Conference Coverage

MRD predicts outcome of HSCT in ALL, study suggests


 

HSCT preparation Photo by Chad McNeeley Photo by Chad McNeeley

HSCT preparation

ORLANDO, FL—Minimal residual disease (MRD) measurements before and after hematopoietic stem cell transplant (HSCT) can help predict outcomes in patients with childhood acute lymphoblastic leukemia (ALL), according to researchers.

Their work also suggests several other factors can be used to predict event-free survival (EFS) in this patient population, and the team developed risk scores incorporating these factors.

Michael A. Pulsipher, MD, of Children’s Hospital Los Angeles in California, presented this work as one of the “Best Abstracts” at the 2017 BMT Tandem Meetings (abstract 4*).

“The new risk scores that we were able to develop very nicely predict outcomes post-transplant and can guide study planning,” Dr Pulsipher said.

“MRD pre-transplant was a very powerful predictor of outcome, and MRD post-transplant highlights individual patients at risk.”

For this study, Dr Pulsipher and his colleagues retrospectively analyzed 747 patients treated in Europe, North America, and Australia. The patients received transplants between September 1999 and May 2015.

Most patients had pre-B ALL (78%, n=586), 19% (n=145) had T-cell ALL, 2% had “other” ALLs (n=8) or no data on ALL type (n=8). Sixty-two percent (n=466) were male.

Nearly half of patients were between the ages of 2 and 10 (49%, n=365), 47% (n=351) were older than 10, and 4% (n=31) were younger than 2.

Transplant details

Patients received grafts from matched unrelated donors (42%, n=314), matched sibling donors (30%, n=227), mismatched donors (10%, n=75), and cord blood from unrelated donors (17%, n=128). There was no data on donor type for 3 patients.

Most patients received bone marrow transplants (61%, n=458), 20% (n=147) received cord blood, and 18% (n=131) received peripheral blood stem cells. Eight patients received “other” types of transplants, and 3 patients had no data on stem cell source.

More than half of the patients (55%, n=410) were in their second complete remission (CR) at transplant. Thirty-seven percent were in their first CR (n=275), 7% were in their third or greater CR (n=53), and 1% were not in remission (n=7). Two patients had no data on remission status.

MRD

MRD was assessed before HSCT as well as after—on or near days 30, 60, 90, 180, 365, and beyond.

There were 4 MRD categories:

  • MRD negative: No signal
  • MRD low: >0 to <10-4 (<0.01%)
  • MRD high: ≥10-4 to <10-3 (0.01 to 0.1%)
  • MRD very high: ≥10-3 to <10-2 (>0.1%).

Dr Pulsipher noted that, when analyzing MRD pre-HSCT or at 30 days after HSCT, the estimated 5-year EFS was similar for patients in the MRD-negative and MRD-low groups. However, as time went on (at days 90, 180, and 365), any detectable level of MRD was associated with a poor prognosis.

“And patients arriving at day 365 with no detectable MRD had an exceptional prognosis, with survival approaching 90%,” Dr Pulsipher said.

He also pointed out an interaction between acute graft-vs-host disease (aGVHD) and MRD post-HSCT. He and his colleagues observed better survival for MRD-positive patients with aGVHD (grade 1-2) than for MRD-positive patients without aGVHD.

Pre-HSCT risk score

Via an adjusted Cox regression analysis, the researchers identified several pre-transplant factors that predicted EFS at 18 months.

These included remission status, donor type, immunophenotype, and MRD. The researchers assigned points to each of these factors to create a risk score.

Compared to patients in first CR, the hazard ratio (HR) for patients in early second CR was 2.53, and the score was 3. For patients in third CR or greater, the HR was 1.95, and the score was 2.

Compared to patients with a matched sibling donor, the HR for patients with a mismatched donor was 1.41, and the score was 1. For patients who received cord blood from an unrelated donor, the HR was 1.48, and the score was 1.

Compared to patients with T-cell ALL, the HR for patients with pre-B ALL was 1.35, and the score was 1.

Compared to patients with MRD <10-4, the HR for patients with MRD ≥10-4 was 2.32, and the score was 2.

The probability of EFS at 18 months was 78% ± 2% for patients with 0 to 1 points, 54% ± 3% for those with 2 to 3 points, and 46% ± 5% for patients with 4 or more points.

Day 30 post-HSCT risk score

When considering patients at day 30 post-HSCT, factors that predicted 18-month EFS included remission status, donor type, immunophenotype, aGVHD status, and MRD.

The HR for patients in early second CR was 2.51, and the score was 3. For patients in third CR or greater, the HR was 2.09, and the score was 2.

The HR for patients with a mismatched donor was 1.75, and the score was 2. The HR for patients with pre-B ALL was 1.40, and the score was 1.

Compared to patients with grade 1-2 aGVHD, the HR was 2.02 for patients with grade 0 aGVHD, and the score was 2. For patients with grade 3 aGVHD, the HR was 1.44, and the score was 1. For patients with grade 4 aGVHD, the HR was 7.12, and the score was 7.

The researchers evaluated MRD prior to HSCT and MRD at day 30, using a reference of MRD <10-4 at both time points. For patients with MRD <10-4 pre-HSCT and ≥10-4 at day 30, the HR was 2.29, and the score was 2.

For patients with MRD ≥10-4 pre-HSCT and <10-4 at day 30, the HR was 3.17, and the score was 3. For patients with MRD ≥10-4 pre-HSCT and at day 30, the HR was 3.63, and the score was 4.

The probability of EFS at 18 months was 80% ± 2% for patients with 0 to 3 points, 54% ± 4% for those with 4 to 6 points, and 25% ± 6% for those with 7 or more points.

Day 90 post-HSCT risk score

When considering patients at day 90 post-HSCT, factors that predicted 18-month EFS included remission status, aGVHD status, and MRD.

For patients in early second CR, the HR was 2.81, and the score was 3. For those in third CR or greater, the HR was 1.85, and the score was 2.

Compared to patients with grade 1-2 aGVHD, the HR was 1.60 for patients with grade 0 aGVHD, and the score was 2. For patients with grade 4 aGVHD, the HR was 2.49, and the score was 2.

The researchers assessed MRD prior to HSCT and MRD at day 90, using a reference of MRD <10-4 at both time points. For patients with MRD <10-4 pre-HSCT and ≥10-4 at day 90, the HR was 6.03, and the score was 6.

For patients with MRD ≥10-4 pre-HSCT and <10-4 at day 90, the HR was 3.11, and the score was 3. For patients with MRD ≥10-4 pre-HSCT and at day 90, the HR was 4.59, and the score was 5.

The probability of EFS at 18 months was 83% ± 2% for patients with 0 to 2 points, 60% ± 4% for those with 3 to 5 points, and 17% ± 11 for those with 6 or more points.

*Information in the abstract differs from the presentation.

Next Article: