You are here

Setting the record straight

Reproductive BioMedicine Online, Volume 33, Issue 6, December 2016, Pages 657 - 658

This editorial comments on a paper from Chen et al. (2016), published in this issue of RBM Online, in an attempt to set the record straight after some shoddy scientific journalism. The paper describes the long-term outcomes of experimental treatments involving ooplasmic transfers conducted between 1996 and 2001 at Saint Barnabas Medical Centre, Livingston, New Jersey, USA (Barritt et al., 2001a). It is timely to report these findings in light of (i) a recent report of the birth of a baby after ‘mitochondrial transfer’, and (ii) because the Human Fertilisation and Embryology Authority (HFEA) in the UK is deciding whether to issue a license for treatment by mitochondrial transfer (

First, the report of the birth of a baby after meiotic metaphase II spindle transfer to alleviate mitochondrial disease came not in a standard journal article, but in the form of an abstract submitted to the recent American Society for Reproductive Medicine (ASRM) meeting (Zhang et al., 2016a). In choosing this route, the authors have unfortunate form – the same authors having used the same approach previously (Zhang et al., 2003) when describing the adverse outcome of a pregnancy involving transfer of embryos generated by the related technique of pronuclear transfer, the full details of which were not published until earlier this year (Zhang et al., 2016b). It is to be hoped that it will not take 13 years for the full details of this case to be made available to the medical community. However, the point at issue is that, although the work described in the paper by Chen et al. (2016) is referred to in press reports on Zhang et al. (2016a), the former differed markedly in intent, methods and outcomes. The intention of cytoplasmic transfer was not to prevent mitochondrial disease or to transfer mitochondria per se, but to make up for a presumptive ‘ooplasmic deficiency’ in women who had experienced repeated poor embryo development and implantation failure. Indeed, only a minority of the offspring (2/8 tested) showed evidence of the presence of donor ooplasmic mitochondria (Barritt et al., 2001b), making the widely used epithet of ‘three parent’ offspring particularly inappropriate for these babies (Cohen, Alikani, 2013 and Johnson, 2013). Likewise, the method used, ooplasmic injection as per ICSI (Cohen et al, 1997 and Cohen et al, 1998), was far less invasive than the wholesale transfer of the second meiotic spindle by a method that is not described in Zhang et al. (2016a), but presumably involved membrane relaxants and cell fusion. But perhaps the major difference, especially in the context of mis-reporting, relates to the outcomes. An early but much-quoted inaccurate report in New Scientist magazine (Hamzelou, 2016) claimed that ‘Last time embryologists tried to create a baby using DNA from three people was in the 1990s, when they injected mitochondrial DNA from a donor into another woman's egg, along with sperm from her partner. Two of the babies developed genetic disorders, and the technique was banned. (The latter sentence was changed the day after first publication online to ‘Two of the fetuses developed genetic disorders, and the technique was halted by the US Food and Drug Administration.) The report further stated that ‘The problem may have arisen from the fetuses having mitochondria from two sources.’ Several of the multiple inaccuracies in this ‘summary’ of the work will be evident immediately, but there is one to which attention must be drawn because, before any corrections were made, it was cited widely in the press. Thus, there is a suggestion that the work was banned by the US Food and Drug Administration (FDA) because of genetic abnormalities in two of the babies. This suggestion rests on several erroneous readings of the work. Firstly, as acknowledged in a correction made on 28 September, the XO genetic abnormalities were in two fetuses and not in babies. Secondly, the FDA did not ban the work. They halted it, as is made clear in a letter from Zoon (2001) requiring clinics to obtain ‘an Investigational New Drug (IND)’ permit. The XO pregnancies were not mentioned in this letter. No specific reason was given for this action, the XO pregnancies, which were unlikely to have been caused by the technique, not being mentioned in this letter.

The paper by Chen et al. (2016) is also of potential relevance to the deliberations by the HFEA as to whether to license mitochondrial transfer, but more from ethico-social perspectives rather than for technical or medical reasons. It will be clear that the technique of cytoplasmic injection as performed in the late 1990s is so different from those of pronuclear or spindle transfer as to make the apparent normality of the offspring born through the former technique of little relevance in the context of spindle transfer, if superficially reassuring:

  • Mitochondrial replacement therapy (MRT) patients are likely to be fertile, whereas those patients selected for cytoplasmic transfer (CT) were infertile, having had multiple cycles of IVF with implantation failure.
  • Reconstitution during MRT requires cell fusion, and involves multiple extra steps, which were absent in CT, which used ICSI.
  • CT was done at a time when IVF was considerably less successful than it is today. Blastocyst formation rate was low, multiple embryos were often transferred and freezing was still being optimized.
  • MRT is being introduced at a time when blastocyst biopsy, screening for 24 chromosomes by next-generation sequencing and mtDNA finger-printing can be undertaken before embryo transfer. These extra safety measures are optional, but were not available as adjunct therapies in the early 1990s.
  • When offsite labs are used to confirm the proportion of affected mtDNA carried over during spindle or pronuclear transfer, embryos can be vitrified successfully at the blastocyst stage, and warmed embryos can be re-biopsied in case more assays are necessary. Cleavage-stage embryo biopsy followed by slow freezing yielded poor results with CT embryos.
  • When CT was performed, unused donor eggs could only be frozen after fertilization for future attempts at pregnancy with donated eggs if the patient so desired; egg vitrification was not yet available. In MRT cycles, surplus donor eggs can be vitrified and used later for either egg donation or in another MRT cycle.

Thus, the HFEA can take only limited medico-scientific information of relevance to their licensing decision from the paper by Chen et al. (2016). However, there are two ethico-social aspects raised by the paper that may be of concern to the HFEA.

The first arises from a subsequent report in New Scientist that there are currently two ongoing pregnancies derived by pronuclear transfer in the Ukraine (Coghlan, 2016). It is reported that these pregnancies were initiated not for reasons of MRT, but to treat longstanding infertility problems –exactly the motivation behind the CT experimental treatments described in Chen et al. (2016), and a treatment not allowed under UK law. This report – if substantiated, and given the apparent normality of the offspring reported in the Chen et al. study – means that the HFEA must consider whether there is a case to be made to the UK Government to extend the use of pronuclear or spindle transfer to include such infertility situations, and if it does not, then it should as a matter of some urgency issue convincing advice to intractable infertility patients as to why they should not go to the Ukraine or Mexico to seek treatment there, because many of them will be considering this course of action now.

Second, the observation that most of the parents declined to allow direct follow up of their offspring by the team of Chen et al. (2016), because all bar one couple had not told their children of the nature and circumstances of their conception, is pertinent to the deliberations of the HFEA. The necessity to gain information about the children indirectly through the parents reduces the reliability of the data reported by Chen et al. (2016), an issue that led to critical referees' reports. Notwithstanding these reports, the decision was taken to publish the paper, where these concerns are acknowledged and the very real difficulties faced by the authors in ethically collecting the desired objective information are also described. The point to be taken by the HFEA is that with the best of intentions, follow up of babies born through mitochondrial transfer, or any other novel experimental technique, whilst desirable, may not be as easy as this regulatory body may hope.