Microvascular reconstruction of the skull base: Indications and procedures

J Oral

Maxlllofac

Surg

57:233-239,

1999

Microvascular Reconstruction of the Skull Base: Indications and Procedures Henning Schliephake, MD, DDS, PbD, * Rainer Schmelzeisen, MD, DDS, PbD, t Madjid Samii, MD, PbD,f and WolfPeter Sollmann, MD, DDS, PbDJ The aim of the current study was to review the use of free tissue transfer for reconstruction Purpose: the skull base and for coverage of intracranial contents.

of

Patients and Methods: From 1990 until 1996, revascularized flaps were transferred to the skull and the skull base in 11 patients in whom intracranial/extracranial resection of tumors of the skull base was performed in cooperation with the Department of Neurosurgery. The defects resulted from removal of squamous cell carcinomas (n = 4) basal cell carcinomas (n = 4) malignant melanoma, malignant schwannoma, and malignant meningioma. Defect repair was accomplished by revascularized transfer of latissimus dorsi muscle flaps in seven cases and rectus abdominis flaps and forearm flaps in two cases each. In five patients with extensive intracranial tumor spread, reconstruction was performed for palliative reasons.

A safe soft tissue closure of the intracranial and intradural space was achieved in all patients, Results: whereas the contour of the facial skull and the neurocranium was satisfactorily restored at the same time. By using the entire length of the grafted muscle, the vascular pedicle could be positioned next to the external carotid artery and conveniently connected to the cervical vessels. The mean survival time of the patients with palliative treatment was 8.4 months, with an average duration of hospital stay of 24.5 days. Despite the increased surgical effort of revascularized tissue transfer, microvascular of large skull base defects appears to be justified, even as a palliative measure.

Conclusions:

reconstruction

Reconstruction of the skull base and the facial skeleton in patients with intracranially/extracranially growing malignancies constitutes an interdisciplinary challenge. Single-stage removal of intracranial tumors with extension into the facial soft tissues, or large malignancies of the craniofacial area with intradural invasion, require primary repair together with the resection in a joint effort for aesthetic and functional reasons. Extensive resections in the area of the anterior skull base are associated with cerebrospinal fluid (CSF) leakage and exposure of the intradural contents to the bacterialaiden paranasal sinuses and the nasal cavity. Addition*Associate Professor, Department

of Oral and Maxihofacial

gery, Hannover Medical School, Medizinische

Hochschule

ally, ablation in this area commonly results in gross disfigurement of the patient. For such reasons, reconstruction in these cases should effectively protect the dural space against bacterial invasion, seal off CSF leakage, and, at the same time, restore facial aesthetics as much as possible. Resection of the posterior area of the skull creates large composite tissue defects, which necessitate immediate and safe closure to provide reliable coverage of the dura or the intradural contents and to protect the brain from direct mechanical damage. A variety of measures have been described for reconstruction of the skull base, ranging from local galeo-pericranial flaps to free tissue transfer, depending on the defect size. l-3 However, because large composite defects resulting from intracranial/extracranial ablation usually involve the facial skin and bones, and the skull base, repair of these defects is commonly beyond the capacity of local flaps and requires free tissue transfer to accomplish safe closure. Because most of the patients with intraduraVextradura1 growing tumors present with advanced tumor stages, the question may arise whether the performance of extensive microvascular procedures in such patients is justified with regard to the limited long-term prognosis. Therefore, the aim of the current study was to review the indications and the results of microvascular reconstructions of the skull base and the facial skull

Sur-

Hanno-

ver, Hannover, Germany. tprofessor,

Department of Oral and MaxiUofacial Surgery, Albert-

Ludwig+University, *Professor,

Freiburg, Germany.

Department

of Neurosurgery,

Hannover

Medical

School, Medizinische Hochschule Hannover, Hannover, Germany. SProfessor, Department Braunschweig,

of Neurosurgery,

Braunschweig,

Address correspondence Department

and reprint requests to Dr Schliephake:

of Oral and Maxillofacial

School, Medizinische

Stadtisches Klinikum

Germany.

