Newman J, O’Malley BW, Chalian A, Brown MT. Microvascular Reconstruction of Cranial Base DefectsAn Evaluation of Complication and Survival Rates to Justify the Use of This Repair. Arch Otolaryngol Head Neck Surg. 2006;132(4):381-384. doi:10.1001/archotol.132.4.381
To describe our experience, including selection criteria and complication rates, using microvascular free tissue transfer to repair large skull base defects and to determine if complication rate and posttreatment survival justify the use of this technique.
Retrospective review of clinical cohort.
Tertiary care medical center.
A consecutive sample of patients undergoing ablative surgery with repair of cranial base defects with free tissue transfer from 1995 to 2004. All the patients' defects involved intracranial exposure.
Main Outcome Measures
Rate of local and systemic complications, postoperative survival, and recurrence rate.
The study population comprised 40 patients. Fifteen (38%) of the patients' defects were in the anterior cranial base, and 26 (65%) were in the middle cranial base. We used 5 types of free tissue flap, with a success rate of 95%. Our rate of perioperative mortality, meningitis, stroke, cerebrospinal fluid leak, epidural abscess, and osteomyelitis was 0%. We had 7 local complications and 4 systemic complications requiring increased length of hospital stay. Including microvascular problems, 12 patients had complications, for an overall complication rate of 30%. Follow-up ranged from 1 to 96 months, with a mean of 24 months. The tumor recurrence rate was 30%, and disease-specific survival was 81% at a mean 24-month follow-up.
We did not experience any perioperative mortality or intracranial morbidity. Our low complication rate in combination with our tumor recurrence rate and rate of patient survival justify the use of free tissue transfer as an option in the closure of appropriate cranial base defects.
In the 1950s and 1960s, Smith et al1 and Ketcham et al2 both described an intracranial and extracranial approach to tumors involving the paranasal sinuses. It was at this time when the 2-team approach to lesions at the skull base was born. Full-thickness skin grafts were initially their described method of defect closure. In the 1970s, pericranial,3 galeal, and galeopericranial flaps became popular as vascularized pedicled flaps in the closure of these defects. Pedicled myocutaneous flaps also gained popularity in the head and neck region, but because of the distance of most of these flaps from the skull base, they never became as popular for the closure of skull base defects.
Over the past decade, skull base surgical techniques have expanded the limits of resectability to include lesions that were previously considered unresectable. The skills of neurosurgeons and head and neck surgeons used in combination provides both intracranial and extracranial exposure, minimizes morbidity, and allows for more complete surgical resection.1 While small defects can often be adequately closed using local flaps (ie, pericranial and temporalis muscle) or distant pedicled flaps (ie, latissimus, trapezius, and pectoralis), these reconstructions often required multiple stages with prolonged healing. These prolonged periods can lead to delays in radiation therapy and substantial psychological stress to the patient and family.4
The evolution of ablative techniques has provided for more complete extirpation of these tumors. However, it has also resulted in larger and more complex surgical defects. Free tissue transfer provides reconstructive surgeons with a versatile tool for these situations. In fact, some argue that the safe, efficient, and reliable transfer of free flaps makes more extensive extirpation possible.5 The development of single-stage reconstructions, which include the option of using the pericranial, temporalis muscle, and free tissue flaps, substantially improves reconstructive efforts and decreases the associated morbidity of skull base procedures.3,4
Patient data were collected retrospectively. Inpatient and outpatient medical charts were reviewed, and patients were identified who fit our study criteria. Patients had to have a cranial base defect that involved exposure of intracranial contents, reconstruction had to involve microvascular technique, and adequate and accurate inpatient and outpatient follow-up must have been recorded. Patient data were recorded on a secure computer, with data stored under patient initials.
In this study, a cranial base defect refers to exposed intracranial contents to the skin, paranasal sinuses, nasopharynx, oropharynx, or oral cavity. These defects resulted from resections of neoplasms that involved the anterior or lateral skull base. Not all patients—particularly those with middle skull base defects—required a combined intracranial/extracranial approach. All reconstructive efforts in this cohort sought to (1) create a clear division between intracranial and extracranial cavities, (2) obliterate dead space, and (3) optimize cosmesis by reestablishing facial contour.
