Recently, we have observed three cases with infarction of regenerative nodules, one patient with no variceal bleeding. Similar cases were reported as single case reports [
1‐
4]. Kim
et al. pointed out that especially in patients with a history of substantial gastrointestinal bleeding, infarcted regenerative nodules must be included in the list of differential diagnoses of focal liver lesions [
3]. Microscopically, in our three cases, the degree of infarction was extensive. Hemorrhage and inflammatory changes were also observed around the infarcted regenerative nodules (Figure
2). In order to differentiate these lesions from malignant tumors, serial imaging seems to be necessary. Normally on unenhanced CT, typical regenerative nodules in cirrhosis are either not visible or appear with higher attenuation than adjacent parenchyma when they contain iron [
3‐
6]. Such nodules are usually not visible on contrast-enhanced CT scans and appear isoattenuating to enhanced surrounding liver parenchyma. Similarly, only the siderotic regenerative nodules are visible as hypointense lesions on T2-weighted MR images [
7‐
9]. In our patients, the lesions of infarcted regenerative nodules were depicted as different-appearing nodular lesions of low attenuation on unenhanced CT and as heterogeneous enhancement with regions of iso- and hypoattenuation relative to the surrounding liver on contrast-enhanced CT scans. In ultrasound, the image is like an abscess. In all of our patients the radiologists suggested liver abscesses. In MR imaging the appearance of such lesions is known to be different from that of regenerating nodules, showing high signal intensity on T2-weighted spin-echo MR images. Thus, these lesions can have different findings in cirrhotic livers and can be mistaken for a malignancy or even an abscess [
10]. There are no data from autopsies and it is not known how often these infarcted regenerative nodules occur. Kang
et al. reported on a vascular situation by regenerative nodules [
1,
2,
5]. Hypotension can lead to a reduction of portal and arterial liver flow. Since variceal bleeding or septic shock can induce hypotension, as observed in our patients, we conclude that this could lead to infarction of nodules.