Capital is moving into women’s health. But the science may give it even more reason to stay. Across the field, substantial unmet need is converging with rapidly advancing science and enormous areas of unexplored biology. Advances in genetics, molecular biology, imaging, computational science, and AI are making it possible to investigate questions involving immune regulation, reproductive aging, cardiovascular biology, neuroinflammation, the maternal brain, endometriosis, and other areas in ways that were difficult or impossible only a decade ago (1).
What makes this moment particularly exciting is not simply what we now know, but how much remains to be discovered. Unanswered biological questions create room for scientists to explore new mechanisms and for industry to translate those discoveries into new diagnostics, therapeutics, technologies, and companies. The growing flow of capital and expertise into women’s health gives the field something it has not always had: the resources to follow those questions wherever they lead.
However, while the opportunity is enormous, it comes with development challenges. When the science is new, we cannot assume that established models reproduce the biology that matters, that familiar endpoints capture meaningful patient benefit, or that an efficacious intervention will become a viable product unless we also consider fertility, hormonal life stage, chronic safety, and patient burden.
The next phase of women’s health is therefore not simply about filling the pipeline. It is about building the translational engine capable of navigating what we do discover. Importantly, many of the discoveries emerging from women’s health will not remain confined to conditions unique to women. These discoveries are expanding our understanding of immune function, cardiovascular biology, neuroscience, aging, and disease across the lifespan and genders.
Building that translational engine therefore starts with a deceptively simple question: what exactly are we trying to learn?
The model must match the question
Answering that question requires defining the evidence needed to test the therapeutic hypothesis and choosing the experimental system best suited to generate it. Simple enough. The challenge is that different features of the same disease may require very different kinds of evidence.
Endometriosis makes this particularly clear. Rodent models can reproduce aspects of lesion biology, inflammation, innervation, and pain-associated behavior, while patient-derived tissues and organoids provide complementary insight into human biology (2). Yet no single system captures a chronic disease in which lesions, pain, fertility, and physiology intersect. Even commonly used pain assays such as thermal testing or von Frey measure responses to defined stimuli that only partially reflect the complex, chronic pain experienced by patients.
The challenge is not to find a perfect model. Mice do not menstruate (2,3), while naturally occurring endometriosis in menstruating non-human primates is relatively uncommon and difficult to study at scale (3,4). Progress therefore depends on fitness for purpose: using the right model for the right question, understanding its limitations, and looking for convergence across complementary systems. The strength of a translational strategy lies in knowing what each model can tell us, what it cannot, and what additional evidence is needed before moving forward.
However, there is also a point at which the limitation is not the model, but how much certainty we expect it to provide. Early-stage developers know the cycle well: promising data leads to requests for another model, another endpoint, another species, another layer of validation. Each request may be reasonable on its own; however, collectively they can create the illusion that enough preclinical work will eliminate translational risk. Spoiler: It will not. Animals are not humans, and organoids are not patients. Good translation therefore requires knowing what additional evidence is needed and recognizing when the remaining question can only be answered in the clinic.
What should we measure once we get to the clinic is the next question. Regulatory development necessarily requires validated endpoints, and those endpoints provide essential standards for determining efficacy. Yet an endpoint that is accepted for regulatory decision-making may capture only one dimension of a patient’s disease or treatment experience.
In endometriosis, lesion burden, dysmenorrhea, non-menstrual pelvic pain, dyspareunia, fertility, and function each capture something different. Pain-related endpoints are central to current drug development, but pain itself is heterogeneous and subjective, and improvement in one measure may not capture changes in other symptoms, function, fertility, or underlying disease biology (5,6).
The endpoint must match the intended benefit
Validated clinical endpoints remain essential. For drug developers, the regulatory endpoint is ultimately the bar a pivotal trial must clear. But regulatory validity and completeness of the patient experience are not necessarily the same thing. An endpoint can be reliable, interpretable, and appropriate for demonstrating efficacy while still capturing only one dimension of a complex disease.
