an internal silicone rubber disc that abuts against the gastric mucosa to keep it in position. Other gastrostomy tubes have larger balloons or buttons and are therefore less likely to become epithelialised. When placing PEGs care should be taken not to pull the internal flange too tightly against the gastric mucosa as this may also contribute to epithelialisation and subsequent blockage. We now recommend that gastrostomies of this type should be assessed by a gastroenterologist every 6 months, and if blockages occur frequently and are not easily cleared by flushing, replacement of the PEG should be considered prophylactically rather than when complete blockage has occurred.
Lipscomb, C M Brown, T Wardle, W D W Rees
Department of Medicine, University of Manchester School of Medicine, Hope Hospital, Salford M6 8HD, UK
Hull MA, Rawlings J, Murray FE, et al. Audit of outcome of long-term enteral nutrition by percutaneous endoscopic gastrostomy. Lancet 1993; 341: 869-71. Moran BJ, Frost RA. Percutaneous endoscopic gastrostomy in 41 patients: indications and clinical outcome. J R Soc Med 1992; 85: 320-21.
Early and rapid prenatal exclusion of Down’s syndrome SIR—We excluded fetal Down’s syndrome early in pregnancy with an overnight assay that combines sampling of amniotic fluid by a filter technique1 and fluorescence in-situ hybridisation (FISH) with a large DNA probe specific for a unique sequence on chromosome 21.2 Filtration and reinjection of amniotic fluid permits amniocentesis to be done early in pregnancy when the fluid volume is small and the cell number is low. FISH with a chromosome-specific probe renders a small area of a specific chromosome visible in interphase cell nuclei. FISH with a chromosome-21-specific probe on a cell sample from a person with a normal chromosome complement and on a sample from a person with Down’s syndrome results in most nuclei having 2 and 3 signals,
The FISH-signal profile of the disomic samples obtained by filter or standard sampling was similar-67-100% of the nuclei showed two signals whereas only 0-33 % of the nuclei had three signals (table). By contrast, the trisomic samples obtained by standard sampling showed a distinctly different profile, with fewer nuclei with two signals than with three signals. We avoided the previously described uncoupling between hybridisation signal number and chromosome number due to heteromorphisms of the repetitive centromeric areas on chromosome 213-5 by using a unique sequence probe rather than a centromeric sequence. We succeeded in doing FISH and standard cytogenetic techniques in parallel on amniotic fluid samples obtained by filter or standard amniocentesis. The results by FISH on samples obtained by either amniocentesis technique were similar. We were unable to find a trisomy 21 sample early in gestation during the consecutive sampling. However, it should be possible not only to exclude but also to detect Down’s syndrome overnight by FISH on samples of amniotic fluid obtained early in pregnancy by the filter technique.
Bryndorf, Karin Sundberg, Britta Christensen, John Philip, Kathy Yokobata, Candy Gaiser Thue
Chromosome Laboratory, Section 4051, Section of Clinical Genetics, Department of Obstetrics and Gynaecology, Rigshospitalet, 2100 Copenhagen, Denmark; and Becton Dickinson Immunocytometry Systems, San Jose, California, USA
Sundberg K, Smidt-Jensen S, Lundsteen C, Agerbaek K, Philip J. Filtration and recirculation of early amniotic fluid: evaluation of cell cultures from 100 diagnostic cases. Prenat Diagn 1993; 13: 1101-10. Bryndorf T, Christensen B, Xiang Y, et al. Fluorescence in situ hybridization with a chromosome 21 specific cosmid contig: 1 day detection of trisomy 21 in uncultured mesenchymal chorionic villus cells. Prenat Diagn 1994; 14: 87-96. Verma RS. Luke S. Variations in alphoid DNA sequences escape detection of aneuploidy at interphase by FISH technique. Genomics 1992; 14: 113-16. Weier H-UG, Gray JW. A degenerate alpha satellite probe, detecting a centromeric deletion on chromosome 21 in an apparently normal human male, shows limitations of the use satellite DNA probes for interphase ploidy analysis. Analyt Cell Pathol 1992; 4: 81-86. Seres-Santamaria A, Catala V, Cuatrecasas E, Villanueva R. Fluorescent in-situ hybridisation and Down’s syndrome. Lancet 1993; 341: 1544.
respectively. 10 consecutive samples of amniotic fluid cells were obtained by the filter method from pregnancies ranging from 11 to 13 weeks’ gestation (median 12). As controls we obtained 210 consecutive amniotic-fluid samples by standard means from pregnancies ranging from 13 to 33 weeks’ gestation (median 16). Given the incidence of Down’s syndrome occurring in the screened population, a large number of controls was necessary to acquire a satisfactory number of samples from trisomy 21 fetuses. All samples were analysed by FISH and standard cytogenetic methods in a blinded fashion, with approximately one-quarter of each sample used for FISH.