Hochschule

Surgery, Hannover

Hannover,

Medical

Carl Neuberg Str 1,

30625 Hannover, Germany. o 1999

American

Association

027%2391/99/5703-0003$3

of Oral

and Maxillofacial

Surgeons

00/O

233

234

Patient No. 1 2 3 4 2 7 8 9 10 11

MICROSURGICAL SKULL BASE RECONSTRUCTION

As

50 53 E 34 57 77 45 20 58 55

Gender

Diagnosis

Resection

Graft

Male Male Male Male Male Male Male Male Male Female Male

Rec. SSC BCC BCC ssc MM BCC ssc Rec. SSC Malignant NF BCC Astrocytoma

S, B S, B S, B, D S, B S, B, D S, B, D S, B, 03 S, B S, f-4D S, B S, B, D

Latissimus dorsi Latissimus dorsi Latissimus dorsi Latissimus dorsi Latissimus dorsi Latissimus dorsi Latissimus dorsi Rectus myocut. Rectus muscle Forearm flap Forearm flap

Abbreviations: Ret, recurrent; SSC,squamous cell carcinoma; BCC, basal cell carcinoma; MM, malignant melanoma; NF, neurofibroma; S, skin; B, bone; D, dura.

in large intracranially/extracranially growing tumors with regard to curative and palliative treatment strategies.

Patients

and Methods

From 1990 until 1996, revascularized flaps were transferred to the skull and the skull base in 11 patients with intracranial/extracranial lesions extending into the maxillofacial area (Table 1). Most of the lesions were of epithelial’ origin, such as squamous cell carcinomas of the parietal skin (n = 2) and the maxilla (n = 2), basal cell carcinomas (n = 4), and one melanoma that had penetrated the adjacent skull bones. In five of these patients; a full-thickness resection of the dura was performed because of tumor invasion, and in one patient only the outer layer of the dura was removed. In the remaining two patients, the skull bones were resected without opening the dural space. In two patients, intracranial tumors (malignant neurofibroma and anaplastic astrocytoma) had invaded the extracranial portions of the facial skull from intracranially. All ablative procedures were performed in cooperation with the Department of Neurosurgery. Location and extent of the defects are shown in Figure 1. Because of the extensive intracranial spread of the tumor, surgical resection had failed to gain tumor-free margins in five patients, and hence ablation and subsequent reconstruction were performed as palliative measures of tumor reduction. Two patients are presented in greater detail to elucidate the surgical management of ablation and reconstruction.

Report

of Cases

Case 1 A 64-year-old man presented with a large, recurrent basal cell carcinoma of the parietal and frontal skin. He had a

complete facial paralysis on the right side caused by cutting of the facial nerve during previous ablative surgery. The magnetic resonance imaging (MRI) scans showed destruction of the right frontal bone and the orbital roof and suspected invasion of the dura, with a diameter of approximately 4 cm (Fig 2A). In a first step, the parotid gland, the infratemporal fossa, and the submandibular triangle were dissected, with ablation of an angulated portion of the parietal and forehead skin measuring 17 X 15 cm. Then,

Lat. dorsi (pat.1)

Lat. dorsi (pat.2)

Lat. dorsi (pat.3)

Lat. dorsi (pat.4)

Lat. dorsi (pat.5)

Lat. dorsi (pat.6)

Lat. dorsi (pat.7)

Radial forearm (pat.10) FIGURE 1. location ihe skull base.

Recks

abd. (pat.8)

Rectus abd. (pat.9)

Radial forearm + PMMA (pat.1 1) and extent of composite

defects of ihe skull and

SCHLIEPHAKE

ET AL

235

FIGURE 2. A, Preoperative MRI scan showing invasion and penetration of the right frontal bone and orbital roof. 6, Resection of the parietal skin and dissection of the preauricular and submandibular area. C, ResectIon of the frontal bone and dura. D, Myocutaneous latissimus dorsi graft with skin portion shaped according to the defect. E, Reconstruction completed with satisfactory contour restoration.

