Osteopetrosis

Osteopetrosis (marble bone disease, Albers-Schonberg disease) is a group of rare inherited skeletal disorders characterised by defective osteoclast-mediated bone resorption, which leaves the skeleton generally sclerotic. Despite its radiographic density the bone is structurally abnormal and paradoxically brittle.

Three forms. Inheritance, age at onset and severity separate them:

• Autosomal recessive osteopetrosis (ARO): 1 in 250,000 births, severe, infantile onset
• Autosomal dominant osteopetrosis (ADO): 1 in 20,000, mild, adult onset. ADO Type II is the most common form overall
• Intermediate autosomal recessive: variable, childhood onset

Males and females are affected equally, and ARO has a higher incidence in consanguineous populations.

Genetics. Recessive disease is caused by mutations in TCIRG1 (50%), CLCN7 (15%) or OSTM1, or in TNFSF11 (RANKL) or TNFRSF11A (RANK), which produce an osteoclast-poor form. In dominant disease CLCN7 accounts for 70% and is the most common cause of ADO Type II, while LRP5 is associated with ADO Type I. What each gene does is set out under Pathophysiology.

The core defect. Osteopetrosis is a disease of osteoclast dysfunction. In most forms the osteoclasts are present but cannot resorb bone effectively, while osteoblast function is normal.

How an osteoclast resorbs bone. The normal sequence:

1. The osteoclast attaches to the bone surface (sealing zone)
2. The proton pump (H+-ATPase) acidifies the resorption lacuna
3. The chloride channel (CLCN7) provides charge balance
4. Acid dissolves the hydroxyapatite mineral
5. Cathepsin K digests the collagen matrix
6. The resorption products are endocytosed

Where it fails. The genes, and what their mutations break:

TCIRG1 encodes the a3 subunit of the vacuolar H+-ATPase, and OSTM1 is the osteopetrosis-associated transmembrane protein. With no resorption, calcified cartilage cores accumulate and primary spongiosa is never converted to mature bone, so the sclerotic skeleton has an abnormal architecture.

Why dense bone is weak. The bone lacks normal architecture. Calcified cartilage is retained within it, the trabecular microstructure is disorganised, woven rather than mature lamellar bone predominates, and without remodelling microdamage is never repaired. Its mineral content is increased, which is why it looks dense, but its elastic modulus is decreased and its ultimate strength reduced, so it breaks at lower loads, like chalk, with a transverse, shattering fracture pattern.

Failed modelling. Tubulation of the metaphysis requires resorption. Without it the metaphyses stay flared, the Erlenmeyer flask deformity, and the diaphysis remains wider than normal.

Marrow obliteration. The medullary canal fills with abnormal bone and replaces the haematopoietic tissue. The result is pancytopenia: anaemia from reduced erythropoiesis and haemolysis, thrombocytopenia with bleeding and bruising, and leucopenia with recurrent infections. Extramedullary haematopoiesis enlarges the liver and spleen.

The cranial foramina. The foramina of the skull base cannot enlarge normally because resorption fails, and the structures passing through them are compressed:

Teeth and jaw. Teeth erupt late because the overlying bone is not resorbed. Dental abscesses and osteomyelitis of the jaw follow from its poor blood supply, and mandibular osteomyelitis is a major complication.

Calcium. In some forms the bone cannot release calcium, causing hypocalcaemia with secondary hyperparathyroidism, and rickets-like features are paradoxically possible. The mechanism has its own section, Osteopetrorickets, below.

The two main forms side by side:

Blood. The full blood count shows pancytopenia in ARO and is usually normal in ADO. The film shows nucleated red cells and immature white cells from extramedullary haematopoiesis, and the reticulocyte count may be raised by haemolysis. The biochemistry:

• Calcium: variable - normal, low or high
• Phosphate: usually normal
• Alkaline phosphatase: may be elevated (osteoblast activity)
• Acid phosphatase (TRAP): elevated (osteoclast marker)
• Creatine kinase BB isoenzyme: elevated (osteoclast marker in ARO)
• PTH: may be elevated (secondary hyperparathyroidism)

Genetic testing. The panel covers TCIRG1, CLCN7, OSTM1, TNFSF11 and TNFRSF11A. The result matters for prognosis and family counselling, and it determines eligibility for BMT.

Bone marrow. Aspiration is difficult through sclerotic bone. A trephine biopsy may show abnormal architecture, with reduced or absent haematopoietic tissue.

