临床見解
文章
2026年6月10日
When the Scan is Clear: How MRD is Changing Cancer Surveillance After Treatment
Cancer follow-up has long depended on what can be seen. But recurrence can begin below the threshold of imaging, symptoms, or routine markers. MRD adds another question: not only whether disease is visible, but whether a molecular signal may still be present, and how that signal changes over time.
The tumor has been removed and the scan is clear. The patient enters surveillance. This can bring relief for many patients. But for over 15% of patients, recurrence may begin before it can be seen. [1–6]
Recurrence is also associated with poorer outcomes, with mortality increases reported from 2- to 18-fold in selected settings. [7–9]
These numbers are not meant to alarm. Many patients do remain disease-free. However, they do explain why surveillance matters.
What Current Surveillance can Miss
After curative-intent treatment, follow-up is designed to detect recurrence early enough to guide the next clinical step. For many patients, this involves regular clinical review, protein tumor marker tests, and scheduled imaging over several years.
These tools are essential. But each has a threshold.
Imaging (CT, MRI, and PET scans) can only detect disease once it reaches a visible size. Very small lesions may still sit below the threshold of imaging, even when they already contain many cancer cells.
Tumor markers (e.g. CEA, CA 19-9, PSA) can also be difficult to interpret. They may rise because of inflammation, infection, or benign conditions, and some tumors produce little or no marker protein at all. [10]
A patient may appear disease-free by conventional assessment while disease persists at a molecular level. That is where Minimal Residual Disease (MRD) is changing the surveillance conversation.
What MRD Adds
MRD refers to molecular evidence of cancer that may remain after treatment, even when imaging and examination are clear. In solid tumors, it is most often studied through circulating tumor DNA (ctDNA): small fragments of tumor-derived DNA detectable in blood. In many cases, this residual ctDNA after treatments such as surgery is highly predictive of recurrence. This helps physicians decide whether a patient needs additional therapy. [11, 12]
Current evidence is strongest in selected cancer types and continues to expand across others.
Colorectal Cancer Surveillance: Serial plasma sequencing identified cancer recurrence at an average 8.7 months earlier than radiologic imaging. [13]
TRACERx Project (Lung Cancer): ctDNA within 4 months of surgery identified 49% of patients who later relapsed. Serial testing caught a further 20% before imaging. [14]
Across tumor types, earlier molecular detection has repeatedly preceded imaging. [15] That lead time is not automatically an outcome benefit. It is an opportunity for better context.
The Two Arms of MRD
MRD testing generally takes one of two approaches. [16]
Tumor-informed. A personalized approach that starts with tumor tissue. The tumor is sequenced first, then a personalized blood test is designed to track the mutations found in that tumor. This can improve confidence that the signal being tracked is tumor-specific.
Tumor-naive. A universal approach that does not require tissue upfront. It looks directly in blood for predefined cancer-associated alterations. This may be useful when tissue is unavailable, insufficient, or not practical to obtain, though interpretation may require additional clinical context.
The recommended approach is the method that fits the patient’s available sample, cancer type, and clinical question.
From One Result to Meaningful Pattern
MRD is not only about whether a signal is detected, but how it changes over time.
A single result can be informative, but incomplete. Too early after surgery, cell-free DNA from tissue injury may add background noise. During treatment, tumor shedding may fluctuate or become temporarily undetectable.
That is why serial testing matters.
Longitudinal tracking can show whether the signal is moving in a way that fits the clinical picture. A repeatedly undetectable signal may offer reassurance, although it cannot guarantee that recurrence will not happen. A rising signal may prompt closer attention, repeat testing, or further investigation.
This is also where the field is still maturing.
For many solid tumors, MRD is not yet fully embedded into routine treatment guidelines. A positive MRD result may indicate higher recurrence risk, but it also raises the next clinical question: what should the physician do with that information?
That question is beginning to have clearer answers in selected settings. In 2026, the first FDA approval associated with MRD in solid tumors was granted in muscle-invasive bladder cancer. For the first time, a blood-based MRD result helped define who could receive a post-surgery treatment. [17]
That does not mean every MRD-positive result has an approved treatment pathway today. It means the clinical model is changing.
MRD is moving from a marker of recurrence risk toward a tool that may help guide earlier intervention in defined patient groups. In some cancers, evidence and treatment pathways are emerging. In others, the field still needs prospective data, clearer guidance, and careful clinical judgment.
MRD should neither be oversold as a universal answer nor dismissed as just another test. Its value is context: information that may help physicians decide whether to observe, investigate, repeat testing, escalate care, or reassure.
Turning longitudinal tracking into a meaningful, actionable pattern.
References
Nors J, Iversen LH, Erichsen R, et al. JAMA Oncol. 2024; 10(1): 54–62.
Lyngholm C, Christiansen P, Damsgaard TE, et al. Acta Oncol. 2016; 55(8): 983–992.
Karacz CM, Yan J, Zhu H, et al. Clin Lung Cancer. 2019; 21(2): 127–135.e3.
Kim JH, Sinn DH, Gwak GY, et al. Clin Mol Hepatol. 2020; 26(4): 516–528.
Murakawa T, et al. World J Surg Oncol. 2023; 21: 263.
Komaya K, et al. Surgery. 2018; 163(4): 732–738.
Balboa-Barreiro V, et al. Dig Liver Dis. 2024; 56(7): 1229–1236.
Dent R, Trudeau M, Pritchard KI, et al. Curr Oncol. 2014; 21(3): e418–e425.
Consonni D, Pierobon M, Gail MH, et al. J Natl Cancer Inst. 2015; 107(6): djv059.
Duffy MJ. Tumor markers in clinical practice: a review focusing on common solid cancers. Med Princ Pract. 2013; 22(1): 4–11.
Chin RI, Chen K, Usmani A, et al. Mol Diagn Ther. 2019; 23(3): 311–331.
Zhu L, Xu R, Yang L, et al. Front Genet. 2023; 14: 1172108.
Reinert T, Henriksen TV, Christensen E, et al. JAMA Oncol. 2019; 5(8): 1124–1131.
Abbosh C, Frankell AM, Harrison T, et al. Nature. 2023; 616(7957): 553–562.
Moding EJ, Nabet BY, Alizadeh AA, Diehn M. Cancer Discov. 2021; 11(12): 2968–2986.
Honoré N, Galot R, van Marcke C, et al. Cancers (Basel). 2021; 13(21): 5361.
U.S. Food and Drug Administration. FDA approves atezolizumab for adjuvant treatment of muscle-invasive bladder cancer in patients with molecular residual disease. Published May 15, 2026.