Non-Invasive prenatal diagnosis encouraged to see our method for non-invasive prenatal diagnosisl used to determine the sex of fetuses as early as 4 weeks 5 days (Thomas and colleagues, Feb 12, p 413). This method, using nested polymerase chain reaction (PCR), has been extensively investigated by many groups. Results from over 600 cases2 gave an accuracy of 80%. The problems of false-positives and false-negatives are well documented, and clearly the method needs improvement before it can be routinely used for early sex determination or other prenatal genetic diagnoses. We now present an improved methodology for sex determination, as a further step towards that goal. This SiR-We were
method is faster, less labour intensive and more robust previous methods with significant anti-contamination measures incorporated. The key components are the prevention of pre-PCR mispriming with the "hot start" process,3 dUTP incorporation, and uracil N-glycosylase (UNG) sterilisation.4 Nested PCR is very sensitive but there is a risk of introducing contamination between the two rounds of amplification. Hot Start PCR allows us to achieve single molecule sensitivity without nested PCR. The new system can detect a single fetal cell even when the feto-maternal ratio is as low as 1 in 106. The anti-contamination component can sterilise up to 20 000 molecules of carryover PCR product and is much more efficient than restriction enzyme digestion.1
(range) per sample. Table: FISH results from amnlocyte samples obtained method and by ordinary sampling 802
Figure: Prenatal sex determination from maternal peripheral blood using PCR to the TSPY locus’
Table : PGA score In three groups of alcoholic according to their liver status
PGA score In diagnosis of alcoholic fibrosis
(table). No difference was seen between scores in mild moderate alcoholic hepatitis and fatty liver. Receiveroperating characteristics (ROC) curves were calculated. In our hands, the optimum cut-off was 6, the sensitivity and specificity, respectively, of PGA score for detection of alcoholic cirrhosis being 85% and 74% for the whole group and 64% and 74% when the Child B and C cirrhosis groups were excluded. In the 32 patients (28 non-cirrhotics and 4 cirrhotics) in whom the PGA score was measured again one week after alcohol withdrawal, there was no significant variation in the overall mean PGA score (5-8 [2.8] at day 0 vs 5.9 [1.8] at day 7). However, the PGA score increased by one unit in 14 patients, two units in 8, three units in 2, and four units in 1, and fell by one unit in 1. With the proposed cut-off of 6,15 non-cirrhotics (15/28,54%) would have been regarded as cirrhotics if the PGA score had been measured at day 7, versus 8 (29%) when measured at admission. Our results suggest that the clinical usefulness of PGA score is limited. First, in patients presumed to be non-cirrhotic, the individual PGA score does not seem to help to determine the type or severity of underlying disease; in such cases liver biopsy remains the best diagnostic tool. Second, the ability of PGA score to identify cirrhosis depends on the cut-off value. When the score was 2 or less, the probability of cirrhosis was zero; conversely, all patients with scores of 9 or more had cirrhosis, which was complicated in 11 of 12 cases. However, complicated cirrhosis is easy to diagnose clinically, and PGA would be most useful for the early detection of cirrhosis. From the ROC curves, and at variance with the value of 3 suggested by Teare and colleagues, the optimum cut-off was 6, with fairly low sensitivity and specificity. Third, when the PGA score is measured is important, since individual variations after alcohol withdrawal might greatly modify the diagnostic value of the test. In our study, the score increased one week after alcohol withdrawal by at least one unit in about three-quarters of patients. Consequently, with a cut-off of 6, the false-positive rate (36% at admission) rose to 50% one week later. Contrary to other workerswe suggest that the PGA score should not be used after alcohol withdrawal.
