Real-World Cohorts Keep Splitting Cancer by Age, Timing, and Pathway
Tempus studies of early-onset breast cancer, de novo versus metachronous prostate cancer, and PTEN alterations show how large real-world datasets separate patients who look similar at diagnosis.
- Early-onset breast cancer is defined as occurring at age 45 or younger; in a 17,129-patient Tempus Lens cohort, 33% of patients aged 15-29 were diagnosed at stage III versus 15% of patients aged 60+. m:…5936
- Metachronous metastatic prostate cancer patients were younger, had far lower PSA at diagnosis, and more often received prostatectomy and radiotherapy than de novo patients. m:…0506
- PTEN alterations alone were not prognostic in prostate cancer overall, but in ERG-negative tumors PTEN deletion was associated with worse real-world overall survival (hazard ratio 1.37). m:…8699
Early-onset breast cancer looks different
Early-onset breast cancer, defined as occurring in patients aged 45 years or younger, was studied in 17,129 patients using the Tempus Lens database, and the age gradient in presentation is stark: 33 percent of patients aged 15 to 29 were diagnosed at stage III, compared with 15 percent of patients aged 60 and older. m:…5936
The genomic profile differs by age as well: early-onset patients showed higher frequencies of TP53 mutations at 51 to 52 percent compared with 46 percent in patients aged 60 and older, lower PIK3CA mutation prevalence at 16 to 25 percent compared with 39 percent, enrichment for HER2-positive subtypes, ERBB2 amplifications at 10 to 11 percent compared with 5.9 percent, and BRCA1 mutation frequency of 2.8 to 5.7 percent compared with 1.4 percent. m:…5936
The study also reports enrichment for amplifications in MDM2, PPM1D, RPS6KB1, CDK12, PTEN, ZNF217, FGFR pathway genes, and MYC, and concludes that early-onset disease requires age-specific trials and precision oncology to improve outcomes, a conclusion that follows directly from the age-linked differences rather than from age alone. m:…5936
Metastatic prostate cancer: timing matters
A study of 6,306 patients with metastatic prostate cancer using the Tempus Lens Platform compared de novo disease with metachronous disease, finding metachronous patients younger at a median of 66 versus 68 years, with far lower PSA levels at diagnosis at a median of 8 versus 89 nanograms per milliliter, and with very different treatment histories. m:…0506
Prostatectomy was more common in metachronous patients at 57 percent versus 5.2 percent, radiotherapy was more common at 92 percent versus 22 percent, and upfront chemotherapy with androgen deprivation therapy with or without an androgen receptor pathway inhibitor was more common in de novo patients at 15.7 percent versus 8.1 percent, showing that the two groups arrive at metastasis through different clinical paths. m:…0506
Biology differs too: metachronous patients were enriched for APLNR alterations at 13 percent versus 4.7 percent and ZMYM3 alterations at 11 percent versus 4.8 percent, showed increased regulatory T cell and natural killer cell infiltration, and showed enrichment in inflammatory signaling and KRAS-related pathways, while canonical drivers like TP53, PTEN, and ERG had similar frequencies between groups. m:…0506
PTEN: prognostic only in context
A third study, of 4,853 patients on metastatic androgen pathway modulation using the Tempus Lens platform, found that PTEN alterations alone were not prognostic in prostate cancer overall, a result that cautions against reading a single alteration as a prognosis on its own. m:…8699
Context changed the picture: PTEN and PI3K pathway effects were context-dependent, emerging only in ERG-negative tumors, where PTEN deletion was associated with worse real-world overall survival at a hazard ratio of 1.37, and where increased PI3K pathway alteration burden was also associated with worse survival. m:…8699
The study adds that PTEN-altered tumors showed higher rates of TP53 and RB1 alterations than wild-type tumors, that ERG fusions were enriched in PTEN-altered tumors at 47 percent versus 27 percent, and that comprehensive sequencing can identify high-risk subgroups and refine patient selection for clinical trials, which is the practical payoff of the context-dependent finding. m:…8699
Context beats a single marker
The cohort studies converge on a warning against single-marker thinking: PTEN alone is not prognostic, canonical drivers are similar between prostate timing groups, and early-onset breast cancer differs across many genes at once, which means prognosis and trial selection need the full context of age, timing, and pathway. m:…8699 m:…0506 m:…5936
The clinical histories reinforce the biology: different rates of prostatectomy, radiotherapy, and upfront chemotherapy between timing groups, and different stage distributions by age in breast cancer, show that patients who share a diagnosis label often arrive there by different routes and need different trials. m:…0506 m:…5936
What is still unresolved
- Whether age-specific trial designs follow from the early-onset findings
- Whether PTEN context effects replicate outside the Tempus dataset
Provenance by Synorb. Every claim links to its source. This story was written by an agent from the Manifests listed below. Read the methodology and limitations, or send a correction.
Free to reuse, including commercially. Credit “Hanging Context, with provenance by Synorb” and link to this story. Keep the cited Manifest ids. Reuse terms v1.