Endometriosis again illustrates the distinction. The current trial endpoint is pain (5,6). However, improvements in pain can demonstrate meaningful symptomatic benefit but do not necessarily establish effects on lesion biology, fertility, or disease progression. Conversely, lesion regression or changes in inflammatory markers may demonstrate biological activity without proving that patients experience less pain or improved function (5,6).
The same issue appears elsewhere. In menopause, improvement in vasomotor symptoms cannot establish benefit across cognition, bone, sleep, or cardiovascular health, each of which requires appropriate evidence (7). A measurable effect is not automatically the intended clinical benefit. Translation requires defining that benefit clearly and selecting endpoints capable of demonstrating it, while remaining open to the possibility that emerging biology may require new ways of measuring success (8).
Disease biology is not static
Efficacy alone does not determine whether an intervention becomes a useful product. In diseases that span decades, the biology being treated, the patient’s priorities, and the acceptable benefit-risk profile may all change over time. The question is therefore not simply whether a therapy works, but for whom, when, and for how long.
Endometriosis makes the product-development consequences tangible. A 16-year-old seeking relief from debilitating pain, a 28-year-old actively trying to conceive, and a 43-year-old managing recurrent disease may share the same diagnosis but have very different treatment goals. A therapy intended for years of use also faces different requirements for reproductive safety, bone health, tolerability, route of administration, and treatment burden than one intended for short-term use. Existing GnRH-directed therapies illustrate the tradeoff: meaningful pain relief must be balanced against hypoestrogenic effects, including loss of bone mineral density, which has influenced treatment duration and the use of add-back therapy (9). Women should not have to choose between tricking their bodies into thinking they are pregnant versus forcing their bodies through menopause.
Biology itself may also change with life stage. In multiple sclerosis, relapse activity falls markedly during late pregnancy and rebounds postpartum (10), while menopause has been associated with greater disability accumulation and neuroaxonal injury (11). Rheumatoid arthritis often improves during pregnancy and flares postpartum (12). These observations underscore that pregnancy, postpartum, and menopause are not simply demographic characteristics. In some diseases, they represent biologically distinct states that may alter disease activity, therapeutic priorities, and potentially the benefit-risk profile of treatment. Interestingly, they may also give clues into potential targets and allow us to explore how to harness these benefits.
These considerations belong upstream in development. If a therapy may be used for decades, chronic safety should influence candidate selection. If patients may want to conceive, reproductive effects cannot wait until late-stage development. And if disease biology changes across hormonal life stages, clinical development should consider whether treatment response, dose, endpoints, or benefit-risk might change with it. The target product profile is therefore not simply a description of the product we hope to develop. Used early, it becomes a scientific tool for defining what intervention actually needs to accomplish, for whom, and at what point in their lives.
Building the translational engine
Perhaps the most exciting part of this moment is how much we still do not know. Every new insight into disease biology, sex differences, hormonal life stage, and patient heterogeneity has the potential to change the questions we ask and reveal new ways to intervene. New biology can uncover new targets, identify previously unrecognized patient populations, enable new diagnostic strategies, and create opportunities for products that simply could not have been envisioned when our understanding of the disease was more limited.
The advantage of working in a scientific white space is that the development playbook is not yet fixed. Researchers can build better models as the biology emerges. Clinical developers can incorporate new measures alongside validated endpoints. Founders can design target product profiles around how patients will actually use an intervention, and investors and strategic partners can identify opportunities before categories and competitive landscapes become mature (and crowded).
Importantly, these principles are not confined to women’s health. Matching models to biological questions, endpoints to meaningful benefit, and products to the patients who will use them is simply good translational science. The discoveries now emerging from women’s health are already informing our understanding of immune function, cardiovascular biology, neuroscience, aging, pain, and disease across the lifespan.
For now, the category of women’s health remains necessary because the scientific and clinical gaps remain real. Its success, however, should not be measured by how permanently it stands apart. Success will come when research and development is sex-informed and life-course-aware across medicine.
Women’s health is health. The more we learn, the more opportunities we create to improve health for all. And we may only be getting started.
Author Bio
