BMT 2014

GRAPEVINE, TEXAS—An analysis of more than 8000 patients provides insights regarding HLA disparity that may help optimize outcomes in individuals undergoing unrelated-donor hematopoietic stem cell transplant.

The study showed that a single allele- or antigen-level HLA mismatch (7/8) increased the risk for acute and chronic graft-vs-host disease (GVHD) and worsened survival rates.

However, there were no locus-specific effects on survival, and there was no impact of allele- vs antigen-level mismatch on survival.

In addition, patients with an 8/8 matched graft had an increased risk of acute GVHD if they had a DQB1 mismatch or a DPB1 mismatch. DPB1 mismatch also decreased the risk of relapse, and nonpermissive DPB1 mismatch was associated with worse survival.

“Thus, we believe that consideration of DPB1 in donor selection may permit skewing toward donors with permissive DPB1 mismatch and may improve outcomes,” said study investigator Joseph Pidala, MD, of the H. Lee Moffitt Cancer Center in Tampa, Florida.

He presented these findings at the 2014 BMT Tandem Meetings as abstract 5, which was designated one of the meeting’s “Best Abstracts.”

Patient characteristics

Dr Pidala and his colleagues evaluated data from 8003 adult and pediatric patients who had undergone their first myeloablative, unrelated transplant between 1999 and 2011. The patients had been diagnosed with acute myeloid leukemia, acute lymphoblastic leukemia, chronic myeloid leukemia, or myelodysplastic syndrome.

Patients and their donors had high-resolution typing for HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1. Most cases were 8/8 matched (n=5449), followed by 7/8 (n=2071) and 6/8 (n=483).

“Those with a greater extent of HLA mismatch were younger, less likely to be Caucasian, they differed according to disease indication for transplant, and had a greater proportion of advanced disease stage,” Dr Pidala noted.

“Additionally, those with a greater extent of HLA mismatch were more likely to receive bone marrow grafts in comparison to peripheral blood, total body irradiation, and also to receive in vivo T-cell depletion.”

Furthermore, there was a declining proportion of 6/8 match over time. It decreased from 49% (1999-2002) to 37% (2003-2006) to 14% (2007-2011).

Effects of HLA mismatch

In all analyses, the researchers considered findings significant if the P value was less than 0.01.

Compared to an 8/8 matched graft, receiving a 7/8 graft was significantly associated with an increase in acute grade 2-4, acute grade 3-4, and chronic GVHD; higher transplant-related mortality (TRM); lower disease-free survival (DFS) among patients with early stage or intermediate (but not advanced) disease; and lower overall survival (OS) regardless of disease stage.

Receiving a 6/8 graft was significantly associated with an increase in acute grade 2-4 and 3-4 GVHD, increased TRM, decreased DFS (early stage or intermediate disease), and decreased OS (early stage or intermediate disease).

“In no cases did we find [that mismatch had] an impact on the incidence of primary disease relapse,” Dr Pidala said. “Comparing 7/8 to 6/8 cases, we found that those with 7/8 had improved transplant-related mortality and disease-free and overall survival [only] in the context of early stage disease.”

The team also confirmed these findings in an analysis of 5846 unique cases. In this cohort, 2528 patients were 8/8 matched, 882 were 7/8, and 157 were 6/8.

Locus-specific effects

Mismatch at HLA-A (n=743) was significantly associated with an increase in acute grade 2-4/3-4 and chronic GVHD, increased TRM, decreased DFS, and decreased OS.

Mismatch at HLA-B (n=345) was significantly associated with an increase in grade 2-4/3-4 acute GHVD, chronic GVHD, and TRM. And mismatch at HLA-C (n=766) was significantly associated with an increase in acute grade 3-4 GVHD and TRM and a decrease in DFS and OS.

There were no significant differences in the case of DRB1 mismatch. However, Dr Pidala noted that this group had the smallest number of patients, at 217.

“I’d also like to point out that in direct, pair-wise comparisons across the mismatched loci, we did not observe any significant differences in overall survival,” Dr Pidala said. “Similarly, we found no impact of allele- vs antigen-level mismatch on overall survival.”

DP and DQ mismatch

Among 8/8 matched cases, DPB1 mismatch led to a significantly increased risk for acute grade 2-4 and 3-4 GHVD, as well as a significant reduction in the risk of relapse. This was the case whether patients had a single- or double-allele mismatch.

The addition of DQB1 mismatch significantly increased the risk of grade 2-4 acute GVHD among 8/8 matched cases only if it was a single-antigen mismatch.

Neither DPB1 nor DQB1 mismatch had a significant effect on OS, DFS, or TRM in 8/8 matched cases. And neither DPB1 nor DQB1 mismatch had a significant effect on outcomes of patients who received 7/8 matched grafts.

Among 8/8 matched cases, those with permissive DPB1 mismatch had a significantly lower risk of acute grade 2-4 and 3-4 GVHD and a significantly higher risk of relapse than patients with nonpermissive DPB1 mismatch. Patients with nonpermissive mismatch also had significantly higher TRM and significantly lower DFS and OS.

However, there were no significant differences between permissive and nonpermissive mismatches for patients with 7/8 matched grafts.

Treatment implications

Based on these results and those of previous studies, Dr Pidala said we can conclude that 8/8 matched grafts confer better survival than mismatched grafts. And a permissive DPB1 mismatch can further improve survival in 8/8 cases.

Among patients receiving a 7/8 matched graft, it is preferable to identify those with HLA-C 03:03/03:04 mismatches, as these patients appear to have mortality rates comparable to 8/8 matched cases.

Among 7/8 matched cases, we should avoid nonpermissive DPB1 mismatch, having 3 or more low-expression loci mismatches (DP, DQ, DRB3/4/5), and nonpermissive allele combinations and amino acid substitutions.

GRAPEVINE, TEXAS—An analysis of more than 8000 patients provides insights regarding HLA disparity that may help optimize outcomes in individuals undergoing unrelated-donor hematopoietic stem cell transplant.

The study showed that a single allele- or antigen-level HLA mismatch (7/8) increased the risk for acute and chronic graft-vs-host disease (GVHD) and worsened survival rates.

However, there were no locus-specific effects on survival, and there was no impact of allele- vs antigen-level mismatch on survival.

In addition, patients with an 8/8 matched graft had an increased risk of acute GVHD if they had a DQB1 mismatch or a DPB1 mismatch. DPB1 mismatch also decreased the risk of relapse, and nonpermissive DPB1 mismatch was associated with worse survival.

“Thus, we believe that consideration of DPB1 in donor selection may permit skewing toward donors with permissive DPB1 mismatch and may improve outcomes,” said study investigator Joseph Pidala, MD, of the H. Lee Moffitt Cancer Center in Tampa, Florida.

He presented these findings at the 2014 BMT Tandem Meetings as abstract 5, which was designated one of the meeting’s “Best Abstracts.”

Patient characteristics

Dr Pidala and his colleagues evaluated data from 8003 adult and pediatric patients who had undergone their first myeloablative, unrelated transplant between 1999 and 2011. The patients had been diagnosed with acute myeloid leukemia, acute lymphoblastic leukemia, chronic myeloid leukemia, or myelodysplastic syndrome.

Patients and their donors had high-resolution typing for HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1. Most cases were 8/8 matched (n=5449), followed by 7/8 (n=2071) and 6/8 (n=483).

“Those with a greater extent of HLA mismatch were younger, less likely to be Caucasian, they differed according to disease indication for transplant, and had a greater proportion of advanced disease stage,” Dr Pidala noted.

“Additionally, those with a greater extent of HLA mismatch were more likely to receive bone marrow grafts in comparison to peripheral blood, total body irradiation, and also to receive in vivo T-cell depletion.”

Furthermore, there was a declining proportion of 6/8 match over time. It decreased from 49% (1999-2002) to 37% (2003-2006) to 14% (2007-2011).

Effects of HLA mismatch

In all analyses, the researchers considered findings significant if the P value was less than 0.01.

Compared to an 8/8 matched graft, receiving a 7/8 graft was significantly associated with an increase in acute grade 2-4, acute grade 3-4, and chronic GVHD; higher transplant-related mortality (TRM); lower disease-free survival (DFS) among patients with early stage or intermediate (but not advanced) disease; and lower overall survival (OS) regardless of disease stage.

Receiving a 6/8 graft was significantly associated with an increase in acute grade 2-4 and 3-4 GVHD, increased TRM, decreased DFS (early stage or intermediate disease), and decreased OS (early stage or intermediate disease).

“In no cases did we find [that mismatch had] an impact on the incidence of primary disease relapse,” Dr Pidala said. “Comparing 7/8 to 6/8 cases, we found that those with 7/8 had improved transplant-related mortality and disease-free and overall survival [only] in the context of early stage disease.”

The team also confirmed these findings in an analysis of 5846 unique cases. In this cohort, 2528 patients were 8/8 matched, 882 were 7/8, and 157 were 6/8.

Locus-specific effects

Mismatch at HLA-A (n=743) was significantly associated with an increase in acute grade 2-4/3-4 and chronic GVHD, increased TRM, decreased DFS, and decreased OS.

Mismatch at HLA-B (n=345) was significantly associated with an increase in grade 2-4/3-4 acute GHVD, chronic GVHD, and TRM. And mismatch at HLA-C (n=766) was significantly associated with an increase in acute grade 3-4 GVHD and TRM and a decrease in DFS and OS.

There were no significant differences in the case of DRB1 mismatch. However, Dr Pidala noted that this group had the smallest number of patients, at 217.

“I’d also like to point out that in direct, pair-wise comparisons across the mismatched loci, we did not observe any significant differences in overall survival,” Dr Pidala said. “Similarly, we found no impact of allele- vs antigen-level mismatch on overall survival.”

DP and DQ mismatch

Among 8/8 matched cases, DPB1 mismatch led to a significantly increased risk for acute grade 2-4 and 3-4 GHVD, as well as a significant reduction in the risk of relapse. This was the case whether patients had a single- or double-allele mismatch.

The addition of DQB1 mismatch significantly increased the risk of grade 2-4 acute GVHD among 8/8 matched cases only if it was a single-antigen mismatch.

Neither DPB1 nor DQB1 mismatch had a significant effect on OS, DFS, or TRM in 8/8 matched cases. And neither DPB1 nor DQB1 mismatch had a significant effect on outcomes of patients who received 7/8 matched grafts.

Among 8/8 matched cases, those with permissive DPB1 mismatch had a significantly lower risk of acute grade 2-4 and 3-4 GVHD and a significantly higher risk of relapse than patients with nonpermissive DPB1 mismatch. Patients with nonpermissive mismatch also had significantly higher TRM and significantly lower DFS and OS.

However, there were no significant differences between permissive and nonpermissive mismatches for patients with 7/8 matched grafts.

Treatment implications

Based on these results and those of previous studies, Dr Pidala said we can conclude that 8/8 matched grafts confer better survival than mismatched grafts. And a permissive DPB1 mismatch can further improve survival in 8/8 cases.

Among patients receiving a 7/8 matched graft, it is preferable to identify those with HLA-C 03:03/03:04 mismatches, as these patients appear to have mortality rates comparable to 8/8 matched cases.

Among 7/8 matched cases, we should avoid nonpermissive DPB1 mismatch, having 3 or more low-expression loci mismatches (DP, DQ, DRB3/4/5), and nonpermissive allele combinations and amino acid substitutions.

GRAPEVINE, TEXAS—An analysis of more than 8000 patients provides insights regarding HLA disparity that may help optimize outcomes in individuals undergoing unrelated-donor hematopoietic stem cell transplant.

The study showed that a single allele- or antigen-level HLA mismatch (7/8) increased the risk for acute and chronic graft-vs-host disease (GVHD) and worsened survival rates.

However, there were no locus-specific effects on survival, and there was no impact of allele- vs antigen-level mismatch on survival.

In addition, patients with an 8/8 matched graft had an increased risk of acute GVHD if they had a DQB1 mismatch or a DPB1 mismatch. DPB1 mismatch also decreased the risk of relapse, and nonpermissive DPB1 mismatch was associated with worse survival.

“Thus, we believe that consideration of DPB1 in donor selection may permit skewing toward donors with permissive DPB1 mismatch and may improve outcomes,” said study investigator Joseph Pidala, MD, of the H. Lee Moffitt Cancer Center in Tampa, Florida.

He presented these findings at the 2014 BMT Tandem Meetings as abstract 5, which was designated one of the meeting’s “Best Abstracts.”

Patient characteristics

Dr Pidala and his colleagues evaluated data from 8003 adult and pediatric patients who had undergone their first myeloablative, unrelated transplant between 1999 and 2011. The patients had been diagnosed with acute myeloid leukemia, acute lymphoblastic leukemia, chronic myeloid leukemia, or myelodysplastic syndrome.

Patients and their donors had high-resolution typing for HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1. Most cases were 8/8 matched (n=5449), followed by 7/8 (n=2071) and 6/8 (n=483).

“Those with a greater extent of HLA mismatch were younger, less likely to be Caucasian, they differed according to disease indication for transplant, and had a greater proportion of advanced disease stage,” Dr Pidala noted.

“Additionally, those with a greater extent of HLA mismatch were more likely to receive bone marrow grafts in comparison to peripheral blood, total body irradiation, and also to receive in vivo T-cell depletion.”

