Bone Island of the Shoulder
A bone island is a small patch of dense, compact bone within the spongy interior of a bone—a developmental variation rather than a disease, causing no symptoms and requiring no treatment. Around the shoulder they occur in the head and upper shaft of the arm bone, the shoulder blade, and the collarbone. They are found here regularly, because shoulder imaging is performed frequently for rotator cuff and joint problems and the bone is examined in detail along the way. The practical point in this region is almost always the same: shoulder pain is common and has many genuine causes, and a bone island is not one of them, so finding one should not end the search for the actual problem.
What is it?
Bone exists in two forms: compact bone, dense and solid, forming the outer shell, and cancellous bone, the open lattice of fine struts filling the interior with marrow between them.
During growth, solid bone is resorbed and reshaped into this internal lattice. A bone island occurs where a small area of that remodelling never happened, leaving a patch of compact bone within the spongy interior.
The medical term is enostosis. It is a developmental variation rather than a disease—not a tumour, not capable of becoming malignant, not a cause of weakness or symptoms. It is present from early life and remains essentially unchanged.
Around the shoulder, bone islands occur in the head and upper shaft of the arm bone, in the shoulder blade, and in the collarbone. The upper arm bone is the usual site.
They are identified here regularly, not because the shoulder is particularly prone to them, but because shoulder imaging is performed so frequently. Rotator cuff problems, impingement, frozen shoulder, labral injuries, and arthritis generate large numbers of X-rays and MRI scans, and the bone is examined in detail each time. A person being scanned for a suspected tendon tear has their humerus imaged thoroughly, and incidental findings follow.
The features identifying a bone island are the same throughout the skeleton.
The border carries the most weight. A bone island does not have a smooth, sharply drawn edge. Its margin shows fine, thorn-like spicules radiating outward and blending seamlessly into the surrounding bony lattice—an appearance described as a brush border. This reflects what the lesion is: part of the surrounding bone that was never remodelled, rather than something deposited into it.
Orientation along the direction of the surrounding bony struts is typical.
Density is uniform and matches the outer shell of the bone.
On MRI, the lesion is uniformly dark on every sequence and there is no swelling of the surrounding marrow. This absence of marrow change is strongly reassuring and separates it from most alternatives.
On bone scan there is typically no increased activity, though giant examples may show mild uptake.
Slow enlargement is compatible with the diagnosis, particularly in young people, with published criteria accepting an increase of up to a quarter in diameter over a year in a lesion that otherwise appears classic. Growth alone does not indicate malignancy; loss of the characteristic border, development of marrow changes, rapid growth, or new pain are the features that matter.
Several other conditions produce dense areas in this region, and knowing them helps interpret a report.
Secondary cancer deposits are the principal alternative in an adult. The upper arm bone retains some active blood-forming marrow into adult life, making it a recognised site for spread, though less frequently involved than the spine, pelvis, and ribs. Prostate cancer characteristically produces dense deposits and breast cancer frequently does. The distinguishing features are those described above, with the border, the absence of marrow change on MRI, and the absence of bone scan activity being decisive. Where previous imaging exists showing the same lesion unchanged before any cancer was diagnosed, the question is settled conclusively.
Osteoid osteoma occurs in the upper arm bone, producing a dense area with a small central lucent core. The distinction from a bone island is clinical: osteoid osteoma causes pain, characteristically worse at night and classically relieved by anti-inflammatory medication. Since bone islands cause no pain, a dense lesion in someone with night pain in the shoulder or upper arm requires a different explanation. It is readily treatable, most often by a minimally invasive technique that destroys the central core.
Bone infarcts occur in the upper arm bone, producing a winding serpentine dense outline rather than a solid focus. They are associated with corticosteroid use, excess alcohol, sickle cell disease, and decompression sickness, and are frequently asymptomatic. Their relevance is that the same process affecting the head of the arm bone beneath the joint surface is avascular necrosis, which has very different consequences, so finding an infarct may prompt assessment of the joint itself.
Sclerotic osteosarcoma occurs in the upper arm bone, this being the second commonest site after the knee, in adolescents and young adults. Its appearance is entirely different from a bone island—an ill-defined margin, cortical destruction, aggressive bone surface reaction, soft tissue extension, and surrounding marrow changes—and it causes persistent pain including at night, frequently with swelling. The two are not genuinely confusable, but the coexistence of both in the same bone is why dense lesions here are described carefully in young people.
Dense areas following previous fracture, surgery, infection, or radiotherapy are explicable from the history.
Osteopoikilosis, the inherited condition producing numerous bone islands throughout the skeleton, clusters them around the shoulder along with other large joints in a symmetrical distribution, and is entirely benign.
Symptoms are the simplest aspect of this topic: bone islands cause none.
This matters more at the shoulder than at most sites, because shoulder pain is extremely common and the genuine causes are numerous. Rotator cuff tendon problems and tears account for the majority. Frozen shoulder produces severe pain and stiffness with normal bones. Arthritis of the main joint or the small joint at the top of the shoulder causes pain. Labral injuries, instability, biceps tendon problems, and referred pain from the neck all contribute.
