Part II: Metabolomics in Cell and Gene Therapy Product Development

Introduction

Today, Cell and Gene Therapies (CGTs) are safer and more efficacious than ever, with the FDA having approved more CGTs in 2023 than any previous year¹. Despite these advancements, developing a safe, specific, and potent CGT remains far from easy due to challenges associated with identifying therapeutic targets, evaluating product safety, and determining which patients will benefit from the therapy^2-5. One currently underutilized technology that can help address these challenges is metabolomics.

Background

Metabolomics is the study of metabolites: the small molecule reactants, intermediates, and products of metabolism that are organized into biochemical pathways. Metabolites indicate disease activity through changes in their abundance, which can be measured using liquid chromatography-tandem mass spectrometry (LC-MS). By reporting on biochemical changes, metabolomics can provide deep phenotypic insight into cellular and molecular processes that regulate disease onset, progression, and response to treatment. This insight, which cannot be gained from genomics or proteomics alone, is crucial to developing efficacious CGTs. The first blog in this series discussed the benefits of including metabolomics in Potency Assurance strategies for CGTs⁶. Here, we broaden our discussion of CGT product development, and examine studies that demonstrate the ability of metabolomics to identify targets of GCT products, assess product safety, and select the best patient candidates for the therapy.

Identifying Therapeutic Targets

Historically, metabolomics was only used to identify therapeutic targets for inborn errors of metabolism⁷. While the contributions of metabolomics to broadening therapeutic strategies for metabolic diseases cannot be overstated, our discussion will focus on the proven utility of metabolomics in identifying therapeutic targets for non-metabolic diseases.

One such disease is Retinitis Pigmentosa (RP), an inherited progressive deterioration of rods followed by secondary death of cones that eventually leads to irreversible blindness. Since RP is caused by 3100+ mutations in more than 80 genes⁸, gene therapies that target common disease pathways rather than individual mutations (gene-agnostic gene therapies) are seen as the ideal treatment solution.

Finding gene-agnostic targets for RP is a topic of intense research, which has been significantly advanced by metabolomics. Early findings on gene expression from four different mouse models of RP suggested that cone death is caused by aberrant activation of the mTOR pathway, leading to a shortage of glucose in the retina⁹. One research group characterized these findings using a genetic mouse model of RP in which disease pathology was induced through chronic upregulation of mTOR signaling¹⁰. To test the glucose shortage hypothesis, retinal tissue from RP mice and wildtype littermates were analyzed using untargeted metabolomic profiling. RP mice had significantly increased levels of several compounds in the glycolysis pathway and oxidative arm of the pentose phosphate pathway. This phenotype aligned with upregulation of several genes that encode enzymes critical to the glycolysis and pentose phosphate pathways, as well as de novo synthesis of sterols, isoprenoids, and fatty acids. These findings suggested that aberrant mTOR signaling may upset the balance between anabolic and catabolic processes, leading to a metabolic favoring of catabolic reactions which could potentially shorten the outer segment of photoreceptors and cause their death. A follow up study performed by a different research group further characterized these findings by showing oxidative stress to be a key disrupter of glucose metabolism leading to its depletion in the retina¹¹.

Other studies expanded upon this early work by identifying microglia, the innate immune cells of the CNS, as a major driver of RP progression through changes in their metabolic tuning. In the resting state microglia derive energy through glycolysis, which generates ATP through oxidative phosphorylation. Yet, in response to oxidative stress, they shift from using oxidative phosphorylation to generate ATP to using aerobic glycolysis, which promotes nutrient uptake and proliferation¹². This metabolic switch is associated with the chronic release of neuroinflammatory factors, which may underlie photoreceptor death. These findings suggested that loss of rods and cones in RP could be prevented by reverting microglia back to oxidative metabolism. Transforming growth factor-β1 (TGF-β1) has been shown to suppress the metabolic switch from oxidative phosphorylation to aerobic glycolysis in microglia^13,14. A recent study showed that AAV-mediated delivery of TGF-β1-expressing genes prevented the metabolic switch and in turn, rescued cone degeneration in three mouse models of RP that each carried a different pathogenic mutation¹⁵. This study also showed that depleting microglia or blocking TGF-β receptors reversed this metabolic rescue, resulting in photoreceptor death.