Furthermore, there was a declining proportion of 6/8 match over time. It decreased from 49% (1999-2002) to 37% (2003-2006) to 14% (2007-2011).

Effects of HLA mismatch

In all analyses, the researchers considered findings significant if the P value was less than 0.01.

Compared to an 8/8 matched graft, receiving a 7/8 graft was significantly associated with an increase in acute grade 2-4, acute grade 3-4, and chronic GVHD; higher transplant-related mortality (TRM); lower disease-free survival (DFS) among patients with early stage or intermediate (but not advanced) disease; and lower overall survival (OS) regardless of disease stage.

Receiving a 6/8 graft was significantly associated with an increase in acute grade 2-4 and 3-4 GVHD, increased TRM, decreased DFS (early stage or intermediate disease), and decreased OS (early stage or intermediate disease).

“In no cases did we find [that mismatch had] an impact on the incidence of primary disease relapse,” Dr Pidala said. “Comparing 7/8 to 6/8 cases, we found that those with 7/8 had improved transplant-related mortality and disease-free and overall survival [only] in the context of early stage disease.”

The team also confirmed these findings in an analysis of 5846 unique cases. In this cohort, 2528 patients were 8/8 matched, 882 were 7/8, and 157 were 6/8.

Locus-specific effects

Mismatch at HLA-A (n=743) was significantly associated with an increase in acute grade 2-4/3-4 and chronic GVHD, increased TRM, decreased DFS, and decreased OS.

Mismatch at HLA-B (n=345) was significantly associated with an increase in grade 2-4/3-4 acute GHVD, chronic GVHD, and TRM. And mismatch at HLA-C (n=766) was significantly associated with an increase in acute grade 3-4 GVHD and TRM and a decrease in DFS and OS.

There were no significant differences in the case of DRB1 mismatch. However, Dr Pidala noted that this group had the smallest number of patients, at 217.

“I’d also like to point out that in direct, pair-wise comparisons across the mismatched loci, we did not observe any significant differences in overall survival,” Dr Pidala said. “Similarly, we found no impact of allele- vs antigen-level mismatch on overall survival.”

DP and DQ mismatch

Among 8/8 matched cases, DPB1 mismatch led to a significantly increased risk for acute grade 2-4 and 3-4 GHVD, as well as a significant reduction in the risk of relapse. This was the case whether patients had a single- or double-allele mismatch.

The addition of DQB1 mismatch significantly increased the risk of grade 2-4 acute GVHD among 8/8 matched cases only if it was a single-antigen mismatch.

Neither DPB1 nor DQB1 mismatch had a significant effect on OS, DFS, or TRM in 8/8 matched cases. And neither DPB1 nor DQB1 mismatch had a significant effect on outcomes of patients who received 7/8 matched grafts.

Among 8/8 matched cases, those with permissive DPB1 mismatch had a significantly lower risk of acute grade 2-4 and 3-4 GVHD and a significantly higher risk of relapse than patients with nonpermissive DPB1 mismatch. Patients with nonpermissive mismatch also had significantly higher TRM and significantly lower DFS and OS.

However, there were no significant differences between permissive and nonpermissive mismatches for patients with 7/8 matched grafts.

Treatment implications

Based on these results and those of previous studies, Dr Pidala said we can conclude that 8/8 matched grafts confer better survival than mismatched grafts. And a permissive DPB1 mismatch can further improve survival in 8/8 cases.

Among patients receiving a 7/8 matched graft, it is preferable to identify those with HLA-C 03:03/03:04 mismatches, as these patients appear to have mortality rates comparable to 8/8 matched cases.

Among 7/8 matched cases, we should avoid nonpermissive DPB1 mismatch, having 3 or more low-expression loci mismatches (DP, DQ, DRB3/4/5), and nonpermissive allele combinations and amino acid substitutions.

GRAPEVINE, TEXAS—Adding specialized T cells to allogeneic bone marrow transplants can prevent graft-vs-host disease (GVHD) while preserving the graft-vs-tumor (GVT) effect, preclinical research suggests.

Donor-derived CD4+ invariant natural killer T cells (iNKT cells) helped protect mice from developing GVHD and lymphoma, thereby improving their survival.

As human CD4+ iNKT cells can be expanded in vitro, these findings may translate to the clinic, said study investigator Dominik Schneidawind, MD, of Stanford University in California.

Dr Schneidawind discussed this research at the 2014 BMT Tandem Meetings as abstract 2*, which was designated one of the meeting’s “Best Abstracts.”

Improved survival

Dr Schneidawind and his colleagues created a model of allogeneic bone marrow transplant in Balb/c (H-2Kd) mice. The mice were first irradiated with 8 Gy and then received T cell-depleted bone marrow, along with 1 x 10⁶ CD4/CD8 T cells (Tcon) from C57Bl/6 (H-2Kb) donor mice and varying doses of CD4+ iNKT cells: 1 x 10⁴ to 1 x 10⁵.

Mice that received 100,000 CD4+ iNKT cells had significantly improved survival (P=0.002)—and better weight and GVHD scores—compared to mice that received bone marrow grafts alone. But a CD4+ iNKT dose as low as 10,000 cells also conferred a significant improvement in survival (P=0.03).

“In contrast, you need roughly 500,000 to 1 million donor-derived regulatory T cells to achieve a comparable survival benefit in this model of bone marrow transplantation,” Dr Schneidawind noted.

GVT effect

Dr Schneidawind and his colleagues also evaluated the impact of CD4+ iNKT cells on the GVT effect. They used luciferase-transfected BCL1 cells to induce lymphoma development, injecting the cells into mice 1 day before transplant.

Mice that received BCL1 cells, with or without CD4+ iNKT cells, showed increasing bioluminescence imaging (BLI) signals and all died from progressive lymphoma. The mice that received BLC1 and Tcon did not show an increase in BLI signals, but they all died from GVHD.

However, the mice that received BCL1, Tcon, and CD4+ iNKT cells didn’t show any sign of lymphoma or GVHD and survived through the whole experiment (P=0.002).

The investigators conducted the same experiment with the luciferase-positive A20 cell line. And they observed similar results. Mice that received A20 cells, Tcon, and CD4+ iNKT cells showed no sign of lymphoma and had significantly improved survival (P=0.002) over animals that received A20 and Tcon.

“Interestingly, animals that received A20 cells and 50,000 CD4+ iNKT cells only, without Tcon, showed a significant decrease in BLI signal and significantly improved survival,” Dr Schneidawind noted.

“So we conclude from these experiments that CD4+ iNKT cells do not abrogate the Tcon-mediated [GVT] effect but might exert some distinct antitumor activity by themselves.”

Mechanism of GVHD prevention

Lastly, Dr Schneidawind and his colleagues wanted to determine exactly how CD4+ iNKT cells prevent GVHD. The group’s experiments showed that the cells inhibit the proliferation of alloreactive T cells and promote the expansion of donor-derived CD4+ FoxP3+ Tregs.

The investigators decided to take a closer look at the Tregs, evaluating their role in GVHD-related death.

To do this, the team evaluated survival in 4 groups of mice: (1) irradiated controls; (2) mice that received only bone marrow; (3) mice that received Tcon, CD4+ iNKT cells, and a graft depleted of CD4+ FoxP3+ Tregs; and (4) mice that received Tcon, CD4+ iNKT cells, and a non-depleted graft.

“Animals that received a Treg-depleted graft and were treated with CD4+ iNKT [plus Tcon] did not show a significant improvement in survival [over controls],” Dr Schneidawind said.

“In contrast, animals that received a Treg-non-depleted graft showed a significant improvement in survival [P=0.006]. So we conclude that the expansion of these Tregs is necessary to protect from GVHD lethality and that they are required in this context.”

In closing, Dr Schneidawind noted that, although human CD4+ iNKT cells are rare, they can be isolated from peripheral blood and expanded in vitro. This suggests his group’s findings might ultimately prove useful for patients.

*Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—Adding specialized T cells to allogeneic bone marrow transplants can prevent graft-vs-host disease (GVHD) while preserving the graft-vs-tumor (GVT) effect, preclinical research suggests.

Donor-derived CD4+ invariant natural killer T cells (iNKT cells) helped protect mice from developing GVHD and lymphoma, thereby improving their survival.

As human CD4+ iNKT cells can be expanded in vitro, these findings may translate to the clinic, said study investigator Dominik Schneidawind, MD, of Stanford University in California.

Dr Schneidawind discussed this research at the 2014 BMT Tandem Meetings as abstract 2*, which was designated one of the meeting’s “Best Abstracts.”

Improved survival

Dr Schneidawind and his colleagues created a model of allogeneic bone marrow transplant in Balb/c (H-2Kd) mice. The mice were first irradiated with 8 Gy and then received T cell-depleted bone marrow, along with 1 x 10⁶ CD4/CD8 T cells (Tcon) from C57Bl/6 (H-2Kb) donor mice and varying doses of CD4+ iNKT cells: 1 x 10⁴ to 1 x 10⁵.

Mice that received 100,000 CD4+ iNKT cells had significantly improved survival (P=0.002)—and better weight and GVHD scores—compared to mice that received bone marrow grafts alone. But a CD4+ iNKT dose as low as 10,000 cells also conferred a significant improvement in survival (P=0.03).

“In contrast, you need roughly 500,000 to 1 million donor-derived regulatory T cells to achieve a comparable survival benefit in this model of bone marrow transplantation,” Dr Schneidawind noted.

GVT effect

Dr Schneidawind and his colleagues also evaluated the impact of CD4+ iNKT cells on the GVT effect. They used luciferase-transfected BCL1 cells to induce lymphoma development, injecting the cells into mice 1 day before transplant.

Mice that received BCL1 cells, with or without CD4+ iNKT cells, showed increasing bioluminescence imaging (BLI) signals and all died from progressive lymphoma. The mice that received BLC1 and Tcon did not show an increase in BLI signals, but they all died from GVHD.

However, the mice that received BCL1, Tcon, and CD4+ iNKT cells didn’t show any sign of lymphoma or GVHD and survived through the whole experiment (P=0.002).

The investigators conducted the same experiment with the luciferase-positive A20 cell line. And they observed similar results. Mice that received A20 cells, Tcon, and CD4+ iNKT cells showed no sign of lymphoma and had significantly improved survival (P=0.002) over animals that received A20 and Tcon.

“Interestingly, animals that received A20 cells and 50,000 CD4+ iNKT cells only, without Tcon, showed a significant decrease in BLI signal and significantly improved survival,” Dr Schneidawind noted.

“So we conclude from these experiments that CD4+ iNKT cells do not abrogate the Tcon-mediated [GVT] effect but might exert some distinct antitumor activity by themselves.”

Mechanism of GVHD prevention

Lastly, Dr Schneidawind and his colleagues wanted to determine exactly how CD4+ iNKT cells prevent GVHD. The group’s experiments showed that the cells inhibit the proliferation of alloreactive T cells and promote the expansion of donor-derived CD4+ FoxP3+ Tregs.

The investigators decided to take a closer look at the Tregs, evaluating their role in GVHD-related death.

To do this, the team evaluated survival in 4 groups of mice: (1) irradiated controls; (2) mice that received only bone marrow; (3) mice that received Tcon, CD4+ iNKT cells, and a graft depleted of CD4+ FoxP3+ Tregs; and (4) mice that received Tcon, CD4+ iNKT cells, and a non-depleted graft.

“Animals that received a Treg-depleted graft and were treated with CD4+ iNKT [plus Tcon] did not show a significant improvement in survival [over controls],” Dr Schneidawind said.

“In contrast, animals that received a Treg-non-depleted graft showed a significant improvement in survival [P=0.006]. So we conclude that the expansion of these Tregs is necessary to protect from GVHD lethality and that they are required in this context.”

In closing, Dr Schneidawind noted that, although human CD4+ iNKT cells are rare, they can be isolated from peripheral blood and expanded in vitro. This suggests his group’s findings might ultimately prove useful for patients.

*Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—Adding specialized T cells to allogeneic bone marrow transplants can prevent graft-vs-host disease (GVHD) while preserving the graft-vs-tumor (GVT) effect, preclinical research suggests.

Donor-derived CD4+ invariant natural killer T cells (iNKT cells) helped protect mice from developing GVHD and lymphoma, thereby improving their survival.

As human CD4+ iNKT cells can be expanded in vitro, these findings may translate to the clinic, said study investigator Dominik Schneidawind, MD, of Stanford University in California.

Dr Schneidawind discussed this research at the 2014 BMT Tandem Meetings as abstract 2*, which was designated one of the meeting’s “Best Abstracts.”

Improved survival

Dr Schneidawind and his colleagues created a model of allogeneic bone marrow transplant in Balb/c (H-2Kd) mice. The mice were first irradiated with 8 Gy and then received T cell-depleted bone marrow, along with 1 x 10⁶ CD4/CD8 T cells (Tcon) from C57Bl/6 (H-2Kb) donor mice and varying doses of CD4+ iNKT cells: 1 x 10⁴ to 1 x 10⁵.

Mice that received 100,000 CD4+ iNKT cells had significantly improved survival (P=0.002)—and better weight and GVHD scores—compared to mice that received bone marrow grafts alone. But a CD4+ iNKT dose as low as 10,000 cells also conferred a significant improvement in survival (P=0.03).