Radiographs. The plain film findings are pathognomonic:

• Sandwich vertebrae (rugger jersey spine): sclerotic endplates with a relatively lucent centre, alternating bands of sclerosis and lucency. Diagnostic for osteopetrosis
• Erlenmeyer flask deformity: flask-shaped widening of the metaphysis from failed tubulation, most visible at the distal femur
• Bone-in-bone appearance (endobone phenomenon): a ghost outline of the previous bone within the current one, representing retained growth lines, seen in the vertebrae, phalanges and iliac wings
• Diffuse osteosclerosis: generalised increased density through the entire skeleton, with loss of medullary cavity distinction

Elsewhere, the skull base is thickened with frontal bossing, the iliac wings are dense, the long bones show transverse fractures with variable callus formation, and the spine shows vertebral compression fractures.

View larger

Look-alikes. Other sclerosing bone disorders are the trap on a film of dense bone, and the differential is tabulated under Differential Diagnosis below. The three figures that follow come from a cohort of osteosclerotic disorders that excluded osteopetrosis, and are shown for comparison only.

View larger

View larger

View larger

CT and MRI. CT gives better visualisation of foraminal stenosis and cranial nerve canal narrowing, and helps surgical planning. MRI assesses marrow replacement and cranial nerve compression, and identifies extramedullary haematopoiesis.

Eyes and ears. Ophthalmology uses visual evoked potentials (VEP) and fundoscopy for optic atrophy, with regular monitoring for optic nerve compression. Audiology uses brainstem auditory evoked responses (BAER) and pure tone audiometry to monitor progression of hearing loss.

Bone density. DEXA shows a paradoxically high BMD despite weak bone. Do not use BMD to guide treatment in osteopetrosis: the quality of the bone matters more than its quantity.

Bone biopsy. Rarely needed for diagnosis. It shows retained calcified cartilage cores and, in most forms, osteoclasts that are present but ineffective, with no resorption lacunae.

HLA typing. Needed for transplant donor matching. Sibling donors are preferred, with a search of the unrelated donor registry otherwise.

Sclerosing Bone Disorders

Key Distinguishing Points:

• Pyknodysostosis: Toulouse-Lautrec had this; characterized by short stature, fragile bones, but with ACRO-OSTEOLYSIS (absent in osteopetrosis) and open fontanelles
• Engelmann Disease: Affects diaphyses primarily, causes pain and weakness, autosomal dominant
• Melorheostosis: Unilateral, follows sclerotome/dermatomal pattern, looks like dripping candle wax

The disease-related complications of each form are described under Clinical Presentation. The operative and transplant complications:

Early. The complications of the operation itself:

• Intraoperative fracture of the brittle bone during manipulation
• Drill bit breakage, the broken bits usually left in situ
• Prolonged operative time, increasing the infection risk
• Bleeding, which may be significant from the abnormal bone
• Thermal necrosis from prolonged drilling

Late. Those that emerge after the operation:

• Delayed union and nonunion (see Healing, above)
• Implant failure: screw loosening, plate breakage
• Refracture after implant removal or adjacent to hardware
• Infection, with increased susceptibility and difficult to eradicate

After BMT. The transplant carries its own risks:

• Graft-versus-host disease (GVHD)
• Graft failure
• Infection during the immunocompromised period
• Veno-occlusive disease
• Late effects of conditioning chemotherapy

Global Epidemiology

There is no high-level (RCT) guidance for this rare disease; practice rests on registry data and expert consensus.

Consensus Positions Across Societies

Registries and Networks

HSCT outcomes are pooled through international transplant registries (CIBMTR in North America, EBMT in Europe), which together generated the largest osteopetrosis transplant cohort to date. Unrelated-donor matching relies on national and international marrow donor registries. Rare-bone-disease reference networks (for example ERN-BOND in Europe) coordinate diagnosis, genetic confirmation and multidisciplinary care.

High- vs Limited-Resource Practice Variation

• Well-resourced settings: next-generation sequencing gene panels, early HSCT in infancy, skull-base and ophthalmology services, and locked-plate plus external-fixation inventories for fracture surgery.
• Limited-resource settings: diagnosis is often clinical and radiographic; transplant access may be delayed or unavailable, shifting care toward transfusion support, infection control and conservative fracture management. Consanguinity raises ARO incidence, so genetic counselling and antenatal diagnosis carry high value where available.

