The new study, published in the British Journal of Cancer, focuses on a kinase called ATR, short for ataxia telangiectasia and Rad3-related protein. ATR sits at the heart of the cellular response to replication stress, the potentially lethal situation in which the molecular machinery that copies DNA stalls or breaks down. When chemotherapy drugs such as mitomycin C damage DNA, dividing cells rely heavily on ATR signalling to pause the cell cycle, stabilise stalled replication forks and coordinate repair. Block ATR pharmacologically, and cells exposed to DNA-damaging agents lose their safety net: replication forks collapse, DNA double-strand breaks accumulate, and the cell is pushed toward catastrophe. This concept, often described as exploiting a vulnerability created by the tumour’s own dependence on DNA damage checkpoints, has already shown promise in clinical trials of ATR inhibitors such as berzosertib in combination with platinum chemotherapy for advanced solid tumours.

What makes the Utrecht study distinctive is its model system. Rather than relying on immortalised cancer cell lines grown in two dimensions, which often fail to capture the biology of real tumours, the team used patient-derived organoids, miniature three-dimensional tumour cultures grown from tissue of six patients with non-muscle invasive bladder cancer. Organoids preserve many of the genetic and molecular features of the original tumours, including the expression of urothelial carcinoma markers, making them a far more faithful testing ground for new drug combinations. The researchers confirmed that their organoid lines expressed characteristic bladder cancer markers, validating them as genuine representatives of the disease they were designed to model.

The experimental design cleverly mimicked clinical practice. In patients, mitomycin C is delivered directly into the bladder as an instillation that remains in contact with the tumour tissue for roughly one to two hours before being drained. The researchers therefore exposed the organoids to mitomycin C for just two hours, replicating the transient exposure that tumour cells experience in the bladder, and only afterwards did they add ATR inhibitors, which the cells encountered for a prolonged 72-hour period. Three clinically relevant ATR inhibitors were tested: berzosertib, ceralasertib and tuvusertib, all of which have entered clinical trials in various cancers. The team also examined combinations with gemcitabine and epirubicin, two further agents used in intravesical chemotherapy regimens, in one organoid line.

The results were striking. Organoids treated with mitomycin C alone, or with an ATR inhibitor alone, eventually recovered: when the researchers followed the cultures for six weeks after treatment, the surviving cells proliferated at rates similar to untreated controls, demonstrating that neither agent on its own could eliminate the tumour cell population. In sharp contrast, organoids that received the sequential combination of mitomycin C followed by an ATR inhibitor showed severely impaired viability, and crucially, this effect persisted throughout the six-week observation period. The combination did not merely slow the cancer cells down; it appeared to destroy their capacity to regrow, which is precisely the property needed for a therapy intended to prevent recurrence after tumour resection.

Delving into the mechanism, the researchers showed that berzosertib potently suppressed the ATR signalling that mitomycin C normally triggers. DNA damage induced by the chemotherapy was marked by phosphorylated H2AX, a well-established molecular beacon of DNA double-strand breaks, and blocking ATR prevented the checkpoint response that would normally allow cells to survive this damage. Consistent with catastrophic, irreparable DNA damage, the combination treatment drove the organoid cells into apoptosis, the controlled programme of cell death. Quantitative analysis of the drug interaction using synergy scoring frameworks confirmed that the effect was genuinely synergistic rather than merely additive, meaning the two drugs together killed far more cells than would be predicted from their individual activities.

The implications for patients are considerable. Recurrence after intravesical therapy remains the central clinical challenge in non-muscle invasive bladder cancer, driving repeated surgeries, lifelong surveillance and, in a substantial minority of cases, progression to life-threatening muscle-invasive disease. The economic burden of bladder cancer across Europe is among the highest of any malignancy, largely because of the intensity of monitoring and repeat treatment that recurrence entails. A regimen that converts transient chemotherapy exposure into durable eradication of residual tumour cells could reduce recurrence rates, spare patients repeated interventions and delay or prevent progression. Because ATR inhibitors such as berzosertib, ceralasertib and tuvusertib are already in clinical development, the path from laboratory finding to clinical testing is shorter than for an entirely novel drug class.