“In contrast, you need roughly 500,000 to 1 million donor-derived regulatory T cells to achieve a comparable survival benefit in this model of bone marrow transplantation,” Dr Schneidawind noted.

GVT effect

Dr Schneidawind and his colleagues also evaluated the impact of CD4+ iNKT cells on the GVT effect. They used luciferase-transfected BCL1 cells to induce lymphoma development, injecting the cells into mice 1 day before transplant.

Mice that received BCL1 cells, with or without CD4+ iNKT cells, showed increasing bioluminescence imaging (BLI) signals and all died from progressive lymphoma. The mice that received BLC1 and Tcon did not show an increase in BLI signals, but they all died from GVHD.

However, the mice that received BCL1, Tcon, and CD4+ iNKT cells didn’t show any sign of lymphoma or GVHD and survived through the whole experiment (P=0.002).

The investigators conducted the same experiment with the luciferase-positive A20 cell line. And they observed similar results. Mice that received A20 cells, Tcon, and CD4+ iNKT cells showed no sign of lymphoma and had significantly improved survival (P=0.002) over animals that received A20 and Tcon.

“Interestingly, animals that received A20 cells and 50,000 CD4+ iNKT cells only, without Tcon, showed a significant decrease in BLI signal and significantly improved survival,” Dr Schneidawind noted.

“So we conclude from these experiments that CD4+ iNKT cells do not abrogate the Tcon-mediated [GVT] effect but might exert some distinct antitumor activity by themselves.”

Mechanism of GVHD prevention

Lastly, Dr Schneidawind and his colleagues wanted to determine exactly how CD4+ iNKT cells prevent GVHD. The group’s experiments showed that the cells inhibit the proliferation of alloreactive T cells and promote the expansion of donor-derived CD4+ FoxP3+ Tregs.

The investigators decided to take a closer look at the Tregs, evaluating their role in GVHD-related death.

To do this, the team evaluated survival in 4 groups of mice: (1) irradiated controls; (2) mice that received only bone marrow; (3) mice that received Tcon, CD4+ iNKT cells, and a graft depleted of CD4+ FoxP3+ Tregs; and (4) mice that received Tcon, CD4+ iNKT cells, and a non-depleted graft.

“Animals that received a Treg-depleted graft and were treated with CD4+ iNKT [plus Tcon] did not show a significant improvement in survival [over controls],” Dr Schneidawind said.

“In contrast, animals that received a Treg-non-depleted graft showed a significant improvement in survival [P=0.006]. So we conclude that the expansion of these Tregs is necessary to protect from GVHD lethality and that they are required in this context.”

In closing, Dr Schneidawind noted that, although human CD4+ iNKT cells are rare, they can be isolated from peripheral blood and expanded in vitro. This suggests his group’s findings might ultimately prove useful for patients.

*Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—The order in which patients receive cyclophosphamide (Cy) and total body irradiation (TBI) does not affect the outcome of hematopoietic stem cell transplant (HSCT), researchers have reported.

In a large, retrospective study, the rates of relapse, survival, chronic graft-vs-host disease (GVHD), and other complications were similar whether patients received Cy-TBI or TBI-Cy.

However, receiving Cy-TBI was associated with a slight decrease in the risk of grade 2-4 acute GVHD.

Jennifer L. Holter-Chakrabarty, MD, of the University of Oklahoma in Oklahoma City, presented these findings at the 2014 BMT Tandem Meetings as abstract 13.

She noted that other researchers have investigated the impact of TBI/Cy order, but the results have not provided definitive answers.

“Mouse models show that TBI-Cy is superior, with reduced lung toxicity and increased incidence of bone marrow damage,” Dr Holter-Chakrabarty said.

“Cy-TBI, however, in a retrospective study, showed improved anti-leukemic effect, as well as increased incidence of sinusoidal obstructive syndrome. So for this reason, we wanted to look at the CIBMTR database and compare the order of cyclophosphamide and TBI.”

She and her colleagues analyzed data from 1769 HSCT recipients who were younger than 60 years of age and had been reported to the CIBMTR from 2003 to 2010. Patients had been diagnosed with acute myeloid leukemia (AML, n=945) or acute lymphoblastic leukemia (ALL, n=824) and were in their first or second remission.

They had received TBI doses of at least 1200 cGy, followed by related or unrelated bone marrow or peripheral blood stem cell grafts. Patients who had received cord blood, haploidentical, or T-cell-depleted grafts were excluded.

In all, 948 patients received Cy-TBI, and 821 received TBI-Cy. The 2 cohorts had comparable patient-, disease- and transplant-related characteristics.

The sequence of TBI and Cy did not significantly affect the rates of relapse, leukemia-free survival, or overall survival. And it had no significant impact on transplant-related complications, with the exception of acute grade 2-4 GVHD.

At 100 days, the rate of grade 2-4 acute GVHD was 39% in the Cy-TBI group and 45% in the TBI-Cy group (P=0.01). But the rates of grade 3-4 acute GVHD were 16% and 15%, respectively (P=0.62).

At 100 days, the incidence of veno-occlusive disease and sinusoidal obstructive syndrome was 4% in the Cy-TBI group and 6% in the TBI-Cy group (P=0.082). At 1 year, the incidence of interstitial pneumonia syndrome was 6% and 5%, respectively (P=0.370).

At 3 years, the relapse rate was 27% in the Cy-TBI group and 29% in the TBI-Cy group (P=0.34). Leukemia-free survival was 48% and 49%, respectively (P=0.27). And overall survival was 53% and 52%, respectively (P=0.62).

Transplant-related mortality at 3 years was 24% and 23%, respectively (P=0.67). And the rate of chronic GVHD was 45% and 47%, respectively (P=0.39).

“So, in this population, we see no difference,” Dr Holter-Chakrabarty said. “Cy-TBI or TBI-Cy—it seems to not make a difference either way in relapse, treatment-related mortality, [chronic] graft-vs-host disease, acute 3 and 4 [GVHD], and survival. However, it does give us an idea that 2-4 acute GVHD may be lower in Cy-TBI.”

GRAPEVINE, TEXAS—The order in which patients receive cyclophosphamide (Cy) and total body irradiation (TBI) does not affect the outcome of hematopoietic stem cell transplant (HSCT), researchers have reported.

In a large, retrospective study, the rates of relapse, survival, chronic graft-vs-host disease (GVHD), and other complications were similar whether patients received Cy-TBI or TBI-Cy.

However, receiving Cy-TBI was associated with a slight decrease in the risk of grade 2-4 acute GVHD.

Jennifer L. Holter-Chakrabarty, MD, of the University of Oklahoma in Oklahoma City, presented these findings at the 2014 BMT Tandem Meetings as abstract 13.

She noted that other researchers have investigated the impact of TBI/Cy order, but the results have not provided definitive answers.

“Mouse models show that TBI-Cy is superior, with reduced lung toxicity and increased incidence of bone marrow damage,” Dr Holter-Chakrabarty said.

“Cy-TBI, however, in a retrospective study, showed improved anti-leukemic effect, as well as increased incidence of sinusoidal obstructive syndrome. So for this reason, we wanted to look at the CIBMTR database and compare the order of cyclophosphamide and TBI.”

She and her colleagues analyzed data from 1769 HSCT recipients who were younger than 60 years of age and had been reported to the CIBMTR from 2003 to 2010. Patients had been diagnosed with acute myeloid leukemia (AML, n=945) or acute lymphoblastic leukemia (ALL, n=824) and were in their first or second remission.

They had received TBI doses of at least 1200 cGy, followed by related or unrelated bone marrow or peripheral blood stem cell grafts. Patients who had received cord blood, haploidentical, or T-cell-depleted grafts were excluded.

In all, 948 patients received Cy-TBI, and 821 received TBI-Cy. The 2 cohorts had comparable patient-, disease- and transplant-related characteristics.

The sequence of TBI and Cy did not significantly affect the rates of relapse, leukemia-free survival, or overall survival. And it had no significant impact on transplant-related complications, with the exception of acute grade 2-4 GVHD.

At 100 days, the rate of grade 2-4 acute GVHD was 39% in the Cy-TBI group and 45% in the TBI-Cy group (P=0.01). But the rates of grade 3-4 acute GVHD were 16% and 15%, respectively (P=0.62).

At 100 days, the incidence of veno-occlusive disease and sinusoidal obstructive syndrome was 4% in the Cy-TBI group and 6% in the TBI-Cy group (P=0.082). At 1 year, the incidence of interstitial pneumonia syndrome was 6% and 5%, respectively (P=0.370).

At 3 years, the relapse rate was 27% in the Cy-TBI group and 29% in the TBI-Cy group (P=0.34). Leukemia-free survival was 48% and 49%, respectively (P=0.27). And overall survival was 53% and 52%, respectively (P=0.62).

Transplant-related mortality at 3 years was 24% and 23%, respectively (P=0.67). And the rate of chronic GVHD was 45% and 47%, respectively (P=0.39).

“So, in this population, we see no difference,” Dr Holter-Chakrabarty said. “Cy-TBI or TBI-Cy—it seems to not make a difference either way in relapse, treatment-related mortality, [chronic] graft-vs-host disease, acute 3 and 4 [GVHD], and survival. However, it does give us an idea that 2-4 acute GVHD may be lower in Cy-TBI.”

GRAPEVINE, TEXAS—The order in which patients receive cyclophosphamide (Cy) and total body irradiation (TBI) does not affect the outcome of hematopoietic stem cell transplant (HSCT), researchers have reported.

In a large, retrospective study, the rates of relapse, survival, chronic graft-vs-host disease (GVHD), and other complications were similar whether patients received Cy-TBI or TBI-Cy.

However, receiving Cy-TBI was associated with a slight decrease in the risk of grade 2-4 acute GVHD.

Jennifer L. Holter-Chakrabarty, MD, of the University of Oklahoma in Oklahoma City, presented these findings at the 2014 BMT Tandem Meetings as abstract 13.

She noted that other researchers have investigated the impact of TBI/Cy order, but the results have not provided definitive answers.

“Mouse models show that TBI-Cy is superior, with reduced lung toxicity and increased incidence of bone marrow damage,” Dr Holter-Chakrabarty said.

“Cy-TBI, however, in a retrospective study, showed improved anti-leukemic effect, as well as increased incidence of sinusoidal obstructive syndrome. So for this reason, we wanted to look at the CIBMTR database and compare the order of cyclophosphamide and TBI.”

She and her colleagues analyzed data from 1769 HSCT recipients who were younger than 60 years of age and had been reported to the CIBMTR from 2003 to 2010. Patients had been diagnosed with acute myeloid leukemia (AML, n=945) or acute lymphoblastic leukemia (ALL, n=824) and were in their first or second remission.

They had received TBI doses of at least 1200 cGy, followed by related or unrelated bone marrow or peripheral blood stem cell grafts. Patients who had received cord blood, haploidentical, or T-cell-depleted grafts were excluded.

In all, 948 patients received Cy-TBI, and 821 received TBI-Cy. The 2 cohorts had comparable patient-, disease- and transplant-related characteristics.

The sequence of TBI and Cy did not significantly affect the rates of relapse, leukemia-free survival, or overall survival. And it had no significant impact on transplant-related complications, with the exception of acute grade 2-4 GVHD.

At 100 days, the rate of grade 2-4 acute GVHD was 39% in the Cy-TBI group and 45% in the TBI-Cy group (P=0.01). But the rates of grade 3-4 acute GVHD were 16% and 15%, respectively (P=0.62).

At 100 days, the incidence of veno-occlusive disease and sinusoidal obstructive syndrome was 4% in the Cy-TBI group and 6% in the TBI-Cy group (P=0.082). At 1 year, the incidence of interstitial pneumonia syndrome was 6% and 5%, respectively (P=0.370).

At 3 years, the relapse rate was 27% in the Cy-TBI group and 29% in the TBI-Cy group (P=0.34). Leukemia-free survival was 48% and 49%, respectively (P=0.27). And overall survival was 53% and 52%, respectively (P=0.62).

Transplant-related mortality at 3 years was 24% and 23%, respectively (P=0.67). And the rate of chronic GVHD was 45% and 47%, respectively (P=0.39).

“So, in this population, we see no difference,” Dr Holter-Chakrabarty said. “Cy-TBI or TBI-Cy—it seems to not make a difference either way in relapse, treatment-related mortality, [chronic] graft-vs-host disease, acute 3 and 4 [GVHD], and survival. However, it does give us an idea that 2-4 acute GVHD may be lower in Cy-TBI.”

GRAPEVINE, TEXAS—Administering radioimmunotherapy (RIT) prior to non-myeloablative allogeneic transplant (NMAT) can improve survival in patients with persistent disease, according to a speaker at the 2014 BMT Tandem Meetings.

Ryan Cassaday, MD, of the University of Washington in Seattle, noted that RIT-augmented NMAT can produce long-term remissions in patients with relapsed or refractory B-cell non-Hodgkin lymphoma (B-NHL) or chronic lymphocytic leukemia (CLL).

But outcomes for patients with persistent disease are “underdescribed.”

So he and his colleagues set out to describe outcomes of NMAT for patients with persistent indolent B-NHL or CLL and estimate the impact of RIT in these patients.

Treatment details

The researchers retrospectively analyzed data from 89 patients who underwent NMAT from December 1998 to April 2009 and were followed until September 2013. Eighteen of the patients had received RIT as part of a prospective study (AK Gopal et al, Blood 2011).

