Discussion

The advantages of reconstructive techniques for the closure of surgically induced defects of the skull base enabled important progress in endonasal skull base surgery. The ultimate goals of reconstruction techniques in the skull base region include: the stable separation between the nose and the cranial cavity, the protection of neurovascular structures, the conservation or reconstruction of cosmoses, the preservation or reconstitution of function, and the avoidance of dead spaces. In this respect, the separation of the nasal and cranial cavities is of utmost importance, because it prevents post-operative CSF leaks, pneumocephalus and intracranial infections, and protects cranial nerves and large vessels from infection and trauma. The effects of post-operative radiotherapy also have to be considered.

Although a number of different techniques have been used successfully to endoscopically treat CSF leaks that occur after trauma, iatrogenic injury or in spontaneous rhinoliquorrhoea, these methods have proven inadequate for the reconstruction of large defects in extended endonasal skull base surgery.^{Reference Stammberger, Lund and Gleeson5–Reference Schick and Draf8}

As a neuro-endoscope is utilised, especially after an endoscopic endonasal approach, the incidence of complications, such as CSF leak, infection and encephalocele formation, is high. Carrabba et al. reported that the incidence of CSF leak was 24 per cent after an endoscopic endonasal approach.^{Reference Carrabba, Dehdashti and Gentili9} Therefore, skull base defect reconstruction is of paramount importance to prevent a CSF leak after endoscopic endonasal surgery.^{Reference El-Banhawy, El-Dien, Zolfakar, Halaka and Ayad10,Reference El-Banhawy, Halaka, Altuwaijri, Ayad and El- Sharnoby11} In patients with large dural defects of the anterior and ventral skull base, there is a significant risk of a post-operative CSF leak after reconstruction.

Reconstruction with vascularised tissue is desirable, to facilitate rapid healing, especially in irradiated patients. The Hadad–Bassagasteguy flap, a neurovascular pedicled flap of nasal septum mucoperiosteum and mucoperichondrium based on the nasoseptal artery, was first developed at the National University of Rosario, Argentina, for the reconstruction of ventral skull base dural defects.^{Reference El-Banhawy, Halaka, Altuwaijri, Ayad and El- Sharnoby11}

It was not until local vascularised flaps, in particular the Hadad–Bassagasteguy flap,^{Reference Locatelli, Rampa, Acchiardi, Bignami, De Bernardi and Castelnuovo6} were developed that the rate of post-operative CSF leaks, even after expanded skull base resections, could be reduced to below 5 per cent.^{Reference Eloy, Patel, Shula, Choudhary and Liu12} In the landmark study that introduced the Hadad–Bassagasteguy flap, Hadad et al. reported a post-operative leak rate of 4.5 per cent, without any partial or complete flap loss.^{Reference Hadad, Bassagasteguy, Carrau, Mataza, Kassam and Snyderman1} Recently, Eloy et al. reported a post-operative CSF leak rate of 3.1 per cent in a study of 96 skull base defects with high-flow CSF leaks that included 47 sellar defects (including 2 revision cases: 1 recurrence and 1 residual tumour) (Table 7).^{Reference Eloy, Patel, Shula, Choudhary and Liu12}

Table 7. Presenting outcome profile of posterior nasoseptal flap in previous studies

CSF = cerebrospinal fluid

A post-operative CSF leak rate of 2 per cent (2 out of 100) was observed in our study following the posterior nasoseptal flap repair of 100 sellar defects, with the use of an external lumbar drain in 86 per cent of cases. This is in line with the published results of post-operative CSF leak rates after posterior nasoseptal flap repair. The relatively higher rate (10.66 per cent) of post-operative CSF leaks observed by Kassam et al. was predominantly found in those patients who required intra-arachnoidal dissection, and the authors partly attributed it to their initial experience with the technique.^{Reference Kassam, Thomas, Carrau, Snyderman, Vescan and Prevedello4}

