Introduction

Circumferential laryngopharyngectomy and cervical oesophagectomy present a unique reconstructive challenge due to the complete loss of both the anterior and posterior pharyngeal walls. The radial forearm free flap (RFFF), the latissimus dorsi flap, the anterolateral thigh flap, the free jejunal flap (FJF) and the pectoralis major myocutaneous flap (PMMC) are more commonly described in the literature for reconstruction used either tubed or sutured in a ‘horseshoe-shaped’ fashion to the prevertebral fascia. [1] In the early surgical series, the use of free flaps was associated with a higher rate of post-operative fistula formation (up to 67%) compared to PMMC flap reconstruction (22%) [2, 3] but the percentage of such complications in the former groups has significantly improved in more recent studies (11–14% fistula;14–16% stenosis). [4] However, the biggest challenge is regarding the long-term functional outcomes of swallow and speech, which remain poor and difficult to produce good outcomes consistently.

Despite the fact that free tissue transfer remains the first option for reconstruction of circumferential pharyngeal defects, pedicled flaps are a valid alternative in regions where there is no availability of microsurgical team. Over the past 18 months, many head and neck units have opted to reconstruct oncological defects using pedicled flaps to shorten operative time and the need for post-operative intensive care unit to allow patients to proceed with their operations and reconstruction despite the scarce hospital resources during the covid-19 pandemic. [5, 6] During the first wave of the pandemic some units completely stopped performing free flap reconstruction and opted for pedicled and/or staged defect reconstruction. [7] During the Covid-19 pandemic, we have continued to perform laryngectomies with pedicled flap reconstruction as needed following the peri-operative protocol recommended by the British Association of Head and Neck Oncologist [8], we have also published our local hospital peri-operative protocol and our experience with the management of a patient found to be covid-19 positive during the pre-operative assessment stage. This required postponing the procedure until he was negative. [9]

In our unit, it was common practice prior to the pandemic to reconstruct the circumferential pharyngeal defects with a single staged procedure using a deltopectoral flap for reconstruction of the posterior pharyngeal wall and another flap (PMMC or supraclavicular flap in our case series) to complete the anterior reconstruction of the circumferential pharyngo-oesophageal defects. This is the first time that this single stage approach is described in the literature. In this paper, we are presenting our experience with the technique, including a detailed presentation of the surgical steps and our surgical outcomes.

Methods

Patient Population

A retrospective case series review was conducted in the head and neck unit of a district general hospital, from January 2005 (when the senior author SK started his consultancy post), until December 2020. This identified all patients who had undergone circumferential laryngopharyngectomy ± cervical oesophagectomy with dual flap reconstruction using the deltopectoral (DP) flap to reconstruct the posterior wall and another flap to cover the defect anterolaterally. Eight patients were identified during the study period.

Ethical Considerations

Institutional approval was attained following discussion at the local Head and Neck Oncology multidisciplinary team meeting and Caldicott Guardian approval was granted.

Surgical Technique

Reconstruction is performed following laryngectomy and circumferential pharyngectomy + /– cervical oesophagectomy. A DP flap is raised in the subfascial plane. The blood supply from the 2nd and 3rd internal mammary perforator arteries allows harvesting of a broad tissue flap with a width extending from the clavicle to the 5th intercostal space, and a length reliably extending towards the shoulder tip. The flap is sutured onto the posterior oropharyngeal mucosa superiorly, and the proximal oesophagus inferiorly (Fig. 1). In one individual, the superior extent of the flap repair extended to the level of the soft palate. This individual had a short and broad neck that allowed the DP flap to reach the soft palate. Nevertheless, this can be challenging in patients with long neck or in cachectic patients with poor tissue quality. In these occasions or when the resection is so extensive to require removal of tissue up to the level of the soft palate, there is no need to over-stretch the DP flap to reach and be sutured to the soft palate line. It is important for the DP flap to reach the tongue base and good closure to be achieved at the lateral superior part of the flap to the tongue base. This will prevent saliva leak at this point and create the broad funnel needed for direction of the saliva via the neopharynx to the oesophagus without causing stenosis. The superior edge of the posterior pharyngeal wall can be left to heal by granulation in the long-necked patients that the DP flap cannot completely reach the mucosa line. This modification is rarely required but should be mentioned as it can accommodate for anatomical variations and differences in the extend of cancer requiring resection. Total time required to raise the DP flap and suture it to the inferior edge of the oropharynx is approximately 30 min.