SiR-Teare and co-workers’
J J Jiang, M Salvucci, V Thepot, S Pol, O G Ekindjian, B Nalpas
Lane 1 = fetal sex M (gestation 8 wk); lane 2 = M (10 wk); lane 3 =F (10 wk); lane 4 = M (10 wk); lane 5 =F (12 wk); lane 6 = M (14 wk). None of the 14 blanks used was positive. Lane M = pBR322 DNA digested with Haelll (marker). Results shown after 60 cycles of Hot Start PCR (94°C 1 min, 63°C 1 min, 72°C 1 min) using primers TSPY1 and TSPY2L. All reactions were treated with 1 unit of UNG before DNA amplification. TSPY1 =5’ GCCAATGTTGTATCCTTCTCAGTG 3’ TSPY2L=5’ GTTrCTCTGCCGCATGCAGGACA 3’
The use of this sensitive system for prenatal sex determination from maternal blood is shown in the figure. We hope that these improvements in amplification technology will complement those in other areas such as fetal cell enrichment and fluorescence in situ-hybridisation, to allow the early clinical usage of non-invasive prenatal diagnosis. YMDL is supported by a career development fellowship from the Wellcome Trust.
Y-M D Lo, J S Wainscoat, K A Fleming Nuffield Department of Clinical Medicine, Department of Haematology, and Nuffield Department of Pathology, John Radcliffe Hospital, Oxford OX3 9DU, UK
1 Lo YMD, Patel P, Wainscoat JS, Sampietro M, Gillmer MDG, Fleming KA. Prenatal sex determination by DNA amplification from maternal peripheral blood. Lancet 1989; ii: 1363-65. 2 Simpson JL, Elias S. Isolating fetal cells from maternal blood: advances in prenatal diagnosis through molecular technology. JAMA 1993; 270: 2357-61. 3 Chou Q, Russell M, Birch DE, Raymond J, Bloch W. Prevention of pre-PCR mis-priming and primer dimerization improves low-copynumber amplifications. Nucl Acids Res 1992; 20: 1717-23. 4 Longo MC, Berninger MS, Hartley JL. Use of uracil DNA glycosylase to control carry-over contamination in polymerase chain reactions. Gene 1990; 93: 125-28. 5 Arnemann J, Jakubiczka S, Turing S, Schmidtke J. Cloning and sequence analysis of a human Y-chromosome-derived, testicular cDNA, TSPY. Genomics 1991; 11: 108-14.
that the PGA index
(combination of apolipoprotein A[apoAl] measurement with
gamma-glutamyl transpeptidase activity and prothrombin time) is a sensitive and specific test for the detection of fibrosis in alcoholic liver disease. Our experience is different. We investigated 61 alcoholic patients (48 men and 13 women, aged 29-70 years, mean [SD] 48-0 [10 9]). The mean duration of alcohol abuse was 23-7 (12-0) years and the quantity ingested daily was 198 (93) g. On the basis of histological findings (in 35 cases) and clinical, laboratory, and ultrasonographic results, the patients were classified as: mild to moderate alcoholic hepatitis (16), fatty liver (18), and Child-Pugh A, B, or C cirrhosis (A 11, B 8, C 8). Serum apoAlconcentrations were measured by a kit (Tina-Quant, Boehringer Mannheim, Mannheim, Germany). The three PGA score indices were tested at admission in every case, and one week after alcohol withdrawal in 32. The PGA score was calculated according to Poynard et al.2 The distribution of PGA scores differed between the three
d’Hépatologie, Hôpital Necker, 75747 Paris, France; Laboratoire de Biochimie, Hôpital Laennec, Paris; and INSERM U-99
Teare JP, Sherman D, Greenfield SM, et al. Comparison of serum procollagen III peptide concentrations and PGA index for assessment of hepatic fibrosis. Lancet 1993; 342: 891-94. Poynard T, Aubert A, Bedossa P, et al. A simple biological index for detection of alcoholic liver disease in drinkers. Gastroenterology 1991; 100: 1397-402.
Filgrastim combined with tretinoin In acute promyelocytic leukaemia SIR—Nakajima and colleagues (Jan 15, p 173) describe the successful treatment of a case of acute promyelocytic leukaemia (APL) by tretinoin all-trans retinoic acid in combination with granulocyte-colony stimulating factor (G-CSF). We present a similar case. 803