Metabolic changes in T cells may limit CAR potential in kids

David Barrett, MD, PhD

Researchers analyzed peripheral blood T cells from 157 pediatric cancer patients at diagnosis and after chemotherapy and found the potential to produce effective chimeric antigen receptor (CAR) T cells declined with each cycle of chemotherapy.

This was also true for acute lymphoblastic leukemia (ALL) and Wilms’ tumor, which had high CAR T-cell manufacturing potential in the pre-chemotherapy samples.

Children younger than 3 years particularly showed a significant decline in CAR T-cell potential with cumulative cycles of chemotherapy.

“Everybody knows that chemotherapy is really bad for your T cells, and the more chemo you get, the less likely you are to have healthy T cells,” David M. Barrett, MD, PhD, of Children’s Hospital of Philadelphia in Pennsylvania, said at a press preview of research to be presented at the AACR Annual Meeting 2018.

“We know a lot about what a highly active, highly successful CAR T cell looks like right before it goes back into the patient after it’s finished manufacturing,” Dr Barrett added.

But he and his colleagues wanted to determine what goes into producing high-quality cells from a patient and the difference between cells that were good starting material and cells that weren’t.

The investigators analyzed blood samples from pediatric patients with ALL, non-Hodgkin lymphoma, neuroblastoma, osteosarcoma, rhabdomyosarcoma, Wilms’ tumor, Hodgkin lymphoma, chronic myeloid leukemia, and Ewing sarcoma. The team collected samples at diagnosis and after every cycle of chemotherapy.

Using flow cytometry, they quantified the CD3+ cell population and expanded the T cells using CD3 and CD28 stimulatory beads, “the backbone of pretty much every center’s way to make CAR T cells in the lab,” Dr Barrett said.

And the researchers found poor CAR T-cell manufacturing potential in all tumor types at diagnosis except for ALL and Wilms’ tumor. In standard-risk and high-risk ALL, more than 90% of patients had high-quality T cells at diagnosis.

The team report the findings in abstract 1631, which is scheduled to be presented at the AACR Annual Meeting on April 15.

“This may have played into why pediatric ALL is one of the great successes with CAR T-cell therapy,” Dr Barrett explained. “We may have actually been working with uniquely well-suited, good starting material to build a CAR T cell.”

T cells from lymphoma patients—Burkitt lymphoma, diffuse large B-cell lymphoma, primary mediastinal B-cell lymphoma, and Hodgkin lymphoma—were actually quite poor in their potential to become good CAR T cells, Dr Barrett noted.

“This may be reflected clinically in pediatrics at Children’s Hospital of Philadelphia,” he said. “We’ve only been able to successfully treat 3 children with lymphoma, as opposed to more than 200 children with leukemia.”

The only other type of tumor that seemed to have good CAR T potential was Wilms’ tumor.

“I don’t have a CAR T cell for Wilms’ tumor yet,” Dr Barrett said, “but, if I wanted to make one, I would at least have a degree of confidence that the cells gotten from a patient would at least be able to be successfully made into a highly functional T cell that can go back into a patient.”

The investigators also observed that cumulative chemotherapy alters the metabolic profile in T cells, “gradually turning them by cycle 6 into something that doesn’t work anymore,” Dr Barrett said.

The researchers then looked into what differences there were in the quality of collected T cells and found that metabolic changes varied with tumor and treatment.

T cells with poor CAR T-cell potential were biased toward using glycolysis as their energy source instead of using fatty acids.

“Normal, healthy donor T cells cluster together in terms of metabolic pathways that are active or inactive,” Dr Barrett explained.

“[P]atients who had leukemia and the Wilms’ tumor patients could make successful CAR T cells from those samples. On the other hand, solid tumors and a Hodgkin disease patient look like they have a very different metabolic profile. And that is associated with failure to make a good CAR T cell.”

The investigators were able to get the T cells to shift metabolic pathways by “essentially force-feeding T cells things like fatty acids so they don’t use as much glucose,” Dr Barrett said.

“We’ve had some success in force-feeding them essentially neutral amino acids and others like arginine. And so you can actually potentially provide a T cell with an attractive alternative fuel source.”

Dr Barrett noted that the findings have already altered practice for children at his institution.

They now collect T cells early even if the patient is not eligible for a CAR trial, “simply because we know that cumulative chemotherapy is going to progressively deteriorate the likelihood that those cells will make a functional CAR product, and we’ve been recommending that to other centers,” Dr Barrett said.

“We’re trying to understand what goes into making the best starting material so that we can alter our approaches to make sure that we make a highly functional CAR T-cell product not only for kids with leukemia and CART19, but also potentially for solid tumor CARs as we try to develop those in the future.” 