Altogether, metabolomics has shown that response of microglia to oxidative stress both depletes glucose from the retina and increases inflammation, to result in photoreceptor death characteristic of RP. The insight that metabolomics has revealed about RP progression and pathophysiology in mouse models has identified at least one promising gene-agnostic therapeutic target that can be pursued in future human studies. Additionally, these findings demonstrate the capability of metabolomics in characterizing therapeutic targets of non-metabolic disease.

Evaluating Product Safety

CGT product safety is multifaceted, encompassing toxicity, off-target effects, and side effects of the treatment. Metabolomics can inform product safety, not only by identifying a safety concern, but also by revealing mechanisms that trigger and regulate the biological process that underpins the concern.

For example, anti-CD19 chimeric antigen receptor (CAR) T-cell therapy is a highly effective treatment for specific types of cancer¹⁶. However, its use in the clinic is tempered by a side effect of the treatment known as clinically significant acute inflammatory toxicities. These toxicities, which include cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), induce significant morbidity, and are unpredictable in both onset and symptom severity. To address this safety concern, one research group aimed to better predict which patients are most likely to develop CRS or ICANS so that proper monitoring and prompt mitigation practices can be deployed¹⁷. They theorized that since elevated levels of glucose support higher cytokine levels and stronger T-cell metabolism, preexisting metabolic profiles may be predictive of CRS or ICANS in response to CAR T-cell therapy. Untargeted metabolomics was performed on EDTA plasma collected from patients before they started CAR T-cell therapy. These data along with follow-up targeted metabolomics profiling showed that higher levels of glucose and lower levels of cholesterol and glutamate were highly associated with faster onset of CRS. High severity CRS was associated with lower levels of the amino acids, proline and glycine, and the secondary bile acid isoursdeoxycholate. Findings also showed that lower levels of the amino acid hydroxyproline were associated with a more severe and faster onset of ICANS, while delayed onset correlated with low levels of glutamine. Altogether, these data revealed key metabolite biomarkers that were shown to be predictive of the onset and severity of a serious side effect of CAR T-cell therapy. By indicating which patients require the closest monitoring and mitigation strategies to be at the ready, these markers can significantly improve patient outcomes.

The contributions that metabolomics can make towards evaluating CGT product safety was shown in a second study that examined the occurrence of CRS in two groups of patients that each received a different chemotherapy drug as part of the standard preparation for CAR T-cell treatment.

Before starting CAR T-cell therapy, patients must take chemotherapy drugs to deplete their lymphocyte supply. Lymphodepletion is essential to favorable outcomes of CAR T-cell therapy because it reduces tumor cells and creates a niche for incoming CAR T infusions¹⁸. Several lymphodepletion drugs are available, but recommended dosages are different for different CAR T products, and safety data on various combinations of these drugs with CAR T products is lacking^19,20. This study evaluated the outcomes of CAR T-cell therapy in response to two lymphodepletion regimens given to patients with large B-cell non-Hodgkin lymphoma²¹. Two groups of patients were followed. One group received the lymphodepletion drug bendamustine (Benda) while the second received the drug cocktail fludarabine/cyclophosphamide (Flu/Cy). Serum samples collected before and after lymphodepletion were analyzed for cytokine levels and global metabolomic changes. Even though CAR T-cell efficacy was the same between groups, any-grade CRS occurred in 91.9% of patients that received Flu/Cy compared to 72.7% of patients who received Benda (p < 0.05), and any-grade ICANS occurred in 54.9% of the Flu/Cy patient group as opposed to 18.2% of patients in the Benda group (p < 0.05). Cytokine and metabolomics analyses revealed that patients who received Flu/Cy had a significantly higher inflammatory cytokine burden and reduced levels of nicotinamide ribose and NAD+/NADH, metabolites that are essential for counterbalancing oxidative stress. This group also showed reduced levels of carnitine-esters, which are essential for biosynthesis of lipids and amino acids. These findings suggest that low levels nicotinamide ribose and NAD+/NADH may facilitate the onset of cytokine release syndrome, and the resulting elevation in cytotoxicity removes lipids and amino acids from circulation to promote hematopoiesis. The authors note that the mechanistic insights derived from the metabolomics analysis suggest that supplementation of nicotinamide ribose during CAR T-cell therapy may mitigate CRS and ICANS. Overall, these findings demonstrate the benefits of including metabolomics in safety assessments for CGTs.