The remaining 71 patients did not receive RIT but met eligibility criteria for that study. Specifically, they had a CD20+ B-cell malignancy, an HLA-matched peripheral blood stem cell donor, and persistent disease at NMAT. These control subjects received fludarabine (30 mg/m² on days -7, -6, and -5) and 2 Gy of total body irradiation prior to NMAT.

Patients in the RIT group received the same treatment following RIT. On day -21, they received 250 mg/m² of rituximab before an imaging dose of ¹¹¹In-ibritumomab tiuxetan. And on day -14, they received 250 mg/m² of rituximab and 0.4 mCi/kg of ⁹⁰Y-ibritumomab tiuxetan.

Patient characteristics

In the RIT group, 10 patients had CLL/small lymphocytic lymphoma (SLL), 6 had follicular lymphoma (FL), 1 had marginal zone lymphoma (MZL), and 1 had hairy cell leukemia. As for controls, 52 had CLL/SLL, 18 had FL, and 1 had MZL.

“The majority of patients were male [74%] and a relatively young age, given the diseases being treated [median of 56 years],” Dr Cassaday said. “The majority of patients had previously received rituximab [88%], and patients were heavily pretreated [median of 4 prior therapies, range 1-12].”

There were no significant differences between the 2 treatment groups with regard to the aforementioned characteristics. However, there were some “striking differences” between the 2 groups, Dr Cassaday said, including characteristics that portend worse prognosis.

Specifically, RIT-treated patients had more bulky disease (> 5 cm) than controls (61% vs 15%, P<0.001) and more chemoresistant disease (81% vs 39%, P=0.003). And RIT patients were more likely to have HCT comorbidity index scores of 3 or higher (72% vs 37%, P=0.006), as well as pre-NMAT platelet counts less than 25k/μL (33% vs 7%, P=0.002).

RIT improves PFS, OS

The researchers conducted a multivariate analysis including the factors that differed significantly between the 2 treatment groups. And they found that only RIT was significantly associated with both progression-free survival (PFS) and overall survival (OS).

When calculating survival curves, the researchers adjusted for the imbalance in covariates between the treatment groups.

“[The adjusted survival rate] is essentially what one might expect had the RIT group had similar baseline characteristics as the control group,” Dr Cassaday explained.

Control subjects had a 3-year OS of 55%. For the RIT-treated patients, the unadjusted 3-year OS was 78% (P=0.20), and the adjusted OS was 87% (P=0.008).

The 3-year PFS was 44% for controls. For the RIT group, the unadjusted 3-year PFS was 56% (P=0.36), and the adjusted PFS was 71% (P=0.02).

The researchers also found that RIT did not increase the rate of non-relapse mortality. The unadjusted hazard ratio was 0.5 (P=0.32), and the adjusted hazard ratio was 0.4 (P=0.18).

“This analysis does have some limitations,” Dr Cassaday conceded. “Clearly, it does not replace the strength of evidence that would come from a randomized, controlled trial. And the relatively small sample size does limit our ability to look at a lot of different subgroups.”

In addition, the findings may not apply to other non-myeloablative regimens. And, due to the time frame of the study, the researchers could not account for the potential impact of newer agents.

Nevertheless, Dr Cassaday said the data suggest that RIT can improve the outcome of NMAT in patients with persistent indolent B-NHL or CLL. And a prospective, randomized study evaluating this approach is warranted.

Dr Cassaday presented this research at the 2014 BMT Tandem Meetings as abstract 75. Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—Administering radioimmunotherapy (RIT) prior to non-myeloablative allogeneic transplant (NMAT) can improve survival in patients with persistent disease, according to a speaker at the 2014 BMT Tandem Meetings.

Ryan Cassaday, MD, of the University of Washington in Seattle, noted that RIT-augmented NMAT can produce long-term remissions in patients with relapsed or refractory B-cell non-Hodgkin lymphoma (B-NHL) or chronic lymphocytic leukemia (CLL).

But outcomes for patients with persistent disease are “underdescribed.”

So he and his colleagues set out to describe outcomes of NMAT for patients with persistent indolent B-NHL or CLL and estimate the impact of RIT in these patients.

Treatment details

The researchers retrospectively analyzed data from 89 patients who underwent NMAT from December 1998 to April 2009 and were followed until September 2013. Eighteen of the patients had received RIT as part of a prospective study (AK Gopal et al, Blood 2011).

The remaining 71 patients did not receive RIT but met eligibility criteria for that study. Specifically, they had a CD20+ B-cell malignancy, an HLA-matched peripheral blood stem cell donor, and persistent disease at NMAT. These control subjects received fludarabine (30 mg/m² on days -7, -6, and -5) and 2 Gy of total body irradiation prior to NMAT.

Patients in the RIT group received the same treatment following RIT. On day -21, they received 250 mg/m² of rituximab before an imaging dose of ¹¹¹In-ibritumomab tiuxetan. And on day -14, they received 250 mg/m² of rituximab and 0.4 mCi/kg of ⁹⁰Y-ibritumomab tiuxetan.

Patient characteristics

In the RIT group, 10 patients had CLL/small lymphocytic lymphoma (SLL), 6 had follicular lymphoma (FL), 1 had marginal zone lymphoma (MZL), and 1 had hairy cell leukemia. As for controls, 52 had CLL/SLL, 18 had FL, and 1 had MZL.

“The majority of patients were male [74%] and a relatively young age, given the diseases being treated [median of 56 years],” Dr Cassaday said. “The majority of patients had previously received rituximab [88%], and patients were heavily pretreated [median of 4 prior therapies, range 1-12].”

There were no significant differences between the 2 treatment groups with regard to the aforementioned characteristics. However, there were some “striking differences” between the 2 groups, Dr Cassaday said, including characteristics that portend worse prognosis.

Specifically, RIT-treated patients had more bulky disease (> 5 cm) than controls (61% vs 15%, P<0.001) and more chemoresistant disease (81% vs 39%, P=0.003). And RIT patients were more likely to have HCT comorbidity index scores of 3 or higher (72% vs 37%, P=0.006), as well as pre-NMAT platelet counts less than 25k/μL (33% vs 7%, P=0.002).

RIT improves PFS, OS

The researchers conducted a multivariate analysis including the factors that differed significantly between the 2 treatment groups. And they found that only RIT was significantly associated with both progression-free survival (PFS) and overall survival (OS).

When calculating survival curves, the researchers adjusted for the imbalance in covariates between the treatment groups.

“[The adjusted survival rate] is essentially what one might expect had the RIT group had similar baseline characteristics as the control group,” Dr Cassaday explained.

Control subjects had a 3-year OS of 55%. For the RIT-treated patients, the unadjusted 3-year OS was 78% (P=0.20), and the adjusted OS was 87% (P=0.008).

The 3-year PFS was 44% for controls. For the RIT group, the unadjusted 3-year PFS was 56% (P=0.36), and the adjusted PFS was 71% (P=0.02).

The researchers also found that RIT did not increase the rate of non-relapse mortality. The unadjusted hazard ratio was 0.5 (P=0.32), and the adjusted hazard ratio was 0.4 (P=0.18).

“This analysis does have some limitations,” Dr Cassaday conceded. “Clearly, it does not replace the strength of evidence that would come from a randomized, controlled trial. And the relatively small sample size does limit our ability to look at a lot of different subgroups.”

In addition, the findings may not apply to other non-myeloablative regimens. And, due to the time frame of the study, the researchers could not account for the potential impact of newer agents.

Nevertheless, Dr Cassaday said the data suggest that RIT can improve the outcome of NMAT in patients with persistent indolent B-NHL or CLL. And a prospective, randomized study evaluating this approach is warranted.

Dr Cassaday presented this research at the 2014 BMT Tandem Meetings as abstract 75. Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—Administering radioimmunotherapy (RIT) prior to non-myeloablative allogeneic transplant (NMAT) can improve survival in patients with persistent disease, according to a speaker at the 2014 BMT Tandem Meetings.

Ryan Cassaday, MD, of the University of Washington in Seattle, noted that RIT-augmented NMAT can produce long-term remissions in patients with relapsed or refractory B-cell non-Hodgkin lymphoma (B-NHL) or chronic lymphocytic leukemia (CLL).

But outcomes for patients with persistent disease are “underdescribed.”

So he and his colleagues set out to describe outcomes of NMAT for patients with persistent indolent B-NHL or CLL and estimate the impact of RIT in these patients.

Treatment details

The researchers retrospectively analyzed data from 89 patients who underwent NMAT from December 1998 to April 2009 and were followed until September 2013. Eighteen of the patients had received RIT as part of a prospective study (AK Gopal et al, Blood 2011).

The remaining 71 patients did not receive RIT but met eligibility criteria for that study. Specifically, they had a CD20+ B-cell malignancy, an HLA-matched peripheral blood stem cell donor, and persistent disease at NMAT. These control subjects received fludarabine (30 mg/m² on days -7, -6, and -5) and 2 Gy of total body irradiation prior to NMAT.

Patients in the RIT group received the same treatment following RIT. On day -21, they received 250 mg/m² of rituximab before an imaging dose of ¹¹¹In-ibritumomab tiuxetan. And on day -14, they received 250 mg/m² of rituximab and 0.4 mCi/kg of ⁹⁰Y-ibritumomab tiuxetan.

Patient characteristics

In the RIT group, 10 patients had CLL/small lymphocytic lymphoma (SLL), 6 had follicular lymphoma (FL), 1 had marginal zone lymphoma (MZL), and 1 had hairy cell leukemia. As for controls, 52 had CLL/SLL, 18 had FL, and 1 had MZL.

“The majority of patients were male [74%] and a relatively young age, given the diseases being treated [median of 56 years],” Dr Cassaday said. “The majority of patients had previously received rituximab [88%], and patients were heavily pretreated [median of 4 prior therapies, range 1-12].”

There were no significant differences between the 2 treatment groups with regard to the aforementioned characteristics. However, there were some “striking differences” between the 2 groups, Dr Cassaday said, including characteristics that portend worse prognosis.

Specifically, RIT-treated patients had more bulky disease (> 5 cm) than controls (61% vs 15%, P<0.001) and more chemoresistant disease (81% vs 39%, P=0.003). And RIT patients were more likely to have HCT comorbidity index scores of 3 or higher (72% vs 37%, P=0.006), as well as pre-NMAT platelet counts less than 25k/μL (33% vs 7%, P=0.002).

RIT improves PFS, OS

The researchers conducted a multivariate analysis including the factors that differed significantly between the 2 treatment groups. And they found that only RIT was significantly associated with both progression-free survival (PFS) and overall survival (OS).

When calculating survival curves, the researchers adjusted for the imbalance in covariates between the treatment groups.

“[The adjusted survival rate] is essentially what one might expect had the RIT group had similar baseline characteristics as the control group,” Dr Cassaday explained.

Control subjects had a 3-year OS of 55%. For the RIT-treated patients, the unadjusted 3-year OS was 78% (P=0.20), and the adjusted OS was 87% (P=0.008).

The 3-year PFS was 44% for controls. For the RIT group, the unadjusted 3-year PFS was 56% (P=0.36), and the adjusted PFS was 71% (P=0.02).

The researchers also found that RIT did not increase the rate of non-relapse mortality. The unadjusted hazard ratio was 0.5 (P=0.32), and the adjusted hazard ratio was 0.4 (P=0.18).

“This analysis does have some limitations,” Dr Cassaday conceded. “Clearly, it does not replace the strength of evidence that would come from a randomized, controlled trial. And the relatively small sample size does limit our ability to look at a lot of different subgroups.”

In addition, the findings may not apply to other non-myeloablative regimens. And, due to the time frame of the study, the researchers could not account for the potential impact of newer agents.

Nevertheless, Dr Cassaday said the data suggest that RIT can improve the outcome of NMAT in patients with persistent indolent B-NHL or CLL. And a prospective, randomized study evaluating this approach is warranted.

Dr Cassaday presented this research at the 2014 BMT Tandem Meetings as abstract 75. Information in the abstract differs from that presented.

GRAPEVINE, TEXAS—Acute leukemia patients who undergo cord blood (CB) transplant have longer hospital stays than patients who receive other types of transplant, new research indicates.

The study also suggests the length of stay (LOS) is similar whether patients receive double or single CB grafts.

So it seems strategies are needed to decrease hospital stay after CB transplant, particularly as LOS drives the cost of care, said Karen K. Ballen, MD, of Massachusetts General Hospital in Boston.

Dr Ballen presented this research at the 2014 BMT Tandem Meetings as abstract 104.*

She and her colleagues studied patients diagnosed with acute leukemias who were transplanted at US centers and reported to the CIBMTR between 2008 and 2011.

Patients were eligible if they received an unrelated single or double CB transplant, an 8/8 matched unrelated donor (MUD) transplant with peripheral blood (PB) or bone marrow (BM), or a 7/8 MUD PB or BM graft.

In all, 1796 patients met these criteria. The researchers evaluated patients’ total hospital LOS in the first 100 days after transplant, compared LOS among graft sources, and looked for predictors of LOS in the first 100 days.

The team stratified patients according to age and conditioning regimen. Pediatric patients were classified as those aged 18 and younger, and they only received myeloablative conditioning (MAC). Adults received either MAC or reduced-intensity conditioning (RIC).

