Nuclear Medicine

Tc-99m MDP Bone Scintigraphy

Mechanism. Tc-99m MDP is injected intravenously and carried to the skeleton in the bloodstream, where it binds calcium hydroxyapatite by chemisorption. Uptake is proportional to two things: local blood flow, which delivers the tracer, and osteoblastic activity, which binds it. Any process that increases either produces increased uptake, which is why the scan is so sensitive and so unspecific.

The three-phase protocol. The phases move from seconds to hours after the injection:

• Phase 1, flow: dynamic images acquired immediately after injection, over the first 30-60 seconds, showing arterial blood flow to the region; increased flow means hypervascularity (infection, tumour, fracture)
• Phase 2, blood pool: a static image at 5-10 minutes, showing soft tissue distribution and venous pooling; increased blood pool means soft tissue hyperaemia (infection, inflammation)
• Phase 3, delayed bone: static images at 2-4 hours, by which time approximately 50% of the tracer has been taken up by bone and background soft tissue activity has cleared; the MDP is now bound to hydroxyapatite at sites of active turnover, so focal uptake means osteoblastic activity

Reading the phases together. The pattern across the three phases is the discriminator:

• All three phases positive: osteomyelitis, acute fracture, active tumour, CRPS/RSD
• Phase 3 positive only: stress fracture (subacute), established metastasis, degenerative change
• Phases 1 and 2 positive with phase 3 negative: cellulitis, a soft tissue infection without bone involvement

View larger

Applications in orthopaedics. The metastatic skeletal survey, with sensitivity of approximately 95% for blastic metastases; occult fracture detection; stress fracture diagnosis; the extent and activity of Paget disease; and complex regional pain syndrome, which gives diffuse periarticular uptake.

View larger

Read that 95% carefully. It is the figure for blastic lesions, and the denominator matters. Measured per patient rather than per blastic lesion, planar scintigraphy performs less well: in 131 breast and prostate patients at high risk of metastases, patient-level sensitivity was 75% for planar imaging, rising to 87% when SPECT/CT was added. Every patient found on planar imaging was also found on SPECT/CT, and SPECT/CT found three more, so SPECT/CT changes who gets diagnosed rather than merely refining the report, and reprojected SPECT/CT images can replace the separate planar acquisition altogether. The trade-off to state honestly is that specificity of the reprojected images was lower than planar, so correlation with cross-sectional imaging stays essential.

The tumour biology decides whether the test works at all. MDP uptake reflects the osteoblastic response, not the presence of tumour, so the scan is only as good as the bone's reaction to the tumour. In purely lytic disease it fails: compared against radiographs across 562 skeletal sites in 51 myeloma patients, scintigraphy missed or underestimated disease at 27% of sites. That is not a technique problem to be solved with better cameras; myeloma provokes little osteoblastic response, so there is nothing for the tracer to bind. Bone scan is therefore the wrong test to screen or monitor myeloma, and whole-body low-dose CT, MRI or FDG-PET-CT are the modern standards.

The same logic explains why FDG-PET-CT outperformed scintigraphy for skeletal staging in lung cancer (pooled 0.92 sensitivity / 0.98 specificity versus 0.86 / 0.88), while bone scintigraphy retains a strong role in predominantly blastic disease such as prostate cancer.

View larger

What makes a hot spot, and what makes a cold one. Anything that increases bone turnover produces a hot spot; the cold spots are the false negatives. Both lists are worth carrying verbatim.

F-18 FDG PET-CT

Mechanism. F-18 fluorodeoxyglucose is a glucose analogue taken into cells through glucose transporters (GLUT-1) and phosphorylated by hexokinase. Unlike glucose, FDG-6-phosphate cannot be metabolised further and is trapped within the cell, where the F-18 label emits positrons that the PET scanner detects. FDG therefore accumulates in metabolically active cells: tumour cells (the Warburg effect, increased glycolysis), activated inflammatory cells (macrophages, neutrophils) and healing tissue. The CT component gives the functional PET findings their anatomical localisation.