Determining Patient Candidacy

Determining a patient’s candidacy for a given CGT requires knowing if that patient is likely to respond to the treatment. Metabolomics can provide this insight into CGT products that are both in development and available for use.

For example, anti-CD19 CAR T-cell therapy is an FDA-approved treatment for relapsed or refractory (r/r) large B cell lymphoma (LBCL). Despite the high overall response rate, long-term response rate to CAR T-cell therapy is low in LBCL tumors that overexpress MYC, an oncogene that encodes the MYC transcription factor. Traditionally, MYC overexpression is detected through protein analysis of a tumor biopsy. One research group aimed to develop an easier way to identify biomarkers of MYC overexpression to determine patient suitability for CAR T-cell therapy²². In this study they evaluated the plasma metabolomic profiles of patients with high-expressing MYC tumors verse normal-expressing MYC tumors. Elevated levels of certain acetylated polyamines were highly associated with elevated MYC expression and a short duration response to treatment. As a result of these findings a 6-metabolite panel consisting of acetylspermidine, diacetylspermidine, and several lysophospholipids was validated in a separate patient cohort and shown to predict short-term response to CAR T-cell therapy. This study demonstrates the utility of metabolomics in predicting a patient’s response to a CGT product and the impact that such insight can have on clinical decision making and patient outcomes.

Conclusions

The studies discussed in this blog demonstrate the benefits of using metabolomics to identify therapeutic targets of CGTs, evaluate their safety, and determine the patient population most likely to benefit from them. Overall, these examples show that including metabolomics assessments in certain stages of CGT product development may help produce a highly targeted, safe, and efficacious treatment.