Pediatric patients

In a univariate analysis of the 368 pediatric patients, there was no significant difference in 100-day survival according to graft source (P=0.13).

However, patients who received single or double CB grafts had a significantly higher median total LOS by day 100 than patients who received 8/8 MUD BM, which was the only other graft source in this patient group (P=0.03).

Patients who received CB grafts also had significantly fewer days in which they were alive and not in the hospital (P=0.005).

“We wanted to account for patients whose length of stay was short because they actually died early after transplant,” Dr Ballen explained. “Therefore, we did an analysis of days alive and not in the hospital.”

In a multivariate analysis, pediatric patients who received CB grafts had significantly fewer days alive and out of the hospital than those who received 8/8 MUD BM (P=0.03).

Other factors associated with fewer days alive and out of the hospital were CMV positivity (P=0.01), black race (P=0.01), and a Karnofsky performance score of less than 80 (P=0.03).

Adults on MAC

In a univariate analysis of the 768 adults who received MAC, recipients of CB grafts had significantly worse 100-day survival than their peers (P<0.001), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received MAC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD BM (P<0.001), 8/8 MUD PB (P<0.001), or 7/8 MUD PB (P=0.01), but not 7/8 MUD BM (P=0.49).

Other factors associated with fewer days alive and out of the hospital were black race (P=0.04), having acute lymphocytic leukemia rather than acute myeloid leukemia (P=0.01), and age 18-25 (P=0.01).

“We were a little surprised at these results—that the older patients actually spent more time alive and out of the hospital,” Dr Ballen said.

Adults on RIC

In a univariate analysis of the 660 adults who received RIC, recipients of CB grafts had significantly worse 100-day survival than their peers (P=0.017), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received RIC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD PB (P<0.001) or 7/8 MUD PB (P<0.001).

No other factors were associated with the number of days these patients were alive and out of the hospital.

These results, when taken together, suggest that CB grafts are associated with longer hospital stays, independent of other factors.

“The majority of cost appears to be driven by the number of days in the hospital,” Dr Ballen noted. “So these data may be important for resource allocation, especially given the recent changes in the US healthcare system.”

*Information in the abstract differs from that presented at the meeting.

GRAPEVINE, TEXAS—Acute leukemia patients who undergo cord blood (CB) transplant have longer hospital stays than patients who receive other types of transplant, new research indicates.

The study also suggests the length of stay (LOS) is similar whether patients receive double or single CB grafts.

So it seems strategies are needed to decrease hospital stay after CB transplant, particularly as LOS drives the cost of care, said Karen K. Ballen, MD, of Massachusetts General Hospital in Boston.

Dr Ballen presented this research at the 2014 BMT Tandem Meetings as abstract 104.*

She and her colleagues studied patients diagnosed with acute leukemias who were transplanted at US centers and reported to the CIBMTR between 2008 and 2011.

Patients were eligible if they received an unrelated single or double CB transplant, an 8/8 matched unrelated donor (MUD) transplant with peripheral blood (PB) or bone marrow (BM), or a 7/8 MUD PB or BM graft.

In all, 1796 patients met these criteria. The researchers evaluated patients’ total hospital LOS in the first 100 days after transplant, compared LOS among graft sources, and looked for predictors of LOS in the first 100 days.

The team stratified patients according to age and conditioning regimen. Pediatric patients were classified as those aged 18 and younger, and they only received myeloablative conditioning (MAC). Adults received either MAC or reduced-intensity conditioning (RIC).

Pediatric patients

In a univariate analysis of the 368 pediatric patients, there was no significant difference in 100-day survival according to graft source (P=0.13).

However, patients who received single or double CB grafts had a significantly higher median total LOS by day 100 than patients who received 8/8 MUD BM, which was the only other graft source in this patient group (P=0.03).

Patients who received CB grafts also had significantly fewer days in which they were alive and not in the hospital (P=0.005).

“We wanted to account for patients whose length of stay was short because they actually died early after transplant,” Dr Ballen explained. “Therefore, we did an analysis of days alive and not in the hospital.”

In a multivariate analysis, pediatric patients who received CB grafts had significantly fewer days alive and out of the hospital than those who received 8/8 MUD BM (P=0.03).

Other factors associated with fewer days alive and out of the hospital were CMV positivity (P=0.01), black race (P=0.01), and a Karnofsky performance score of less than 80 (P=0.03).

Adults on MAC

In a univariate analysis of the 768 adults who received MAC, recipients of CB grafts had significantly worse 100-day survival than their peers (P<0.001), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received MAC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD BM (P<0.001), 8/8 MUD PB (P<0.001), or 7/8 MUD PB (P=0.01), but not 7/8 MUD BM (P=0.49).

Other factors associated with fewer days alive and out of the hospital were black race (P=0.04), having acute lymphocytic leukemia rather than acute myeloid leukemia (P=0.01), and age 18-25 (P=0.01).

“We were a little surprised at these results—that the older patients actually spent more time alive and out of the hospital,” Dr Ballen said.

Adults on RIC

In a univariate analysis of the 660 adults who received RIC, recipients of CB grafts had significantly worse 100-day survival than their peers (P=0.017), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received RIC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD PB (P<0.001) or 7/8 MUD PB (P<0.001).

No other factors were associated with the number of days these patients were alive and out of the hospital.

These results, when taken together, suggest that CB grafts are associated with longer hospital stays, independent of other factors.

“The majority of cost appears to be driven by the number of days in the hospital,” Dr Ballen noted. “So these data may be important for resource allocation, especially given the recent changes in the US healthcare system.”

*Information in the abstract differs from that presented at the meeting.

GRAPEVINE, TEXAS—Acute leukemia patients who undergo cord blood (CB) transplant have longer hospital stays than patients who receive other types of transplant, new research indicates.

The study also suggests the length of stay (LOS) is similar whether patients receive double or single CB grafts.

So it seems strategies are needed to decrease hospital stay after CB transplant, particularly as LOS drives the cost of care, said Karen K. Ballen, MD, of Massachusetts General Hospital in Boston.

Dr Ballen presented this research at the 2014 BMT Tandem Meetings as abstract 104.*

She and her colleagues studied patients diagnosed with acute leukemias who were transplanted at US centers and reported to the CIBMTR between 2008 and 2011.

Patients were eligible if they received an unrelated single or double CB transplant, an 8/8 matched unrelated donor (MUD) transplant with peripheral blood (PB) or bone marrow (BM), or a 7/8 MUD PB or BM graft.

In all, 1796 patients met these criteria. The researchers evaluated patients’ total hospital LOS in the first 100 days after transplant, compared LOS among graft sources, and looked for predictors of LOS in the first 100 days.

The team stratified patients according to age and conditioning regimen. Pediatric patients were classified as those aged 18 and younger, and they only received myeloablative conditioning (MAC). Adults received either MAC or reduced-intensity conditioning (RIC).

Pediatric patients

In a univariate analysis of the 368 pediatric patients, there was no significant difference in 100-day survival according to graft source (P=0.13).

However, patients who received single or double CB grafts had a significantly higher median total LOS by day 100 than patients who received 8/8 MUD BM, which was the only other graft source in this patient group (P=0.03).

Patients who received CB grafts also had significantly fewer days in which they were alive and not in the hospital (P=0.005).

“We wanted to account for patients whose length of stay was short because they actually died early after transplant,” Dr Ballen explained. “Therefore, we did an analysis of days alive and not in the hospital.”

In a multivariate analysis, pediatric patients who received CB grafts had significantly fewer days alive and out of the hospital than those who received 8/8 MUD BM (P=0.03).

Other factors associated with fewer days alive and out of the hospital were CMV positivity (P=0.01), black race (P=0.01), and a Karnofsky performance score of less than 80 (P=0.03).

Adults on MAC

In a univariate analysis of the 768 adults who received MAC, recipients of CB grafts had significantly worse 100-day survival than their peers (P<0.001), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received MAC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD BM (P<0.001), 8/8 MUD PB (P<0.001), or 7/8 MUD PB (P=0.01), but not 7/8 MUD BM (P=0.49).

Other factors associated with fewer days alive and out of the hospital were black race (P=0.04), having acute lymphocytic leukemia rather than acute myeloid leukemia (P=0.01), and age 18-25 (P=0.01).

“We were a little surprised at these results—that the older patients actually spent more time alive and out of the hospital,” Dr Ballen said.

Adults on RIC

In a univariate analysis of the 660 adults who received RIC, recipients of CB grafts had significantly worse 100-day survival than their peers (P=0.017), as well as a longer median LOS by day 100 (P<0.001) and fewer days alive and not in the hospital (P<0.001).

In a multivariate analysis, adults who received RIC had significantly fewer days alive and out of the hospital if they received CB grafts than if they received 8/8 MUD PB (P<0.001) or 7/8 MUD PB (P<0.001).

No other factors were associated with the number of days these patients were alive and out of the hospital.

These results, when taken together, suggest that CB grafts are associated with longer hospital stays, independent of other factors.

“The majority of cost appears to be driven by the number of days in the hospital,” Dr Ballen noted. “So these data may be important for resource allocation, especially given the recent changes in the US healthcare system.”

*Information in the abstract differs from that presented at the meeting.

GRAPEVINE, TEXAS—Results of a large, retrospective study suggest that allogeneic stem cell transplant (allo-SCT) can overcome the poor prognosis associated with central nervous system (CNS) involvement in acute myeloid leukemia (AML).

By analyzing transplant outcomes in more than 5000 patients, researchers found that subjects with CNS AML had rates of relapse and survival that were similar to those of patients without CNS involvement.

The team also identified factors that can predict for survival in CNS AML, including cytogenetic risk group, the presence of chronic GVHD, and whether a patient was in complete response at transplant.

Jun Aoki, MD, of Tokyo Metropolitan Komagome Hospital in Japan, presented these findings at the 2014 BMT Tandem Meetings as abstract 68.

Dr Aoki pointed out that CNS involvement is rare in adult AML, occurring in about 5% of patients. However, these patients generally have poor prognosis. And although allo-SCT is one of the options used to treat CNS AML, exactly how CNS involvement impacts transplant outcomes remains unclear.

So Dr Aoki and his colleagues conducted a nationwide, retrospective study to gain some insight.  They collected data from the registry database of the Japan Society for Hematopoietic Cell Transplantation.

Patients had to be older than 15 years of age, have their first allo-SCT between 2006 and 2011, and not have acute promyelocytic leukemia.

The researchers identified 5068 patients who met these criteria, and 157 of them had CNS AML. CNS involvement was defined as infiltration of leukemia cells into CNS or myeloid sarcoma in CNS that were identified at any time from diagnosis to transplant.

No difference in relapse, survival

There were no significant differences between CNS patients and controls with regard to the estimated overall survival (OS), leukemia-free survival, cumulative incidence of relapse, or non-relapse mortality at 5 years.

OS was 39.9% among controls and 38.5% among CNS patients (P=0.847). Leukemia-free survival was 41.2% and 41.5%, respectively (P=0.82).

The cumulative incidence of relapse was 29.8% among controls and 31.8% among CNS patients (P=0.418). And non-relapse mortality was 22.5% and 26.5%, respectively (P=0.142).

Factors predicting OS

To determine the impact of patient and treatment characteristics on OS, the researchers conducted a multivariate analysis. This confirmed that CNS involvement was not a risk factor for OS.

But it revealed a number of other factors that adversely affect OS, including age of 50 or older (P<0.001), lack of a complete response at allo-SCT (P<0.001), a donor source of unrelated cord blood (P=0.005), having a prognostic score of 2-4 (P<0.001), unfavorable cytogenetics (P<0.001), and the absence of acute or chronic GVHD (P<0.001 for both).

When the researchers analyzed only CNS patients, they discovered that not all of these factors retained significance. Only the absence of chronic GVHD (P=0.002), lack of complete response at transplant (P<0.001), and having either intermediate (P=0.003) or unfavorable cytogenetics (P=0.011) were adversely associated with OS in these patients.

GRAPEVINE, TEXAS—Results of a large, retrospective study suggest that allogeneic stem cell transplant (allo-SCT) can overcome the poor prognosis associated with central nervous system (CNS) involvement in acute myeloid leukemia (AML).

By analyzing transplant outcomes in more than 5000 patients, researchers found that subjects with CNS AML had rates of relapse and survival that were similar to those of patients without CNS involvement.

The team also identified factors that can predict for survival in CNS AML, including cytogenetic risk group, the presence of chronic GVHD, and whether a patient was in complete response at transplant.

Jun Aoki, MD, of Tokyo Metropolitan Komagome Hospital in Japan, presented these findings at the 2014 BMT Tandem Meetings as abstract 68.

Dr Aoki pointed out that CNS involvement is rare in adult AML, occurring in about 5% of patients. However, these patients generally have poor prognosis. And although allo-SCT is one of the options used to treat CNS AML, exactly how CNS involvement impacts transplant outcomes remains unclear.

So Dr Aoki and his colleagues conducted a nationwide, retrospective study to gain some insight.  They collected data from the registry database of the Japan Society for Hematopoietic Cell Transplantation.

Patients had to be older than 15 years of age, have their first allo-SCT between 2006 and 2011, and not have acute promyelocytic leukemia.

