Mirels Score for Impending Pathological Fracture

View larger

View larger

View larger

Each of the four variables is scored 1, 2, or 3. The total (4–12) stratifies fracture risk.

The Mirels score answers "will this metastasis fracture?" — but it silently assumes the lesion is a metastasis. The single most important safety point an examiner will test is that a solitary bone lesion of unknown origin must never be internally fixed before it has been worked up and biopsied. Internally fixing (especially intramedullary nailing) a primary bone sarcoma that has been mistaken for a metastasis disseminates tumour along the entire bone and soft-tissue planes, converting a potentially limb-salvageable, curable tumour into one that costs the patient the limb or their life.

Bone is one of the commonest sites of metastasis, and a focused work-up identifies the primary in the large majority of cases.

When the Mirels score does indicate fixation, the operation follows principles that differ from fixing an ordinary fracture, because the construct must outlive a diseased, non-healing bone in a patient who may have limited time and physiological reserve.

• Protect the whole bone, not just the lesion. Metastatic disease is rarely confined to one spot, and a second deposit can fracture distal to a short implant. Use a device that spans the entire bone — a long cephalomedullary nail for femoral and most long-bone lesions — so the whole bone is protected at one operation.
• Build a durable construct that does not rely on bone healing. The bone is unlikely to unite, so achieve immediate mechanical stability with cement augmentation (PMMA) of defects and screws, allowing full weight-bearing immediately. The implant should be expected to last the patient's lifetime.
• Replace, do not fix, when destruction is too extensive. For a destroyed femoral head/neck or extensive peri-articular bone loss, an endoprosthetic (tumour) replacement or arthroplasty is more reliable than internal fixation — fixation into bone that crumbles will fail.
• Embolise hypervascular metastases pre-operatively. Renal cell and thyroid metastases can bleed catastrophically; angiographic embolisation 24–48 hours before surgery dramatically reduces intra-operative blood loss.
• Coordinate adjuvant therapy. Give post-operative radiotherapy to the whole operated bone to control local tumour and reduce implant loosening — but fix first, then irradiate. Add bone-targeted agents (bisphosphonates or denosumab) to reduce skeletal events, and manage the patient through a multidisciplinary team with the oncologists, factoring expected survival into how aggressive the reconstruction should be.

• It over-predicts fixation. In Damron's multi-observer study the score was highly sensitive (about 91 percent) but poorly specific (about 35 percent), so many lesions are fixed that would never have fractured. A score of 8 in particular is unreliable in isolation.
• The original threshold was actually 8, not 9. Mirels' 1989 paper recommended prophylactic fixation for scores of 8 or higher and irradiation for 7 or lower. Because the score over-predicts fracture, modern practice has largely shifted the operative threshold to 9 or more, treating 8 as an individualised grey zone — a common examiner "gotcha," so be ready to state both the original and the contemporary cut-off.
• The pain variable is subjective and has the poorest inter-observer agreement (Damron et al.); size and site are more reproducible. Confirm cortical involvement on CT, not plain film, before finalising the size points.
• CT-based structural rigidity analysis (and finite-element estimates of bone strength) predicts fracture more accurately than Mirels in research settings, but is not yet routine.
• Modern systemic therapy lowers the fracture rate at any given Mirels score, so interpret the number alongside treatment response and life expectancy. A very short prognosis may favour radiotherapy or a non-operative approach even at a high score.
• Mirels is validated for long-bone (mainly femoral) lesions; it does not apply to the spine (use SINS) or flat bones.
• It is not the only system. The older Harrington criteria define an impending fracture qualitatively (over 50 percent cortical destruction, a lytic lesion over 2.5 cm, a painful lesion persisting after radiotherapy, or avulsion of the lesser trochanter) and are still quoted alongside Mirels. Newer prognostic and machine-learning tools (for example PATHFx) estimate patient survival to help decide whether — and how aggressively — to operate, since there is no point in a lifetime-durable reconstruction in a patient with weeks to live. Mirels tells you the bone risk; survival tools tell you whether surgery is worthwhile.