Tip:
Highlight text to annotate it
X
Hello, I'm Frank Gaillard form Radiopaedia.org. In this, our 6th episode in our imaging of
stroke series we will review haemorrhagic transformation of ischaemic stroke. Before
we begin lets have a look at three cases. Which of these do your think represents haemorrhagic
transformation of a stroke? By the end of this episode you should be able to make the
distinction with confidence. In episode 4 of this series we saw the natural progression
of a bland ischaemic stroke from the earliest signs, such as a hyperdense artery, all the
way through to encephalomalacia. Haemorrhagic transformation is usually seen in the first
4 days following infarction but is rare in the first 6 hours. Haemorrhagic transformation
in one form or another is seen over half of all infarcts (although the reported rates
vary widely depending on definition and modality). Two distinct processes fall under this term.
These are petichial haemorrhage and secondary haematoma. As we will see they not only appear
different but have different prognostic implications. As such it is important in reports to ensure
that when haemorrhagic transformation is present, one clear distinguishes between the two. Lets
first discuss petichial haemorrhages. Petichial haemorrhages, which pathologists refer to
this as "red softening" macroscopically, account for the vast majority of cases with haemorrhagic
transformation and result in increased attenuation of affected brain. In this case a CT obtained
acutely does not convincingly demonstrate change, but a follow up study 2 weeks later
demonstrates increased attenuation of the cortex. This appearance and timing could represent
fogging phenomenon we discussed in episode 5 which typically occurs 2 to 3 weeks after
infarction. If however we review an MRI obtained one day after onset of symptoms, after tPA
administration we can see a large infarct and gyriform susceptibility induced signal
drop out consistent with petichial haemorrhagic transformation, which typically occurs within
a day of thrombolysis. In this case a large right MCA infarct is again not clearly visible
on the initial scan, although minor blurring of grey white matter interface is present.
The next day on MRI the infarct is easily seen on DWI and on echo planar imaging signal
loss is seen in the lentiform nucleus. Two days later still this region demonstrates
increased attenuation. Petchicial haemorrhages result of small about of blood seeping out
of vessels into the near by tissues. They do not have mass effect and do not generally
impact on prognosis or treatment. Now onto the far more sinister secondary haematomas.
As we have seen the vast majority of ischaemic strokes, without intervention, undergo a non-haemorrhagic
progression to encephalomalacia. A small minority, say 5% will however spontaneously haemorrhage
and only a subset of these are symptomatic. This is easy to understand when you look at
this cases. The haematomas are relatively small and embedded in brain that is already
infarcted. The cause of haemorrhagic transformation is thought to be due to in the majority of
cases to the early reperfusion of infarcted tissue. The damaged vessels are unable to
withstand arterial pressures and rupture. Collateral flow is also implicated in some
patients, as secondary haemorrhage can be seen in patients without recanalisation. The
rate is significantly higher when reperfusion therapy is employed, such as intravenous or
intra-arterial thrombolysis or clot retrieval. The rate of symptomatic secondary haematoma
formation in these patients is variably reported, but is up to 6% of those treated with IV tPA.
The importance of haemorrhagic transformation is that it is associated with a much poorer
outcome. This is particularly troublesome in the context of active reperfusion therapy
as in the process of trying to improve outcome one can inadvertently make the patient much
worse. As a result a great deal of research into acute stroke management has been focused
on trying to accurately select the subgroup of patients who will get the most benefit
form therapy and are at the lowest risk of haemorrhagic transformation. We will cover
this in some detail in our final episode in this series. Secondary haematomas tend to
occur spontaneously within the first 4 days after infarction (but are rare in the first
6 hours) and typically within a day after reperfusion therapy. In most cases the diagnosis
is obvious: the fact that the patient has had a previous ischaemic stroke is either
written on the request card or is evident form recent CT scans as is the case here where
a day earlier the patient had a hyperdense MCA sign, a distal M1 occlusion and perfusion
defect. Some difficulty can occur when the first study is obtained some time after onset
of symptoms, at which time haemorrhage is already present. If this was the first CT
you had available to you, one needs to decide whether this represents a lobar haemorrhage
or haemorrhagic transformation. Although this is a straight forward example, it illustrates
many features typical of secondary hematomas. Firstly there is nothing particular about
the haematomas themselves although usually they are multifocal. The key to the diagnosis
is the surrounding brain. In most cases one can see evidence of an established but non
haemorrhagic component, which conforms to an expected vascular territory. The cortex
is involved as well as the white matter indicating cytotoxic rather than vasogenic oedema. So
looking at our three cases, you can see that case A has these features. There is a region
of cytotoxic oedema involving most of the MCA territory but sparing the PCA territory.
Haemorrhage accounts for only part of the abnormality. Case B has large haematoma with
only a small amount of vasogenic edema surrounding it. It also seems to involved both the MCA
and ACA territory but no abnormality of the rest of the MCA territory is visible. This
represents a primary lobar haemorrhage. Case C Is a small circumscribed mass with only
some vasogenic oedema in the adjacent white matter. The adjacent cortex is normal in appearance.
This represents a hemorrhagic metastasis. You can of course find out much more about
haemorrhagic transformation on Radiopaedia.org.