Keywords

FormalPara Key Points

  • Aggressive lymphomas exhibit the Warburg effect.

  • Lactic acidosis is a result of the overproduction of lactate and leads to a fatal prognosis.

  • Mutation of p53 helps cancer cells survive glutamine deprivation.

  • PI3K regulates fatty acid synthesis (FAS) in primary effusion lymphoma (PEL) and other B-NHLs.

  • AMPK regulates NADPH balance for fatty acid oxidation (FAO) to supplement the TCA cycle.

  • PRPS2 couples protein and nucleotide biosynthesis to drive lymphomagenesis.

  • mTOR activation promotes fatty acid synthesis (FAS).

  • MYC regulates cancer cell metabolism under glucose-deprived and hypoxic conditions.

  • HIF-1 acts as a regulator in hypoxia adaption and the related metabolic changes.

  • Knowledge of metabolic phenotypes in cancer can be used in tandem with genetic alterations to develop effective treatment strategies.

  • [18F]FDG PET/CT is a valuable tool to visualize tumor glycolytic activity and characterize metabolic heterogeneity.

1 Introduction

Non-Hodgkin lymphomas (NHLs) are a heterogeneous group of lymphoid neoplasms with different biological characteristics. About 90% of all lymphomas in the United States originate from B lymphocytes, while the remaining originate from T cells [1]. The treatment of NHLs depends on the neoplastic histology and stage of the tumor, which will indicate whether radiotherapy, chemotherapy, or a combination is the best suitable treatment [2]. The American Cancer Society describes the staging of lymphoma as follows: Stage I is lymphoma in a single node or area. Stage II is when that lymphoma has spread to another node or organ tissue. Stage III is when it has spread to lymph nodes on two sides of the diaphragm. Stage IV is when cancer has significantly spread to organs outside the lymph system. Radiation therapy is the traditional therapeutic route for localized follicular and mucosa-associated lymphomas. Chemotherapy is utilized for the treatment of large-cell lymphomas and high-grade lymphomas [2]. However, the treatment of indolent lymphomas remains problematic as the patients often have metastasis, for which no standard approach exists [2].

Follicular lymphoma (FL), a form of non-Hodgkin lymphoma, is the second most common form of B-cell lymphoma and remains incurable in the majority of cases, despite recent advances, including anti-CD20 antibodies (rituximab) and kinase inhibitors (ibrutinib) [3]. Following an indolent phase, 50% of patients suffer from disease transformation to an aggressive form of lymphoma (transformed FL; tFL) [4]. This dramatic switch in disease behavior typically culminates in rapid deterioration and is usually fatal. Accordingly, much effort has been focused on understanding the genetics of transformation, which has resulted in the identification of key genetic lesions (e.g., MYC activation, loss of p53, activation of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), loss of tumor necrosis factor-alpha-induced protein 3 A20 (TNFAIP3/A20)) [5,6,7,8]. However, exactly how tumor metabolism, which is altered by these genetic lesions, contributes to disease aggressiveness is not known. Therefore, the metabolic changes that occur during FL transformation are poorly understood. We need to understand the biological and metabolic changes upon disease transformation in order to develop effective treatment strategies.

Malignant cells have metabolic adaptations supporting bioenergetics, biosynthesis, and redox homeostasis in response to the development of the tumor microenvironment [9]. Metabolic heterogeneity is present in the tumor microenvironment, where concentrations of key resources can be spatially (localization) and temporally (stage of the diseases) varied [10], creating the heterogeneity of metabolism within the same tumor [11]. Cancer metabolism is influenced by tumor localization and vascularization status. Cancer cells can uptake nutrients and oxygen from the blood supply, which results in the production of ATP via aerobic oxidative phosphorylation (OXPHOS) as well as through anabolic pathways, supporting rapid cell proliferation.

In this chapter, we describe the intricacies of NHLs’ metabolism resulting from alterations in gene expressions, which subsequently lead to poor prognosis. Furthermore, we explore how metabolomics technologies [12] and analysis can be applied to treatment strategies.

2 Lymphoma Metabolism Exhibits Multifaceted Characteristic Features Which Are Correlated to Poor Prognosis

2.1 Aggressive Lymphomas Exhibit the Warburg Effect

As described in the previous chapters, the drastic increase in glucose uptake of cancer cells is a feature of the distinctive metabolic rewiring known as the Warburg effect [13]. Recently, researchers have taken advantage of this metabolic shift to clinically detect localized glucose uptake of cancer cells using 18F-fluorodeoxyglucose positron-emission tomography ([18F]FDG PET) . High-grade NHL patients and intermediate-grade NHL patients with poor prognoses showed a high accumulation of [18F]FDG in their tumors [14, 15].

