Sacrectomy for Primary Sacral Tumours

Primary sacral tumours amenable to sacrectomy are dominated by three diagnoses.

Chordoma. The commonest primary malignant sacral tumour. It arises from notochordal remnants in the midline, typically in patients in the sixth and seventh decades, and is slow-growing and locally destructive. Histology shows physaliphorous cells, and the tumour is brachyury-positive on immunohistochemistry. It is radioresistant and chemoresistant at conventional doses, so wide en bloc resection is the only reliably curative treatment.

View larger

Chondrosarcoma. Eccentric rather than midline, and also radio- and chemoresistant, so the surgical margin is again the decisive variable.

Giant cell tumour (GCT). A tumour of younger patients, in the third and fourth decades, which is eccentric, lytic and expansile. It is benign but locally aggressive, and sacrificing continence for a benign tumour is rarely justified; its management is set out under the decision thresholds below.

Margin is destiny. In sacral chordoma an intralesional or contaminated margin roughly doubles local recurrence against a wide (R0) resection, and local recurrence is the strongest predictor of disease-specific death. Never debulk a resectable chordoma.

Plan backwards from the margin. For chordoma and chondrosarcoma the bony osteotomy level, the roots sacrificed, the amount of gluteus maximus and pelvic floor taken, and whether the rectum must be resected are all dictated by achieving a wide margin, never by the desire to preserve function. Trading margin for function is an oncologic error; trading function without consent is a medicolegal one.

Distant relapse. Local resection does not remove the risk of distant relapse, which mandates surveillance imaging.

View larger

Biopsy

The biopsy is a CT-guided core needle biopsy through a posterior midline transtumoural tract, placed in the line of the future incision, with the tract tattooed or marked. The tract is considered contaminated and is excised en bloc with the specimen.

The roots and what they carry. The sacral roots exit their foramina and merge to form the sacral plexus on the pelvic surface of piriformis. The pelvic splanchnic nerves (S2-S4) carry parasympathetic supply to bladder, rectum and erectile tissue, and the pudendal nerve (S2-S4) supplies the external sphincters and perineal sensation. The level at which the foramina and their exiting roots are divided determines bladder, bowel and sexual function after resection.

View larger

View larger

Roots, not bony level. The deficit is dictated by which roots are sacrificed, not by the bony level per se. Unilateral S1-S5 sacrifice (hemisacrectomy) leaves the contralateral pelvic splanchnic and pudendal supply intact, so bladder, bowel and sexual function are usually preserved by contralateral compensation, at the cost of possible ipsilateral motor and sensory loss. Bilateral sacrifice is graded root by root in the table.

Consent follows the roots. Plan the highest root that must be divided, then predict the deficit from that level. Consent must state the expected bladder, bowel and sexual outcome for the planned root sacrifice. The case most often under-warned is bilateral S3 sacrifice with both S2 roots preserved, because "S2 preserved" sounds reassuring.

• Both S3 roots divided, S2 kept: consent explicitly for a high probability of deficit, not a minor one, and arrange bladder retraining, intermittent catheterisation teaching and a bowel regimen in advance
• Bilateral S2-S5 sacrificed: mandatory pre-operative counselling and consent for intermittent self-catheterisation, a regimented bowel programme or colostomy, with stoma and continence nurse review
• Extension to S1 and/or L5: add counselling for an orthosis, gait rehabilitation and lumbopelvic reconstruction

Posterior-only. Feasible when the anterior tumour extent is modest, the rectal plane is free and vascular control from behind is realistic, which generally means low and mid lesions. Root sacrifice is limited to the planned margin and no anterior stage is needed.

Combined staged anterior-posterior surgery. Chosen for any of the following, with reconstruction and flap cover planned from the outset:

• High or total sacrectomy
• Bulky presacral disease
• Rectal involvement, with en bloc rectal resection and colostomy
• Re-resection after contamination
• Post-radiotherapy tissue

View larger

Sacral giant cell tumour. Denosumab and serial embolisation have shifted many sacral GCTs away from sacrectomy toward intralesional surgery or non-operative control. First-line is usually denosumab with or without serial embolisation and, if surgery is needed, nerve-sparing intralesional excision. Sacrectomy with sphincter sacrifice for a benign tumour is reserved for exceptional, refractory, destructive disease.

Resectable chordoma or chondrosarcoma. Where resection carries acceptable morbidity, en bloc wide resection is the standard of care.

Chordoma needing bilateral S2 sacrifice or S1 body resection. Discuss it in a multidisciplinary sarcoma meeting. The options are total sacrectomy accepting a complete deficit, or definitive carbon-ion or high-dose proton radiotherapy, which achieves durable local control in a meaningful proportion of patients and is favoured for surgically unfit patients, patients declining sphincter loss, and some high sacral tumours.