The researchers identified 5068 patients who met these criteria, and 157 of them had CNS AML. CNS involvement was defined as infiltration of leukemia cells into CNS or myeloid sarcoma in CNS that were identified at any time from diagnosis to transplant.

No difference in relapse, survival

There were no significant differences between CNS patients and controls with regard to the estimated overall survival (OS), leukemia-free survival, cumulative incidence of relapse, or non-relapse mortality at 5 years.

OS was 39.9% among controls and 38.5% among CNS patients (P=0.847). Leukemia-free survival was 41.2% and 41.5%, respectively (P=0.82).

The cumulative incidence of relapse was 29.8% among controls and 31.8% among CNS patients (P=0.418). And non-relapse mortality was 22.5% and 26.5%, respectively (P=0.142).

Factors predicting OS

To determine the impact of patient and treatment characteristics on OS, the researchers conducted a multivariate analysis. This confirmed that CNS involvement was not a risk factor for OS.

But it revealed a number of other factors that adversely affect OS, including age of 50 or older (P<0.001), lack of a complete response at allo-SCT (P<0.001), a donor source of unrelated cord blood (P=0.005), having a prognostic score of 2-4 (P<0.001), unfavorable cytogenetics (P<0.001), and the absence of acute or chronic GVHD (P<0.001 for both).

When the researchers analyzed only CNS patients, they discovered that not all of these factors retained significance. Only the absence of chronic GVHD (P=0.002), lack of complete response at transplant (P<0.001), and having either intermediate (P=0.003) or unfavorable cytogenetics (P=0.011) were adversely associated with OS in these patients.

GRAPEVINE, TEXAS—Results of a large, retrospective study suggest that allogeneic stem cell transplant (allo-SCT) can overcome the poor prognosis associated with central nervous system (CNS) involvement in acute myeloid leukemia (AML).

By analyzing transplant outcomes in more than 5000 patients, researchers found that subjects with CNS AML had rates of relapse and survival that were similar to those of patients without CNS involvement.

The team also identified factors that can predict for survival in CNS AML, including cytogenetic risk group, the presence of chronic GVHD, and whether a patient was in complete response at transplant.

Jun Aoki, MD, of Tokyo Metropolitan Komagome Hospital in Japan, presented these findings at the 2014 BMT Tandem Meetings as abstract 68.

Dr Aoki pointed out that CNS involvement is rare in adult AML, occurring in about 5% of patients. However, these patients generally have poor prognosis. And although allo-SCT is one of the options used to treat CNS AML, exactly how CNS involvement impacts transplant outcomes remains unclear.

So Dr Aoki and his colleagues conducted a nationwide, retrospective study to gain some insight.  They collected data from the registry database of the Japan Society for Hematopoietic Cell Transplantation.

Patients had to be older than 15 years of age, have their first allo-SCT between 2006 and 2011, and not have acute promyelocytic leukemia.

The researchers identified 5068 patients who met these criteria, and 157 of them had CNS AML. CNS involvement was defined as infiltration of leukemia cells into CNS or myeloid sarcoma in CNS that were identified at any time from diagnosis to transplant.

No difference in relapse, survival

There were no significant differences between CNS patients and controls with regard to the estimated overall survival (OS), leukemia-free survival, cumulative incidence of relapse, or non-relapse mortality at 5 years.

OS was 39.9% among controls and 38.5% among CNS patients (P=0.847). Leukemia-free survival was 41.2% and 41.5%, respectively (P=0.82).

The cumulative incidence of relapse was 29.8% among controls and 31.8% among CNS patients (P=0.418). And non-relapse mortality was 22.5% and 26.5%, respectively (P=0.142).

Factors predicting OS

To determine the impact of patient and treatment characteristics on OS, the researchers conducted a multivariate analysis. This confirmed that CNS involvement was not a risk factor for OS.

But it revealed a number of other factors that adversely affect OS, including age of 50 or older (P<0.001), lack of a complete response at allo-SCT (P<0.001), a donor source of unrelated cord blood (P=0.005), having a prognostic score of 2-4 (P<0.001), unfavorable cytogenetics (P<0.001), and the absence of acute or chronic GVHD (P<0.001 for both).

When the researchers analyzed only CNS patients, they discovered that not all of these factors retained significance. Only the absence of chronic GVHD (P=0.002), lack of complete response at transplant (P<0.001), and having either intermediate (P=0.003) or unfavorable cytogenetics (P=0.011) were adversely associated with OS in these patients.

GRAPEVINE, TEXAS—Intestinal bacteria can offer protection from death related to graft-vs-host disease (GVHD), according to research presented at the 2014 BMT Tandem Meetings.

Experiments showed that Blautia, commensal bacteria found in the intestinal tract, can protect against GVHD-related mortality in mice and in humans.

So efforts to support Blautia survival—such as restricting the use of antibiotics and promoting better nutrition—may

prevent GVHD-related death, according to researchers.

Robert Jenq, MD, of Memorial Sloan-Kettering Cancer Center in New York, discussed this possibility when presenting this research, which was designated one of the “Best Abstracts” at the meeting (abstract 1*).

Dr Jenq noted that researchers have been trying for decades to determine whether the intestinal flora impact GVHD. Clinical studies have suggested that prophylaxis against anaerobes and gram-positive bacteria can reduce GVHD.

And murine studies have indicated that prophylaxis against gram-negative bacteria can reduce GVHD, that Lactobacillus can reduce GVHD, and that donor microbiota do not impact GVHD.

“If you’re confused, so are we,” Dr Jenq said. “It seems like it’s a mixed picture.”

So he and his colleagues conducted a series of experiments in an attempt to determine if any bacterial subgroups impact the risk of gut GVHD in mice and humans.

Bacteria seem to impact GVHD

The researchers first studied 76 adult transplant patients, analyzing stool samples taken at roughly 10 days after transplant (+/- 4 days). The team performed 16S gene sequencing using the Roche 454 platform.

This revealed the presence of several types of bacteria, including 6 gram-positive Firmicutes, 2 gram-negative Proteobacteria, and 2 gram-negative Bacteroidetes.

The researchers then used a computational assay to determine which of these bacteria might be associated with protection from GVHD. And they identified 2 possibilities—Lactobacillus and Blautia.

Additional analyses revealed that Blautia and Lactobacillus were significantly associated with GVHD-related mortality at 1500 days after transplant (P=0.03 and 0.01, respectively). But there was no significant association with Bacteroides (P=0.6), Enterobacteriales (P=0.2), or Enterococcus (P=0.3).

Blautia appears to affect GVHD-related mortality

To confirm their initial findings, Dr Jenq and his colleagues analyzed a second cohort of 50 adult transplant patients. The team analyzed stool samples for the abundance of bacterial subgroups using a different sequencing platform, Illumina miseq.

This time, they found that Blautia abundance predicted GVHD-related mortality at more than 500 days after transplant, but the abundance of Lactobacillus did not (P=0.01 and 1, respectively).

“Not enough Blautia in your gut seems to lead to an increase in GVHD-related mortality,” Dr Jenq said. “So what does this do to overall survival? In the first cohort, there’s a big difference in overall survival between the ‘haves’ and ‘have nots’ with Blautia [P=0.0008]. And this also holds up in the second cohort [P=0.04].”

Further analyses of data from both cohorts suggested that Blautia abundance was associated with GVHD-related mortality (P=0.004) and relapse-related mortality (P=0.01) but not non-relapse- and non-GVHD-related mortality (P=0.4).

“I don’t have a good explanation for [the relationship between Blautia and relapse-related death],” Dr Jenq said. “This was a surprise finding.”

The researchers also looked at Blautia’s ability to predict GVHD-related mortality. They found that, around day 10 after transplant, Blautia abundance predicts “very strongly” for GVHD-related death.

Another question was whether known GVHD risk factors—such as donor type, race, gender, and performance status—impact Blautia abundance. But an analysis revealed that Blautia is an independent risk factor for GVHD-related mortality.

A possible mechanism

To gain more insight into the association between Blautia and GVHD-related death, Dr Jenq and his colleagues decided to study it in mice.

The team killed off Blautia in mice using vancomycin and ampicillin, then introduced either murine Blautia or murine Enterococcus, transplanted the mice with MHC-disparate T cells, and monitored them for GVHD.

Mice that received Blautia had significantly better overall survival (at more than 80 days after transplant) than mice that received Enterococcus (P<0.001).

“So how is this happening?” Dr Jenq asked. “We think, potentially, it might be due to short-chain fatty acids . . . butyric acid, propionic acid, and acetic acid. These are metabolites that bacteria produce when they ferment glucose and other sugars.”

To test this theory, the researchers treated mice with antibiotics and introduced Blautia or Enterococcus.

Blautia increased the level of short-chain fatty acids (butyrate and propionate) when compared to Enterococcus, although levels were not as high as those observed in mice that did not receive antibiotics. Nevertheless, these results point to a possible mechanism, according to Dr Jenq.

Explaining Blautia reduction

Dr Jenq also noted that antibiotics may contribute to the decrease in Blautia observed in transplant patients. When patients come in for transplant, they often have more than 25% Blautia in their stool. But the bacteria decrease to negligible levels by day 2 after transplant.

To determine the role of antibiotics, the researchers treated mice with 4 different antibiotics and looked at the levels of different bacteria.

They found that aztreonam and cefepime increased the levels of Bacteroidales and Clostridiales (the family to which Blautia belongs), but imipenem and metronidazole decreased bacteria levels.

So antibiotics do affect Blautia levels, Dr Jenq said, but they’re only part of the problem. He noted that patients’ Blautia levels start to decrease before antibiotics are administered. So he and his colleagues believe nutrition might also play a part.

The team found a significant difference in Blautia abundance between patients who received total parenteral nutrition and those who did not (P<0.001).

The researchers also discovered that reduced caloric intake led to a loss of Blautia and other Clostridiales. They analyzed 50 samples from 5 patients and found that patients who consumed fewer than 500 calories had a marked reduction in Blautia (P<0.0001).

And experiments in mice confirmed this association. A week of calorie restriction significantly reduced the abundance of Blautia and other Clostridiales (P=0.0002).

“In GVHD, as we all know, patients and mice eat less because of the nausea,” Dr Jenq said. “And we found that GVHD itself can also lead to a reduction in Clostridiales, both in humans [P=0.02] and in mice [P=0.01].”

Protecting Blautia to prevent GVHD

Having confirmed the role of nutrition in Blautia reduction, the researchers set out to identify a nutrition-based intervention to support Blautia in transplant recipients.

They settled on a sugar called raffinose, which is found in beans, cruciferous vegetables, and whole grains. It passes undigested through the upper intestine but is fermented in the lower intestine and metabolized to produce short-chain fatty acids.

The team tested raffinose in mice by introducing it into their drinking water. At 100 days after transplant, mice that received raffinose had significantly better overall survival than controls (P<0.001).

Based on these results, Dr Jenq and his colleagues believe nutritional intervention can protect Blautia and, therefore, may prevent GVHD and related death. The team thinks encouraging eating, gastric nutritional supplementation, and flora-targeted nutritional supplements might all prove effective.

But other interventions might work as well, such as reintroducing endogenous flora (via autologous fecal microbiota transplant), reintroducing select bacteria with beneficial potential, selecting antibiotics that spare bacteria with beneficial potential, and identifying and introducing bacterial metabolites that mediate anti-inflammatory effects.

*Data in the abstract differ from data presented.

GRAPEVINE, TEXAS—Intestinal bacteria can offer protection from death related to graft-vs-host disease (GVHD), according to research presented at the 2014 BMT Tandem Meetings.

Experiments showed that Blautia, commensal bacteria found in the intestinal tract, can protect against GVHD-related mortality in mice and in humans.

So efforts to support Blautia survival—such as restricting the use of antibiotics and promoting better nutrition—may

prevent GVHD-related death, according to researchers.

Robert Jenq, MD, of Memorial Sloan-Kettering Cancer Center in New York, discussed this possibility when presenting this research, which was designated one of the “Best Abstracts” at the meeting (abstract 1*).

Dr Jenq noted that researchers have been trying for decades to determine whether the intestinal flora impact GVHD. Clinical studies have suggested that prophylaxis against anaerobes and gram-positive bacteria can reduce GVHD.

And murine studies have indicated that prophylaxis against gram-negative bacteria can reduce GVHD, that Lactobacillus can reduce GVHD, and that donor microbiota do not impact GVHD.

“If you’re confused, so are we,” Dr Jenq said. “It seems like it’s a mixed picture.”

So he and his colleagues conducted a series of experiments in an attempt to determine if any bacterial subgroups impact the risk of gut GVHD in mice and humans.

Bacteria seem to impact GVHD

The researchers first studied 76 adult transplant patients, analyzing stool samples taken at roughly 10 days after transplant (+/- 4 days). The team performed 16S gene sequencing using the Roche 454 platform.

This revealed the presence of several types of bacteria, including 6 gram-positive Firmicutes, 2 gram-negative Proteobacteria, and 2 gram-negative Bacteroidetes.

The researchers then used a computational assay to determine which of these bacteria might be associated with protection from GVHD. And they identified 2 possibilities—Lactobacillus and Blautia.

Additional analyses revealed that Blautia and Lactobacillus were significantly associated with GVHD-related mortality at 1500 days after transplant (P=0.03 and 0.01, respectively). But there was no significant association with Bacteroides (P=0.6), Enterobacteriales (P=0.2), or Enterococcus (P=0.3).

