749

reduction in child smokers after introducing its advertisingban in 1975. As the Coronary Prevention Group sets out inthe revised text of its Case Against Tobacco Advertising,which will be published next week, non-binding voluntaryagreements with the tobacco industry may control certainelements of advertising style but do little to reduce theamount or effectiveness of advertising.The news from Brussels, where officials met again last

week, was that Germany was beginning to wobble. Only oneof the four states opposing the ban has to bend to allow thenew controls to come into force. Voting is based on

population. There are 76 votes in all, and 54 are needed toget it through. The EC health ministers, who discussed thedraft at their spring get-together, meet again in November toratify it. Britain is probably the most hardline opponent ofthe move. The presidency at present is held by the Dutch,but if the decision is delayed beyond November thenPortugal, which has already banned advertising at home,will be in the chair. Meanwhile a new organisation, Doctorsfor Tobacco Law, has emerged with the backing of 26medical organisations including the Royal Colleges. It isalready claiming to be the largest medical coalition in theUK. One of its leading members, Prof John Moxham ofKing’s College Hospital, London, declared: "The cure fortobacco disease cannot be found in doctors’ surgeries-it liesin MPs’ surgeries". The coalition is sending MPs the grislydetails of the cancer and coronary deaths caused by tobacco.

Malcolm Dean

Round the World

USA: Problems with power lines

Scientists at the Environmental Protection Agency whoare reviewing the health effects of electromagnetic fields(EMF) from power lines and other sources at first wanted tocall the radiation a "probable carcinogen". Overruled bytheir superiors, they toned it down in the published draft to a"possible but not proven cause of cancer in people".Currently, non-government scientists on a panel of theEPA’s science advisory board advise even greater caution.They believe the final report should say: "There is limitedinformation on the potential carcinogenicity of EMFs inhumans".Given the high standing within the EPA of its science

advisory board, the authors are bound to take thisrecommendation seriously when they rewrite the draft.Such a conclusion should come close to meeting theobjections of D. Allan Bromley, President Bush’s scienceadviser. He has strongly opposed even the suggestion of alink between EMF and cancer (Lancet 1991; 337: 544).Robert Adair, his fellow physicist at Yale University, claimsthat "An examination of the physical interaction of [EMF]with the body shows that such interactions are too weak tohave a significant effect on human biology at the cell level"(Physical Review A, Jan 15, 1991, p 1039).That judgment has made an impression on

epidemiologist Genevieve Matanoski, of Johns HopkinsUniversity, chair of the EPA panel which is reviewing the1990 draft of the report, called Evaluation of the PotentialCarcinogenicity of Electromagnetic Fields. Dr Matanoski, theprincipal author of an EMF study that found unexpectedly

high cancer rates among linemen for the New York

Telephone Co, says that the authors of the EPA report needto come to grips with Dr Adair’s objections. "The theory ofphysics", she told me, "shows none of this can happen". Butthe panel also said that cancer incidence could be affected byan agent that does not produce mutations, citing as examplesthe influence of hormonal imbalance and nutrition on cancer

promotion.The Matanoski panel also had reservations about the

evidence on which reports of links between EMF and cancerare based. Here the principal objection was to the absence ofquantified exposures to EMF. Another drawback was saidto be the "limited understanding of possible biologicalmechanisms" that could cause cancer. Still, said the panel,"the evidence cannot be dismissed". The whole question ofEMF effects, it said "is important and exceptionallychallenging".

It could take at least a year for the EPA to rewrite and

publish this document. Aside from wanting to make sure thescience is accurately presented and interpreted, it is clearthat President Bush’s advisers and the science advisoryboard panel also hope to avoid any language that would, asthe Matanoski panel expressed it, "reinforce the skewed andsomewhat sensationalized picture presented to the public inrecent years by the news media".

J. B. Sibbison

Italy: An end to the fun and games?