The anterior cranial fossa (ACF) supports the frontal lobes of the brain intracranially and comprises the roof of both orbits. Most tumors involving the ACF arise within the nasal and ethmoid cavities and extend through the anterior skull base (ie, squamous cell carcinoma, melanoma, and sinonasal undifferentiated carcinoma). Occasionally, primary intracranial tumors such as meningiomas breech the anterior skull base to involve the superior ethmoids or orbit. Esthesioneuroblastoma arises at the skull base proper, the cribriform, and ethmoid roof. In our reconstruction of these bony skull base defects we aimed to (1) seal the intracranial space from the exterior and upper aerodigestive tract, (2) provide support for the brain, (3) obliterate often geometrically complex dead space, and (4) create a cosmetically acceptable result. Pericranial and temporalis flaps3,4,6 are the workhorses for smaller basicranial defects, often used alone or in combination with free bone grafts, alloplastic materials, or skin grafts. Larger cranial base defects or defects in radiated fields may need the bulk and vascularity of microvascular transfers for successful closure.
The middle cranial fossa (MCF) provides a bony division between the temporal lobes of the brain and the infratemporal fossa. Critical anatomic structures found within or bordering the MCF are the optic canal and superior orbital fissure anteriorly, the cavernous sinus and trigeminal ganglion medially, and the temporal bone posteriorly.
Intracranial primary tumors of the MCF, such as sphenoid wing meningiomas, are uncommon. More often seen are metastases and tumors that directly extend into the MCF from other primary sites. Primary tumors of the paranasal sinuses, palate, tonsil, and lateral oropharyngeal wall (ie, squamous cell and adenoid cystic carcinomas) often involve the infratemporal fossa because the pterygoid musculature provides little barrier to tumor spread. The tumors then track along the extracranial branches of the trigeminal nerve to enter the MCF through the foramen rotundum (V2) or foramen ovale (V3).
Removal of infratemporal fossa tumor exposes intracranial contents or skull base neurovascular structures to mucosal or external contamination. Similarly, lateral pharyngeal wall defects may expose the carotid canal. These defects require reliable mucosal or cutaneous coverage to avoid cross-contamination. Parotid lesions requiring total resection of the gland and infratemporal soft tissues (with or without mandibulectomy) often expose the great vessels and nerves. Covering this area with muscle or myocutaneous tissue provides a soft tissue buffer over vital structures for subsequent therapies.7 An added requirement for the successful reconstruction of defects involving the oral and nasal cavities is the successful reintroduction of speech and swallowing function.
Lateral temporal defects can be challenging to reconstruct. Large cutaneous defects combined with exposure of temporal lobe dura and possibly cerebrospinal fluid (CSF) require addressing the intracranial as well as cutaneous components of the problem. Typically, radiation therapy is planned postoperatively, and soft tissue coverage of the exposed cranium is necessary to decrease wound complications. Obturation of the contour defect may improve postoperative cosmesis as well.
For most anterior cranial base defects, we continue to use pedicled flaps (pericranial, temporalis) in the repair. However, in selected cases, we have found free tissue transfer to be more suitable. These include anterior defects larger than 25 cm2 and defects that involve significant exposure of the paranasal sinuses in conjunction with the orbit. We also use free tissue transfer for cases in which previous surgery has precluded the use of pericranial or temporalis flaps. Although in some cases it is technically feasible to harvest a large pericranial flap, there is emerging evidence based on blood supply and volume studies of pericranial flaps that defects greater than 40 cm3 in volume exceed the reliable vascular supply.8 Thus, for defects of this size or greater involving the ACF, infratemporal fossa, maxilla, and lateral skull base, we use myocutaneous and fasciocutaneous free flaps to provide tissue bulk, protect vital structures, and decrease postoperative complications. If there is bony mandibular discontinuity, fibula flaps are useful to ultimately reconstruct masticatory function as well. In patients who have received a full course of radiation (60-70 Gy [6000-7000 rad]) to the anterior skull base, we have been more aggressive with the use of microvascular repair, largely because of the concern of poor vascularity of pedicled flaps taken from the radiated field. Several other authors corroborate this finding.9,10
Between August 1995 and June 2004 there were 40 skull base resections reconstructed with free tissue transfer. The patients included 32 male and 8 female patients with an age range of 17 to 84 years (median age, 63 years). There were 15 anterior cranial base defects (38%) and 26 middle cranial base and infratemporal fossa defects (65%) (one of the cases was combined anterior and middle cranial base defects).