There are important caveats. The study is preclinical, conducted in organoids rather than in patients, and although organoids are among the most clinically predictive laboratory models available, they cannot fully reproduce the immune system, the bladder wall architecture or the complex urine environment that shapes drug activity in vivo. The number of organoid lines tested, six for the mitomycin C combinations, is modest, and the gemcitabine and epirubicin experiments were limited to a single line, so the generality of the synergy across the molecular diversity of bladder cancer remains to be established. Questions also remain about the optimal sequencing, dosing and delivery of ATR inhibitors in the bladder, and about whether systemic administration would be needed or whether the inhibitors could themselves be delivered intravesically to limit side effects.

Nevertheless, the study provides a compelling proof of principle that the DNA damage response is a druggable Achilles heel of non-muscle invasive bladder cancer, and it establishes patient-derived organoids as a practical platform for optimising intravesical combination therapies before they are tested in the clinic. The findings build on a growing body of evidence that ATR inhibition sensitises bladder tumours to DNA-targeted agents, including earlier work showing enhanced cisplatin and gemcitabine activity in bladder cancer cell lines and clinical trial data combining berzosertib with platinum chemotherapy in advanced urothelial carcinoma. If the synergy observed in these miniature tumours translates to patients, the humble bladder instillation, a treatment whose basic design has changed little in decades, could be transformed into a precision strike that leaves behind not just damaged cancer cells, but none at all.

Subject of Research: Combining ATR kinase inhibitors with intravesical chemotherapy to prevent recurrence in non-muscle invasive bladder cancer, tested in patient-derived organoids.

Article Title: ATR inhibitors synergise with mitomycin C to enhance cytotoxicity in patient-derived non-muscle invasive bladder cancer organoids

Article References: Zuidema, A., Nijland, L., van Megesen, K., Vosjan, M. M., Viergever, B. J., Kranenburg, O., & Meijer, R. P. (2026). ATR inhibitors synergise with mitomycin C to enhance cytotoxicity in patient-derived non-muscle invasive bladder cancer organoids. British Journal of Cancer. https://doi.org/10.1038/s41416-026-03581-0

Image Credits: AI Generated

DOI: 10.1038/s41416-026-03581-0

Keywords: bladder cancer, ATR inhibitors, mitomycin C, patient-derived organoids, DNA damage response, non-muscle invasive bladder cancer, berzosertib, ceralasertib, tuvusertib, intravesical chemotherapy, drug synergy, cancer recurrence

Steroid cell tumors are unusual because their biology is defined not only by where they grow, but also by the hormones they can manufacture. They arise from ovarian cells capable of producing steroid hormones, including testosterone and other androgens. When a tumor releases substantial amounts of testosterone, a patient may develop virilizing symptoms: increased facial or body hair, acne, deepening of the voice, male-pattern hair loss, menstrual disruption, or enlargement of the clitoris. These changes can appear gradually and may be mistaken for more common endocrine conditions, such as polycystic ovary syndrome. In the new study, elevated serum testosterone accompanied by virilizing manifestations was present in all 19 patients who underwent fertility-sparing surgery. It was also found in five of the 11 patients who underwent more extensive, non-fertility-sparing procedures. This hormonal signature can provide an important diagnostic clue, although not every steroid cell tumor produces enough hormones to cause obvious symptoms.

The investigators retrospectively reviewed patients who underwent surgery at the Beijing hospital between January 2008 and August 2025. They divided the group according to whether surgeons attempted fertility-sparing surgery, meaning that the uterus and at least part of the reproductive potential were retained. Nineteen patients were placed in the fertility-sparing group, while 11 underwent non-fertility-sparing surgery. The age difference between the groups was striking. Patients receiving fertility-sparing treatment had a median age of 27 years, with ages ranging from 4 to 42. The non-fertility-sparing group had a median age of 55, ranging from 25 to 80. That contrast reflects both medical decision-making and the realities of ovarian cancer care: younger patients are more likely to seek preservation of fertility, while older patients may have completed childbearing or face different surgical considerations. It also means the two groups cannot be interpreted as directly comparable treatment arms in a clinical trial.

Cancer stage emerged as one of the strongest signals in the analysis. The International Federation of Gynecology and Obstetrics, or FIGO, classifies ovarian cancer according to how far it has spread. Stage IA generally indicates that the tumor is limited to a single ovary, with the ovarian capsule intact and no malignant cells detected in relevant fluid or washings. In the fertility-sparing group, 12 of 19 patients, or 63.2 percent, had stage IA disease. Five patients, representing 26.3 percent, had stage IC1 disease, while two patients, or 10.5 percent, had stage II disease. Stage IC1 typically involves surgical spill or rupture of the tumor, whereas stage II indicates extension to pelvic structures. Every patient in the non-fertility-sparing group had stage IA disease, a distribution that highlights the complexity of the surgical choices. Despite the apparent severity of some cases in the fertility-sparing group, no patient with stage IA disease experienced a recurrence during the study’s follow-up period.