A bone island found on imaging performed for any of these is an unrelated incidental finding. Accepting it as the explanation leaves the actual problem untreated, and since most of these conditions respond well to appropriate treatment, the cost of that misattribution is real.
Assessment is usually straightforward.
X-ray frequently suffices, with the characteristic appearance allowing confident diagnosis.
CT demonstrates the brush border and uniform density best.
MRI shows the uniformly dark appearance with absent marrow change, and is frequently the study on which the lesion is first identified, since it is simultaneously assessing the rotator cuff, labrum, and joint—which is usually the reason the scan was requested.
Bone scan contributes where distinguishing from a metastatic deposit matters, particularly in someone with a known cancer.
Comparison with previous imaging is the most valuable single step, and given how frequently shoulders are imaged, earlier studies are often available.
Biopsy is essentially never required for a lesion with classic features.
Important to Know
A bone island at the shoulder is a normal variant requiring no action. Where anything is needed, it is confirmation of the diagnosis rather than treatment of anything. Care, in the rare cases requiring any, involves radiology and orthopaedic services.
For a lesion with classic features, the appropriate response is documentation and reassurance—no treatment, no biopsy, no activity restriction, and no follow-up.
Documentation is the one genuinely useful action. Bone islands persist for life and will appear on every future scan of the shoulder, and given how often shoulders are imaged, that may be several times over the years. Where the finding has not been clearly recorded as benign, it can trigger fresh investigation each time, and most consequentially if the person later develops a cancer and staging imaging is performed. Having an earlier scan available for comparison resolves the question immediately, so noting where and when previous imaging was performed is worthwhile.
The most important practical message in this region concerns symptoms, and it is worth stating plainly. Shoulder pain almost always comes from the rotator cuff, the joint, the capsule, the biceps tendon, or the neck. These are the structures that generate pain, and most of them are treatable. A bone island generates none.
Where a scan performed for shoulder pain identifies a bone island, the relevant findings are what the scan shows about the tendons, the joint surfaces, the labrum, and the subacromial space. The bone island is incidental, and the assessment should continue on its own terms. This is worth asking about directly—what the scan showed about the rotator cuff and the joint—rather than focusing on the bone finding.
One symptom pattern points away from a bone island and toward something treatable. Persistent shoulder or upper arm pain that is constant rather than movement-related, is worse at night, and is relieved markedly by anti-inflammatory medication should prompt consideration of osteoid osteoma. This is a genuinely painful lesion that occurs in the upper arm bone, is frequently misdiagnosed for long periods, and responds well to treatment. The small central core can be easy to overlook, and CT is the best means of demonstrating it.
Persistent upper arm pain with swelling in an adolescent or young adult, particularly where pain is constant or wakes them at night, warrants imaging rather than being managed as a training problem. The upper arm bone is the second commonest site for osteosarcoma, and the typical history before diagnosis involves months of symptoms attributed to sport.
Where someone has a known cancer and a dense lesion is found in the upper arm bone, the same distinguishing features apply with full reliability. Bone islands are common, so coexistence is frequently coincidental. The border, the absence of marrow change on MRI, and the absence of bone scan activity are decisive, and previous imaging showing the lesion unchanged before the cancer resolves the question conclusively. It is reasonable to ask whether earlier imaging has been reviewed.
Bone infarcts, if found in the upper arm bone, usually require nothing. Their main relevance is prompting consideration of an underlying cause and assessment of the head of the arm bone for avascular necrosis, which affects the joint and is managed quite differently.
Where appearances are atypical—an indistinct margin, surrounding marrow change, marked bone scan activity, rapid growth, cortical involvement, or associated symptoms—further assessment is appropriate rather than reassurance.
Osteopoikilosis requires nothing beyond recognition, and identifying it in one person explains the same appearance in relatives.
Care, where required at all, involves radiology and orthopaedics, though in most cases no referral is needed. Imaging features are definitive in classic cases, and comparison with previous studies is the most valuable investigation available.
Patient education plays an important role, largely because any lesion reported in bone causes concern regardless of its nature. Understanding that a bone island is a developmental variant rather than a disease, that it causes no symptoms and never will, that it does not become cancerous, that shoulder pain requires a separate explanation, and that night pain points away from a bone island and toward something treatable all contribute to appropriate care.
Red flag symptoms include shoulder or upper arm pain that is constant rather than movement-related, particularly if it wakes you at night, which is not explained by a bone island; pain relieved markedly by anti-inflammatory medication, characteristic of osteoid osteoma; persistent upper arm pain with swelling in a child, adolescent, or young adult; progressive swelling or a palpable lump over the bone; rapid change in a known bone lesion on sequential imaging; a lesion reported with an indistinct border, cortical destruction, aggressive periosteal reaction, or surrounding marrow changes; a new dense bone lesion in someone with a history of cancer where no earlier imaging is available for comparison; and persistent shoulder pain, which warrants assessment of the tendons, joint, and neck on its own merits rather than being attributed to an incidental bone island. These warrant prompt medical evaluation.