References  

1. Senior M. Fresh from the biotech pipeline: record-breaking FDA approvals. Nat Biotechnol. Mar 2024;42(3):355-361. doi:10.1038/s41587-024-02166-7
2. Kohn DB, Chen YY, Spencer MJ. Successes and challenges in clinical gene therapy. Gene Ther. Nov 2023;30(10-11):738-746. doi:10.1038/s41434-023-00390-5
3. Weber T. Anti-AAV Antibodies in AAV Gene Therapy: Current Challenges and Possible Solutions. Front Immunol. 2021;12:658399. doi:10.3389/fimmu.2021.658399
4. Sterner RC, Sterner RM. CAR-T cell therapy: current limitations and potential strategies. Blood Cancer J. Apr 6 2021;11(4):69. doi:10.1038/s41408-021-00459-7
5. Chohan KL, Siegler EL, Kenderian SS. CAR-T Cell Therapy: the Efficacy and Toxicity Balance. Curr Hematol Malig Rep. Apr 2023;18(2):9-18. doi:10.1007/s11899-023-00687-7
6. Sommerville L. Metabolomics for Cell and Gene Therapy. https://www.metabolon.com/blog/metabolomics-cell-gene-therapy/
7. Yilmaz BS, Gurung S, Perocheau D, Counsell J, Baruteau J. Gene therapy for inherited metabolic diseases. J Mother Child. Nov 10 2020;24(2):53-64. doi:10.34763/jmotherandchild.20202402si.2004.000009
8. Daiger SP, Sullivan LS, Bowne SJ. Genes and mutations causing retinitis pigmentosa. Clin Genet. Aug 2013;84(2):132-41. doi:10.1111/cge.12203
9. Punzo C, Kornacker K, Cepko CL. Stimulation of the insulin/mTOR pathway delays cone death in a mouse model of retinitis pigmentosa. Nat Neurosci. Jan 2009;12(1):44-52. doi:10.1038/nn.2234
10. Duvel K, Yecies JL, Menon S, et al. Activation of a metabolic gene regulatory network downstream of mTOR complex 1. Mol Cell. Jul 30 2010;39(2):171-83. doi:10.1016/j.molcel.2010.06.022
11. Kanan Y, Hackett SF, Taneja K, Khan M, Campochiaro PA. Oxidative stress-induced alterations in retinal glucose metabolism in Retinitis Pigmentosa. Free Radic Biol Med. Mar 2022;181:143-153. doi:10.1016/j.freeradbiomed.2022.01.032
12. Orihuela R, McPherson CA, Harry GJ. Microglial M1/M2 polarization and metabolic states. Br J Pharmacol. Feb 2016;173(4):649-65. doi:10.1111/bph.13139
13. Mia S, Warnecke A, Zhang XM, Malmstrom V, Harris RA. An optimized protocol for human M2 macrophages using M-CSF and IL-4/IL-10/TGF-beta yields a dominant immunosuppressive phenotype. Scand J Immunol. May 2014;79(5):305-14. doi:10.1111/sji.12162
14. Zhang F, Wang H, Wang X, et al. TGF-beta induces M2-like macrophage polarization via SNAIL-mediated suppression of a pro-inflammatory phenotype. Oncotarget. Aug 9 2016;7(32):52294-52306. doi:10.18632/oncotarget.10561
15. Wang SK, Xue Y, Cepko CL. Microglia modulation by TGF-beta1 protects cones in mouse models of retinal degeneration. J Clin Invest. Aug 3 2020;130(8):4360-4369. doi:10.1172/JCI136160
16. Institute NNC. CAR T Cells: Engineering Patients’ Immune Cells to Treat Their Cancers. https://www.cancer.gov/about-cancer/treatment/research/car-t-cells
17. Jalota A, Hershberger CE, Patel MS, et al. Host metabolome predicts the severity and onset of acute toxicities induced by CAR T-cell therapy. Blood Adv. Sep 12 2023;7(17):4690-4700. doi:10.1182/bloodadvances.2022007456
18. Lickefett B, Chu L, Ortiz-Maldonado V, et al. Lymphodepletion – an essential but undervalued part of the chimeric antigen receptor T-cell therapy cycle. Front Immunol. 2023;14:1303935. doi:10.3389/fimmu.2023.1303935
19. Neelapu SS. CAR-T efficacy: is conditioning the key? Blood. Apr 25 2019;133(17):1799-1800. doi:10.1182/blood-2019-03-900928
20. Amini L, Silbert SK, Maude SL, et al. Preparing for CAR T cell therapy: patient selection, bridging therapies and lymphodepletion. Nat Rev Clin Oncol. May 2022;19(5):342-355. doi:10.1038/s41571-022-00607-3
21. Ghilardi G, Paruzzo L, Svoboda J, et al. Bendamustine lymphodepletion before axicabtagene ciloleucel is safe and associates with reduced inflammatory cytokines. Blood Adv. Feb 13 2024;8(3):653-666. doi:10.1182/bloodadvances.2023011492
22. Fahrmann JF, Saini NY, Chia-Chi C, et al. A polyamine-centric, blood-based metabolite panel predictive of poor response to CAR-T cell therapy in large B cell lymphoma. Cell Rep Med. Nov 15 2022;3(11):100720. doi:10.1016/j.xcrm.2022.100720