New insights into management of poor prognosis patients (POSEIDON Classification)

The core of the talk is the division of patients into four POSEIDON groups. It rests on three parameters: the woman's age (with a conventional cut-off around 35 years), ovarian reserve indices (AMH and antral follicle count) and, where available, the actual response to a previous stimulation. Combining these features separates so-called unexpected poor responders with normal reserve from predictably difficult patients with diminished reserve, allowing a more precise choice of protocol and gonadotropin dose.

Particular attention is given to the conceptual shift from response to stimulation toward prognosis of outcome. POSEIDON introduces the notion of low prognosis rather than low response and proposes, as a reference point, the ability to obtain at least one euploid embryo suitable for transfer. This links laboratory and clinical goals: what matters is not the follicle count in itself, but the probability of live birth that this count supports given the patient's age.

From this logic follow practical implications for management. For each group the lecture discusses a rational starting dose, the case for adding LH activity, strategies aimed at increasing the number of oocytes retrieved, and the role of accumulating embryos across several cycles. The emphasis is on individualising the plan so as to obtain enough mature oocytes to give a realistic chance of success.

The closing part deals with the validation of the POSEIDON criteria, their clinical value and their limitations. The classification has improved the consistency of terminology and planning, yet the choice of threshold values and the reproducibility of reserve markers remain under discussion. Fischer stresses that POSEIDON is a framework for decision-making rather than a ready-made protocol, and calls for critical interpretation at the bedside.

The practical takeaway for the clinic is that the POSEIDON classification helps stratify difficult patients, set realistic expectations and build a personalised stimulation strategy oriented toward the final outcome rather than the follicle count in a single cycle.