Flap selection was as follows: 24 rectus free flaps, 9 radial forearm free flaps, 2 serratus anterior free flaps, 3 anterolateral thigh free flaps, and 2 latissimus dorsi free flaps. One patient required a vein graft for access to donor vessels in the neck.
There were 14 different underlying pathologic tumor types among the 40 cases. Squamous cell carcinoma was present in 17 cases, basal cell carcinoma in 6 cases, and adenoid cystic carcinoma in 4 cases. The rest of the tumor types were uncommon and are listed in Table 1.
The flap success rate was 95% (38/40). We experienced 1 case of arterial insufficiency and 1 case of venous congestion. We experienced 7 local wound complications and 4 systemic complications in 10 patients (Table 2). We had no cases of CSF leak, meningitis, osteomyelitis, or stroke. In total, 12 patients experienced complications (including flap, local, and systemic), for an overall complication rate of 30%.
Follow-up ranged from 1 to 96 months (median, 24 months). Of the 37 patients with a malignant neoplasm, 30 (81%) are alive and 26 (70%) are alive without evidence of disease. Eleven patients (30%) have been diagnosed as having recurrence (4 are alive with disease and 7 died of cancer). Three patients died of causes unrelated to their cancer or surgical procedure, with no evidence of recurrence. The 3 patients with benign disease are considered cured. In the group of patients with recurrence of the cancer, 5 were diagnosed as having metastatic disease and 6 as having locoregional recurrence.
Vascularized tissue reconstruction of cranial base defects is critical for separation of intracranial contents and vital neurovascular structures from external or mucosal contamination. Prior to the introduction of free tissue transfer, defects created by the removal of large or complex cranial base tumors were repaired using a combination of pedicled flaps and skin grafts. In many circumstances, these flaps did not provide adequate support, vascularity, or bulk to reconstruct the defects. Complication rates in these settings were high and included high incidences of CSF leak, meningitis, skull base osteomyelitis, and stroke. The use of microvascular tissue transfer in the reconstruction of these defects has played a major role in the advancement of the field of cranial base surgery. It has allowed us to increase the limits of resectability without compromising the rate of complications. In our study, we found that the use of free tissue transfer allowed for the closure of large and often complex defects. The rate of local and systemic complications was low.
Regardless of the method used, effective reconstruction reduces the incidence of postoperative complications including CSF leak, meningitis, tension pneumocephalus, and great vessel exposure.11,12 Bony defects can often be reconstructed with titanium mesh or split calvarial bone grafts. The most convenient vascularized tissues available for regional tissue transfer are pericranium3 and temporalis muscle.13 Under certain circumstances, soft tissue and bony defects are not adequately addressed by the use of bone grafts and regional tissue flaps alone. In the case of large craniofacial and basicranial defects, free tissue transfer alone or in combination with regional techniques provides reliable closure and coverage of vital structures.4 In patients who have undergone radiation or in whom postresection radiation may interfere with regional flap survival, the free tissue transfer technique is appropriate.
Protection of infratemporal fossa defects from external exposure is accomplished with a variety of flaps. For tonsil and lateral pharyngeal wall defects extending into the infratemporal fossa exposing skull base neurovascular structures, we have found that the radial forearm flap provides an excellent, easily inset, functional reconstruction14 and protection from salivary contamination. In these defects, a pedicled flap may ultimately dehisce, at least partially, resulting in potential treatment delay due to prolonged healing. For larger defects, including orbital exenterations or large lateral skull base defects with significant soft tissue involvement, we have used rectus, latissimus, and more recently anterolateral thigh flaps to reconstruct these areas.