David Barrett, MD, PhD

Researchers analyzed peripheral blood T cells from 157 pediatric cancer patients at diagnosis and after chemotherapy and found the potential to produce effective chimeric antigen receptor (CAR) T cells declined with each cycle of chemotherapy.

This was also true for acute lymphoblastic leukemia (ALL) and Wilms’ tumor, which had high CAR T-cell manufacturing potential in the pre-chemotherapy samples.

Children younger than 3 years particularly showed a significant decline in CAR T-cell potential with cumulative cycles of chemotherapy.

“Everybody knows that chemotherapy is really bad for your T cells, and the more chemo you get, the less likely you are to have healthy T cells,” David M. Barrett, MD, PhD, of Children’s Hospital of Philadelphia in Pennsylvania, said at a press preview of research to be presented at the AACR Annual Meeting 2018.

“We know a lot about what a highly active, highly successful CAR T cell looks like right before it goes back into the patient after it’s finished manufacturing,” Dr Barrett added.

But he and his colleagues wanted to determine what goes into producing high-quality cells from a patient and the difference between cells that were good starting material and cells that weren’t.

The investigators analyzed blood samples from pediatric patients with ALL, non-Hodgkin lymphoma, neuroblastoma, osteosarcoma, rhabdomyosarcoma, Wilms’ tumor, Hodgkin lymphoma, chronic myeloid leukemia, and Ewing sarcoma. The team collected samples at diagnosis and after every cycle of chemotherapy.

Using flow cytometry, they quantified the CD3+ cell population and expanded the T cells using CD3 and CD28 stimulatory beads, “the backbone of pretty much every center’s way to make CAR T cells in the lab,” Dr Barrett said.

And the researchers found poor CAR T-cell manufacturing potential in all tumor types at diagnosis except for ALL and Wilms’ tumor. In standard-risk and high-risk ALL, more than 90% of patients had high-quality T cells at diagnosis.

The team report the findings in abstract 1631, which is scheduled to be presented at the AACR Annual Meeting on April 15.

“This may have played into why pediatric ALL is one of the great successes with CAR T-cell therapy,” Dr Barrett explained. “We may have actually been working with uniquely well-suited, good starting material to build a CAR T cell.”

T cells from lymphoma patients—Burkitt lymphoma, diffuse large B-cell lymphoma, primary mediastinal B-cell lymphoma, and Hodgkin lymphoma—were actually quite poor in their potential to become good CAR T cells, Dr Barrett noted.

“This may be reflected clinically in pediatrics at Children’s Hospital of Philadelphia,” he said. “We’ve only been able to successfully treat 3 children with lymphoma, as opposed to more than 200 children with leukemia.”

The only other type of tumor that seemed to have good CAR T potential was Wilms’ tumor.

“I don’t have a CAR T cell for Wilms’ tumor yet,” Dr Barrett said, “but, if I wanted to make one, I would at least have a degree of confidence that the cells gotten from a patient would at least be able to be successfully made into a highly functional T cell that can go back into a patient.”

The investigators also observed that cumulative chemotherapy alters the metabolic profile in T cells, “gradually turning them by cycle 6 into something that doesn’t work anymore,” Dr Barrett said.

The researchers then looked into what differences there were in the quality of collected T cells and found that metabolic changes varied with tumor and treatment.

T cells with poor CAR T-cell potential were biased toward using glycolysis as their energy source instead of using fatty acids.

“Normal, healthy donor T cells cluster together in terms of metabolic pathways that are active or inactive,” Dr Barrett explained.

“[P]atients who had leukemia and the Wilms’ tumor patients could make successful CAR T cells from those samples. On the other hand, solid tumors and a Hodgkin disease patient look like they have a very different metabolic profile. And that is associated with failure to make a good CAR T cell.”

The investigators were able to get the T cells to shift metabolic pathways by “essentially force-feeding T cells things like fatty acids so they don’t use as much glucose,” Dr Barrett said.

“We’ve had some success in force-feeding them essentially neutral amino acids and others like arginine. And so you can actually potentially provide a T cell with an attractive alternative fuel source.”

Dr Barrett noted that the findings have already altered practice for children at his institution.

They now collect T cells early even if the patient is not eligible for a CAR trial, “simply because we know that cumulative chemotherapy is going to progressively deteriorate the likelihood that those cells will make a functional CAR product, and we’ve been recommending that to other centers,” Dr Barrett said.

“We’re trying to understand what goes into making the best starting material so that we can alter our approaches to make sure that we make a highly functional CAR T-cell product not only for kids with leukemia and CART19, but also potentially for solid tumor CARs as we try to develop those in the future.” 

