Free tissue transfer has been the gold standard of extensive skull base reconstruction, but the onlay of free flaps onto skull base defects carries the risk of cerebrospinal fluid (CSF) leakage. The purpose of this study was the evaluation of a novel technique of a combined sub- and onlay concept with a partially intracranially positioned folded free fasciocutaneous flap in terms of flap applicability, versatility and complication rate. Within 5 years, 7 patients with anterior ( n = 4), middle ( n = 2) or posterior ( n = 1) skull base defects were reconstructed with free extended lateral arm ( n = 3) or anterolateral thigh ( n = 4) flaps. The flaps were partially intracranially positioned and fixed with osteo-dermal sutures. Both flaps proved to be applicable in terms of sealing efficiency, minimizing intracranial flap volume and folding. No flap loss was observed. Specific complications consisted of one pneumocranium via an accessory frontal sinus and one cerebellar herniation due to lumbar CSF loss. No flap failure or haematoma of the intracranial flap part occurred. This new concept of intracranial positioning of fasciocutaneous flaps in a sandwich technique using osteo-dermal sutures should be considered as a primary treatment for skull base reconstruction rather than merely as a salvage manoeuvre.
Surgical treatment of skull base neoplasm often requires extensive resection, which results in large 3-dimensional (3D) tissue defects exposing bone, dura and intradural contents. Skull base reconstruction is challenging, due to anatomic location, functional requirements and potentially life-threatening complications, such as cerebrospinal fluid (CSF) leakage, secondary meningitis, or osteomyelitis, which may arise in case of reconstructive failure . Therapeutic protocol calls for a watertight, safe and functional reconstruction. Primary pericranial, galeal and galeal-myofascial scalping, forehead, glabellar and temporalis have been recommended for local flap options. Distant pedicled muscle, myocutaneous, or fasciocutaneous flaps, such as latissimus dorsi, trapezius, and pectoralis flaps, have been proposed for regional flaps . The versatility of local or regional flaps is limited in restoring extensive defects. In patients who have received high dose radiotherapy or have had previous surgery, these flaps may not provide a watertight seal . Wound edge necrosis in the most distal part of pedicled flaps is not uncommon, and may lead to CSF leakage. The most significant improvement has been the introduction of vascularized free tissue transfer . The transfer of highly vascularized tissue optimizes primary wound healing and thus can also be used in irradiated sites or in patients who will receive postoperative radiotherapy. In addition, virtually any dead space volume can be filled, optimizing the functional and aesthetic outcome. Complications, such as failure of primary wound healing, flap loss, CSF leakage, and secondary complications are significantly lower in free tissue transfer compared with regional flaps . Over the past years, free tissue transfer has become the treatment of choice in extensive skull base reconstruction .
Abdominal muscle and myocutaneous flaps have been the workhorses in skull base reconstruction, even though their harvest often results in high donor site morbidity, such as abdominal wall herniation . Sealing the skull base defect using a flap onlay-technique carries the risk of CSF leakage. Intracranial positioning of the flap in a sublay technique calls for a thin, unbulky flap with a large surface area. With the aim of improving the safety of onlay free flap reconstruction of skull base defects, the authors created a new technique by combining the sub- and onlay flap concept by folding a flap of low volume and large surface area. Fasciocutaneous flaps suit these parameters best. They are thin and pliable and remain of constant volume. Extended lateral arm flaps (ELAF) and anterior lateral thigh (ALT) flaps provide the best versatility .
The aim of this prospective study was to report on a consecutive series of procedures to evaluate the use of free fasciocutaneous flaps in a novel sandwich technique for skull base reconstruction, in terms of flap applicability and versatility and complication rate over a 5-year period.
Patients and methods
A prospective study was carried out on 7 consecutive patients requiring skull base reconstruction with free fasciocutaneous ELAF or ALT-flap transfer following trauma or tumour extirpation between December 2002 and January 2008. An intracranial application of free ELAF and ALT-flaps was performed on all patients. Skull base defect sites in all anatomic regions (I–III) were considered. Further inclusion criteria were extensive defects not treatable with local or regional flaps, a history of previous radiotherapy or surgery at the defect site, persistent CSF leak, osteomyelitis, abscess formation, or bone necrosis.
Skull base surgery was performed interdisciplinarily with neuro- and craniomaxillofacial-surgeons as well as plastic and reconstructive surgeons. The preoperative planning phase consisted of a determination of surgical approach, estimation of resection size, exposure of anatomic structures, flap selection, skin island and pedicle positioning, recipient vessels and site of anastomoses, requirement of dead space volume and implant positioning. Preoperatively, all patients underwent an MRI and CT scan.
Either the neurosurgeons or the craniomaxillofacial surgeons carried out the tumour ablation. The plastic and craniomaxillofacial surgeons proceeded with the defect reconstruction as soon as the immediate histological section of the resected tumour margins showed no carcinogenic tissue. Following tumour ablation, dural tack-up sutures that would be tied later in the procedure were positioned. The dural defect was repaired if necessary with a neuropatch. A titanium 3D mesh (0.3, 04. or 0.6 mm thick) (Synthes, Oberdorf, Switzerland) was positioned to restore the 3D structure of the cranial base. No bony reconstructions were performed. The decision whether to use an ELAF or an ALT-flap to cover the skull base defect and the reconstructed dura was made based on a predetermined template. In the case of a high flap surface-to-volume ratio, an ELAF was applied . If a large dead space had to be filled, an ALT-flap was harvested. The preoperatively located vessels for vascular anastomoses were dissected. After flap harvest, a partial de-epithelialization was performed on the flap. For the flap’s vascular supply, microsurgical anastomoses were made to the temporal and facial vessels.