Success rates of free tissue transfer for cranial base defects ranges from 86% to 100% in other series.12,15- 17 These free flap survival rates compare very favorably to pedicled flaps used for similar indications.18- 20 Many of the patients treated for these tumors have had prior surgery, prior radiation, or both. This, of course, has the potential to increase both the difficulty of the surgery as well as the rate of complications. In addition, the location of many of the defects can make microvascular anastomosis more challenging compared with similarly sized defects in the larynx and pharynx. They tend to be more remote from appropriate vasculature. We used a combination of neck vessels as well as superficial temporal vessels to reconstruct the defects. On only 1 occasion did we require the use of an interposition vein graft to lengthen the pedicle.
Rectus abdominis21 and latissimus22 flaps were initially described as the flap of choice for skull base reconstruction. Radial forearm, omental, scapular, anterolateral thigh, and iliac crest flaps also have utility. The ability to completely seal CSF intracranially and cover the carotid artery with muscle makes the rectus very versatile. Its long pedicle rarely requires a vein graft, the use of which increases the risk of thrombosis at the microvascular anastomosis. We have found the anterolateral thigh flap to also be an excellent choice for closure of large defects involving skin in the lateral cranial base.
There have been several classification schemes designed to evaluate complication rates associated with skull base surgery.15 None are accepted universally. In a review of the literature, the rate of complications in the repair of skull base defects has been reported from 16% to 64%.15,17,23,24 However, the lack of uniform reporting criteria leads to a lack of uniform method of reporting results. In much of the literature, local and systemic complications are not distinguished. The defect repaired is also not reported in a standardized fashion. In other articles, the method of repair is not specified. The lack of uniformity in the reporting of complications makes comparisons among studies difficult; however, complication rates have been shown to be similar between local and free tissue flaps used to repair cranial base defects.14,18 One can argue that the tendency to close smaller and simpler defects with local flaps and larger and more complex flaps with free tissue transfer unfairly compares the 2 techniques.
The treatment of patients with cranial base tumors poses a challenge in many respects. Anatomically, this area is very difficult to treat, and often these patients have been treated several times for their tumors prior to referral for cranial base surgery. Studies have shown high rates of prior treatment in patients being seen for malignant skull base lesions. One would expect this to have negative consequences on both complication rates and survival. Although determining cause-specific survival rates in a population with such a diverse pathologic presentation would require larger numbers than those available in our study, we had an 81% rate of survival at a mean of 24 months of follow-up in patients with malignant tumors. We believed that this at least justifies the undertaking of the complex surgery and repair.
In conclusion, our flap survival rate (95%) and overall complication rate (30%) argue strongly for the routine use of free tissue transfer in cranial base surgery. We successfully closed a variety of skull base defects that included the anterior and lateral cranial fossae and the infratemporal fossa. The flaps have provided reliable segregation of intracranial and extracranial compartments, competent separation of mucosal surfaces from vital structures, obliteration of geometrically complex dead space, and acceptable cosmesis. We believe that in any skull base defect too large or complex for local flap closure, free tissue transfer is indicated. Also, in many previously radiated areas, or in areas in which postsurgical radiation may compromise the function of local or regional flaps, free tissue transfer may reduce the rate of complications. These techniques offer a reliable source of well-vascularized tissue with an acceptable rate of complications and morbidity.
Correspondence: Jason Newman, MD, Department of Otorhinolaryngology, Hospital of the University of Pennsylvania, 3400 Spruce St, Ravdin 5, Philadelphia, PA 19104 (firstname.lastname@example.org).
Submitted for Publication: June 15, 2005; final revision received October 12, 2005; accepted November 2, 2005.
Financial Disclosure: None.
Previous Presentation: Some of these data were presented at the North American Skull Base Society Meeting; April 7, 2005; Toronto, Ontario.