.

Currently, elective neck dissection is indicated for an expected occult metastatic probability of 15% to 20% ([456], [457], [458], [459], [460], [461], [462]). If the expected occult metastasis rate is lower, watchful waiting can also be performed; if a neck dissection is nevertheless performed, the potential patient benefit and the expected morbidity must be carefully weighed against each other. Relevant complications that must be taken into consideration are, in addition to bleeding events, primarily sensory or motor nerve damage (especially of the accessorius nerve), chronic lymphedema, chyle fistulas and wound healing disorders ([463], [464], [465], [451], , [452], [466], [467], [223], [468], [469]).

T1/2 cN0 oropharyngeal carcinoma

For oropharyngeal carcinomas in the early stages T1 and T2N0M0, the rates of occult lymph node metastasis described in the literature range from approximately 12 - 39% [458], [467], [470], [471], [472], [473], [474], [475], , [476], [477], [478], [479], [480], [481]. The difference between patients with tumour size T1 compared to T2 is remarkable: For T1 tumours, the risk of occult lymph node metastases is less than 15%, while rates of around 30% are described for T2 and above. The latter corresponds to the occult metastasis rates in patients with primary tumours in the advanced tumour sizes T3 and T4. In the case of primary tumours clearly located laterally to the midline, there is also a difference in the lateral distribution of metastases. While ipsilateral occult metastases are found in 17 - 33% of patients with T1 and T2 oropharyngeal carcinomas across all stages, the rate for contralateral occult metastases is 0 - 29% [460], [471], [477], [482], [483], [484], [Koyfman, SA et al. 2019], [485], [486], [O'Brien, CJ et al. 2000], [488], [489], [490], [491], [492].

Occult contralateral metastases are found in 0 - 20% of cases of T1 tumours and in 5 - 35% of cases of T2 tumours ([460], [472], [485], [492], [493]). Anatomical differences within the oropharynx must also be taken into account: When the soft palate and the base of the tongue are affected, the rates of contralateral occult metastasis are higher than for tumours in the tonsil area [Koyfman, SA et al. 2019], [485], [494], [495], [O'Sullivan, B et al. 2001], [497], [498], [499] due to the intersecting lymphatic drainage pathways. Occult bilateral metastasis rates of approximately 14% have been described for soft palate carcinomas as early as stage T1, and up to 20% for tongue base carcinomas. From stage T2, the rates are approximately 32% for soft palate carcinomas and approximately 35% for tongue base carcinomas [460], [Koyfman, SA et al. 2019], [O'Sullivan, B et al. 2001], [498].

There are also clear differences with regard to the involvement of the lymph node levels: With cN0 status, occult lymph node metastases predominantly manifest themselves in the ipsilateral levels II and III. The distribution according to lymph node regions is as follows ipsilaterally: Level Ib: 0 - 9%, Level IIa: 26 - 75%, Level IIb: 5 - 14.9%, Level III: 0% - 41%, Level IV: 0% - 9%, Level V: 0 - 12% [458], [481], [485], [492], [500], [501], [502], [503]. A higher T-stage, extracapsular growth, clinically ipsilateral lymph node involvement, lymphangiosis carcinomatosa or oropharyngeal carcinoma of the tonsil increase the probability of ipsilateral level IIb involvement [458].

Overall, HPV-associated oropharyngeal carcinomas in tumour stages T1 and T2 metastasize lymphogenously more frequently than non-HPV-associated tumours [504], [43]. With regard to the influence of HPV association on the presence of occult metastases, no clear data are currently available: For example, Shoustare and Kato et al. describe a higher probability of occult contralateral metastases compared to non-HPV-associated tumours; in contrast, Shah, Amsbbaugh and Tritter et al. describe lower metastasis rates [481], [497], [498], [500], [505].

There are currently no randomized controlled phase III studies on the issue under discussion. Taking into account the above-mentioned study results, elective ipsilateral selective neck dissection is always indicated in patients with oropharyngeal carcinomas of the tonsil and the lateral pharyngeal wall that are clearly lateral to the midline, i.e. already at tumour size T1. In patients with stage T2, bilateral elective selective neck dissection should be performed. In patients with tumours close to the midline, soft palate and tongue base carcinomas, bilateral selective neck dissection is already recommended at stage T1. This applies to patients with both HPV-positive and HPV-negative oropharyngeal carcinomas. Both ipsilateral and contralateral elective selective neck dissection can be limited to levels II, III and IV [Koyfman, SA et al. 2019], [494], [506].

