Microinjections = macroproblems: the viewpoint of a reproductive biologist

Abstract

Intracytoplasmic sperm injection (ICSI) has revolutionized the treatment of male infertility ( 8). Thanks to this technique, it is now possible to achieve fertilization and pregnancy when the number of spermatozoa in the ejaculate is extremely low. I therefore understand the enthusiasm of clinicians such as Dr Baker, whose practice puts them in a position from which they can appreciate the benefits brought by this technique to patients affected by the most severe forms of infertility. I also appreciate the value of the efforts made by different in vitro fertilization (IVF) groups, starting with that of Professor van Steirteghem and his collaborators, who are doing their best to develop the necessary follow-up of the children born after the ICSI procedure. My major concern is with the way in which assisted reproduction techniques generally emerge, including those that may develop in the wake of ICSI. These new microassisted human fertilization procedures are: testicular sperm extraction (TESE) to alleviate male factor infertility due to obstructive azoospermia ( 2) or due to azoospermia resulting from spermatogenetic maturation arrest ( 3; 5; 9); elongating and elongated spermatid injection (ELSI; 14) when spermatozoa are absent;␣round spermatid injection (ROSI; 13; 14) and round spermatid nuclei injection (ROSNI; 6; Sofikitis et al., 1995), when late spermatids are missing; and, very recently, secondary spermatocyte injection (SESI; 11) when spermatids are not available. These new techniques bypass the sperm selection processes that have evolved to regulate vertebrate fertilization (ICSI; 1), various steps of spermiogenesis (ROSI and ROSNI), the complete spermiogenetic process and the equatorial stage of division of meiotic cells (SESI). The first aspect that is deeply disturbing is that the development and use of these techniques were not originally published in the form of true scientific papers, i.e. papers that are peer-reviewed, that contain proof the assertions made (figures, pictures), and that provide all the information required for verification (detailed protocols). Instead, the first information on these revolutionary technologies was published in the form of press releases and short `letters to the Editor' of the Lancet (ELSI; 5; ICSI; 8) or of the New England Journal ofMedicine (ROSI; 13) or in the form of conference reports (ROSNI, SESI; 10, 11). These major distortions of the accepted method of scientific communication, concerning an issue as important as assisted reproduction which always receives tremendous media attention, raises the␣question of the responsibility of some of the clinicians in␣the IVF field. They also raise the question of the deontology of the editorial boards of some medical journals which seem to chase the scoop, rather than rigorously informing the scientific community on the conditions under which these human experiments were permitted (ethical committees?) and performed (reality, security, repeatability?) and how the outcome of these experiments is being followed up. There is another reason for disquiet. These new techniques are used to create human beings. However, in most cases they do not result from the logical extension of thorough animal experimentation designed to prepare the ground for the transfer to the clinical field, or of clearly planned and well-thought-out clinical decisions. Although I agree with Dr Baker that many important medical breakthroughs have developed fortuitously, in the field of reproduction it is striking that the fortuitousness is systematic, or almost. Instead, the emergence of the microinjection techniques has been presented as the consequence of an `accident' in the case of ICSI ( 8), or as totally opportunistic in the case of ROSI (Tesarick et al., 1995) and ELSI ( 5). SESI ( 11) was presented as the extrapolation of animal studies. In fact, these studies had only been performed using fertile animals, not animals affected by spermatogenic arrest, and were performed shortly before experimentation with human subjects, and by a single laboratory. There was therefore no opportunity for validation by the scientific community. As a consequence, not only do the practitioners of these techniques lack the minimum biological background that is always required before experimenting on human beings but, in most cases, no ethical approval was sought and therefore granted to develop these extremely audacious techniques. A further consequence is that the practitioners were and still are, de facto, unable to provide their patients with even the minimal information required for obtaining `informed consent', in particular when cells other than spermatozoa are used. Several arguments are often raised when the value of animal experimentation in assisted reproduction is discussed with IVF practitioners. The first is that several previous major advances in assisted reproduction technology were not preceded by any detailed animal studies (e.g. embryo freezing) and have not had any apparent negative consequence. This leads to the view that new technological jumps can be made in the absence of prior animal experimentation. Secondly, it is often said that, for technical reasons, it is not possible to develop animal models and/or that there is no adequate animal model. The third is that `scientists have not proved that these new techniques are