Carbonic anhydrase II deficiency is the one osteopetrosis variant with a signature systemic triad, which makes it highly examinable: the diagnosis can be reached from the combination rather than from the bone alone.

• A distinct autosomal-recessive triad. Carbonic anhydrase II (CA2) deficiency causes a recognisable syndrome of (1) osteopetrosis + (2) renal tubular acidosis (RTA) + (3) cerebral (basal ganglia) calcification - often remembered together, with developmental delay/intellectual disability and short stature completing the picture. It is also called Guibaud-Vainsel syndrome or "marble brain disease".
• Why one enzyme does all three. Carbonic anhydrase II generates the protons (H⁺) that the osteoclast's proton pump secretes to acidify the resorption lacuna - so its loss impairs bone resorption (osteopetrosis). The same enzyme is needed for renal tubular acid handling (both proximal and distal), producing a mixed proximal + distal RTA; and it is expressed in the brain, where its deficiency is associated with the characteristic basal-ganglia calcification. One shared biochemical step, three organ systems.
• How it differs from "classic" osteopetrosis - and why it matters. Unlike TCIRG1/CLCN7 disease, CA2 deficiency is milder in its skeletal severity (marrow failure is usually NOT life-threatening), the sclerosis may even diminish with age, and crucially it is NOT curable by haematopoietic stem-cell transplant - because the defect is a systemic enzyme deficiency, HSCT corrects the marrow-derived osteoclast but not the renal or cerebral disease. Management is therefore supportive: treat the RTA (alkali therapy), manage fractures, and monitor growth/development.

How a disease of too much bone can coexist with rickets is a classic exam curveball, and thickened growth plates with rickets-like features are genuinely described in osteopetrosis.

• It is real, and it makes mechanistic sense. "Osteopetrorickets" is the coexistence of rachitic (unmineralised) growth-plate changes on a background of dense osteopetrotic bone - seen particularly in severe infantile ARO. The key is that osteopetrosis is a defect of resorption, not of mineral supply, and the two problems attack different compartments.
• Why the osteoclast defect causes a mineralisation failure. Normal osteoclasts liberate calcium (and phosphate) from bone into the blood; when they cannot resorb, the skeleton behaves as a calcium "sink" that traps mineral and cannot release it, so despite radiodense bones the child can become hypocalcaemic. The hypocalcaemia drives secondary hyperparathyroidism, and the low available mineral at the growth plate produces genuine rickets (widened, frayed, cupped physes) - hence dense metaphyses with rachitic growth plates side by side.
• Clinical consequences to recognise. The hypocalcaemia can be symptomatic - tetany and seizures in infancy (a feature specifically noted in ARO) - and the picture can be worsened by coexisting nutritional vitamin-D deficiency. Treatment is calcium and vitamin D/calcitriol to correct the mineral deficit (calcitriol may additionally stimulate residual osteoclasts); this is one reason mineral status is monitored closely, and it does not contradict the "no antiresorptives" rule.

The rarity of osteopetrosis means most management rests on registries and expert opinion rather than randomised trials. Several genuine areas of debate persist:

• Optimal timing and conditioning for HSCT. Earlier transplantation in infancy is associated with better outcomes, but the ideal age threshold, conditioning regimen, and donor hierarchy beyond matched siblings remain unsettled. Graft failure and hepatic/pulmonary toxicity are the dominant limitations rather than the transplant decision itself.

• Reversibility of neurosensory deficits. Whether established optic atrophy or hearing loss can be salvaged by transplant or surgical decompression is uncertain; most evidence suggests vision rarely improves once atrophy is fixed, which is why decompression is considered only before established damage.

• Role of optic-canal decompression. Indications and benefit are inconsistent across centres, and it is reserved for selected, progressive cases rather than offered routinely.

• Pharmacological alternatives to transplant. Interferon-gamma-1b offers only modest, mainly supportive benefit; recombinant RANKL for osteoclast-poor (TNFSF11) disease and gene therapy for TCIRG1/CLCN7 remain investigational, with limited human data.

• Denosumab in osteopetrosis. As a disease therapy it is contraindicated (it suppresses osteoclasts), yet it has been explored narrowly for transplant-related hypercalcaemia - a nuance that is easy to misstate.

• Fracture fixation strategy. There is no consensus trial comparing plating, intramedullary fixation, and external fixation; choice is individualised, and the often-quoted precise nonunion percentages are extrapolated from small series rather than robust cohort data.