Fig. 1
figure 1

DP flap raised and sutured into position proximally and distally. Note the conical shape of the neopharynx generated naturally upon positioning the DP flap. A Region of DP flap forming neopharynx. B Region of DP flap to be de-epithelialise

Full size image

A second flap is then harvested. This can be a free flap or a pedicled flap. In our series, a PMMC flap or a supraclavicular fasciocutaneous flap on the opposite side was used. This is to form the anterior and lateral pharyngeal wall. This second flap is raised and inserted in a similar manner to a partial pharyngeal patch repair. [3]

The distal part of the deltopectoral cutaneous tissue is left to form the posterior pharyngeal wall and skin de-epithelialisation is performed to a small middle segment of the flap, so that the only cutaneous tissue buried within the neck is the neopharynx itself (Fig. 2). The skin of the deltopectoral flap near the stump of the oesophagus is sutured, thereby ensuring the integrity of the neo-posterior pharyngeal wall from oropharynx to oesophagus.

Fig. 2
figure 2

Neopharynx being created from the DP flap A and SC flap B around a salivary bypass tube. Laryngeal stoma is notable inferiorly. C De-epithelialised SC flap. D De-epithelialised DP flap. E Pharyngeal remnant

Full size image

The second flap can then be sutured to the anterior mucosa of the proximal oesophagus, the DP flap laterally, and the tongue base superiorly, creating a conical neopharynx. Flap harvest sites are then closed primarily. A salivary bypass tube is positioned in the reconstructed neopharynx to help the healing process splinting it open, which is subsequently removed 3 weeks later, prior to commencement of oral diet. Six-month post-operative outcome is displayed in Fig. 3.

Fig. 3
figure 3

Post-operative outcome – an external view of the neck and stoma. A Deltopectoral flap

Full size image

Results

The case series includes 8 patients (Male:Female = 4:4 with a mean age of 68.6 years (range 55–82). Five procedures were performed as primary circumferential laryngopharyngectomies, and three patients underwent salvage operations following previous radical chemoradiotherapy (7 years, 9 years and 28 years prior to surgery). Primary tumour maximal diameters ranged from 17–75 mm (mean: 41.6 mm).

The anterior pharyngeal wall was reconstructed with a PMMC flap for 5 patients and a supraclavicular flap for the remaining 3 patients. All patients developed a small pharyngocutaneous fistula, laterally in the neck, over the 3-point junction of the DP flap. In all patients this required no intervention, did not result in wound dehiscence, and closed spontaneously in a mean of 15 days post-operatively. For the three patients requiring adjuvant (chemo)radiotherapy, there was no delay in treatment delivery.

Functional outcomes were favourable in the majority of patients. All patients have achieved normal (n = 2) or soft (n = 6) diet, although one individual continues to require PEG tube dietary supplementation. Of the 6 patients that manage soft diet, one developed a low neopharyngeal stricture requiring repeated dilatations. A speech valve has been inserted in 4 patients, with all achieving intelligible speech, 2 patients are awaiting tracheo-oeosophageal puncture which has been postponed during the COVID-19 pandemic. Speech valve insertion was not possible for 2 patients, as the oesophagectomy level is significantly below the level of the stoma. Attempt of insertion of a valve when the oesophageal segment is low can compromise stoma and flap reconstruction healing as the valve can cause tension around the flap and stoma reconstruction. The PMMC is very bulky, so no attempt is made to insert the valve through the PMMC. During the months to follow, the PMMC starts to atrophy and become less bulky, and it is due to this process that our patients achieve good long term voice outcomes. Two patients died during follow up, one following tumour recurrence, and one unrelated to their malignancy. All results are summarised in Table 1.

Table 1 Table displaying individual patient demographics and clinical information
Full size table

Discussion

This case series presents an alternative method for reconstruction of circumferential pharyngeal defects using a dual flap technique, with the rotated deltopectoral flap as the posterior wall of the neopharynx. This technique reconstructs the neopharynx with minimal soft tissue bulk in the central compartment space, creating a wider conduit for food passage. A single tubed flap, with the exception of the jejunal free flap, incorporates all the flap bulk in the central compartment, limiting the calibre of the neopharynx. Positioning of the dual flaps resulted in a conical shaped neopharynx (Fig. 2), resembling the natural pharynx more than a tubed flap reconstruction. The authors theorise that this wider, conical neopharynx would both enhance the swallow function, and allow improved air passage for speech. Our limited series seems to support this, with good functional outcomes reported.

