Focal nodular and diffuse haematopoietic marrow hyperplasia in patients with underlying malignancies: a radiological mimic of malignancy in need of recognition.

AIM: To report the authors' experience of focal nodular haematopoietic marrow hyperplasia (FNHMH) and diffuse haematopoietic marrow hyperplasia (DHMH) clinically masquerading as skip, distant, or disseminated metastasis in seven patients with underlying malignant neoplasms. MATERIALS AND METHODS: Five patients with FNHMH and two with DHMH mistaken radiologically as skip and disseminated metastasis, respectively, were compared and contrasted with four patients with osteosarcomas and two with chondrosarcomas harbouring skip metastasis, noting the temporal relationship with their haematological profile. RESULTS: FNHMH and DHMH were undetectable by plain radiography and computed tomography (CT) except one showing subtle sclerosis on CT. They showed either isointense or hyperintense, but not hypointense, attenuation at T1-weighted imaging, and all showed hyperintense attenuation at T2-weighted MRI relative to skeletal muscle. Of the five patients who underwent bone scintigraphy, one showed mildly increased uptake, and one out of two showed markedly increased 2-[18F]-fluoro-2-deoxy-d-glucose (FDG)-positron-emission tomography (PET) uptake. The rates for sarcoma skip metastasis by plain radiography, CT, MRI, and bone scintigraphy were 40%, 66.7%, 100%, and 66.7%, respectively. At MRI, 60% showed hypointense and 40% isointense attenuation at T1-weighted, 80% hyperintense and 20% hypointense attenuation at T2-weighted imaging. Combined FDG-PET and CT, which was performed in only one patient, failed to show the skip metastasis. Not every patient with FNHMH or DHMH received granulocyte colony-stimulating factor (GCSF), but all had low or falling haemoglobin levels, which may thus be the prime cause for HMH. CONCLUSIONS: Due to overlapping radiological features, FNHMH and DHMH are great radiological mimics of malignancy. In some cases, needle biopsy is required for their definitive differentiation.

Assessing the evidence for an association between circumferential tumour clearance and local recurrence after resection of rectal cancer.

OBJECTIVE: Circumferential resection margin involvement (CRMI) after resection of rectal cancer is regarded as a risk factor for local recurrence. We have been able to identify only nine peer-reviewed English-language publications which focus primarily on this association, and they report widely differing rates of local recurrence. The aims of this study were to review possible reasons for this variability and to assess the evidence for the micrometrically measured threshold defining CRMI. METHOD: Methodological and statistical evaluation of relevant literature. RESULTS: Several factors which could account for this variability are discussed including the nature of the patient series, surgical technique, curative vs palliative resections, pathology technique, the definition of CRMI, adjuvant therapy, tumour stage, definition and ascertainment of local recurrence, length of follow-up and method of analysis. The objective evidence for the conventional definition of CRMI as tumour 1 mm or less from a circumferential margin is considered along with the evidence supporting a recent proposal that the margin be extended to 2 mm or less. The evidence is numerically weak in both cases and we believe that neither definition should be set in concrete at this stage. CONCLUSION: Pending further research, we recommend that routine pathology reports should record frank tumour transection, if present, or otherwise report the histological width of the margin between the tumour and the nearest circumferential line of resection in millimetres. The definition of CRMI should be simply histological evidence of tumour in a line of resection, that is, a margin of 0 mm. The definition of CRMI as a margin of