When Breast Cancer Comes Back: Recurrence Questions

This page is for people facing a possible or confirmed breast cancer recurrence (cancer that comes back after treatment). It explains how recurrence is usually found, which tests come next, and which questions help shape the plan.

People are often surprised that breast cancer follow-up involves so little testing. That is deliberate, and it was decided by trials.

NCI states the position directly. Randomized trials tested periodic bone scans, liver ultrasound, chest x-rays and liver blood tests. None of it improved survival. None of it improved quality of life. A routine physical exam did just as well.

NCI goes further. Even when those tests find recurrent disease earlier, survival is unaffected.

On that basis, NCI says acceptable follow-up for people without symptoms after stage I to III treatment can be limited to two things:

• A physical exam
• A yearly mammogram

So a recurrence is usually found by a symptom or by a mammogram, not by a surveillance scan. That is the system working as designed, not a gap in care.

It also changes which questions are useful. Not "why was nothing scanned," but "which symptoms should prompt a call between mammograms." That is a question the team can answer specifically.

Two steps come before any treatment decision.

The first is tissue. NCI says recurrent disease is confirmed by looking at cells or tissue under a microscope (cytological or histological documentation) whenever possible. A scan showing something is not the same as a diagnosis.

The second is restaging, which means checking again how far the cancer has spread. NCI calls it indicated before treating a recurrence. The reason shows up in the numbers. Between 9% and 25% of people with a locoregional recurrence (cancer back in the breast, chest wall or nearby lymph nodes) already have distant spread, or locally extensive disease, at that moment.

That is why a local finding still triggers a wider look.

The most useful thing to know about recurrence is that the cancer may not be the same cancer.

NCI records that estrogen receptor status may change at recurrence. It cites a small Cancer and Leukemia Group B study. In it, 36% of hormone receptor-positive tumors came back receptor negative on the recurrence biopsy. Those patients had no treatment in between.

That is a third of a small sample, so it is not a precise figure. It is enough to explain why ER, PR and HER2 get rechecked rather than copied forward from the original report.

Sometimes receptor status cannot be obtained at all. NCI then lists what guides the choice instead:

• The sites of recurrence
• The disease-free interval (how long the person was free of cancer)
• The response to earlier treatment
• Menopausal status

Locoregional recurrence has become less common. NCI cites a meta-analysis suggesting a rate under 3% after breast-conserving surgery with radiation. After mastectomy the rate is somewhat higher, up to 10%.

For a recurrence in the breast after conservation, NCI says further local treatment should be considered, and names mastectomy as the example.

Margins matter in that conversation. A consensus panel reviewed a meta-analysis of 33 studies and 28,162 patients. It set the standard for an adequate margin in invasive cancer at "no ink on tumor." Wider margins did not cut recurrence further. That held for younger patients too, for lobular cancers, and for cancers with an extensive intraductal component.

A recurrence that appears during endocrine therapy raises a specific question: has the cancer found a way around it?

One known route is an ESR1 variant, a change in the gene for the estrogen receptor. NCI explains the mechanism. These variants switch the estrogen receptor on without any estrogen. That causes resistance to aromatase inhibitors. It does not necessarily cause resistance to drugs that degrade the receptor instead.

Elacestrant was built for that gap. The phase III EMERALD trial enrolled 477 patients. All had ER-positive, HER2-negative metastatic breast cancer. All had already had one or two lines of endocrine therapy and a CDK4/6 inhibitor.

ESR1 variants were present in 47.8% of them. Progression-free survival (time living without the cancer growing) favored elacestrant overall, at a hazard ratio of 0.70. In the ESR1-variant group it was 0.55. At 6 months, 40.8% of those patients on elacestrant were progression-free. On standard care it was 19.1%.

NCI also discusses PIK3CA-directed options in this setting, including alpelisib and capivasertib. Whether either applies depends on testing.

A HER2 result that once read simply "negative" now carries more information.

NCI records that 60% of HER2-negative metastatic breast cancers express low levels of HER2. The definition is technical: an immunohistochemistry (a lab stain test) score of 1+, or a score of 2+ with negative in situ hybridization.

Historically, HER2-directed drugs did nothing for this group. Trastuzumab deruxtecan changed that. It is an antibody-drug conjugate: an anti-HER2 antibody linked to a cancer-killing drug (a topoisomerase-I inhibitor payload). NCI notes a bystander effect: the payload is released into nearby HER2-low cells. A randomized phase III trial showed an overall survival benefit here.

The practical point for a recurrence: an old pathology report may say HER2-negative without recording whether it was 0 or 1+. The new biopsy can answer that.

NCI is honest about the ceiling. Recurrent breast cancer often responds to therapy, but treatment is rarely curative at that stage.

It is equally honest about the floor. People with locoregional recurrence may become long-term survivors with appropriate therapy.

Both sentences are true at once, and which one applies depends on where the disease is. That is the reason restaging comes before the treatment plan rather than after it. Local disease and distant disease are treated toward different goals.

Asking the team to name that goal out loud is worth doing. Cure, long-term control, symptom relief, or keeping a particular function are different targets, and they lead to different plans.

For the wider picture, see breast cancer, when cancer comes back and local versus distant recurrence.

Because follow-up here is deliberately quiet, the symptoms you report are what trigger a workup. These are the ones that should not wait for the next appointment.

• Call 911 or go to an emergency department if you have sudden weakness or numbness in both legs, numbness around the groin, or new loss of bladder or bowel control. Call 911 for sudden shortness of breath or chest pain. Do not drive yourself.
• Call your doctor right away or go to the emergency room if you have new back pain (sometimes with pain going down one or both legs), numbness of the legs or belly, leg weakness, or trouble passing urine. The American Cancer Society lists these as signs that cancer may be pressing on the spinal cord, and treatment started early can help prevent paralysis. A new seizure: call 911 or go straight to an emergency department. For a first-time severe headache with vomiting, new confusion, or yellowing of the skin or eyes, call your care team right away; for emergencies call 911 or go to the ER.
• Call your care team within a day or two if you notice a new lump or skin change at the scar, chest wall, armpit or other breast, bone pain that keeps building over days, breathlessness that has crept up, or swelling in the belly.

• NCI PDQ — Breast Cancer Treatment (Health Professional Version)
• NCI — Metastatic Cancer: When Cancer Spreads
• American Cancer Society — Bone Metastases