Radiology Level: FRCR, FRCS, MRCP, ABR, EDiR, Radiology Challenge ++++

A much trickier case! The images provided from the initial Liver/MRCP examination are a T1 fat suppressed non-contrast image (top left), a FIESTA image (T2 equivalent bright blood sequence, top right), and two post contrast T1 fat suppressed images following intravenous gadolinium (second row). Thin and thick slab heavily T2 weighted MRCP images are provided on the third row. The final row demonstrates a post contrast T1 weighted fat suppressed image at 3 months. The FIESTA image has some degradation due to artefact from the lung parenchymal air affecting the right liver lobe.

The original imaging demonstrates a left liver lobe T1 hypointense and heterogenously  T2 hyperintense lesion. The lesion demonstrates both peripheral enhancement but also internal extensive reticular enhancement. In addition the contrast enhanced images demonstrate that there is a tubular low T1 abnormality with peripheral enhancement in the central right liver lobe. The MRCP images demonstrate that there are normal appearances of the gallbladder, common bile duct and pancreatic duct but that the intrahepatic ducts demonstrate small clusters of focal subsegmental duct dilatation.

At 3 months, without any reported therapy the lesion has regressed to a smaller nodule of enhancement with some retraction of the liver capsule. The tubular low T1 abnormality and the associated enhancement are less conspicuous.

The appearances of the tubular abnormality in conjunction with the MRCP images indicate intrahepatic duct dilatation. However, the intrahepatic duct dilatation is discontinuous and not secondary to any obstructive stones or stricture. One could consider the possibility of primary sclerosing cholangitis, however, no typical beading is demonstrated. Additionally the dilatation demonstrates enhancement. This is a feature that indicates inflammation or possibly infection.

What about the enhancing lesion? Well the appearance of this abnormality could indicate a malignant lesion. One could then consider perhaps a case of cholangiocarcinoma with metastases, but again the non-contiguous nature of duct dilatation would not make sense and with regards to the lesion itself it is difficult to believe a malignant lesion regressed with no specific treatment.

These appearances and the regression are typical of infectious processes, in particular recurrent pyogenic cholangitis. The focal lesion is termed an inflammatory pseudotumour and often demonstrates this avid and reticular enhancement. Regression can be spontaneous or in response to antibiotics and recurrence of lesions is common. Healing can result in focal scarring as in this case.
 
Recurrent pyogenic cholangitis was previously referred to as oriental cholangiohepatitis and clinically is often associated with recurrent pyogenic episodes, although may be asymptomatic in the elderly. Intrahepatic marked ductal dilatation with intrahepatic duct stones can develop (hepatolithiasis). The stones are typically calcium bilirubinate. Infection of the bile with E Coli, Clonorchis sinensis or Ascaris Lumbricoides  has been postulated as a cause. The condition remains more common in the far east although is not limited to that population. Long term complications include cholangiocarcinoma and intraductal papilloma formation.

Abdominal pain and fever.
Trickier! Ignore central hepatic signal void artefact (black) on top right image.
Radiology Level: FRCR, FRCS, MRCP, ABR, EDiR, Radiology Challenge ++++

3 months later

Level: ABR, EDiR, FRCR, Radiology Challenge ++++

The first set of images are 2-D time-of-flight non contrast MRI acquisitions through the aortic arch and root of the neck.  The second set of images are coronal MRA acquisitions of the same region following intravenous gadolinium administration.  The final image is the maximum intensity projection of the MRA acquisition.

Both series of images demonstrate a large vessel arising from the posterior aspect of the aortic arch, running to the right side posterior to the oesophagus and then cranial to the right supraclavicular region.  This is an incidental left-sided aortic arch with an aberrant right subclavian artery configuration. The aberrant right subclavian artery is arteriomegalic.  Technically this is not a diverticulum of Kommerrel, from which these can arise as a focal aneurysmal dilatation of the origin, but rather a vessel that is enlarged throughout its length. Patients with a left -sided aortic arch and aberrant right subclavian artery do not have a complete vascular ring, however, as in this case, may have compression of the oesophagus in this region, sometimes referred to as "dysphagia lusoria".

One would expect that with the aberrant right subclavian artery instead of a conventional brachiocephalic artery there would be 4 vessels arising from the aortic arch (in turn: right common carotid, left common carotid, left subclavian, aberrant right subclavian). There are indeed 4 vessels arising off the aortic arch on the 2-D time-of-flight images.  However, on the left side both extend into the neck. There is indeed no flow within the left subclavian artery at any level. What is the fourth vessel? This is a direct origin of the left vertebral artery from the aorta.  This occurs in approximately 5% in patients instead of origin from the proximal left subclavian artery.

However, the MRA contrast enhanced images, including the maximum intensity projection, do demonstrate enhancement of 5 arch vessels including the more peripheral left subclavian artery with a more proximal occlusion of  this vessel, over approximately 1.5 cm.  There are asymmetric tortuous collateral vessels in the left side of the neck collateralising to the subclavian artery.

