The present study was designed to provide women with genuine POR according to the Bologna criteria with prognostic information regarding IVF cycles outcomes and their chances of a live birth. Our data demonstrates that the probabilities of POR patients achieving a live birth are best with younger age and higher number of retrieved oocytes.
The age-related decline in ovarian reserve in women over 40 years is well documented [13, 14]. Nevertheless, for some women the physiologic decrease in ovarian function occurs earlier. The subgroup of patients termed “poor ovarian response” represents the most challenging group of patients to treat, as this population are at high risk for inadequate response during IVF treatment and lower pregnancy and live birth rates .
Several pretreatment diagnostic tests including basal FSH, AFC, inhibin B, and AMH have been used for prognosticating a poor responder to stimulation. More accurate prognostic information can be derived by the completion of ovarian stimulation cycle, as this “stress test” explores the capacity of the ovary to produce enough oocytes.
Previous studies found that recurrence rate of poor ovarian response during the subsequent ovarian stimulation was 54–62% [16, 17].
Dilemmas that physician may encounter while treating patients with POR may arise from the tendency to include this subgroup of patients as a homogenous group and therefore labeling similar treatment and prognosis to all patients. In fact, this is a heterogeneous group composed of different ages and various causes which led to POR. Thus, the question arises whether or not there are any prognostic factors which will mark some of the poor responders' patients with an acceptable prognosis for live birth following IVF treatment. Identifying such prognostic factors will assist fertility providers in counselling POR patients as to whether it is useful or not to start and / or to continue with IVF treatment.
A woman’s age is considered one of the most significant single determinants affecting chances of conception, either naturally or via ART [18–20]. Therefore, while counseling patients regarding their prognosis, age should be strongly considered. Hanoch et al.  conducted a study in which they evaluated differences in pregnancy rates of young versus older low responder patients. They reported significantly higher clinical pregnancy rate among young (20–30 yrs.) low responder patients (19.3% vs 6.5%). Accordingly, they concluded that young age protects from the deleterious effect of POR. In a literature review, Oudendijk et al.  concluded that older poor responders have lower pregnancy rates compared with younger poor responders (1.5%-12.7% vs. 13%-35%).
The findings of the present study are in line with the above-mentioned studies. Woman’s age was found to be negatively associated with the probability of achieving a live birth in lower responder’s patient population.
The distinct effect of female’s age on reproductive outcome is explained by the declining oocytes quality, which coincides with a progressive decrease in the primordial follicle number that occurs with female’s aging [22–24].
The age-induced oocyte quality impairment is closely associated with chromosomal abnormalities, and mitochondrial dysfunction [25, 26]. Women of older age are subjected to a higher number of aneuploid embryos which are more likely to arrest in extended culture . This probably explains the substantially lower live birth rate of women of advanced maternal age.
The primordial follicle pool depletion is common not merely in women of advanced maternal age, but also in most poor ovarian responders, irrespective of age . Thus, significantly limits the success of assisted reproduction treatment .
Since both group of women, advanced maternal age (AMA) and poor ovarian responders, as a subgroup, share the same physiologic follicular pool depletion, the question that arises is whether young poor responders also exhibit a reduction in oocyte quality like AMA women (i.e., high risk of aneuploidy, poor embryo development).
In attempt to address this question, a recent retrospective study published by Morin et al.  found that compared to normal responders, a fertilized oocyte retrieved from a young poor responder patient (< 38 years) is no less likely to form a quality blastocyst, be euploid or produce a live birth. They concluded that an oocyte retrieved from a poor responder patient performs similarly to that from age-matched controls.
The present study demonstrates a significant higher live birth rate in poor ovarian response who are younger than 40 years compared to poor responders > 40 years (7% vs 2%, respectively). This finding as well as the results of the multivariate analysis strongly suggest that a woman’s chronological age is a critical factor which affects IVF outcome and protects against the adverse effect of poor ovarian response.
Another factor which the present study found to be of relevance in determining the prospects of Bologna poor responders for a live birth is the number of oocytes retrieved.
Although the number of oocytes per se is not an indicator of their quality, yet the lower the degree of poor ovarian response, there will be more oocytes to retrieve and subsequently more embryos to transfer, which may improve the chances for pregnancy.
The results of the multivariate analysis which found the number of oocytes retrieved and not the number of dominant follicles visualized ultrasonographically as an independent predictor of live birth could be explained by the fact that oocytes could not be retrieved from some follicles during ovum pick up. The fact that this index cannot be determined in advance limits its value in counseling and preparing POR patients before embarking on IVF treatment.
Our findings are in agreement with previous studies which found positive association between the number of oocytes retrieved and pregnancy/live birth rates in poor responders [15, 31].
This study is not the first study dealing with predicting factors for successful IVF cycles in poor responder patients. However, it is important to stress that when reading through the current published literature there is a lack of uniformity regarding the definition of poor responders (from fewer than two up to five dominant follicles) , which explains the differences in the studies’ results. To the best of our knowledge, the current study is among the largest experiences of poor responders, defined according to the Bologna criteria, up to 3 oocytes retrievals.
The unique health care system in which this study was conducted provides financial coverage for repeated IVF cycles. This enables to collect information on a high number of cycles treatment, which in other places might have been cancelled due to poor response.
The inclusion of high number of cycles treatment strengthens the power of the study results and allows a firm assessment on the markers predicting success in the poor ovarian response population. Therefore, contributing valuable information for both fertility providers and patients alike.
The limitations of the study should also be noted. First, as this was a retrospective study, it had the inherent biases that might affect the results. Although we tried to minimize selection bias by using rigorous inclusion criteria, we cannot exclude unknown confounders that might have affected the results. In addition, AMH values could not be analysed as they are not part of the local routine testing. Of notice, FSH which is considered a reliable test marker to predict ovarian response to stimulation, according to our results was found to have no predictive value on treatment outcome of poor responders.
Improving the cycle outcome is one of the goals of fertility providers.
However, taking into consideration the balance between risks, costs (financial and emotional) and benefits involved in providing IVF treatment for POR, the question emerges is whether additional cycles using the women’s autologous oocytes are justified or continuation of treatment should be discouraged.
Given the present study data, it appears that continuation with IVF treatments seems reasonable in patients younger than 40 years, since in these poor responder’s pregnancy rates were 10%, and live birth rates were 7% per cycle. Yet, for poor responders’ women \(>\)40 there appeared to be hardly any justification for any further IVF treatment, due to very poor outcome (4.5% pregnancy rate, and 2% live birth rate per cycle).
In conclusion, the woman’s age and the number of retrieved oocytes are both independent predicting factors of live birth in poor ovarian responders according to the Bologna criteria.
All poor ovarian responders should be informed regarding the meager probability of achieving a live birth. Specifically, POR women \(>\)40 years should be encouraged to apply for the process of egg donation, which will improve their chances of a live birth substantially.