Advantages over bone scan. PET-CT wins on five counts:

• Better spatial resolution (5mm versus 10-15mm)
• Detects lytic metastases, which may be bone scan negative
• Detects soft tissue metastases, where the bone scan shows only bone
• Better specificity for tumour versus benign lesions
• Completed in a single study, against the 4-hour delay of a bone scan

Orthopaedic applications. Where it is used:

• Tumour staging and restaging, superior to bone scan for most primary bone tumours
• Metastatic survey, increasingly replacing bone scan, particularly for lytic metastases
• Infection evaluation, as an alternative to the white cell scan, because FDG accumulates in activated inflammatory cells
• Prosthetic joint infection, where post-operative uptake limits specificity, although "limited" understates it: on the pooled hip data FDG-PET is comparable to leukocyte scintigraphy rather than inferior to it (figures under WBC Scans)
• Assessment of tumour response to chemotherapy, since the metabolic response precedes the anatomical one

View larger

Spondylodiscitis, where PET against MRI becomes counter-intuitive. Head-to-head in 26 patients with suspected haematogenous spondylodiscitis, MRI was the more sensitive test at 94%, but its specificity was only 38%. FDG-PET/CT was less sensitive at 83% and far more specific at 88%. The instinct that MRI is simply the better spinal test is wrong in a specific and useful way: MRI rarely misses spondylodiscitis, but it labels a great deal of degenerative change as infection, because Modic endplate change and disc degeneration produce oedema signal that looks like early infection. PET separates them because SUVmax discriminates infection from degenerative change.

In practice. MRI first to exclude, PET/CT to confirm when the MRI is equivocal; combining the two detected infection in 100% of confirmed cases. Note the limit before quoting this in a viva: a small single-centre cohort, and patients with recent surgery or antibiotics were excluded, so it does not settle the postoperative spine, which is the harder problem.

Labelled White Cell Scans

Mechanism. The patient's own white cells are harvested by venepuncture, labelled with either Indium-111 oxine or Tc-99m HMPAO, and reinjected. The labelled leucocytes migrate to sites of active infection, where they accumulate and produce a focal area of increased uptake.

View larger

Where it is the standard. The labelled white cell scan is the nuclear medicine standard for detecting prosthetic joint infection, acute osteomyelitis, and diabetic foot osteomyelitis, where it separates infection from neuropathic bone change. Use the pooled figures rather than round numbers, because the trade-off between the tests is the examinable point.

Periprosthetic hip infection. In a meta-analysis of 31 studies and 1753 prostheses, leukocyte scintigraphy alone was 88% sensitive and 92% specific. Adding the marrow scan makes it the most specific option at 96% specificity, but sensitivity falls to 69%. FDG-PET is comparable to leukocyte scintigraphy alone at 86% / 93%. Plain bone scintigraphy is the worst performer, 80% sensitive and only 69% specific, which is why it cannot stand alone around a prosthesis.

Do not miss the direction of that trade. Adding the marrow scan is often taught as simply "better". It is not better across the board: it buys specificity by giving up nearly twenty points of sensitivity. That makes the combined study the right test when you need to confirm infection before committing to revision, and the wrong one to rely on for excluding it in a patient you are already worried about.

The combined white cell and marrow scan. An In-111 white cell scan is paired with a Tc-99m sulphur colloid bone marrow scan. If both show uptake in the same pattern, the concordance represents normal marrow distribution. Discordant uptake, white cells without corresponding marrow activity, indicates infection.

View larger

Axial versus peripheral is a genuine discriminator. In chronic osteomyelitis, leukocyte scintigraphy performed well in the peripheral skeleton (84% sensitive) but collapsed in the axial skeleton (21%): the marrow-rich axial skeleton concentrates labelled cells normally, so infection is drowned out by physiological uptake. FDG-PET was the most accurate modality overall (96% sensitive, 91% specific) and is specifically the one to reach for in the spine and pelvis. Bone scintigraphy had the lowest specificity of any option at 25%.

Limitations. The technique is labour-intensive, with blood harvesting, cell labelling and reinjection, and imaging is delayed by 24-48 hours. It is less widely available than bone scan or PET-CT, and it may be falsely negative in chronic low-grade infection.

Aseptic loosening versus infection. This is a distinct problem from simply detecting uptake. On a plain Tc-99m MDP scan, aseptic loosening classically produces focal uptake at the bone-prosthesis interface, at the stem tip and around the lesser and greater trochanters for a femoral stem, whereas infection tends to produce more diffuse uptake along the whole length of the prosthesis. In practice the distinction is unreliable because the patterns overlap heavily and any loose or remodelling prosthesis is "hot". The exam point is that a bone scan cannot reliably separate aseptic loosening from infection: the combined labelled-leukocyte/marrow scan (or FDG-PET) adds specificity, but the diagnosis of PJI ultimately rests on joint aspiration (cell count, culture), inflammatory markers and histology, not on nuclear imaging alone.

Bone scan sensitivity depends on the osteoblastic response of the underlying lesion, and SPECT/CT meaningfully augments planar imaging where anatomical localisation is needed.

Nuclear medicine infection imaging requires appropriate test selection for the clinical context.