Different diagnostic and prognostic groups of colorectal carcinoma (CRC) have been defined. However, accurate diagnosis and prediction of survival are sometimes difficult. Gene expression profiling might improve these classifications and bring new insights into underlying molecular mechanisms. We profiled 50 cancerous and noncancerous colon tissues using DNA microarrrays consisting of B8000 spotted human cDNA. Global hierarchical clustering was to some extent able to distinguish clinically relevant subgroups, normal versus cancer tissues and metastatic versus nonmetastatic tumours. Supervised analyses improved these segregations by identifying sets of genes that discriminated between normal and tumour tissues, tumours associated or not with lymph node invasion or genetic instability, and tumours from the right or left colon. A similar approach identified a gene set that divided patients with significantly different 5-year survival (100% in one group and 40% in the other group; P ¼ 0.005). Discriminator genes were associated with various cellular processes. An immunohistochemical study on 382 tumour and normal samples deposited onto a tissue microarray subsequently validated the upregulation of NM23 in CRC and a downregulation in poor prognosis tumours. These results suggest that microarrays may provide means to improve the classification of CRC, provide new potential targets against carcinogenesis and new diagnostic and/or prognostic markers and therapeutic targets.

The association between cancer and sarcoidosis is controversial. Some epidemiological studies show an increase of the incidence of cancer in patients with sarcoidosis but only few cases of sarcoidosis following cancer treatment have been reported.We conducted a retrospective case study from internal medicine and oncology departments for patients presenting sarcoidosis after solid cancer treatment. We also performed a literature review to search for patients who developed sarcoidosis after solid cancer. We describe the clinical, biological, and radiological characteristics and outcome of these patients.Twelve patients were included in our study. Various cancers were observed with a predominance of breast cancer. Development of sarcoidosis appeared in the 3 years following cancer and was asymptomatic in half of the patients. The disease was frequently identified after a follow-up positron emission tomography computerized tomography evaluation. Various manifestations were observed but all patients presented lymph node involvement. Half of the patients required systemic therapy. With a median follow-up of 73 months, no patient developed cancer relapse. Review of the literature identified 61 other patients for which the characteristics of both solid cancer and sarcoidosis were similar to those observed in our series.This report demonstrates that sarcoidosis must be considered in the differential diagnosis of patients with a history of malignancy who have developed lymphadenopathy or other lesions on positron emission tomography computerized tomography. Histological confirmation of cancer relapse is mandatory in order to avoid unjustified treatments. This association should be consider as a protective factor against cancer relapse.