Barash et al. (2003) explored the possibility that local injury of the endometrium in the cycle preceding IVF treatment increases the success rate of implantation in a prospective study involving 130 patients who failed to conceive after one or more IVF treatment cycles [15]. Forty five out of 134 subjects were randomised by consent to have repeated endometrial biopsy on days 8, 12, 21 and 26 of the cycle immediately before the IVF treatment cycle. They found that the treatment resulted in a significant improvement (approximately double) in the rates of implantation, clinical pregnancy and live births (27.7, 66.7 and 48.9 % respectively), compared with control subjects who did not have endometrial biopsies (14.2, 30.3 and 22.5 % respectively). Similarly, a further study identified a favourable influence of local injury to the endometrium in intracytoplasmic sperm injection patients with high-order implantation failure [16] . Finally, two recent systematic reviews and meta-analyses of the available literature showed a beneficial effect of local endometrial injury in RIF but advised that further robust randomized trials are required [17, 18].

The mechanism by which endometrial scratch leads to improvement in IVF outcome in women with recurrent implantation failure remains unclear. Many investigators speculate that the repair process following local injury is associated with increased production of various growth factors conducive to implantation [15, 18, 24, 25]. Endometrial gene modulation following endometrial injury has also been hypothesised to increase endometrial receptivity [26]. It is recognised that endometrial development in IVF cycles is more advanced than that of natural cycles by 2–4 days [27, 28]. It has been suggested that repeated IVF failure might be related to asynchrony of the endometrium with the embryo stage [24, 27, 29, 30]. It has also been postulated that local endometrial injury in stimulated cycles delays the endometrial development due to the wound repair process correcting the asynchrony between endometrial and embryo stage [22].

Endometrial injury or scratch could simply be performed by the use of a Pipelle endometrial biopsy sampler (Laboratoire C.C.D. Paris) (Fig. 22.2a–c). After the Pipelle sampler is introduced into the endometrial cavity, the inner shaft is withdrawn to create a negative suction after which the Pipelle sampler is gradually rotated as it is moved up and down the endometrial cavity several times to produce the “scratching” action. In some centres, trial embryo transfers are routinely carried out prior to IVF treatment. In this situation, endometrial scratch and trial embryo transfer may be carried out at the same time.

Apart from the use of a Pipelle endometrial sampler, endometrial biopsy may also be carried out at the time of hysteroscopy. Hysteroscopy is not only a diagnostic tool; it allows therapeutic procedures to be carried out at the time of diagnosis. Hysteroscopy is one of the most important investigations in women with RIF. Current evidence suggests that the incidence of abnormal hysteroscopic findings in women with recurrent IVF failures varies between 25 and 50 % [32]. It is useful to time the hysteroscopy to take place in the luteal phase of the cycle preceding IVF treatment as hysteroscopic directed endometrial biopsy (scratch) may also be performed at the same time to improve the implantation rate [33, 34]. However, when hysteroscopy is performed in the mid-luteal phase, the possibility of disturbing a spontaneously occurring pregnancy should be explained and women advised to consider protected sexual intercourse in the treatment cycle. On the other hand, hysteroscopy performed in the follicular phase has an advantage that the endometrium is thinner and the visibility is better.

Two prospective, randomised controlled studies confirmed the value of hysteroscopy in women with RIF demonstrating significantly increased clinical pregnancy rates when endometrial biopsy and hysteroscopy were carried out in the luteal phase of the cycle preceding IVF treatment [33, 34].

A multi-centre randomised controlled study of pre-IVF outpatient hysteroscopy is currently being performed to evaluate whether outpatient hysteroscopy prior to starting an IVF cycle improves the likelihood of achieving a live birth in women who have experienced two to four IVF implantation failures [35].

A study was performed to determine if a subgroup of women with recurrent implantation failure were more likely to benefit from endometrial scratch than others [36]. It was found that, among the various factors examined, only the concentration of FSH appeared to have a significant prognostic value. Women with FSH ≤10 iu/l were nearly three times more likely to benefit from endometrial scratch than women with FSH > 10 iu/l. Interestingly age, body mass index, antimullerian hormone, free androgen index, total number of embryos previously transferred and prior number of embryo transfer cycles did not appear to have any significant predictive value [36]. It appears that when the underlying reason for recurrent implantation failure is embryological in origin, endometrial scratch is unlikely to improve the outcome [36]. When failure is associated with aberrant endometrial function one may expect that endometrial scratch is more likely to be beneficial. This is consistent with the study findings that an FSH concentration of >10iu/l was associated with a lower chance of pregnancy after endometrial scratch, since this would be a reflection of reduced oocyte quality.

It is unclear if the impact of endometrial scratch varies in women with different underlying aetiologies for their infertility. It would be of particular interest to determine if an endometrial scratch is more or less likely to benefit women with endometriosis, especially as there is conflicting evidence in the literature that the implantation rate in this group of women is adversely affected [37–39].

In conclusion, the available evidence to date points towards a potential benefit to both hysteroscopy or endometrial biopsy in the cycle immediately preceding IVF treatment but further large, prospective, randomised studies are required to conclusively prove this and allow the introduction of this novel procedure into routine clinical practice.