How can we account for the lack of left subclavian artery visualisation on the 2D time of flight sequence but visualisation, albeit incomplete, on the MRA study? What does this tell us? Well the 2-D time-of-flight images are important as they indicate directionality of flow. Note that only cephalad directed arterial flow is recorded. As every slice is acquired sequentially, a saturation pulse is sequentially applied above the next imaged slice to eliminate in-flow signal from structures flowing from above the slice into the slice at the time of acquisition (ie venous blood), allowing only in-flow signal from flowing structures below the slice (arterial). If we wanted insted to visualise venous flow without arterial contamination we would place the saturation pulse below every slice in the 2D time of flight.

So this indicates that there is no cephalad flow in the left subclavian vein. But since there is opacifaction on the contrast enhanced MRA acquisition, the vessel must be partly patent. However, the flow must be retrograde (ie caudally directed). This is a not uncommon situation in vertebral steal phenomenon. A stenosis of the origin of the left subclavian results in poor left subclavian arterial perfusion.  During periods of increased left subclavian arterial demand (exercise), the left subclavian artery fills retrogradely via the left vertebral artery (steal) which anastomoses across the circle of Willis and then connects to the subclavian artery distal to the obstruction.  In this case, however, as the vertebral artery arises directly of the aorta at this collateral pathway is not available.

We can see on the MRA examination that there is an additional tortuous vessel in the left neck anastomosing distally to the left subclavian artery.  This vessel is not visualised on the 2-D time-of-flight acquisitions, indicating that this is downard flow supplying the distal left subclavian artery.  Therefore, in combination these findings indicate that there is only retrograde filling of the left subclavian artery from these left-sided cervical branches.  This is comparable to subclavian steal, but because of separate origin of the vertebral artery the retrograde filling of the left subclavian is not occurring from the vertebral itself.

There is a complex case with two congenital variations of which one, the left vertebral direct aortic origin, is clinically pertinent. Analyzing the difference between 2-D time-of-flight and MRA images assists in determining directionality of flow, and in this case clinical physiology.

Classic Case: Radiology Level - Challenge ++++, FRCR, EDiR, ABR, MRCP

A chest radiograph and an AP lumbar spine radiograph are provided. The chest radiograph demonstrates unfolding of the thoracic aorta which may be aneurysmal although this can be difficult to determine on a plain radiograph alone. There is fine linear calcification running along the right lateral margin of the mediastinum (large arrow), identifying the lateral aspect of the aortic arch. This extends in a continuous fashion to the coarsely calcified ligamentum arteriosum in the AP window (small arrow). There is no coarse calcification of the aorta. The heart size is normal. The lungs are clear.

The AP radiograph of the lumbar spine demonstrate a right total hip replacement and a short segment scoliosis of the mid lumbar spine associated with sclerosis and some fragmentation of the concave aspect as well as new bone formation and loss of normal orientation. There are no congenital variations in this region.

There are several ways in which we can approach the interpretation of these images. A short segment scoliosis has a relatively limited differential. This includes congenital variations such as hemivertebrae related to failures of segmentation such as occurring patients with neurofibromatosis or VACTERL. None of these are present here. Post traumatic appearances can result in short segment scoliosis, however, again not demonstrated here. Tuberculous destruction can result in discitis with partial collapse, however, the endplates, typically destroyed in tuberculous discitis, are well preserved in this instance and the abnormality extends over several levels. Finally one can consider the possibility of a neuropathic (Charcot)spine which is the best fit for the current appearance. Much like other joints a neuropathic spine is associated with the 4 Ds (density, destruction, debris, deformity/dislocation). Causes of a neuropathic spine include spinal cord injury, diabetes and tabes dorsalis (syphilitic).

The most common current cause of a Charcot spine is diabetes, however, in this instance the chest radiograph provides an additional clue. The ascending aortic dilatation and fine ascending aortic peripheral calcification, in contradistinction to the coarse thick calcifications of atherosclerotic aortas arising near the aortic isthmus are typical of syphilitic aortitis. This is the unifying diagnosis in this case–tertiary syphilis with syphilitic aortitis and tabes dorsalis.

Syphilitic aortitis (previously termed luetic aneurysm/aortitis) is a tertiary manifestation of syphilis infection. This is an adventitial disease process resulting in obliterative endarteritis of the vasa vasora resulting in ischaemic injury of the adventitial/media which may progress to aneurysm formation or fine peripheral calcification. Dissection is unusual due to the fibrosis of the medial layer. Coronary calcification, or involvement of the aortic root may also occur. At CT asymmetrical aortic sinus involvement or "tree bark" intimal calcifications can be demonstrated. The disease process is now rare in the Western World due to the reduction of syphilitic infections following introduction of penicillin.

Tabes dorsalis is a form of tertiary neurosyphilis. These presents with dorsal column/nerve root symptomatology relating to weakness and abnormal gait related to lack of proprioception with sensory ataxia (tabetic gait) that as in this case may predispose to falls. Lightening-like paraesthesia pains can occur with altered sensation from joints result in degeneration, as in this case with a right hip replacement. The lower limbs demonstrate hypotonia and diminished reflexes. The differential diagnosis includes other demyelinating pathologies of the cord including multiple sclerosis, transverse myelitis, HIV, Vitamin B12 deficiency resulting in subacute combined degeneration, and focal neoplasms.

This is a classic radiological interpretation to enjoy. A bonus features seen in some cases is calcification of injections in the buttocks due to Bimuth injections predating penicillin treatment.

Classic Case:
Radiology Level - Challenge ++++, FRCR, EDiR, ABR, MRCP

Courtesy Dr A. Brooks