236 together with the neurosurgical department, a portion of the underlying bone approximately 7 cm in diameter was resected (Fig 2B), and the dura was removed (Fig 2C) and replaced by a fascia lata graft. A latissimus dorsi myocutaneous flap was elevated with the cutaneous portion corresponding to the shape and size of the angulated skin defect. Harvesting of the complete length of the thoracodorsal vessels and the position of the skin area on the muscle yielded a length of 13 cm from the edge of the skin portion that enabled the upper cervical vessels to be reached (Fig 2D). The total length of the pedicle allowed for end-to-end anastomoses with the external carotid artery and the facial vein. To avoid compression of the vascular pedicle, the zygomatic arch and parts of the temporalis muscle were resected. Healing was uneventful, with coverage of the dural defect and adequate restoration of the contour of the parietal bone and the forehead (Fig 2E). The patient has been free of recurrence until now.

Case 2 A 16year-old boy presented with facial asymmetry resulting from infraposition of the right orbit and globe and involvement of the skin in the temporal area. The computed tomography (CT) scans showed osseous destruction of the lateral orbital wall and the middle cranial fossa on the right side (Fig 3A). A biopsy specimen taken from the infratemporal fossa led to a diagnosis of malignant neurofibroma. The tumor was approached from a preauricular and coronal incision, with resection of a small portion of temporal skin, exposure of the facial nerve, osteotomy of the zygomatic arch, and dissection of the temporalis muscle (Fig 3B). The tumor was removed after craniotomy and resection of a portion of temporal bone (Fig SC). However, unexpectedly, further exploration indicated that the tumor extended far beyond the surgical margins between the galea and skull bone surface, which had not been apparent in the preoperative CT and MRI scans but was confirmed by frozen sections during the operation. Because of the extent of the lesion, the curative intention was abandoned in favor of palliative tumor reduction. Nevertheless, a revascularized rectus abdominis muscle flap was used for reconstruction to cover the dura and fill the dead space in the infratemporal and (temporal) fossa. Skin coverage was obtained by local rotation flaps (Fig SD). The external carotid artery and the facial vein served as recipient vessels. Healing was uneventful, and postoperative radiation with a total dose of 60 Gy was initiated.

Results In all patients, satisfactory soft tissue coverage of the defect was accomplished. No CSF leaks and no cases of meningitis were encountered. Immediate defect repair was accomplished by revascularized transfer of myocutaneous latissimus dorsi muscle flaps in seven cases and rectus abdominis muscle flaps and fasciocutaneous forearm flaps in two cases each. In one patient, an alloplastic restoration (methacrylate) of parts of the frontal and the parietal bone was performed, because the bone defect was considered to be too large and the soft tissue too thin to protect the brain safely from mechanical damage. No other attempt at osseous reconstruction of the skull bones was made, because the bulk of soft tissue available

MICROSURGICAL SKULL BASE RECONSTRUCTION

was adequate for both the contour restoration and brain protection. The mean size of the skin portion used for surface coverage was 14.3 X 12.6 cm. The myocutaneous flaps were designed in such a way that the skin portion was positioned over the muscle opposite the vascular pedicle, so that the length of the whole muscle could be added to the length of the vascular pedicle. This allowed for the use of the thyroid artery and related veins as recipient vessels for anastomoses of the flap vessels despite the distant locations of some of the defects in relation to eligible vessels in the head and neck area. In one patient, a partial necrosis of superficial layers of the skin portion occurred without jeopardizing the dural coverage or defect closure. No donor site complications were observed in the flaps harvested from the shoulder, the abdominal wall, and the forearm. The mean operation time was 9.1 hours; patients were hospitalized for a mean period of 24.5 days. At the time of this study, 5 of the 11 patients were alive and free of recurrence for a mean period of 3.3 years (9 months to 7.5 years). The remaining six patients had died of their disease (distant metastasis, general cachexia, and secondary complications such as pneumonia) after a mean survival time of 8.4 months (5 to 10 months).

