Retinoblastoma: Caveat or coincidence?

A retinoblastoma is a rare malignant eye-tumor of the retina in young children, especially in the first years of life, and is hereditary in 40 % of the cases. It occurs approximately in 1 in 17.000 children, worldwide. The National Dutch center for retinoblastomas is part of the Free University of Amsterdam, in the past called 'Academic Hospital', currently- in line with the other academic hospitals- called 'University Medical Center' (VUMC). Almost all children in the Netherlands with retinoblastoma are ultimately treated there and, therefore, the center has a virtually complete database, containing data onwards from 1945.

Picture: Retinoblastoma, visible at fundoscopy. It is located in the 'yellow spot' in the center of the retina. On the upper left one can see the 'blind spot', the exit of the optic nerve (photo VUMC) Double click for better picture (36 kb)

All families involved are asked to fill in an extensive medical questionnaire, in which recently questions have been added about the use of assisted procreation such as IVF, ICSI and IUI. Nowadays more than 3600 children are born every year in Holland following IVF, and this figure approaches 2% of all babies born in this country. It turned out that over the last 1.5 years 5 IVF children were diagnosed with retinoblastoma. These children had a different age of the time of diagnosis- they were born between 1997 and 2002, and therefore the data were reanalyzed from 1995 onwards. Dr Annette Moll from the VUMC calculated that the risk for retinoblastoma could potentially be 5 to 7 times higher and this finding was subsequently submitted to, and accepted by, the Lancet as a research letter. The Lancet, however, as on other occasions insisted on a press release on the day before the publication without letting time pass by to investigate the real value of these data.

It is obvious that these calculations need to be checked critically, from both a much larger population and in an international setting. Only then can- if any- conclusions be drawn regarding the validity (as happened before with a so called increased incidence in certain heart defects) or significance of this tumor in relation to IVF. A possible correlation between these two points must be thoroughly analyzed, especially both conditions may coincide, as certain types of infertility as well as of retinoblastoma can be inherited or family related.

In itself it is of the utmost importance that this type of information is being published in the scientific literature. This case, however, represents a typical example of: 'I found a couple of IVF children with a certain kind of malformation, let us see whether the difference of the incidence between babies after IVF and natural conception is statistically significant' The calculation was based (as usual) on the SIR (Standard Incidence ratio) of the number of cases observed in relation to the number of cases expected, with a confidence interval (CI) of 95 %. Thus analyzed the Relative Risk was 4.9 with 95 % CI being 1.6-11.3. As it does not overlap the 1, the difference is considered statistically significant. The problem is, however, that applying this method to 10.000 illnesses, diseases and conditions, there will be a statistically significant difference in 500 cases, solely on the basis of chance (coincidence) and not on a real correlation.

Furthermore in the Omega study, which examined the health of all IVF women treated between the start of IVF in Holland in 1984 and 1995, additional questions were asked about the health of the IVF babies, and not a single case of retinoblastoma was found. There were 9579 IVF children with an incidence of child cancer of 0.07 % (7 cases) in comparison to an incidence of 0,12 % (9 in 7521 control children) after natural conception. One might have envisaged three cases when the risk would have been 5 to 7 times higher. Although the primary focus of this survey was on the women themselves and not on their IVF children, it is highly unlikely that a mother would have 'forgotten' to mention the presence of a retinoblastoma in the child. Because the two time periods nicely join together one can safely say that during the entire period of IVF there were five children with retinoblastoma. Analyzed as such the RR becomes 2.84, with a 95 % confidence interval of 1.18 tot 6.84.

In the mean time a number of other investigators have investigated a possible link, but failed to do so: Karl Nygren has studied the offspring of IVF, 16000 children in total. In his population there was not a single case of retinoblastoma, whilst one might have expected 5 to 7 cases had the contention of the Free University been true. Also the retinoblastoma center in New York has looked at their database but failed to establish a link.

In the early IVF years the same conclusions were drawn about certain congenital heart defects. A pediatric cardiologist had seen a couple of IVF children with a heart defect in a short time span, and suggested a possible correlation with IVF. At the time, this resulted in quite some publicity, but after a larger follow-up this purported difference had vanished, and the condition appeared to occur as often in IVF as after natural conception. This finding, however, is much less newsworthy and therefore never showed up in the lay press

For now we have to wait until larger studies either confirm or refute these suggestions. All parties agree that a careful and complete on-going follow-up of IVF children is imperative to detect possible risks in a timely manner.

Anteby I, Cohen E, Anteby E, BenEzra D. Ocular manifestations in children born after in vitro fertilization. Arch Ophthalmol. 2001;119: 1525- 9.

Moll AC, Imhof SM, Cruysberg JRM, Schouten- van Meeteren AYN, Boers M, van Leeuwen FE. Incidence of retinoblastoma in children born after in-vitro fertilisation. Lancet, 2003; 361- 309-10.

BenEzra D, Commentary: in-vitro fertilization and retinoblastoma, Lancet 2003, 361; 310

Press release Lancet, 24 jan 2003

Persbericht VU, 24 jan 2003

Klip H, Burger CW, de Kraker J, van Leeuwen FE. Risk of cancer in the offspring of women who underwent ovarian stimulation for IVF.Hum Reprod. 2001;16: 2451- 8.

Nygren, K data in preparation 2004

CAM Jansen MD, PhD,  Voorburg, Feb 2, 2003

Please send mail to keesj@rdgg.nl with questions or comments about this Web- Site

Disclaimer:This information is not intended as a substitute for medical advice of physicians. The reader should regularly consult a physician in matters relating to his or her health and particularly with respect to any symptoms that may require diagnosis or medical attention.

© Stichting Medische Voortplanting Voorburg. This material is copyright protected; improper or unauthorized use is an infringement of copyright-laws and is an actionable offense. Original information from this Web-site can only be used if the source is clearly cited.