All of the 100 patients in our study had sellar defects; 13 had macro sellar defects and 87 had micro sellar defects. A substantial, non-traumatised and well-perfused flap is essential for the efficacy of the repair technique. Early harvesting of the flap (before resection) is meant to ensure the quality of the flap and thereby the repair. Like most of the previously published studies, we also employed the pre-harvesting approach in all of our patients. Although the exact size of the expected skull base and dural defect is unknown, a good estimation can be obtained by carefully reviewing the pre-operative imaging scans. Early harvesting of the posterior nasoseptal flap allows a maximally sized flap that can be tucked into the nasopharynx or maxillary sinus, away from inadvertent trauma during the approach and tumour resection. A maximally sized posterior nasoseptal flap that has not been compromised by trauma, with a well-preserved vascular pedicle, may provide the most robust closure. This is even more important in cases where the approach used to access a lesion may involve sacrificing a portion of the posterior nasoseptal flap.^{Reference Eloy, Patel, Shula, Choudhary and Liu12}

Disruption of the surgical barrier between a relatively septic nasal cavity and a highly aseptic arachnoid cavity is inherent in transnasal approaches to the anterior cranial fossa, as is the risk of iatrogenic meningitis. There will also be an inherent risk of olfactory dysfunctions in such surgery. In addition, as with all other surgical procedures, post-operative bleeding, especially in an area of compromised access, remains a possibility. Though rare, such complications are real with endoscopic nasal approaches to the skull base. The true incidence of these complications is unknown given the relatively smaller sample size of the relevant published studies to date. Like most of the reported studies, we observed very few cases of post-operative meningitis or nasal bleeding, even in those cases that had intra-operative bleeding from a cavernous injury (2 out of 100).

Kassam et al. reported post-operative bleeding in 1 out of 75 patients.^{Reference Kassam, Thomas, Carrau, Snyderman, Vescan and Prevedello4} Xuejian reported a single death due to post-operative meningitis out of 20 patients investigated.^{Reference Xuejian, Fan, Xiaobiao, Yong, Ye and Tao13} The latter study, however, did not specify the kind of mucosal vascular flap used in that particular patient. None of our patients had significant post-operative nasal synechiae, nasal obstruction or significant nasal crusting.

Bernal-Sprekelsen et al. conducted a retrospective study of 54 patients who underwent advanced skull base surgery (with large defects (greater than 20 mm)) and 62 patients with CSF leaks of a different origin (with small (2–10 mm) and mid-sized (11–20 mm) defects).^{Reference Bernal-Sprekelsen, Rioja, Enseñat, Enriquez, Viscovich and Agredo-Lemos14} Large defects were reconstructed with a nasoseptal pedicled flap positioned on fat and fascia lata. A lumbar drain was used. In those with small or mid-sized CSF leaks of another origin, intrathecal fluorescein was applied before the surgery to identify the defect; fascia lata in an underlay position was used for reconstruction, covered with mucoperiosteum of either the middle or inferior turbinate. The success rates after the first surgical reconstruction were 91 per cent for large skull base defects and 98 per cent for small or mid-sized skull base defects. After rescue surgery, the closure rate reached 100 per cent. The authors concluded that endoscopic surgery for closure of any type of skull base defect is the ‘gold standard’. Defect size does not seem to have a significant effect on success rate. Fascia lata and mucoperiosteum of the turbinate allow two-layer reconstruction of small and mid-sized defects. For larger skull base defects, a combination of fat, fascia lata and nasoseptal pedicle flaps enable successful reconstruction.^{Reference Bernal-Sprekelsen, Rioja, Enseñat, Enriquez, Viscovich and Agredo-Lemos14}

Some recently published reports have also highlighted the olfactory dysfunction associated with Hadad–Bassagasteguy flap reconstruction. Rotenberg et al. reported a decrease in mean University Of Pennsylvania Smell Identification Test scores, from 37.2 (normosmic) pre-operatively to 30.8 (hyposmic) post-operatively.^{Reference Rotenberg, Saunders and Duggal15} These decreased scores persisted even after complete healing at six months post-surgery. The authors hypothesised that olfactory impairment resulted from use of the Hadad–Bassagasteguy flap. They recommend that the possibility of permanent olfactory changes be included when providing routine patient counselling and obtaining consent for this procedure. They also advised that Hadad–Bassagasteguy flaps be raised judiciously during trans-sphenoidal endoscopic procedures.^{Reference Rotenberg, Saunders and Duggal15} We regret not including such evaluation in our study.

The limitations of the current study include the lack of evaluation of pre- and post-operative olfactory status. In addition, the study did not have a control group. Lastly, the study focused on a single institution only. Consequently, a multi-institutional study, and prospective, randomised controlled, double-blinded studies, would be ideal to validate our results.