Adjuvant radiotherapy. High-dose photon or proton therapy is commonly added for close margins or after any contamination. Pre-operative radiotherapy protocols exist in some centres but raise wound risk.

Recurrent chordoma. Re-resection is rarely curative once the field is contaminated. Particle therapy and systemic options (brachyury-targeted trials, and imatinib historically, with modest activity) are considered.

Template the osteotomies. Pre-operative templating of the osteotomies underpins a negative-margin resection.

View larger

Anterior stagePosterior stageSpinopelvic reconstructionSoft-tissue reconstruction

The patient is supine, and the approach is a midline laparotomy, or laparoscopic or robotic in selected centres.

1. Mobilise the sigmoid colon and rectum and develop the presacral (retrorectal) plane sharply off the tumour pseudocapsule. If the rectum is adherent or involved, resect it en bloc with the specimen and form a colostomy. Identify and protect both ureters.
2. Ligate the median sacral artery. For high resections, ligate or temporarily control both internal iliac arteries and the lateral and median sacral branches, preserving the superior gluteal arteries where possible to protect gluteal flap options. Ligate the internal iliac veins with care: presacral venous plexus injury bleeds torrentially.
3. Perform an anterior discectomy or partial corpectomy at the planned proximal level (L5-S1 for total sacrectomy), and score or start the anterior osteotomies and lateral sacroiliac joint cuts so the posterior surgeon works to a defined plane.
4. Place a silastic sheet (or mesh or spacer) anterior to the sacrum, separating the rectum and iliac vessels from the posterior osteotomy plane. This is the single most protective manoeuvre for the second stage.
5. Consider harvesting a pedicled vertical rectus abdominis myocutaneous (VRAM) flap and tucking it into the pelvis for retrieval at the posterior stage.
6. Close, and perform stage two after physiological recovery.

Set-up. Prone on a four-poster frame with the hips slightly flexed, the entire back, buttocks and perineum prepared, and the anus sutured or excluded if it is being resected. Fluoroscopy or navigation sets the osteotomy level, neuromonitoring protects the preserved roots, and an ultrasonic bone cutter is useful. Cell salvage is debated where tumour spill is a risk, and many units avoid it in malignancy.

Approach. A midline incision excises the biopsy tract en bloc as an ellipse of skin. Raise flaps, detach the gluteus maximus origin (taking a cuff on the tumour side for margin), divide the sacrotuberous and sacrospinous ligaments, release piriformis, and identify and protect the sciatic nerves.

The thecal sac. Through a laminectomy above the tumour, ligate and divide the thecal sac below the last root to be preserved, using a watertight double ligature or oversew to prevent CSF leak. Roots destined for sacrifice are divided sharply after ligation.

Resection. Complete the posterior osteotomies through the pre-scored planes: transverse sacral cuts, or bilateral sacroiliac or iliac cuts where a high sacrectomy must cross the sacroiliac articulation. The silastic sheet confirms the safe anterior plane. Divide the residual pelvic floor, deliver the specimen en bloc, and orient and margin-ink it for the pathologist.

Structures at risk. Those at risk are the presacral venous plexus, internal iliac vessels, rectum, sciatic and preserved sacral roots and superior gluteal neurovascular bundles, with the ureters at the anterior stage.

Pitfalls.

• Entering tumour posteriorly through an unrecognised extraosseous extension
• An osteotomy too low, leaving an R1 margin at S1
• Uncontrolled thecal sac bleeding or CSF leak
• Failing to protect the gluteal vessels, then finding no local flap options

Aftercare. Use drains, avoid direct pressure when lying and grade mobilisation according to the construct, with bladder catheterisation and a bowel regimen from the outset.

After total sacrectomy the load path from spine to pelvis must be re-established; restoring spinopelvic continuity is required once the sacroiliac joints are sacrificed.

Posterior tension band. Pedicle screws from L3 to L5 (usually at least two lumbar levels) are connected with rods to bilateral iliac screws, often two each side. This is the classic Galveston concept modernised as modular screw-rod lumbopelvic fixation.

Anterior column. Vascularised or non-vascularised fibula struts or titanium mesh cages run from the L5 endplate to the ilium and acetabular pillars, triangulated so that axial load is shared and the construct is not a cantilever. Cross-links increase torsional stability, and abundant bone graft aims for a lumbo-iliac fusion mass, because all-metal constructs eventually fatigue.

View larger

When the metal fails. Hardware failure (rod fracture, screw loosening) is common in long-term survivors, occurring in a substantial proportion, but is often tolerable or revisable once partial fusion has occurred. Some units deliberately omit rigid reconstruction in frail patients and let the spine settle onto the pelvis in scar ("soft" reconstruction), accepting trunk shortening for a lower hardware complication rate.

The posterior defect after high sacrectomy is large, rich in dead space and frequently irradiated. Primary closure alone after radiotherapy or total sacrectomy invites wound breakdown, so plan the flap rather than react to the breakdown.