T1/2 cN0 hypopharyngeal carcinomas

Hypopharyngeal carcinomas have a high occult ipsilateral metastasis rate of 24 - 41% regardless of the extent of the primary tumour in stage cN0 [468], [470], [507]. However, a cN0 stage is rarely found in affected patients at initial diagnosis; cervical lymph node metastases are usually already clinically manifest. With a primary tumour size of T1, the occult contralateral metastasis rate is 8% [508]. The location of the primary tumour significantly influences the contralateral metastasis rate: in T1 tumours of the anterior wall of the piriform sinus, the lateral hypopharyngeal wall or the posterior hypopharyngeal wall, the rate is around 5% [508], [509], while occult contralateral metastases are already present in around 13% of T1 cases with infiltration of the medial wall of the piriform sinus. Tumour involvement of the postcricoid region, the medial wall of the piriform sinus, tumour growth close to the midline in the area of the posterior pharyngeal wall or tumour infiltration of more than 1 mm in depth increase the risk of contralateral lymph node metastasis [508], [510], [511], [512], [513].

For primary tumours in stage T2, the rate of occult contralateral metastasis is up to 21% [514], [515]. This already corresponds to the rate for advanced tumour stages T3 and T4 [508]. The distribution of occult lymph node metastases according to metastasis level is as follows: Level Ib: 3.8%, Level IIa: 53.8%, Level IIb: 7.7%, Level III: 50%, Level IV: 19.2%, Level V: 7.7% [516].

There are currently no phase III studies available on the issue in question. Taking into account the above-mentioned study results, an ipsilateral elective neck dissection on the side affected by the tumour can be performed in patients with strictly unilateral tumour localization in the area of the hypopharyngeal side wall or posterior wall in stage T1. In the case of tumour localization close to the midline or indications of deep tumour infiltration or from category T2, bilateral elective neck dissection is recommended. Selective neck dissection of levels II, III and IV should be performed both ipsilaterally and contralaterally. Level VI neck dissections are performed in the hypopharynx as needed to resect the primary tumour and all clinically suspicious neck nodes. Elective dissection depends on the extent and location of the primary tumour.

Based on the basic locoregional metastatic behaviour of oropharyngeal and hypopharyngeal carcinomas in Section 8.6.1 and the current ASCO recommendations [Koyfman, SA et al. 2019], curative neck dissection is indicated as an essential component of definitive surgery in the N+ situation. In the case of unilateral nodal-positive laterally located oropharyngeal and hypopharyngeal carcinoma, at least one ipsilateral curative neck dissection should be performed as part of a primary surgical procedure, regardless of the stage. If the tumour location is close to or even exceeds the midline, a bilateral neck dissection should be performed even in a unilateral cN+ situation. The extent of a curative neck dissection as part of a primary surgical procedure for nodal-positive oropharyngeal or hypopharyngeal carcinomas should depend on the extent of lymph node involvement on the respective side of the neck. The minimum extent includes selective neck dissection of level IIa, III, IV and can be increased to radical neck dissection [Koyfman, SA et al. 2019].

In patients with lateralized oropharyngeal and hypopharyngeal carcinoma undergoing neck dissection simultaneously or prior to transoral endoscopic head and neck surgery, ligation of vulnerable supplying blood vessels should be performed to reduce the severity and incidence of postoperative bleeding. Patients with cN+ disease who have either clear extranodal extension into the surrounding soft tissue or involvement of the carotid artery or cranial nerve should be offered a non-surgical approach. Patients with biopsy-proven distant metastases should not undergo routine surgical resection of metastatic cervical lymph nodes [Koyfman, SA et al. 2019].

For advanced hypopharyngeal carcinoma treated primarily by surgery, level VI dissection (including pretracheal lymph nodes, delphic lymph nodes and unilateral or bilateral paratracheal lymph nodes) and hemithyroidectomy up to total thyroidectomy may be indicated.

The term "salvage surgery" is classified internationally as follows: " Salvage surgery is no longer limited to patients who failed radiotherapy or radiochemotherapy, but also includes patients who previously underwent surgical treatment for tumours located from the base of the skull to the lower neck areas (including thyroid cancer)"[517]. This means that even after sole surgical treatment of the initial carcinoma in the oropharynx or hypopharynx in the event of a new recurrence/second tumour or locoregional metastases in the former resection area, repeat surgery is referred to as salvage surgery. Salvage neck dissection is described in more detail in Section 9.1.1. as part of the more detailed considerations on salvage surgery.

Planned neck dissection after definitive radiochemotherapy for node-positive oropharyngeal and hypopharyngeal carcinomas is a special situation. Here, salvage neck dissection should only be considered immediately after primary therapy if the FDG-PET-CT proves positive. The background and recommendations were explained in detail in chapter 7.3 (see recommendations 7.14 - 7.16, [202] [Koyfman, SA et al. 2019]).