Digital Subtraction Angiography
Digital subtraction angiography (DSA) is a fluoroscopic study of arteries, veins and the vascular supply of organs and lesions. A catheter is passed into a vessel (usually the common femoral artery by the Seldinger technique, or the radial or brachial artery) and iodinated contrast is injected while images are acquired at 1 to 6 frames each second.
A mask image taken before contrast arrives is subtracted from the images taken with contrast, so that bone and soft tissue disappear and only the opacified vessels remain. Selective catheterisation of the carotid, vertebral, coronary, renal, mesenteric, bronchial and other arteries gives high resolution images of small branches, and rotational acquisition gives three dimensional reconstructions.
Contrast volumes range from 3 to 10 mL for a selective injection to 20 to 40 mL for an aortic injection, and carbon dioxide or gadolinium are used when iodinated contrast must be avoided. The effective dose is about 2 to 10 mSv for diagnostic studies and much higher when combined with intervention, and the skin dose is the main concern in long procedures.
CT angiography and MR angiography have replaced diagnostic DSA for most indications, because they are non-invasive and show the vessel wall and surrounding tissue.
DSA remains the reference test for spatial resolution and for flow dynamics, and it is the basis of endovascular treatment: it is the test for small vessel and intracranial disease (aneurysm, arteriovenous malformation, dural fistula, vasculitis), for active bleeding that may be embolised, for planning embolisation, and for problems where non-invasive imaging is inconclusive or cannot be done.
Indications. Intracranial aneurysm, arteriovenous malformation, dural arteriovenous fistula and moyamoya disease when non-invasive imaging is inconclusive or treatment is planned. Cerebral vasculitis and unexplained subarachnoid haemorrhage. Carotid and vertebral stenosis, dissection and occlusion (when non-invasive imaging is unclear).
Peripheral arterial disease: level and extent of occlusion before revascularisation. Acute limb ischaemia. Acute gastrointestinal, bronchial, renal, pelvic or traumatic haemorrhage (for localisation and embolisation). Renal artery stenosis and fibromuscular dysplasia. Aortic and visceral aneurysms, and vascular malformations. Mesenteric ischaemia. Pre-operative mapping of tumour blood supply.
Assessment of dialysis access. Pulmonary angiography for arteriovenous malformation. Not as a first line test for screening, and not when CT or MR angiography answer the question.
Contraindications and precautions. Allergy to iodinated contrast (premedication or carbon dioxide, depending on the territory). Renal impairment (eGFR under 30 mL each minute for 1.73 m2): contrast nephropathy risk is raised, so minimise the volume, hydrate and consider carbon dioxide.
Coagulopathy (international normalised ratio over about 1.5 to 1.8, platelets under 50 000 in each microlitre) and anticoagulant treatment, handled as the local rule requires. Pregnancy: avoid unless essential. Severe uncorrected hypertension, heart failure, or active infection at the puncture site. Metformin: handled as the local rule requires. Thyroid disease (iodine load).
Previous groin surgery or bypass graft, and peripheral occlusion, may need another access route.
Preparation and technique. Informed consent, fasting for 4 to 6 hours (clear fluids allowed), hydration, and a check of renal function, coagulation and medicines. Under local anaesthetic and ultrasound control, the artery is punctured, a sheath is placed, and a wire and catheter are advanced under fluoroscopy to the target.
Contrast is injected by hand or pump, with images in several projections, and the patient is asked to hold still, hold the breath and avoid swallowing. Intravenous heparin is used in long procedures.
After the study the catheter is removed and haemostasis is secured by manual pressure for 10 to 20 minutes, a closure device or a radial compression band, followed by bed rest for 2 to 6 hours (femoral) or 1 to 2 hours (radial) and monitoring of pulses, puncture site, and neurological state.