Medical congresses have widely served to spreadscientific information while at the same time "rewarding"loyal physicians for their attachment to a givenpharmaceutical company. These gifts to physicians may beextremely varied and may include trips to exotic places forthe whole family, cruises after a few lectures, gala dinners,evenings on the town, and expensive presents. The wholeprocess comes under the broad umbrella of "scientificinformation" needed to keep physicians abreast of theprogress of medicine.The Italian Ministry of Health has now decided to put a

stop to these extravaganzas, presumably with the agreementof the Association of the Pharmaceutical Industry(Farmindustria), whose members are alarmed by the

escalating costs of scientific meetings. On Dec 12, 1990, theMinistry issued a decree, which became effective in March,1991, that scientific congresses organised in Italy or abroadby pharmaceutical companies based in Italy or financiallysupported by them, even indirectly, must be free fromadvertising and must aim only at the improvement ofknowledge in medical disciplines. When a company wishesto organise a scientific meeting it must apply to the Ministryof Health at least 60 days in advance, giving informationabout where the meeting will take place, the type ofparticipants, the topic of the meeting, the speakers’qualifications, and an undertaking to avoid any form ofadvertising. In addition the company must give a detailedbudget, which should not include any expenses for traveland lodging for the participants, except the speakers andsome poorly defined "qualified professionals of the

discipline of the meeting, useful for the success of themeeting". Advertising, distribution of free samples, anddisplays of promotional exhibits during the meeting areforbidden. A decree in preparation will probably allow firmsto support meetings without requiring any authorisation if

750

the sum involved does not exceed 10 million lire (aboutL4000) or if the congress is a regular yearly meeting of ascientific biomedical society recognised by the Ministry ofHealth.The Ministry of Health will reply to an application within

30 days and will enforce respect of the rules. If the firm failsto comply with the ministerial directives the advertisedpharmaceutical product can be removed from theProntuario Terapeutico Nazionale, J the formulary of

products admitted for total or partial reimbursement by thehealth service. This form of punishment is not altogetherlogical, because if it involves an essential drug the

punishment may be more harmful to patients than to thecompany concerned. It is of interest that these limitations on

participation of pharmaceutical companies at scientific

meetings do not extend to manufacturers of diagnostic andscientific equipment or medical supplies.

It is too early even to guess the impact of this decree onmedical congresses, but several international medical

meetings that should have taken place in Italy have alreadybeen cancelled.There has been no formal protest from the

pharmaceutical industry, which has apparently accepted thedecree, but the organisers of scientific meetings haveexpressed considerable disappointment.There is no doubt that it was due time to stop this

particular oil slick spreading, but whether this was the bestway of doing it is debatable. It is highly doubtful whether yetanother bureaucratic procedure can be constructively addedto the already inefficient public administration, and it is easyto foresee that uniform decisions will be difficult to

ensure-there have already been disagreements about thedifferences between a scientific congress and an advanced

training meeting.Since the prices of pharmaceutical products in Italy are

established by the State and include a percentage for

advertising; it would have perhaps been simpler to cut thispercentage and therefore to reduce the prices. The Statewould have saved money on bureaucracy and the purchaseof drugs, and the industry would have been forced to reduceits advertising. However, only time will tell whether thedecree was a wise move.

Silvio Garattini

Conference

Early amniocentesis

Two large randomised controlled trials comparingconventional mid-trimester amniocentesis with first-trimester chorion villus sampling (CVS) have shown thatCVS carries a small but significant excess of fetal loss, ahigher rate of false-positive diagnoses, and a higherlaboratory failure rate (Lancet 1989; 337: 1-6 for Canadiantrial and Lancet 1991; 337: 1491 for MRC European trial).Limb deformities in children born after CVS done before 10weeks are also a reason for concern about the procedure(Lancet 1991;337:762-63, 1038-39, 1091-92, 1422-24). Isearly amniocentesis (EA) thus worth pursuing as a possible,perhaps superior, alternative to CVS, and if so what issuesneed to be clarified during the investigation? These

questions were discussed at a workshop on earlyamniocentesis organised by Prof D. V. Coleman and held inJuly at St Mary’s Hospital, London.

