Researchers develop “FORTUNE” model to predict number of euploids from IVF

A 2025 study introduces the FORTUNE model, developed from over 10,000 PGT-A cycles, that uses age, ovarian reserve, partner age, and BMI to predict how many euploid embryos a patient is likely to produce by grouping patients into five prognosis categories.

One of the biggest uncertainties in IVF is how many chromosomally normal (euploid) embryos a cycle will produce. One study showed that transferring 3 euploids (one at a time) led to about a 93% overall chance of live birth.

Traditionally, doctors have predicted IVF success by looking at things like a woman’s age or her ovarian reserve (AMH or AFC). But these are broad categories, and they don’t always capture the full picture. Two people with the same age and AMH level can still have very different chances of making euploids.

A new study by Seli et al. (2025) asked whether a more sophisticated prediction system, one that integrates multiple variables, could provide clearer prognostic groups. Their goal was to classify patients by their likely euploid blastocyst yield before starting IVF, which could help with fertility treatment planning.

⚠️ Remembryo summarizes and interprets IVF research for educational purposes. Posts highlight selected findings and may simplify or omit study details, including methods, analyses, author interpretations, limitations, and protocol specifics (such as timing, dosing, or eligibility criteria). These summaries are not a substitute for the original study. Always review the full publication before treatment decisions.

🔗 Original studies are referenced in this post or within the linked Remembryo posts.

💡 Reminder: Terms underlined with a dotted black line are linked to glossary entries. Clicking these does not count toward your paywall limit.

FORTUNE model development

The team analyzed over 10,000 IVF cycles performed between 2020 and 2023 at a single center where every blastocyst was tested with PGT-A. Using advanced regression techniques and neural networks, they tested how well different variables predicted the number of euploid embryos after stimulation and ranked them:

  • Female age – the single strongest factor; embryo numbers dropped sharply with increasing age.
  • Ovarian reserve – both AMH and AFC strongly predicted higher euploid embryo counts (because higher AMH/AFC indicate more eggs and embryos produced).
  • Partner’s age – older male age reduced euploid embryos slightly.
  • Sperm source – donor sperm cycles had a small advantage.
  • BMI (weight status) – no significant effect.
  • Gonadotropin starting dose – no effect.

From this, they developed the FORTUNE model (Forecasting Outcomes of Assisted Reproductive Treatments Using Artificial Networks) that could predict euploid numbers based on female age, partner age, AMH, AFC, and BMI. BMI on its own wasn’t strongly linked, but including it alongside the other factors made the model more accurate. They also noted that they excluded sperm source to make it more generalizable

FORTUNE model performance

The model performed well, with high accuracy in predicting whether a cycle would result in at least one, two, or three euploid embryos (AUC ~0.83–0.86). Based on predicted embryo yield, patients could be classified into five prognostic groups: very poor, poor, borderline, good, and very good.

  • Very poor prognosis: These patients were typically older women with a median age of 45.0 years (44.4–45.8), very low AMH (0.4 ng/ml), and low AFC (5 follicles), often with partners also in their mid-40s (45.6 years). Almost all cycles in this group (98%) produced no euploid embryos, and none had more than one.
  • Poor prognosis: Patients here had a median age of 43.0 years (41.7–44.0), with AMH around 0.6 ng/ml and AFC of 7, and partners averaging 43.9 years. About 80% of cycles resulted in no euploid embryos, and only 0.2% achieved three or more.
  • Borderline prognosis: This group had women around 40 years old (38.4–41.7), with AMH at 1.0 ng/ml, AFC of 9, and partners close to 41 years. Just under half (47%) had no euploid embryos, while about 9% managed to produce three or more.
  • Good prognosis: Patients here were typically around 36 years old (33.9–38.2), with AMH of 2.0 ng/ml, AFC of 14, and partners around 37.5 years. Only 16% had no euploid embryos, and nearly 44% ended up with three or more.
  • Very good prognosis: This group included women around 33 years old (30.7–34.6), with strong ovarian reserve (AMH 4.7 ng/ml, AFC 23) and partners averaging 34 years. Just 5% had no euploid embryos, while almost 80% had three or more.

They validated the model with an additional set of more than 2,000 cycles, making sure the predictions held up in a new group of patients. They found that the model performed well (AUC ~0.85) and showed very similar results to the development group:

  • No euploid embryos: 100% (very poor), 82.6% (poor), 47.4% (borderline), 13.1% (good), 4.4% (very good)
  • Three or more euploid embryos: 0% (very poor), 0.6% (poor), 10.7% (borderline), 51.0% (good), 83.5% (very good)

Access the FORTUNE model online

For patients curious about where they might fall with the model, the authors have made the model publicly available:

By entering age, AMH, AFC, partner’s age, and BMI, patients (or their doctors) can see their predicted prognosis group.

Conclusions

This study developed and validated the FORTUNE model, which combines age, ovarian reserve, partner age, and BMI to predict how many euploids a patient might get from IVF. It divided patients into five prognosis groups, from very poor prognosis to very good prognosis.

The model showed strong performance, with an AUC of about 0.85, meaning it could correctly distinguish between patients with higher versus lower chances of producing euploid embryos about 85% of the time.

The authors explain that even though the model was built using PGT-A data, it simply predicts how many healthy embryos a cycle might produce, so it can still be useful whether or not PGT-A is done.

The model gives a way to estimate how many euploids a patient might get, which can help with planning cycles and setting expectations. But it was tested in just one clinic, so more studies are needed before it’s widely used.

Limitations: Single center study, only one PGT-A platform was used, didn’t include different diagnoses, focuses on AMH and AFC (measurements that may not be available to everyone).

Want to read more about factors that influence euploid numbers or transfer success?

Reference

Seli E, Kalafat E, Reig A, Whitehead C, Ata B, Garcia-Velasco J. Forecasting Outcomes of Assisted Reproductive Treatments Using Artificial Networks (FORTUNE) classification system: a new prognostic model to predict euploid blastocyst yield in patients undergoing IVF. Hum Reprod. 2025 Sep 1:deaf163. doi: 10.1093/humrep/deaf163. Epub ahead of print. PMID: 40889782.

If you liked this post and want to support what I do, please consider a paid subscription, Patreon or donate through PayPal!

 


About Embryoman

Embryoman (Sean Lauber) is a former embryologist and the founder of Remembryo, an IVF research and fertility education website. After working in an IVF lab in the US, he returned to Canada and now focuses on making fertility research more accessible. He holds a Master’s in Immunology and launched Remembryo in 2018 to help patients and professionals make sense of IVF research. Sean shares weekly study updates on Facebook, Instagram, and Reddit regularly. He also answers questions on Reddit or in his private Facebook group.