Screening for Fragile X Syndrome (Murray et al.) - NIHR Journals ...

More documents

Recommendations

Info

Modelling allele dynamicsfrequency. This requires the assumption that theFM allele is in a steady state in the populationinsofar as the frequency does not change betweengenerations. The allele only arises in the nextgeneration from a mutated X chromosome in awoman, either an FM or an expanding PM (seeFigure 5). Taking the reproductive fitness of FMcarrier women to be 50%, 1 in 16,000 foetusesin the next generation (50% of half of 1 in 4000)will receive an FM from an FM-carrier mother.For the remaining 1 in 5300 (1 in 4000 minus 1in 16,000), the FM will come from an expandedPM allele. Since 1 in 273 women is a PM carrier,the expansion risk must be 10% (1 in 5300 dividedby half of 1 in 273). This calculation relies on theassumption that a steady state exists, but this maynot be so, particularly if there have been interventionstudies aimed at changing reproductivepractice in PM and FM carriers. An evens segregationratio of mutated to normal alleles is alsoassumed, and this too may be incorrect (Shermanet al, 1994). Nonetheless, the estimated valuesof 1 in 16,000 and 1 in 5300 do fit with theobserved ratio of 1:3.5 for FM:PM carriers inmothers of children with an FM in the largeNew South Wales study (Professor G Turner,personal communication).General population modelFrom the above analysis it is possible to completea model for the general population screeningsituation. One caveat is that a range of values needto be considered because of the uncertainty of theprecise rate of expansion from PM to FM. However,of the two sets of values, those generated from therisk calculated in the previous paragraph are themore plausible. Moreover, this second set of valuesare conservative with regard to the potential ofscreening. In the following chapter we use thelatter, as shown in Figure 8.ParentsFM250FM250PM3660Normal955,840Reproduce1253660 955,840Foetalmutation621830FoetalFM31 3192 92Fragile X15 31 46 92FIGURE 8 General population model42
Health Technology Assessment 1997; Vol. 1: No. 4Chapter 11Assessment of screening potentialUltimately the decision whether or not to introduceany of the screening strategies for fragile Xsyndrome in the UK will depend on a variety of differentfactors. Nonetheless, the starting point of thisdecision-making process must be an assessment of thepotential performance of the screening tests involved(Cuckle & Wald, 1984; Wald & Cuckle, 1989).Measures ofscreening performanceThe most important measures of the performanceof a screening test quantify the ability to distinguishaffected from unaffected individuals. The usualmeasures are the sensitivity or detection rate (proportionof affected individuals with positive results)and the false-positive rate (proportion of unaffectedindividuals with positive results). An alternative wayof expressing the latter is the specificity, which is100% minus the false-positive rate.The purpose of screening is to identify the highrisk group for further action and to reassure theremainder that their risk is low. The predictivevalue of the test quantifies these risks. The positivepredictive value is the probability that an individualwith a positive result is indeed affected and thenegative predictive value is the chance of beingunaffected given that the result is negative. Theseparameters are a function of the prevalence of thedisorder in the population being tested as well asof the sensitivity and specificity of the test itself.To be effective, any screening strategy needsto make an impact on one or more outcomemeasures. In addition to the sensitivity of the test,this will depend on the uptake rate of the screeningtest, the acceptability of the diagnostic and otheroptions offered to those with positive results, andthe effect of these on the outcome being measured.The impact of screening can be assessed both inthe population being targeted and overall.Potential of screening forfragile X syndromeAll the screening strategies we have consideredthat aim at preventing affected births, test for aPM or FM in women. If the test indicates that thewoman has such a mutation, it is a positive result.If the woman has a pregnancy affected with fragileX syndrome it is a true-positive result; otherwise,it is a false-positive. The strategies that are aimedat improving prognosis, test for the FM in malesor females. If the individual tested has fragile Xsyndrome, the test is a true-positive result; otherwiseit is a false-positive result. Although screeningfor fragile X syndrome may have other benefits,effectiveness will be judged by the extent to whichit reduces the birth prevalence of the disorder orimproves prognosis.Antenatal screeningNearly all the affected foetuses can be expectedto have a mother who has a PM or FM. Therefore,testing mothers for these mutations can beregarded as having a detection rate and negativepredictive value close to 100%. False-negativesmay occur because of point mutations, deletionsand technical errors but all of these will be rare.There are three ways in which an antenatalscreening test will yield a false-positive result. Amother with an FM or PM may pass her normalallele to the foetus, a mother with a PM may passit to the foetus but it does not expand to an FM,or a female foetus of such a mother may have anFM but is phenotypically normal. The first two willbe resolved by prenatal diagnosis but, with currenttechnology, the last will not.From Figure 8, antenatal screening in a generalpopulation of 1 million couples will yield 184 truepositivesand 3601 (3785 minus 184) false-positives.This is a false-positive rate of 0.4% and a positivepredictive value of 1 in 20. Such results are considerablybetter than those currently achieved byantenatal screening for Down’s syndrome, whichhas a 5% false-positive rate and a 1 in 50 positivepredictive value (Cuckle, 1996).There are no results from studies of general populationantenatal screening for fragile X syndromefrom which to assess the likely uptake of screeningin the UK. The only published data are from teststhat were not free at the point of entry. The43
Page 1 and 2:
Health Technology Assessment 1997;
Page 5 and 6: Screening for fragile X syndromeJen
Page 7: Health Technology Assessment 1997;
Page 11 and 12: Health Technology Assessment 1997;
Page 38 and 39: Screening and diagnosis26extended t
Page 50 and 51: Modelling allele dynamicsshown in T
Page 52 and 53: Modelling allele dynamicsTABLE 12 R
Page 56 and 57: Assessment of screening potentialac
Page 58 and 59: Human and financial costs of screen
Page 60 and 61: Human and financial costs of screen
Page 62 and 63: Recommendations5. A central registr
show all

Screening for Fragile X Syndrome (Murray et al.) - NIHR Journals ...

Create successful ePaper yourself

Delete template?

Save as template?