A trisomy of the distal long arm of chromosome 15(q21 leads to qter) resulting in similar phenotypic and developmental abnormalities in two related children (a boy and a girl) is described. The chromosome defect was due to malsegregation of a balanced translocation (5;15)(p14;q21) in one of the parents. It was inherited in four generations and accompanied by recurrent miscarriages. Comparison of these patients with four previously published cases of trisomy 15q dist reveals a pattern of common features including: microdolichocephaly with characteristic strikingly protuberant occiput and predominance of the visceral over the cerebral cranium; peculiar facial dysmorphism--narrow antimongoloid palpebral fissures; large, malformed, low-set ears; micrognathy; long philtrum; short neck; cardiopathy; profound encephalopathy with lack of suck and swallow reflexes; and no growth retardation.

A 2-year-old girl with a prabable trisomy-22 translocation is described. The principal clinical symptoms described by the authors who have reported cases with proved trisomy 22 are presented. A probable 46,XX,-21,+t(21q;22q) karyotype was established in the patient. The proband's clinical picture is compared with other trisomy 22 cases described in the literature. The incidence of this trisomy among the human population is discussed.

There is increasing evidence that, similar to what is found with other genetic disorders, genomic instability is one of the most general features of cancer. Different forms of manifestation including latent instability have been suggested. To recognize latent chromosomal instability we treated lymphocyte cultures of cancer patients and healthy persons with caffeine, two different doses of bleomycin, and a combination of bleomycin and caffeine. The preliminary results demonstrate that, although the rate of spontaneous chromosomal aberrations is similar in both investigated groups, the lymphocytes of cancer patients display an increased susceptibility to treatment with bleomycin and caffeine. In distinguishing between healthy individuals and those with malignancy, treatment with 30 micrograms/ml bleomycin appears to be most important. Values of chromosomal change above one break per cell, more than 45% cells with chromosomal alterations, and more that two cells with chromosomal rearrangements are suggestive of malignancy. These findings imply that treatment of lymphocyte cultures with bleomycin and caffeine could be a useful assay for monitoring chromosomal instability, and thus detecting a predisposition to malignant disease. In this respect further investigations on a greater amount of material should be performed.

A four-month-old girl with facial dysmorphism, moderate mental retardation, immune deficiency (decreased IgG and IgA and absence of IgM), centromenc instability of chromosomes 1, 9, 16 and very rarely of chromosome 2, and disposition to formation of multibranched chromosomal figures, is described. The case is the fifth described with such chromosomal and immune abnormalities, which prove the existence of a new syndrome. The authors suggest an autosomal recessive inheritance.

The extensive literature on chromosome studies in cancer patients indicates that chromosomal instability is a basic genetic event that explains diverse characteristics of neoplasias, such as predisposition or clonal evolution (Nowel 1986). It is also known that the results obtained with the same protocol can vary greatly and depend, among other things, even on seasonal differences in the break frequency (Mattei et al. 1979).The data provided by the authors of the Letter to the Editor, Toncheva and Nacheva, are an example of this. In two successive evaluations, interestingly denoted as "our study" and "the previous one", they note differences in the chromosomal/chromatid breaks in cancer patients of more than 10% and state that the results of the second study are in agreement with those published by other authors and "contradict" our observations. It should be mentioned that the second "more reliable" study is performed on G-banded chromosomes, notwithstanding the fact that RV staining has previously revealed a higher percentage of breaks and that the results are statistically highly significant (P<0.001). More questions have to be addressed to the selection and heterogeneity of the material and the lack of information about histological type, clinical stage (TNM classification) or the relatively young age of the patientsall below 50 years. As an example, the peak age at presentation of colorectal cancer is around 65 years and of urinary bladder carcinoma between 59 and 65 years. A younger age generally implies the involvement of additional genetic factors in tumorigenesis and thus increased chromosomal instability. In this respect, our attention is attracted by the group of familial adenomatous polyposis colli (APC) patients in which, at the average age of 39 years, only tubular and tubulovillous adenomas are observed, instead of the expected progression towards villous adenomas, dysplasia and carcinomas. The diagnosis of LETTER TO THE EDITOR