Pedicled VRAM flap. Harvested at the anterior stage and delivered through the pelvis, it is the workhorse for total sacrectomy and any irradiated field. Well-vascularised muscle obliterates dead space, covers hardware and dramatically reduces wound breakdown.

Alternatives. Gluteal advancement or V-Y flaps suit smaller low or mid defects when the superior gluteal pedicles were preserved. Free flaps (latissimus) are reserve options.

Haemorrhage. Massive haemorrhage is the leading intraoperative threat. For high resections, anterior-stage ligation of the internal iliac and median sacral vessels, staged surgery and a prepared transfusion protocol are mandatory. The table sets out the mechanism, burden, prevention and salvage of each major hazard.

• Epidemiology: chordoma incidence is roughly one per million per year worldwide; sacrococcygeal disease accounts for about half of chordomas, with male predominance and peak in the sixth to seventh decades. Sacral chondrosarcoma and GCT are rarer at this site.
• Society guidance: ESMO/EURACAN and Chordoma Global Consensus Group guidance emphasise referral of presacral masses to sarcoma centres before biopsy, trocar/core biopsy through a resectable posterior tract, en bloc resection with wide margins as standard for chordoma, and high-dose particle therapy (proton or carbon-ion) as definitive treatment where resection is not feasible or declined. NCCN bone tumour guidance mirrors these principles. BOA/BSG sarcoma pathways in the UK mandate management within designated sarcoma networks.
• Registries and multicentre series: there is no arthroplasty-style implant registry for sacrectomy; evidence rests on multicentre collaborations — AOSpine Knowledge Forum Tumour primary spinal tumour cohorts, Rizzoli Institute and Memorial Sloan Kettering chordoma series, and Japanese and Italian carbon-ion cohorts (NIRS/QST Chiba, CNAO Pavia).
• Practice variation by resource setting: carbon-ion facilities exist in only a handful of countries (Japan, Germany, Italy, Austria, China), so surgical resection remains the only curative option in most of the world; in low-resource settings, staged surgery, flap availability and intensive-care capacity constrain total sacrectomy, and non-operative palliation or referral abroad is common for S1-level disease. Denosumab access similarly determines whether sacral GCT is managed medically or surgically.

• Surgery versus particle therapy for high sacral chordoma: no randomised comparison exists and none is likely, but the field is no longer evidence-free — Yolcu's propensity-matched analysis found no significant difference in overall survival between carbon-ion therapy and en bloc resection (median 68.1 versus 58.6 months, hazard ratio 0.71, p equals 0.53) with markedly less peripheral motor neuropathy after carbon ion (odds ratio 0.13). Read it with its own caveat: the surgical arm was only 54 patients and that confidence interval spans a large benefit either way. Note also that carbon-ion series report ambulation, not continence — Imai's 97 per cent refers to walking, and his paper contains no sphincter data. Long-term insufficiency fracture, late nerve toxicity and the difficulty of salvage after radiotherapy failure remain genuine concerns.
• Rigid spinopelvic reconstruction versus no reconstruction after total sacrectomy: rigid constructs restore load transfer but carry high hardware failure and infection rates; some series report acceptable ambulation without reconstruction.
• Extent of nerve sacrifice for margin: whether a planned marginal (rather than wide) plane over preserved S2/S3 roots is oncologically acceptable in selected chordomas is debated.
• Cell salvage and tumour surgery: traditionally avoided; leucocyte-depletion filters and irradiation of salvaged blood are used in some centres.
• Denosumab in sacral GCT: optimal duration, rebound after cessation, and whether it should replace surgery entirely in the sacrum remain unresolved.
• Prophylactic colostomy in total sacrectomy: reduces perineal wound soiling but adds morbidity; practice varies widely.

Related pages. The tumours that bring a patient to this operation: Chordoma — half of all chordomas are sacrococcygeal — with Chondrosarcoma, Giant Cell Tumour of Bone and its medical alternative Denosumab, Ewing Sarcoma, Osteosarcoma, Metastatic Bone Disease and Primary Bone Tumours of the Spine. Before the knife: Biopsy Principles and Techniques — the tract must be excisable en bloc, which is why the referral goes to the sarcoma centre first — Soft Tissue Masses and Sarcoma Referral and Limb Salvage Surgery Principles for the margin logic that governs the whole plan. The neuroanatomy that decides the consent conversation: Lumbosacral Plexus Anatomy and Cauda Equina Syndrome — the deficit a planned resection deliberately creates is the one an emergency exists to prevent. The mechanical problem after resection: Spinopelvic Dissociation and Spinopelvic Parameters for load transfer, with Sacral Fractures, Sacral Insufficiency Fracture — a recognised late consequence of high-dose particle therapy to the sacrum — and Pelvic Ring Injuries. Recovery and its limits: Nerve Injury and Regeneration and Radiation Safety in Orthopaedics.