Leukemia: Relapsed or Refractory Questions

Refractory means the leukemia never went into remission. Induction chemotherapy was given, and the marrow did not clear. NCI calls this primary refractory disease.

Relapsed means the leukemia did go into remission, and later came back. Both can happen after the same starting treatment. They are not the same problem, and they do not carry the same odds.

Ask which word applies to you, and ask what evidence it rests on: a bone marrow biopsy, a blood count, a flow cytometry result, or a molecular test.

In acute myeloid leukemia (AML), the length of the first remission is one of the strongest predictors of whether salvage treatment will work.

The FLAG regimen shows this plainly. FLAG combines fludarabine, cytarabine and filgrastim. In a multi-center phase II study of 83 patients, complete remission rates split sharply by interval. Among patients whose first remission had lasted 6 months or more, 81 percent reached a second complete remission. Among those whose first remission lasted under 6 months, or who were primarily refractory, 30 percent did.

Same drugs. Same doses. Very different outcome. Ask how long your first remission lasted, in months, and ask what that means for the regimen being proposed.

Remission has a definition with numbers in it. NCI lists the response categories used in AML, and they are worth having in front of you when results are discussed.

• Complete remission (CR): marrow blasts under 5 percent, no leukemia cells circulating in the blood, no Auer rods, no disease outside the marrow, absolute neutrophil count at or above 1,000 per microliter, and platelets at or above 100,000 per microliter.
• CR with incomplete count recovery (CRi): everything above, except the neutrophils or platelets have not yet recovered.
• CR without measurable residual disease: a complete remission where sensitive testing also finds nothing. The tests are RT-qPCR, which looks for a genetic marker, or multicolor flow cytometry.

That last category matters. A marrow can look clear under a microscope and still carry leukemia that a molecular test can detect. Ask whether measurable residual disease testing was done, which method was used, and what the result was.

NCI states this directly: no standard treatment regimen exists for refractory or recurrent AML. That sentence is not a gap in the guidance. It is the guidance, and it is the reason clinical trials matter so much here.

The listed options are chemotherapy, either intensive salvage or reduced-intensity including targeted drugs, and allogeneic stem cell transplant.

Beyond FLAG, the other commonly cited intensive regimen is MEC, which combines mitoxantrone, etoposide and cytarabine. In a study of 74 patients with poor-prognosis AML, MEC produced complete remission in 55 percent. In a randomized ECOG trial of 129 patients with worse disease status, including relapse within 6 months and relapse after transplant, complete response rates were only 17 to 25 percent.

Those two numbers describe the same drugs in different populations. When a remission rate is quoted to you, ask which population it came from.

Some AML carries a change in the FLT3 gene. That opens a specific option.

Gilteritinib is an oral FLT3 inhibitor. It works against both of the main FLT3 subtypes, called ITD and TKD. A phase III trial randomly assigned 371 patients, two to one, to gilteritinib or to salvage chemotherapy chosen in advance. Median overall survival was 9.3 months with gilteritinib and 5.6 months with chemotherapy, with a hazard ratio for death of 0.665. Estimated survival at 2 years was 20.6 percent versus 14.2 percent.

Ask whether your leukemia has been retested at relapse. Genetic features can change between diagnosis and relapse, and a test from a year ago may no longer describe the disease in front of you.

For relapsed or refractory B-cell acute lymphoblastic leukemia, two FDA-approved drugs have randomized evidence against standard chemotherapy.

Blinatumomab is a bispecific antibody. One end grips CD19 on the leukemia cell, the other grips CD3 on a T cell, and it pulls the immune cell onto the target. In a randomized phase III trial against four standard reinduction regimens, remission rates were 43.9 percent with blinatumomab and 24.6 percent with standard treatment.

Inotuzumab ozogamicin is an antibody-drug conjugate. The antibody targets CD22 and carries a toxin called calicheamicin directly into the cell. In a randomized phase III trial of 218 adults receiving first or second salvage treatment, complete remission or remission with incomplete count recovery reached 80.7 percent with inotuzumab, versus 29.4 percent with standard regimens.

Both drugs depend on a marker being present. Ask whether your leukemia cells express CD19 and CD22, and ask for the result rather than the assumption.

One more test belongs in this conversation. The Philadelphia chromosome, which produces the BCR::ABL1 fusion gene, appears in about 20 percent of adults with ALL. NCI notes it can be missed on standard chromosome analysis and may show up only on FISH or RT-PCR testing.

Allogeneic stem cell transplant is on NCI's list for refractory and recurrent leukemia. It is also the option with the longest tail of consequences.

NCI lists the late effects that can follow: chronic fatigue, thyroid and gonadal problems, infertility, ongoing infection risk, accelerated coronary heart disease, thinning bones, cataracts, iron overload, psychological effects and second cancers.

One study puts a number on the broader effect. In the Bone Marrow Transplant Survivor Study, transplant survivors were 8.4 times more likely to be frail than their own siblings. Frailty in turn carried a 2.76-fold increase in the risk of death compared with a non-frail state.

None of that is an argument against transplant. It is an argument for asking, before consenting, what the plan is for fertility, for thyroid and heart monitoring, and for long-term follow-up.

• Is my leukemia refractory or relapsed, and how long did my first remission last?
• Has the leukemia been re-tested at relapse for genetics and for surface markers?
• Was measurable residual disease measured, by which method, and what did it show?
• What is the goal of the treatment you are proposing: getting back to remission, bridging to transplant, or controlling symptoms?
• Am I a transplant candidate, and if so, has a donor search started?
• Which trials are open here, and does eligibility depend on how many prior lines of treatment I have had?

That last question matters more in leukemia than in most cancers. Many trials cap the number of prior treatments, so taking a standard regimen first can close a trial door. Our page on clinical trial versus standard treatment covers how to compare the two.

Salvage treatment causes deep drops in blood counts. NCI notes that myelosuppression and mucositis were common toxicities with FLAG. Bleeding that will not stop is an emergency: call 911 or go to an emergency department, because platelets can be transfused there without waiting on a callback. A fever or shaking chills during FLAG is an emergency in its own right, though the first call is the leukemia team's 24-hour number rather than 911: with counts this low, antibiotics are meant to start within the hour, so go to an emergency department if nobody answers within a few minutes. Reasons to contact the team the same day include mouth sores that stop you drinking, new shortness of breath, or sudden confusion.

Ask for these written down, with a daytime number and an after-hours number, and keep them where a family member can find them.

Symptom-focused care runs alongside treatment rather than after it. Our page on palliative care explains what that team does. For the disease overall, start with leukemia.

• NCI PDQ on adult AML, for doctors
• NCI PDQ on adult ALL, for doctors
• NCI PDQ on adult AML, for patients