Discussion Pericranial flaps constitute a standard measure for reconstruction of the skull base after craniofacial resections in combination with avascular grafts of fascia lata, fat, or split-thickness ~kin.l~~ Although the use of additional avascular grafts appears to be controversial, pericranial, galeo-pericranial, and galeal scalp flaps have proved highly reliable and well suited for reconstruction of the anterior cranial base.5 Defects of the orbit and the cribriform plate also may be covered with this flap if it is based laterally on the temporalis muscle,6 and inclusion of the temporalis muscle into the flap is reported to allow for closure of defects in the middle cranial fossa.7,s However, a recent study on blood supply and volume of galeo-pericranial flaps has shown that it is impossible to delineate the extent of reliable blood supply to these flaps because of the wide range in length of their axial vessels. The authors concluded that the well-vascularized proximal portions of these flaps may adequately serve reconstructive needs, but in instances in which large composite tissue defects and dead space are likely to occur, free tissue transfer should be considered to supply a reliably vascularized and bulky soft tissue flap.9 The extent of the defects in the current study required repair procedures that are supposed to be located on the upper end of the size-dependent

SCHLIEPHAKE ET AL

FIGURE 3.

237

A, Three-dimensional display of reformatted CT scans showing the destruction of the posterror orbital wall and the temporal bone. B, Preauricular/coronal approach with resection of the zygomatic arch and elevation of the temporalis muscle. C, Tumor protruding across the temporal bone. Bur holes for craniotomy have already been placed. D, Postoperative view after primary wound closure with local rotational flaps.

238 cascade of reconstructive strategieslo The use of pedicled myocutaneous flaps such as the pectoralis major island flap or the rhombotrapezius flap has been advocated for the coverage of such large composite defects.2 However, the advantage of easier dissection and tissue transfer may be counteracted by the fact that the length of the pedicle is limited, which precludes the use of these flaps in reconstructions above the level of the orbit. A number of articles dealing with closure of large defects resulting from resections of malignomas with intracranial extension therefore indicate a preference for free tissue transfer of muscle flaps, in particular the rectus abdominis muscle10-*2and the latissimus dorsi flap.13J4 The latissimus dorsi muscle has the advantage of providing a long vascular pedicle, which allows for anastomosis of the flap to the upper cervical vessels. However, the disadvantage of this flap is the need to turn the patient for the harvesting procedure, which is time-consuming and may not be possible in certain neurosurgical situations where the head is preferrably fixed. For these reasons, the rectus abdominis muscle is given preference by some authors,101z because flap elevation is possible simultaneously to the craniofacial procedure in a two-team approach.15 The vascular pedicle of the rectus muscle flap is commonly somewhat shorter than that of the latissimus,lG but use of the whole length of the muscle and positioning of the skin island opposite the pedicle usually allows for anastomosis to the upper cervical vessels, even in peripheral defect locations. Both flaps supply large bulks of tissue, rendering them particularly useful for the protection of the intracranial contents if the osseous component of the defect is not restored by bone grafts or alloplastic material. In cases of shallow but extensive superficial defects of the skull, reconstruction has been accomplished by a free omentum graft with split-thickness skin coverage.” In the current study, those patients with shallow defects and less need for volume after alloplastic reconstruction of the skull bones (cases 10 and 11) have received forearm flaps for soft tissue closure. These flaps have provided a more convenient solution than the revascularized transfer of greater omenturn, because they have a longer pedicle and are easy to harvest in a two-team approach. The possibility of successful free revascularized tissue transfer for reconstruction of the skull base has also extended the strategies of surgical management of intracranially/extracranially growing malignancies because of the reduction of severe postoperative complications such as meningitis and cerebral abscess formation. However, the ability to perform more aggressive treatment is also associated with questions of increased duration of surgery, prolonged periods of