Blautia appears to affect GVHD-related mortality

To confirm their initial findings, Dr Jenq and his colleagues analyzed a second cohort of 50 adult transplant patients. The team analyzed stool samples for the abundance of bacterial subgroups using a different sequencing platform, Illumina miseq.

This time, they found that Blautia abundance predicted GVHD-related mortality at more than 500 days after transplant, but the abundance of Lactobacillus did not (P=0.01 and 1, respectively).

“Not enough Blautia in your gut seems to lead to an increase in GVHD-related mortality,” Dr Jenq said. “So what does this do to overall survival? In the first cohort, there’s a big difference in overall survival between the ‘haves’ and ‘have nots’ with Blautia [P=0.0008]. And this also holds up in the second cohort [P=0.04].”

Further analyses of data from both cohorts suggested that Blautia abundance was associated with GVHD-related mortality (P=0.004) and relapse-related mortality (P=0.01) but not non-relapse- and non-GVHD-related mortality (P=0.4).

“I don’t have a good explanation for [the relationship between Blautia and relapse-related death],” Dr Jenq said. “This was a surprise finding.”

The researchers also looked at Blautia’s ability to predict GVHD-related mortality. They found that, around day 10 after transplant, Blautia abundance predicts “very strongly” for GVHD-related death.

Another question was whether known GVHD risk factors—such as donor type, race, gender, and performance status—impact Blautia abundance. But an analysis revealed that Blautia is an independent risk factor for GVHD-related mortality.

A possible mechanism

To gain more insight into the association between Blautia and GVHD-related death, Dr Jenq and his colleagues decided to study it in mice.

The team killed off Blautia in mice using vancomycin and ampicillin, then introduced either murine Blautia or murine Enterococcus, transplanted the mice with MHC-disparate T cells, and monitored them for GVHD.

Mice that received Blautia had significantly better overall survival (at more than 80 days after transplant) than mice that received Enterococcus (P<0.001).

“So how is this happening?” Dr Jenq asked. “We think, potentially, it might be due to short-chain fatty acids . . . butyric acid, propionic acid, and acetic acid. These are metabolites that bacteria produce when they ferment glucose and other sugars.”

To test this theory, the researchers treated mice with antibiotics and introduced Blautia or Enterococcus.

Blautia increased the level of short-chain fatty acids (butyrate and propionate) when compared to Enterococcus, although levels were not as high as those observed in mice that did not receive antibiotics. Nevertheless, these results point to a possible mechanism, according to Dr Jenq.

Explaining Blautia reduction

Dr Jenq also noted that antibiotics may contribute to the decrease in Blautia observed in transplant patients. When patients come in for transplant, they often have more than 25% Blautia in their stool. But the bacteria decrease to negligible levels by day 2 after transplant.

To determine the role of antibiotics, the researchers treated mice with 4 different antibiotics and looked at the levels of different bacteria.

They found that aztreonam and cefepime increased the levels of Bacteroidales and Clostridiales (the family to which Blautia belongs), but imipenem and metronidazole decreased bacteria levels.

So antibiotics do affect Blautia levels, Dr Jenq said, but they’re only part of the problem. He noted that patients’ Blautia levels start to decrease before antibiotics are administered. So he and his colleagues believe nutrition might also play a part.

The team found a significant difference in Blautia abundance between patients who received total parenteral nutrition and those who did not (P<0.001).

The researchers also discovered that reduced caloric intake led to a loss of Blautia and other Clostridiales. They analyzed 50 samples from 5 patients and found that patients who consumed fewer than 500 calories had a marked reduction in Blautia (P<0.0001).

And experiments in mice confirmed this association. A week of calorie restriction significantly reduced the abundance of Blautia and other Clostridiales (P=0.0002).

“In GVHD, as we all know, patients and mice eat less because of the nausea,” Dr Jenq said. “And we found that GVHD itself can also lead to a reduction in Clostridiales, both in humans [P=0.02] and in mice [P=0.01].”

Protecting Blautia to prevent GVHD

Having confirmed the role of nutrition in Blautia reduction, the researchers set out to identify a nutrition-based intervention to support Blautia in transplant recipients.

They settled on a sugar called raffinose, which is found in beans, cruciferous vegetables, and whole grains. It passes undigested through the upper intestine but is fermented in the lower intestine and metabolized to produce short-chain fatty acids.

The team tested raffinose in mice by introducing it into their drinking water. At 100 days after transplant, mice that received raffinose had significantly better overall survival than controls (P<0.001).

Based on these results, Dr Jenq and his colleagues believe nutritional intervention can protect Blautia and, therefore, may prevent GVHD and related death. The team thinks encouraging eating, gastric nutritional supplementation, and flora-targeted nutritional supplements might all prove effective.

But other interventions might work as well, such as reintroducing endogenous flora (via autologous fecal microbiota transplant), reintroducing select bacteria with beneficial potential, selecting antibiotics that spare bacteria with beneficial potential, and identifying and introducing bacterial metabolites that mediate anti-inflammatory effects.

*Data in the abstract differ from data presented.

GRAPEVINE, TEXAS—Intestinal bacteria can offer protection from death related to graft-vs-host disease (GVHD), according to research presented at the 2014 BMT Tandem Meetings.

Experiments showed that Blautia, commensal bacteria found in the intestinal tract, can protect against GVHD-related mortality in mice and in humans.

So efforts to support Blautia survival—such as restricting the use of antibiotics and promoting better nutrition—may

prevent GVHD-related death, according to researchers.

Robert Jenq, MD, of Memorial Sloan-Kettering Cancer Center in New York, discussed this possibility when presenting this research, which was designated one of the “Best Abstracts” at the meeting (abstract 1*).

Dr Jenq noted that researchers have been trying for decades to determine whether the intestinal flora impact GVHD. Clinical studies have suggested that prophylaxis against anaerobes and gram-positive bacteria can reduce GVHD.

And murine studies have indicated that prophylaxis against gram-negative bacteria can reduce GVHD, that Lactobacillus can reduce GVHD, and that donor microbiota do not impact GVHD.

“If you’re confused, so are we,” Dr Jenq said. “It seems like it’s a mixed picture.”

So he and his colleagues conducted a series of experiments in an attempt to determine if any bacterial subgroups impact the risk of gut GVHD in mice and humans.

Bacteria seem to impact GVHD

The researchers first studied 76 adult transplant patients, analyzing stool samples taken at roughly 10 days after transplant (+/- 4 days). The team performed 16S gene sequencing using the Roche 454 platform.

This revealed the presence of several types of bacteria, including 6 gram-positive Firmicutes, 2 gram-negative Proteobacteria, and 2 gram-negative Bacteroidetes.

The researchers then used a computational assay to determine which of these bacteria might be associated with protection from GVHD. And they identified 2 possibilities—Lactobacillus and Blautia.

Additional analyses revealed that Blautia and Lactobacillus were significantly associated with GVHD-related mortality at 1500 days after transplant (P=0.03 and 0.01, respectively). But there was no significant association with Bacteroides (P=0.6), Enterobacteriales (P=0.2), or Enterococcus (P=0.3).

Blautia appears to affect GVHD-related mortality

To confirm their initial findings, Dr Jenq and his colleagues analyzed a second cohort of 50 adult transplant patients. The team analyzed stool samples for the abundance of bacterial subgroups using a different sequencing platform, Illumina miseq.

This time, they found that Blautia abundance predicted GVHD-related mortality at more than 500 days after transplant, but the abundance of Lactobacillus did not (P=0.01 and 1, respectively).

“Not enough Blautia in your gut seems to lead to an increase in GVHD-related mortality,” Dr Jenq said. “So what does this do to overall survival? In the first cohort, there’s a big difference in overall survival between the ‘haves’ and ‘have nots’ with Blautia [P=0.0008]. And this also holds up in the second cohort [P=0.04].”

Further analyses of data from both cohorts suggested that Blautia abundance was associated with GVHD-related mortality (P=0.004) and relapse-related mortality (P=0.01) but not non-relapse- and non-GVHD-related mortality (P=0.4).

“I don’t have a good explanation for [the relationship between Blautia and relapse-related death],” Dr Jenq said. “This was a surprise finding.”

The researchers also looked at Blautia’s ability to predict GVHD-related mortality. They found that, around day 10 after transplant, Blautia abundance predicts “very strongly” for GVHD-related death.

Another question was whether known GVHD risk factors—such as donor type, race, gender, and performance status—impact Blautia abundance. But an analysis revealed that Blautia is an independent risk factor for GVHD-related mortality.

A possible mechanism

To gain more insight into the association between Blautia and GVHD-related death, Dr Jenq and his colleagues decided to study it in mice.

The team killed off Blautia in mice using vancomycin and ampicillin, then introduced either murine Blautia or murine Enterococcus, transplanted the mice with MHC-disparate T cells, and monitored them for GVHD.

Mice that received Blautia had significantly better overall survival (at more than 80 days after transplant) than mice that received Enterococcus (P<0.001).

“So how is this happening?” Dr Jenq asked. “We think, potentially, it might be due to short-chain fatty acids . . . butyric acid, propionic acid, and acetic acid. These are metabolites that bacteria produce when they ferment glucose and other sugars.”

To test this theory, the researchers treated mice with antibiotics and introduced Blautia or Enterococcus.

Blautia increased the level of short-chain fatty acids (butyrate and propionate) when compared to Enterococcus, although levels were not as high as those observed in mice that did not receive antibiotics. Nevertheless, these results point to a possible mechanism, according to Dr Jenq.

Explaining Blautia reduction

Dr Jenq also noted that antibiotics may contribute to the decrease in Blautia observed in transplant patients. When patients come in for transplant, they often have more than 25% Blautia in their stool. But the bacteria decrease to negligible levels by day 2 after transplant.

To determine the role of antibiotics, the researchers treated mice with 4 different antibiotics and looked at the levels of different bacteria.

They found that aztreonam and cefepime increased the levels of Bacteroidales and Clostridiales (the family to which Blautia belongs), but imipenem and metronidazole decreased bacteria levels.

So antibiotics do affect Blautia levels, Dr Jenq said, but they’re only part of the problem. He noted that patients’ Blautia levels start to decrease before antibiotics are administered. So he and his colleagues believe nutrition might also play a part.

The team found a significant difference in Blautia abundance between patients who received total parenteral nutrition and those who did not (P<0.001).

The researchers also discovered that reduced caloric intake led to a loss of Blautia and other Clostridiales. They analyzed 50 samples from 5 patients and found that patients who consumed fewer than 500 calories had a marked reduction in Blautia (P<0.0001).

And experiments in mice confirmed this association. A week of calorie restriction significantly reduced the abundance of Blautia and other Clostridiales (P=0.0002).

“In GVHD, as we all know, patients and mice eat less because of the nausea,” Dr Jenq said. “And we found that GVHD itself can also lead to a reduction in Clostridiales, both in humans [P=0.02] and in mice [P=0.01].”

Protecting Blautia to prevent GVHD

Having confirmed the role of nutrition in Blautia reduction, the researchers set out to identify a nutrition-based intervention to support Blautia in transplant recipients.

They settled on a sugar called raffinose, which is found in beans, cruciferous vegetables, and whole grains. It passes undigested through the upper intestine but is fermented in the lower intestine and metabolized to produce short-chain fatty acids.

The team tested raffinose in mice by introducing it into their drinking water. At 100 days after transplant, mice that received raffinose had significantly better overall survival than controls (P<0.001).

Based on these results, Dr Jenq and his colleagues believe nutritional intervention can protect Blautia and, therefore, may prevent GVHD and related death. The team thinks encouraging eating, gastric nutritional supplementation, and flora-targeted nutritional supplements might all prove effective.

But other interventions might work as well, such as reintroducing endogenous flora (via autologous fecal microbiota transplant), reintroducing select bacteria with beneficial potential, selecting antibiotics that spare bacteria with beneficial potential, and identifying and introducing bacterial metabolites that mediate anti-inflammatory effects.

*Data in the abstract differ from data presented.

GRAPEVINE, TEXAS—The total nucleated cell (TNC) dose delivered in an allogeneic peripheral blood stem cell transplant (allo-PBSCT) can affect outcomes in certain patients, according to a study presented at the 2014 BMT Tandem Meetings.

Researchers found that a higher TNC dose was associated with better progression-free survival (PFS) and overall survival (OS) among patients who received allo-PBSCT with reduced-intensity conditioning (RIC) and total-body irradiation (TBI).

On the other hand, the dose of CD3+, CD4+, CD8+, or CD34+ cells did not have a significant impact on survival rates in these patients.

And none of the cell doses studied had a significant impact in patients who did not receive TBI or in those who received TBI with myeloablative conditioning.

Michael Burns, of Roswell Park Cancer Institute in Buffalo, New York, presented these findings at the meeting as abstract 12.*

Burns noted that studies have produced conflicting results regarding the correlation between patient outcomes and the dose of CD34+, CD3+, CD4+, or CD8+ cells given in allo-PBSCT. In addition, TNC dose has not been analyzed much in the context of PBSCTs.

Therefore, he and his colleagues retrospectively analyzed graft cell composition in 254 patients who received their first allo-PBSCT from January 2001 to September 2012.