Experience of early amniocentesis (defined for the purpose of themeeting as before 15 weeks’ gestation) has already been

accumulating. Published work both in Europe (reviewed by N. C.Nevin, Belfast) and North America (reviewed by L. Godmillow,Pennsylvania), and experience at King’s College Hospital, London(K. Nicolaides), with most series containing 100-500 cases, indicatefetal loss rates before 28 weeks similar to those obtained withconventional mid-trimester amniocentesis. One European series of395 cases (Stripparo et al, 1990) had a fetal loss rate of 3’ 1 %, but thatincluded 55 cases done before 12 weeks’ gestation and for which theloss rate was 14-8%. At the Pennsylvania Hospital, USA,enthusiasm for EA has waned somewhat because of concern that theloss rate before 28 weeks among 1000 cases done at 11-14 weeks’

gestation was 3-4%, compared with 1-0% for mid-trimesteramniocentesis and 2-6% for CVS. In addition, pregnancy losswithin 2 weeks of the procedure was a little higher for EA (19%)than for CVS (1-3%), loss rate after EA in twin pregnancies wasdiscouraging (with 8 babies lost after sampling of 35 twins), 13 casesof minor orthopaedic deformities were identified among the 1000cases, and there was the impression that leakage of amniotic fluidoccurred in a considerable number of women after EA. In Belfast,where CVS was not introduced, enthusiasm for EA remains.Among 887 done before 15 weeks (275 of these before 12 weeks) thespontaneous abortion rate was 0-8%, with the 7 occurring at 12 (1),13 (2), and 14 (4) weeks. No congenital abnormalities were detectedamong the liveborn babies.

Experience so far has not been sufficient to indicate whether lunghypoplasia will be a serious disadvantage. Mid-trimesteramniocentesis is a risk factor for lung hypoplasia, which in turn ismost probably related to degree of oligohydramnios, although thereis evidence that pressure drop in the amniotic cavity (as occursexperimentally when the cavity is punctured without withdrawal offluid) could by itself affect lung development. Because the

proportion of amniotic fluid withdrawn is greater for EA than formid-pregnancy amniocentesis, lung hypoplasia can be expected tobe commoner with EA.

In Belfast the volume of fluid aspirated at EA has been limited to7 ml, and with the 6 ml sent to the cytogenetic laboratory (1 ml isused for biochemical tests) it has been possible to obtain 98-99%success with cultures for samples obtained before 15 weeks’

gestation (97-5% for gestations under 12 weeks) and to provideresults 11 days after amniocentesis. At King’s College Hospital10 ml of fluid led to successful culture and cytogenetic analysis in98% of cases of EA done between 10 and 13 weeks, whereas in apilot trial culture rates of less than 70% had been obtained with thatvolume for gestations of under 10 weeks.A technique that might reduce the risk of oligohydramnios while

increasing the harvest of cells is amniofiltration, whereby amnioticcells are filtered off as up to about 40 ml fluid is withdrawn and thefiltrate then returned to the amniotic cavity. Early assessment at StMary’s Hospital, London (J. Smith and M. T. Rebello) and at StThomas’ and King’s College Hospitals, London (D. Byrne),indicate that the greater the volume of amniotic fluid filtered, thelarger the number of clones obtained during cell culture, but thereseems to be no great lead in culture time.CVS offers the advantage that sufficient tissue may be obtained

for analysis of direct preparations. Its disadvantage, that mosaicismis commoner than with conventional CVS, is due to the trophoblastcontent of the chorion. The proportion of chorion cells in theamniotic fluid is greater with early than with late pregnancy, but theamount of trophoblastic tissue was thought to be unlikely to bringthe mosaicism rate beyond that obtained with CVS (R. J. Lilford,Leeds), although disagreement on this point was acknowledged. InBelfast and at the Pennsylvania Hospital that rate has been less than1 %; the other European papers reviewed did not give mosaicismrates. The false-negative rate with CVS is no more than thatobtained with amniocentesis if both direct analysis (short-termculture) and long-term cultures are done. Whether theconcentration of chorion cells in fluid obtained at EA will raise the

false-negative rate remains to be seen.