MICROSURGICAL

SKULL BASE RFXONSTRUCTION

intensive care, longer hospitalization, and recovery in general in relation to the expected survival time of the tumor patient. Because transdural extension, which is quite common in lesions with intracranial/extracranial manifestations, has been found to be one of the negative prognostic parameters with regard to survival time,3 complex microvascular procedures may be considered with reluctance. However, malignancies in the head and neck region are different from those elsewhere in the body in that advanced tumor stages are associated with extreme psychological distress, not only because of the severity of the disease, but also because of the high degree of disfigurement associated with extraoral tumor manifestation. Therefore, tumor reduction and reconstruction with a microvascular flap is certainly defensible in otherwise healthy patients in whom local control of the tumor is unlikely to be gamed and further life expectance is short, because continuing tumor progress with ulcerating destruction of the face may constitute a much higher burden for the patients and their relatives during the last months of life than the surgical procedure. The duration of the operation and the period of hospitalization of the patients in the current study appear to be acceptable in relation to a mean survival time of 8.5 months. Because this study is a retrospective evaluation, no measures for assessment of quality of life were applicable to all patients, and there is no information about the quality of life after surgery. Therefore, the effect of surgical palliation is difficult to quantify and remains observer rated. However, the fact that visible tumor in the craniofacial area has been removed and replaced by flap tissue with a reasonable appearance is likely to have added some dignity to the last months of these patients’ lives. The high degree of flap survival and successful closure of large composite defects of the skull base and the scalp in the current study has shown that the free revasculariaed transfer of flaps from the latissimus dorsi muscle, the rectus abdominis muscle, and the forearm can be recommended for primary reconstruction in one-stage resections of intracranially/extracranially growing tumors. The safety of the procedure and the moderate time effort make these measures also acceptable for palliative purposes.

References 1. Price JC, Loury M, Carson B, et al: The pericranial flap for reconstruction of anterior skull base defects. Laryngoscope 98:1159,1988 2. Jackson CG, Netterville JL, Glasscock ME, et al: Defect reconstruction and cerebrospmal fluid management in neurotologic skull base tumors with intracranial extension. Laryngoscope 102:1205,1992 3. Clayman GL, DeMonte F, Jaffe DM, et al: Outcome and complications of extended cranial-base resection requiring

SCHLIEPHAKE

4.

5. 6.

7. 8. 9. 10.

ET AL

microvascular free-tissue transfer. Arch Otolaryngol Head Neck Surg 121:1253, 1995 Sekhar LN, Nanda A, Sen CN, et al: The extended frontal approach to tumors of the anterior, middle, and posterior skull base. J Neurosurg 76:198, 1992 Snyderman CH, Janecka IP, Skhae LN, et al: Anterior cranial base reconstruction: Role of galeal and pericranial flaps. Laryngoscope 100:607,1990 Scher RL, Cantrell RW: Anterior skull base reconstruction with the pericranial flap after craniofacial resection. Ear Nose Throat J 71:210, 1992 Goel A, Gahankari D: Extended subgaleal fascia: Pericranial temporalis flap for skull base reconstruction. Acta Neurochir 135:203, 1995 Goel A: Multilayer reconstruction of the middle cranial fossa. Acta Neurochir 138:584, 1996 Potpraic 2, Fukuta K, Colen B, et al: Galeo-pericranial flaps in the forehead: A study of blood supply and volumes. Br J Plast Surg 49:519, 1996 Krespi YP: Lateral skull base surgery for cancer. Laryngoscope 99:514, 1989

239 11. Johnson GD, Jackson CG, Fisher J, et al: Management of large dural defects in skull base surgery: An update. Laryngoscope 100:200, 1990 12. Hochman M: Reconstruction of midfacial and anterior skullbase defects. Otolaryngol Clin North Am 28: 1269, 1995 13. Yamada A, Harii K, Ueda K, et al: Free rectus abdommis muscle reconstruction of the anterior skull base. Br J Plast Surg 45:302, 1992 14. Bootz F, Gawlowski J: Repair of the anterior base of skull with free latissimus dorsi flap. Acta Neurochir 133: 195, 1995 15. Schliephake H, Schmelzeisen R, Neukam FW: The free revascularized rectus abdominis myocutaneous flap for the repair of tumour-related defects in the head and neck area. Br J Oral Maxillofac Surg 34:18, 1996 16. Urken ML, Cheney ML, Sullivan MJ, et al: Atlas of Regional and Free Flaps for Head and Neck Reconstruction, New York, NY, RavenPress, 1995, p 130 of omentum to a 17. McLean DH, Buncke HJ: Autotransplantation large scalp defect, with microsurgical revascularization. Plast Reconstr Surg 49:268, 1972