Fifty-eight percent of the patients were male, and the median age was 50 (range, 19-73 years). Forty-four percent of patients had acute myeloid leukemia, 18% had myelodysplastic syndromes or myeloproliferative neoplasms, 13% had acute lymphoblastic leukemia, 13% had non-Hodgkin lymphoma, and 12% had other diseases.

Of the 254 patients studied, 93 had received TBI. Among these, 53 received myeloablative conditioning (91% cyclophosphamide, 120 cGy), and 40 received RIC (100% fludarabine and melphalan, 400 cGy).

Of the 161 patients who did not receive TBI, 41 received myeloablative conditioning (88% busulfan and cyclophosphamide), and 120 received RIC (87% fludarabine and melphalan).

Patients received T-cell-replete, G-CSF mobilized, PB allografts. Fifty-six percent had a 6/6 HLA matched related donor, and 44% had an 8/8 HLA matched unrelated donor. Forty-nine percent of patients were in complete remission at the time of transplant.

The researchers analyzed cell doses according to the median dose (above vs below). But they also analyzed CD34+ dose as < 4 x 10⁶ cells/kg vs ≥ 4 x 10⁶ cells/kg and as < 4 x 10⁶ cells/kg vs 4 to 8 x 10⁶ cells/kg vs > 8 x 10⁶ cells/kg. They analyzed TNC as < 8 x 10⁸ cells/kg vs ≥ to 8 x 10⁸ cells/kg.

The team found that a CD34+ cell dose greater than 4 x 10⁶ cells/kg was significantly associated with time to platelet engraftment in all patients. It was also associated with time to neutrophil engraftment in the TBI group, but this was predominantly among patients who received RIC.

On the other hand, CD3+, CD4+, CD8+, and TNC doses were not significantly associated with platelet or neutrophil engraftment in any patients.

CD34+, CD3+, CD4+, and CD8+ cell dose were not associated with OS, PFS, or acute graft-vs-host disease (GVHD). And TNC had no significant effect on acute GVHD.

“However, we did find that the TNC dose did show some pretty interesting survival outcomes,” Burns said.

A higher TNC dose (≥ 8 x10⁸ cells/kg) was associated with significantly better PFS (P=0.027) and OS (P=0.018) in the TBI patients but not in patients who did not receive TBI (P>0.1 for PFS and OS).

When they analyzed patients according to conditioning regimen, the researchers found the association retained significance among patients who received RIC (P=0.01 for PFS and P=0.007 for OS) but not among patients who received myeloablative conditioning (P>0.1 for PFS and OS).

Burns and his colleagues also conducted a multivariate analysis to see if any other factors affected the relationship between TNC and survival. They controlled for patient age, Karnofsky performance status, and body mass index. And they stratified patients into 4 groups according to TBI and conditioning regimen.

The results showed that patients who received TBI and RIC, as well as a TNC dose less than 8 x 10⁸ cells/kg, had a relative risk of 3.3 for PFS (P=0.026) and a relative risk of 3.4 for OS (P=0.021).

“The association of higher TNC dose with better progression-free and overall survival implies there is a population of nucleated cells which mitigate GVHD but enhance the [graft-vs-leukemia] effect after reduced-intensity TBI conditioning,” Burns said.

“Myeloablative conditioning regimens result in more direct tumor killing. Thus, they rely less on the graft-vs-leukemia effect than the RIC regimens.”

He also noted that the lack of an association between TNC dose and survival rates with non-TBI-based regimens implies there are different mechanisms of tumor kill with TBI and non-TBI-containing regimens.

And a more detailed analysis of cell population subsets in apheresis product may allow researchers to identify cell populations that could improve patient outcomes.

*Some data in the abstract differ from data presented at the meeting.

GRAPEVINE, TEXAS—The total nucleated cell (TNC) dose delivered in an allogeneic peripheral blood stem cell transplant (allo-PBSCT) can affect outcomes in certain patients, according to a study presented at the 2014 BMT Tandem Meetings.

Researchers found that a higher TNC dose was associated with better progression-free survival (PFS) and overall survival (OS) among patients who received allo-PBSCT with reduced-intensity conditioning (RIC) and total-body irradiation (TBI).

On the other hand, the dose of CD3+, CD4+, CD8+, or CD34+ cells did not have a significant impact on survival rates in these patients.

And none of the cell doses studied had a significant impact in patients who did not receive TBI or in those who received TBI with myeloablative conditioning.

Michael Burns, of Roswell Park Cancer Institute in Buffalo, New York, presented these findings at the meeting as abstract 12.*

Burns noted that studies have produced conflicting results regarding the correlation between patient outcomes and the dose of CD34+, CD3+, CD4+, or CD8+ cells given in allo-PBSCT. In addition, TNC dose has not been analyzed much in the context of PBSCTs.

Therefore, he and his colleagues retrospectively analyzed graft cell composition in 254 patients who received their first allo-PBSCT from January 2001 to September 2012.

Fifty-eight percent of the patients were male, and the median age was 50 (range, 19-73 years). Forty-four percent of patients had acute myeloid leukemia, 18% had myelodysplastic syndromes or myeloproliferative neoplasms, 13% had acute lymphoblastic leukemia, 13% had non-Hodgkin lymphoma, and 12% had other diseases.

Of the 254 patients studied, 93 had received TBI. Among these, 53 received myeloablative conditioning (91% cyclophosphamide, 120 cGy), and 40 received RIC (100% fludarabine and melphalan, 400 cGy).

Of the 161 patients who did not receive TBI, 41 received myeloablative conditioning (88% busulfan and cyclophosphamide), and 120 received RIC (87% fludarabine and melphalan).

Patients received T-cell-replete, G-CSF mobilized, PB allografts. Fifty-six percent had a 6/6 HLA matched related donor, and 44% had an 8/8 HLA matched unrelated donor. Forty-nine percent of patients were in complete remission at the time of transplant.

The researchers analyzed cell doses according to the median dose (above vs below). But they also analyzed CD34+ dose as < 4 x 10⁶ cells/kg vs ≥ 4 x 10⁶ cells/kg and as < 4 x 10⁶ cells/kg vs 4 to 8 x 10⁶ cells/kg vs > 8 x 10⁶ cells/kg. They analyzed TNC as < 8 x 10⁸ cells/kg vs ≥ to 8 x 10⁸ cells/kg.

The team found that a CD34+ cell dose greater than 4 x 10⁶ cells/kg was significantly associated with time to platelet engraftment in all patients. It was also associated with time to neutrophil engraftment in the TBI group, but this was predominantly among patients who received RIC.

On the other hand, CD3+, CD4+, CD8+, and TNC doses were not significantly associated with platelet or neutrophil engraftment in any patients.

CD34+, CD3+, CD4+, and CD8+ cell dose were not associated with OS, PFS, or acute graft-vs-host disease (GVHD). And TNC had no significant effect on acute GVHD.

“However, we did find that the TNC dose did show some pretty interesting survival outcomes,” Burns said.

A higher TNC dose (≥ 8 x10⁸ cells/kg) was associated with significantly better PFS (P=0.027) and OS (P=0.018) in the TBI patients but not in patients who did not receive TBI (P>0.1 for PFS and OS).

When they analyzed patients according to conditioning regimen, the researchers found the association retained significance among patients who received RIC (P=0.01 for PFS and P=0.007 for OS) but not among patients who received myeloablative conditioning (P>0.1 for PFS and OS).

Burns and his colleagues also conducted a multivariate analysis to see if any other factors affected the relationship between TNC and survival. They controlled for patient age, Karnofsky performance status, and body mass index. And they stratified patients into 4 groups according to TBI and conditioning regimen.

The results showed that patients who received TBI and RIC, as well as a TNC dose less than 8 x 10⁸ cells/kg, had a relative risk of 3.3 for PFS (P=0.026) and a relative risk of 3.4 for OS (P=0.021).

“The association of higher TNC dose with better progression-free and overall survival implies there is a population of nucleated cells which mitigate GVHD but enhance the [graft-vs-leukemia] effect after reduced-intensity TBI conditioning,” Burns said.

“Myeloablative conditioning regimens result in more direct tumor killing. Thus, they rely less on the graft-vs-leukemia effect than the RIC regimens.”

He also noted that the lack of an association between TNC dose and survival rates with non-TBI-based regimens implies there are different mechanisms of tumor kill with TBI and non-TBI-containing regimens.

And a more detailed analysis of cell population subsets in apheresis product may allow researchers to identify cell populations that could improve patient outcomes.

*Some data in the abstract differ from data presented at the meeting.

GRAPEVINE, TEXAS—The total nucleated cell (TNC) dose delivered in an allogeneic peripheral blood stem cell transplant (allo-PBSCT) can affect outcomes in certain patients, according to a study presented at the 2014 BMT Tandem Meetings.

Researchers found that a higher TNC dose was associated with better progression-free survival (PFS) and overall survival (OS) among patients who received allo-PBSCT with reduced-intensity conditioning (RIC) and total-body irradiation (TBI).

On the other hand, the dose of CD3+, CD4+, CD8+, or CD34+ cells did not have a significant impact on survival rates in these patients.

And none of the cell doses studied had a significant impact in patients who did not receive TBI or in those who received TBI with myeloablative conditioning.

Michael Burns, of Roswell Park Cancer Institute in Buffalo, New York, presented these findings at the meeting as abstract 12.*

Burns noted that studies have produced conflicting results regarding the correlation between patient outcomes and the dose of CD34+, CD3+, CD4+, or CD8+ cells given in allo-PBSCT. In addition, TNC dose has not been analyzed much in the context of PBSCTs.

Therefore, he and his colleagues retrospectively analyzed graft cell composition in 254 patients who received their first allo-PBSCT from January 2001 to September 2012.

Fifty-eight percent of the patients were male, and the median age was 50 (range, 19-73 years). Forty-four percent of patients had acute myeloid leukemia, 18% had myelodysplastic syndromes or myeloproliferative neoplasms, 13% had acute lymphoblastic leukemia, 13% had non-Hodgkin lymphoma, and 12% had other diseases.

Of the 254 patients studied, 93 had received TBI. Among these, 53 received myeloablative conditioning (91% cyclophosphamide, 120 cGy), and 40 received RIC (100% fludarabine and melphalan, 400 cGy).

Of the 161 patients who did not receive TBI, 41 received myeloablative conditioning (88% busulfan and cyclophosphamide), and 120 received RIC (87% fludarabine and melphalan).

Patients received T-cell-replete, G-CSF mobilized, PB allografts. Fifty-six percent had a 6/6 HLA matched related donor, and 44% had an 8/8 HLA matched unrelated donor. Forty-nine percent of patients were in complete remission at the time of transplant.

The researchers analyzed cell doses according to the median dose (above vs below). But they also analyzed CD34+ dose as < 4 x 10⁶ cells/kg vs ≥ 4 x 10⁶ cells/kg and as < 4 x 10⁶ cells/kg vs 4 to 8 x 10⁶ cells/kg vs > 8 x 10⁶ cells/kg. They analyzed TNC as < 8 x 10⁸ cells/kg vs ≥ to 8 x 10⁸ cells/kg.

The team found that a CD34+ cell dose greater than 4 x 10⁶ cells/kg was significantly associated with time to platelet engraftment in all patients. It was also associated with time to neutrophil engraftment in the TBI group, but this was predominantly among patients who received RIC.

On the other hand, CD3+, CD4+, CD8+, and TNC doses were not significantly associated with platelet or neutrophil engraftment in any patients.

CD34+, CD3+, CD4+, and CD8+ cell dose were not associated with OS, PFS, or acute graft-vs-host disease (GVHD). And TNC had no significant effect on acute GVHD.

“However, we did find that the TNC dose did show some pretty interesting survival outcomes,” Burns said.

A higher TNC dose (≥ 8 x10⁸ cells/kg) was associated with significantly better PFS (P=0.027) and OS (P=0.018) in the TBI patients but not in patients who did not receive TBI (P>0.1 for PFS and OS).

When they analyzed patients according to conditioning regimen, the researchers found the association retained significance among patients who received RIC (P=0.01 for PFS and P=0.007 for OS) but not among patients who received myeloablative conditioning (P>0.1 for PFS and OS).

Burns and his colleagues also conducted a multivariate analysis to see if any other factors affected the relationship between TNC and survival. They controlled for patient age, Karnofsky performance status, and body mass index. And they stratified patients into 4 groups according to TBI and conditioning regimen.

The results showed that patients who received TBI and RIC, as well as a TNC dose less than 8 x 10⁸ cells/kg, had a relative risk of 3.3 for PFS (P=0.026) and a relative risk of 3.4 for OS (P=0.021).

“The association of higher TNC dose with better progression-free and overall survival implies there is a population of nucleated cells which mitigate GVHD but enhance the [graft-vs-leukemia] effect after reduced-intensity TBI conditioning,” Burns said.

“Myeloablative conditioning regimens result in more direct tumor killing. Thus, they rely less on the graft-vs-leukemia effect than the RIC regimens.”

He also noted that the lack of an association between TNC dose and survival rates with non-TBI-based regimens implies there are different mechanisms of tumor kill with TBI and non-TBI-containing regimens.

And a more detailed analysis of cell population subsets in apheresis product may allow researchers to identify cell populations that could improve patient outcomes.

*Some data in the abstract differ from data presented at the meeting.