在一项为期三年、涵盖142,000名50至77岁人群的试验中，Galleri与常规筛查联合使用，未能减少晚期癌症诊断，说明其核心终点失利。另一项23,000名50岁及以上人群的试验显示，它发现了次年确诊癌症中的约40%（其中一半为早期），但也漏检了60%。

该测试在约每107名参与者中给出1次癌症“信号”，且其中约五分之二最终为误报，提示假阳性负担不小。与此同时，在英格兰一项6,000名有症状转诊患者的研究中，被视为误报者在额外随访15个月后有三分之一被确诊为癌症，暗示部分“误报”可能是超前检出。

生存获益仍具争议：NHS专家委员会在2025年11月反对PSA全民筛查，估计每延长2条生命会伴随约20例无害肿瘤过度诊断和12例常带来阳痿或失禁等伤害的过度治疗。再加上2001至2011年的CA125试验虽增加卵巢癌早期发现却未降低死亡，以及Galleri技术被指较旧（甲基化检测过程会破坏大部分DNA），趋势指向按癌种与风险分层决策，并期待结合全基因组、DNA片段物理特征和AI的新一代多模态检测改进净获益。

In a three-year trial of 142,000 people aged 50-77, adding Galleri to usual screening did not reduce late-stage cancer diagnoses, meaning it failed its key endpoint. A separate trial in 23,000 people aged 50+ showed it detected about 40% of cancers diagnosed over the next year (half early-stage) but missed 60%.

The test produced a cancer “signal” in about 1 in 107 participants, and roughly two in five of those signals were false alarms, indicating a meaningful false-positive burden. Meanwhile, in an England study of 6,000 symptomatic referred patients, one-third of presumed false alarms were diagnosed with cancer after an extra 15 months of follow-up, suggesting some “false positives” were earlier detection.

Survival benefit remains contested: in November 2025, an NHS expert committee advised against universal PSA screening, estimating that every 2 lives extended comes with about 20 overdiagnosed harmless tumors and 12 overtreatments that often cause harms such as impotence or incontinence. Combined with the 2001-2011 CA125 trial that found more early ovarian cancers but no mortality reduction, and claims that Galleri’s methylation method is now outdated because it destroys most DNA, the trend favors cancer-specific, risk-stratified use and next-generation multimodal tests using whole-genome signals, DNA fragment features, and AI to improve net benefit.

Source: Will one-stop blood tests for cancer save lives?

Subtitle: They are increasingly popular

Dateline: 2月 26, 2026 08:24 上午