Search

Paroxysmal nocturnal haemoglobinuria (PNH) cells are deficient in glycosylphosphatidylinositol (GPI) linked antigens due to a somatic mutation of the PIG-A gene in a haemopoietic stem cell. It appears that a PNH clone reaches detectable proportions only when there is selection in its favour. GPI-deficient T lymphocytes have been identified in patients treated with CAMPATH-1H, a monoclonal antibody against the GPI-linked CD52 molecule. CAMPATH-1H selects for cells that are deficient in CD52 (such as PNH-like cells) promoting the development of a PNH-like clone (analogous to PNH). We report that 10/15 patients with chronic lymphocytic leukaemia developed PNH-like lymphocytes after therapy with CAMPATH-1H. The remaining five patients developed no PNH-like cells at any stage, including one patient who received 12 weeks of therapy. The inactivating PIG-A mutation has been identified in one patient. This mutation was detectable by an extremely sensitive mutation-specific PCR-based analysis in the patient's mononuclear cells prior to CAMPATH-1H therapy. The frequency and phenotype of GPI-deficient lymphocytes after CAMPATH-1H and the detection of a PIG-A mutation in the lymphocytes prior to CAMPATH-1H therapy indicated that such mutations were present in a very small proportion of cells prior to selection in their favour by CAMPATH-1H. This suggests that a large proportion of individuals have cells with PIG-A mutations that are not detectable by flow cytometry and thus may have the potential to develop PNH.

The general procedure of conventional clustering is modified for various applications involving problem localization. This modification introduces the concept of clustering criteria which are used for partitioning a training set, and depend upon certain a priori information with regards to the training set. Also, the need of a structure or a mathematical form for the partition boundaries arises naturally from the need to process unknown samples. The general procedure is discussed in detail for applications of piecewise linear classifier design and piecewise linear density estimation. Experimental results are presented for both applications.

n/a