dangerous'. Fourthly, animal experimentation is often presented as costly and time-consuming. This explains in turn why the few laboratories which are at the interface between the basic and clinical fields, and therefore have both the technical and intellectual skills, have not developed long-term animal experiments. The first of these arguments is akin to playing Russian roulette, and is an argument which is completely unacceptable in medical practice. It is also unclear whether enough time has elapsed or if the proper means are available to assess objectively all aspects of the possible consequences of the application of microinjection techniques on the progeny. There is no doubt that the techniques of microinjection of spermatozoa and, a fortiori, of other testicular cells, represents an increased risk to the offspring, because a significant proportion of the men requiring microinjection assistance are infertile for genetic reasons. Furthermore, because different steps of spermatogenesis are bypassed when immature germ cells are used, and because the identification of these cells cannot be accurate, additional and increasing risks are certainly taken with TESE, ELSI, ROSI, ROSNI and SESI. This should, a␣fortiori, lead to even more cautiousness (and modesty?) among IVF practitioners. Finally, the argument that no previous catastrophe has occurred in the field of assisted reproduction totally ignores the unprecedented social demand for safety in the fields of health and medicine resulting from the `contaminated blood scandal' and the other recent medical and scientific problems (e.g. contaminated pituitary extracts and the `mad cow' affairs). I remind those who pretend (or who used to pretend) that it is extremely difficult or impossible to set up animal models that this argument was raised in the past to justify, for example, 8) performing the first ICSI directly on a human being, before 7) succeeded in setting up ICSI in the mouse. This clearly shows that when a topic becomes `hot' enough, skills, time and money can be found to develop animal models. Furthermore, the assertion that `there is no adequate animal model' ignores the fact that animal experimentation is mandatory and extremely constraining for the pharmaceutical industry. Initially, when the first safety assessment guidelines were implemented for drugs, this occurred because of the catastrophic consequences of the thalidomide tragedy, and was not due to a consensus on the existence of an ideal animal model. For those who argue that the onus is on basic scientists to prove that these new technologies are dangerous, I would answer that technology should serve science, not dictate to science. Of course, when circumstances allow it one of the major roles of science is to provide the knowledge that allows understanding of how and why a given therapy works or does not work. Concerning the cost of programmes involving animal experimentation, it should be remembered that the health of human beings is at stake. Furthermore, if we consider the huge profits made by the `assisted reproduction industry', which includes the revenues of clinics, clinicians and industry (e.g. hormone manufacturers), then the absence of investment by these beneficiaries in basic and experimental research on assisted reproduction is striking. By contrast, note, for example, that the pharmaceutical industry invests an estimated 15–20% of its sales in research and development of new drugs. Finally, it is also important to note two major contradictions in the views of some of the pioneers of microinjection techniques: first, after having claimed that animal studies are not really relevant to human trials, they now want to use, or more accurately abuse, the few existing data arising from animal experimentation (e.g. SUZI and ICSI performed in normal animals, not in animals with spermatogenic arrest, and data on genomic imprinting) to justify in retrospect the experimentation performed in humans (see 4); secondly, while advocating caution in the use of microinjection procedures with immature germ cells ( 12), they themselves diffuse these techniques worldwide, including to countries where, a priori, there is very little chance that there will be follow-up studies, for example. Readers of this paper may believe that I am hostile to the development of the technology of assisted reproduction, but this is not the case. My goal is to highlight the singularity of the emergence of the most recent IVF technologies and some universal principles of science and medicine. One of these principles is that animal studies must always precede action on human subjects, and should always be verified before being applied. Being a reproductive biologist, the increasing rift between basic science and technology in the domain of assisted reproduction is a matter of great frustration and anxiety to me. As a simple citizen, I remember that 40 years ago, Hannah Arendt wrote that `if knowledge [in the modern meaning of know-how] and thought separate irretrievably, we will become the playthings and slaves, not so much of our machines, but of our practical knowledge'. I thank Professors M. C. Orgebin-Christ and G. Waites, as well as Professor Kamal Zaki for giving me the opportunity to present the ideas summarized in this article before the participants of the VIth International Congress of Andrology in Salzburg (25–29 May 1997) and the 10th Annual Congress of Andrology in Cairo (17–18 December 1997).