David Barrett, MD, PhD

Researchers analyzed peripheral blood T cells from 157 pediatric cancer patients at diagnosis and after chemotherapy and found the potential to produce effective chimeric antigen receptor (CAR) T cells declined with each cycle of chemotherapy.

This was also true for acute lymphoblastic leukemia (ALL) and Wilms’ tumor, which had high CAR T-cell manufacturing potential in the pre-chemotherapy samples.

Children younger than 3 years particularly showed a significant decline in CAR T-cell potential with cumulative cycles of chemotherapy.

“Everybody knows that chemotherapy is really bad for your T cells, and the more chemo you get, the less likely you are to have healthy T cells,” David M. Barrett, MD, PhD, of Children’s Hospital of Philadelphia in Pennsylvania, said at a press preview of research to be presented at the AACR Annual Meeting 2018.

“We know a lot about what a highly active, highly successful CAR T cell looks like right before it goes back into the patient after it’s finished manufacturing,” Dr Barrett added.

But he and his colleagues wanted to determine what goes into producing high-quality cells from a patient and the difference between cells that were good starting material and cells that weren’t.

The investigators analyzed blood samples from pediatric patients with ALL, non-Hodgkin lymphoma, neuroblastoma, osteosarcoma, rhabdomyosarcoma, Wilms’ tumor, Hodgkin lymphoma, chronic myeloid leukemia, and Ewing sarcoma. The team collected samples at diagnosis and after every cycle of chemotherapy.

Using flow cytometry, they quantified the CD3+ cell population and expanded the T cells using CD3 and CD28 stimulatory beads, “the backbone of pretty much every center’s way to make CAR T cells in the lab,” Dr Barrett said.

And the researchers found poor CAR T-cell manufacturing potential in all tumor types at diagnosis except for ALL and Wilms’ tumor. In standard-risk and high-risk ALL, more than 90% of patients had high-quality T cells at diagnosis.

The team report the findings in abstract 1631, which is scheduled to be presented at the AACR Annual Meeting on April 15.

“This may have played into why pediatric ALL is one of the great successes with CAR T-cell therapy,” Dr Barrett explained. “We may have actually been working with uniquely well-suited, good starting material to build a CAR T cell.”

T cells from lymphoma patients—Burkitt lymphoma, diffuse large B-cell lymphoma, primary mediastinal B-cell lymphoma, and Hodgkin lymphoma—were actually quite poor in their potential to become good CAR T cells, Dr Barrett noted.

“This may be reflected clinically in pediatrics at Children’s Hospital of Philadelphia,” he said. “We’ve only been able to successfully treat 3 children with lymphoma, as opposed to more than 200 children with leukemia.”

The only other type of tumor that seemed to have good CAR T potential was Wilms’ tumor.

“I don’t have a CAR T cell for Wilms’ tumor yet,” Dr Barrett said, “but, if I wanted to make one, I would at least have a degree of confidence that the cells gotten from a patient would at least be able to be successfully made into a highly functional T cell that can go back into a patient.”

The investigators also observed that cumulative chemotherapy alters the metabolic profile in T cells, “gradually turning them by cycle 6 into something that doesn’t work anymore,” Dr Barrett said.

The researchers then looked into what differences there were in the quality of collected T cells and found that metabolic changes varied with tumor and treatment.

T cells with poor CAR T-cell potential were biased toward using glycolysis as their energy source instead of using fatty acids.

“Normal, healthy donor T cells cluster together in terms of metabolic pathways that are active or inactive,” Dr Barrett explained.

“[P]atients who had leukemia and the Wilms’ tumor patients could make successful CAR T cells from those samples. On the other hand, solid tumors and a Hodgkin disease patient look like they have a very different metabolic profile. And that is associated with failure to make a good CAR T cell.”

The investigators were able to get the T cells to shift metabolic pathways by “essentially force-feeding T cells things like fatty acids so they don’t use as much glucose,” Dr Barrett said.

“We’ve had some success in force-feeding them essentially neutral amino acids and others like arginine. And so you can actually potentially provide a T cell with an attractive alternative fuel source.”

Dr Barrett noted that the findings have already altered practice for children at his institution.

They now collect T cells early even if the patient is not eligible for a CAR trial, “simply because we know that cumulative chemotherapy is going to progressively deteriorate the likelihood that those cells will make a functional CAR product, and we’ve been recommending that to other centers,” Dr Barrett said.

“We’re trying to understand what goes into making the best starting material so that we can alter our approaches to make sure that we make a highly functional CAR T-cell product not only for kids with leukemia and CART19, but also potentially for solid tumor CARs as we try to develop those in the future.”