The additional advantage to the DP flap, is that it can be rapidly raised and positioned on the posterior wall, requiring minimal de-epithelialisation. This also converts a complex circumferential defect into a ‘partial’ defect, and consequently, does not contribute significantly to the surgical time. Of course, the surgical time to perform the reconstruction can prove to be longer for the novice surgeon as there is an expected learning curve associated with harvesting and placing the flap, appreciating and taking into consideration the size of the defect that needs to be covered and the quality and characteristics of the surrounding skin and tissues to allow a successful reconstruction. [10].

To the authors knowledge, the combination of a DP and a second flap has never been described before in the literature, nevertheless, the DP flap has been used before to reconstruct the posterior pharyngeal wall in 2 stages procedures from a surgical series reported from the head and neck team in Hong Kong. [11] Various other techniques have been described for reconstructing circumferential defects, including tubed free flaps, tubed pedicled flaps, and suturing of flaps onto the prevertebral fascia. [1, 12] A recent multicentre study concluded that the type of flap used during reconstruction does not impact on the post-operative swallowing outcomes with just over half of patients being able to gain normal diet post-operatively (54% with any flaps; 63% RFFF. 53% PMMC, 58% FJF) with no significant difference between the different flaps being used. [1] Other series have demonstrated superior functional outcomes when free flaps were used (58%—63% adequate oral nutrition; 21–25% stricture rate) compared to tubed PMMC or PMMC with the prevertebral fascia forming the posterior pharyngeal wall (53–91% adequate oral nutrition; 18–40% stricture rate; 43% satisfactory vocal function). [13,14,15,16] In our case series, all patients ultimately achieved oral diet intake, with 7 patients achieving sufficient intake to support their complete nutrition. Only one patient developed a stricture requiring repeated dilatation. All patients who were able to receive a tracheooesophageal puncture attained intelligible speech.

The main weakness of the dual flap technique lies in the likelihood of patients developing a small fistula. This occurs at the de-epithelialised region on the DP flap. While an important consideration in the perioperative period, this has not impacted on subsequent care or adjuvant treatment delivery in any of the patients. Tissue handling is very important during the dual flap reconstruction to ensure that the blood supply to the flaps will not get compromised. This is particularly the case for the fasciocutaneous DP flap. To achieve this, gentle suturing of the junction between the pectoralis major flap and the de-epithelialised segment of the DP flap is of paramount importance. Care should be taken to ensure that the suture line is not too tight in that area, this can make the closure not completely water-tight hence resulting in a minor leak in this segment but on the other hand it ensures the viability of the flap. Overjealous tightening of the suture line can compromise the blood supply to the distal end of the DP flap and strangulate it. The fistula has always remained small and low volume, requiring minimal/no intervention, and the authors are of the opinion that the long-term functional outcomes outweigh the consequences of fistula development. In fact, the small fistula may be advantageous, allowing a means of controlled drainage, as none of the patients have developed significant tissue breakdown. The use of the salivary bypass tube helps to minimise the amount of resultant leak, expediting the healing of the fistula site. The factors that pre-dispose to fistulae have been investigated extensively previously, including comparing between free and pedicled flaps. [17] Despite a recent trend towards increasing free flap use, morbidity profiles appear to be similar between free and pedicled flaps. 17 Although all patients in our case series developed a fistula, the short-lived nature, and the rapid healing indicate that dual flap use appears to be a robust reconstruction technique.

Conclusions

Dual DP/second flap reconstruction of circumferential pharyngeal defects represents a single stage pedicle flaps reconstruction solution to a complex reconstructive problem, with the DP flap converting the circumferential defect into a ‘partial’ defect. This can be considered in areas where there is no availability of microsurgical team or during health crisis like the current covid-19 pandemic. Our series demonstrated the reliability of this technique in providing reasonable quality of life to patients. The authors recommend this as an alternative to the single pedicled or free flap tubed reconstruction for large circumferential pharyngeal defects.