The flap was positioned onto the neuropatch using a sandwich technique ( Figs 1 and 2 ). Watertight sealing was achieved through osteo-dermal sutures. These sutures were placed through the drilled holes in the surrounding skull bone and through the de-epithelialized part of the flap dermis. Before tightening the sutures, the flap was pulled down to the desired position in the defect site using the osteo-dermal threads as guide wires. After appropriate positioning of the flap, all the sutures were tied up. Care was taken to preserve vascularization and to provide closure at the folding site of the flap. All the donor sites were closed primarily. A lumbar CSF drainage was used for 2–5 days.
After surgery, the patient was transferred to the Intensive Care Unit (ICU) for approximately 24 h, then to the Intermediate Care Unit (IMU) for another 3 days. Postoperative flap monitoring depended on whether the flap was buried or exposed. In case of exposure, the flap was clinically monitored. Otherwise, Doppler sonography was applied. A frequent rhinoscopy (colour and capillary refill monitoring) was additionally carried out in intranasally located flaps. If there was any doubt about the free flap status, the patient was brought back to the operating theatre for revision without further delay.
Between December 2002 and January 2008, 7 patients with a mean age of 62 years (range, 35–75 years) underwent skull base reconstruction with either ELAF ( n = 3) or ALT ( n = 4) in a sandwich technique following extensive resection of tumours ( n = 5) involving region I ( n = 4) and region III ( n = 1) of the cranial base or revision following traumatic injury involving region II ( n = 2) ( Table 1 ). Four patients had undergone previous skull base surgery. The tumour types included hemangiopericytoma ( n = 1), meningeoma ( n = 1), squamous cell carcinoma ( n = 2), and melanoma ( n = 1). The patient suffering from hemangiopericytoma ( n = 1) had received preoperative radiotherapy and presented with intracranial abscess formation after tumour extirpation. Multiple revisions, including defect reconstruction with a local temporal flap, were carried out to treat this patient. The patient with meningeoma ( n = 1) presented with recurrent disease. The patients suffering from squamous cell carcinoma and meningeoma underwent free tissue transfer based on the extensive defect dimensions. In the patient suffering from melanoma, resections via two rhinotomies failed to treat the patient curatively. Chronic osteomyelitis was diagnosed in two patients following traumatic injuries. These patients underwent multiple revisions, including a galeal flap for defect covering.
|Pt||Age||Sex||Cause||Localization||Resection extent||Flap type||Suppl material||RT||Complications|
|1||35||M||SCC||Interorbital region, orbital roof and wall, ethmoid sinuses, dura||Medial orbital wall and roof, ethmoid sinuses, anterior skull base, dura||ELAF||Titan-Mesh||Postop||None|
|2||75||M||OM after trauma||Frontal skull base (region II)||FB, frontal skull base up to nasal base and cavity, ethmoidal sinuses, dura||ALT||None||None||None|
|3||44||F||HPC||Temporal and posterior skull base||TB and posterior skull base, dura||ALT||None||Preop||None|
|4||73||M||SCC||Left orbital floor and sinus maxillaris, nasal septum,||Left hemimaxillectomy, nasal septum resection and left orbital exenteration, dura, neck dissection I-III||ALT||Mandibula-plate||Postop||None|
|5||66||M||Recurrent melanoma||Nasal cavity involving the right maxillary, orbit and frontal sinus up to the anterior skull base||Partial maxillectomy, partial orbitectomy, anterior skull base, dura||ALT||Titan-Mesh||Postop||Cerebellar herniation|
|6||65||M||OM after trauma||Anterior skull base, sphenoidal and ethmoidal sinuses||FB, sphenoidal and ethmoidal sinuses, dura||ELAF||None||None||None|
|7||75||M||Meningeoma||Infraorbital fossa||TB, zygomatic arch, lateral orbital wall, pterygoid process, dura||ELAF||Titan-Mesh||None||Pneumocephalus|
Duration of surgery, from induction of general anaesthesia to the end of surgery, averaged 14.5 h (7.5–22 h). Dural reconstruction using a neuropatch was necessary in all patients.
On 3 patients, an ELAF was harvested to restore the tissue defect ( Table 2 ). ALT-flaps were used for reconstruction in 4 patients. All the arterial anastomoses were done end to end, in 4 patients to the temporal artery and in 3 patients to the facial artery. Venous anastomosis was carried out in 6 cases as end-to-end, and in one patient as end-to-side with the external jugular vein. In 3 patients, a 0.6 mm titanium mesh was used to improve structural outcome. All the implants and bone transplants were covered. Tailored skin islands provided internal lining of the nasal cavity and nasopharynx. All the defects had a good volume filling and all donor sites were closed primarily.
|Pt||Flap type||Flap size [cm]||Perforator type||Recipient artery||Recipient vein(s)|
|1||ELAF||16 × 5(10)||Septocutaneous||STA (E-E)||TV (E-E)|
|2||ALT||20 × 11||Musculocutaneous||STA (E-E)||TV (E-E)|
|3||ALT||12 × 8||Musculocutaneous||FA (E-E)||FV (E-E)|
|4||ALT||22 × 10||Musculocutaneous||STA (E-E)||EJV (E-S)|
|5||ALT||8 × 15(20)||Septocutaneous||FA (E-E)||FV (E-E)|
|6||ELAF||17 × 5||Septocutaneous||STA (E-E)||TV (E-E)|
|7||ELAF||15 × 6||Septocutaneous||FA (E-E)||EJV (E-E)|