The less reassuring results appeared among patients with disease beyond stage IA. Recurrence developed in one patient with stage IC disease who had undergone ovarian cystectomy, an operation that removes the tumor or cyst while leaving as much ovarian tissue as possible. Two patients with stage II disease in the fertility-sparing group also experienced recurrence, and one of them died from the disease. These outcomes do not prove that fertility-sparing surgery caused the recurrences; patients with stage IC or stage II tumors already had biological or anatomical features associated with greater risk. But they do show why a decision to preserve reproductive organs cannot be based on age or fertility preference alone. Tumor stage, capsule status, surgical spill, histological features, and the completeness of tumor removal all matter. In a rare cancer for which evidence is drawn largely from small retrospective series, even a few events can substantially influence clinical interpretation.

The surgical approach considered most reasonable for selected patients with stage IA disease was unilateral salpingo-oophorectomy. This procedure removes one ovary and its associated fallopian tube while leaving the opposite ovary and the uterus intact. In principle, the remaining ovary can continue releasing eggs and producing reproductive hormones, and the uterus remains available for pregnancy. The study’s findings support considering this approach when a tumor is restricted to one ovary and the patient strongly desires future fertility. By contrast, fertility preservation in stage IC or stage II disease should be individualized with caution, the authors concluded. Removing only the tumor through cystectomy may preserve more ovarian tissue, but it can also raise concerns about incomplete excision, tumor rupture, or microscopic disease left behind. The balance between oncological safety and reproductive preservation therefore requires multidisciplinary assessment rather than a one-size-fits-all rule.

The reproductive outcomes in the cohort were far less definitive than the cancer outcomes. At the last follow-up, only one patient clearly reported an active desire for childbearing, and that patient subsequently achieved a spontaneous twin pregnancy that resulted in live birth. This is an encouraging result, but it cannot be interpreted as evidence that fertility-sparing surgery reliably restores fertility after treatment. The small number of patients attempting pregnancy makes it impossible to calculate a meaningful pregnancy rate, and the study did not establish how many participants were infertile, had no pregnancy plans, were too young to attempt conception, or faced other reproductive barriers. The ovarian reserve—the remaining supply of eggs—may be affected by surgery, age, the underlying tumor, or other medical factors. Hormonal recovery was assessed as part of the study, but the available results do not establish a predictable pathway from removal of a steroid cell tumor to successful conception.

The patients were followed for a median of 61 months in the fertility-sparing group and 67 months in the non-fertility-sparing group. The ranges extended as long as 184 months and 160 months, respectively, providing an opportunity to detect some late recurrences. Long-term surveillance is particularly important for steroid cell tumors because their malignant potential is difficult to predict. Histological appearance alone may not always capture the behavior of an individual tumor, and recurrence can depend on factors such as size, necrosis, hemorrhage, atypical cellular features, mitotic activity, and extra-ovarian spread. Follow-up may include clinical examinations, imaging, and measurement of hormones such as testosterone when they were elevated before treatment. A return of virilizing symptoms or a renewed rise in androgen levels could signal recurrent hormone-producing tissue, although hormonal markers must be interpreted alongside imaging and pathology rather than used in isolation.

The study’s design places important limits on what can be concluded. It was retrospective, meaning that the researchers analyzed existing medical records rather than assigning patients prospectively to different treatments. All 30 patients came from a single specialized hospital, which may see unusually complex cases and may have surgical expertise not available everywhere. The fertility-sparing and non-fertility-sparing groups also differed greatly in age and clinical context, making direct comparisons vulnerable to confounding. Most importantly, ovarian steroid cell tumors are so rare that even a carefully assembled cohort remains small. The authors therefore call for multi-institutional studies that combine cases across hospitals and countries, use standardized pathological criteria, track ovarian reserve and menstrual recovery, and record pregnancy attempts systematically. For now, the evidence points to a cautiously hopeful message: stage IA disease may allow preservation of reproductive potential without an obvious compromise in cancer control, but patients with stage IC or stage II tumors face a far more precarious trade-off that must be decided case by case.