Primary effusion lymphoma (PEL) exhibits high glycolytic activity due to its hypoxic environment. This form of lymphoma requires aggressive treatment, but no standard therapy exists [16]. PEL is, however, highly sensitive to glucose withdrawal and glycolysis inhibitors, such as 2-deoxyglucose (2-DG) [16]. In this situation, the distinctive metabolic phenotype, glucose dependency, offers hope for effective treatments.

Cancer cells exhibiting the Warburg effect avidly take up glucose. After glucose uptake, cancer cells favor the conversion of glucose-derived pyruvate to lactate. Recent reports showed that NHL patients had elevated plasma lactate and lactate dehydrogenase (LDH) levels, which were linked to poor survival rates [17,18,19]. Furthermore, inhibition of LDHA, the enzyme that catalyzes the conversion of pyruvate to lactate, has yielded positive results in tumor reduction in a MYC-transformed Burkitt lymphoma model in vivo [20, 21]. Taken together, the characteristic metabolic features of the Warburg effect offer diagnostic tools as well as relevant therapeutic targets for NHL.

2.2 Lactic Acidosis Is a Result of Overproduction of Lactate and Leads to a Fatal Prognosis

Following the Warburg effect, lactic acidosis can occur when lactate homeostasis is disproportioned, due to overproduction and/or underutilization. Lactic acidosis is divided into two categories: type A and type B. Type A results from poor oxygenation in the tissue. Type B occurs in normoxic tissue as a result of a drug or toxin [22]. Type B lactic acidosis is the result of the alteration of glycolytic processes and their effects on redox [23, 24]. Type B lactic acidosis is present in many human malignancies, but notably in lymphomas and leukemias [22, 23, 25,26,27]. Once cancers exhibit type B lactic acidosis, these cases show poor prognoses and outcomes if not immediately treated [22].

Another notable cause of lactic acidosis is thiamine deficiency, an observable characteristic connected to type B lactic acidosis in some cancers. Thiamine is a cofactor that is necessary for the conversion of pyruvate into acetyl-CoA via pyruvate dehydrogenase. When malignant cells exhibit thiamine deficiency, pyruvate is heavily converted to lactate [28, 29]. Subsequently, thiamine deficiency leads to lactic acidosis (Fig. 1).

Fig. 1
figure 1

The effect of thiamine on lactate and acetyl-CoA production. Some cancer cells produce lactate even in the presence of oxygen, an effect known as the Warburg effect. Lactic acidosis can result from unregulated lactic acid (lactate) buildup. When thiamine is present, there is a mixture of lactate and acetyl-CoA (left). When thiamine is deficient, acetyl-CoA production is impaired, and pyruvate is mainly converted to lactate resulting in lactate buildup (right)

Full size image

3 Genetic Alterations Lead to Different Metabolic Phenotypes in NHL (Fig. 2)

Fig. 2
figure 2

Genetic alterations lead to different metabolic phenotypes in non-Hopkin lymphoma. Various lymphomas show these alterations in gene expression and the resulting changes in metabolism

Full size image

NHLs often have abnormal activation of the mammalian target of rapamycin complex 1 (mTORC1) that reprograms multiple metabolic pathways, including nucleotide synthesis, amino acid synthesis, fatty acid synthesis, and glutaminolysis. Additionally, MYC is another important trigger for inducing many genes correlated with anabolic growth, including transporters and enzymes involved in glycolysis, mitochondrial biogenesis, fatty acid synthesis, and glutaminolysis [30,31,32,33,34]. MYC is a gene involved in cellular proliferation whose dysregulation was found in B-cell lymphomas [35]. Metabolic reprograming by transcription factors such as MYC and hypoxia-inducible factor 1 (HIF-1) in malignant tissues allows them to better survive the tumor microenvironmental alterations. These various genes can often influence each other; for instance, mTOR can also activate HIF-1 expression even under normoxic states [36] (Table 1).

Table 1 Genetic alterations lead to different metabolic phenotypes in NHL
Full size table

3.1 Mutation of p53 Helps Cancer Cells Survive Glutamine Deprivation

As described in a previous chapter, many cancers depend on glutamine for bioenergy, redox homeostasis, and DNA synthesis, which are essential requirements for cancer cell survival. Therapeutic strategies often target cancer’s glutamine dependency [37]. However, these treatments do not always have the intended impact, as many cancers are resistant to treatment. One such example is that of TP53, a protein responsible for tumor suppression, and its mutant form [38]. Specifically, in lymphoma cell lines, Tran et al. reported that mutp53 proteins could directly bind to the promoters of p53 target genes that regulate the cell cycle, which leads to cell cycle arrest and helps cancer cells survive in glutamine deprivation conditions [38]. Cancer cells expressing mutp53 proteins are able to survive the metabolic stress of glutamine deprivation in poorly vascularized tumor microenvironments, whereas p53-deficient cells and wtp53-expressing cells experience impaired proliferation and increased cell death [38]. The resistance to glutamine deprivation in mutp53-expressing malignant cells allows these cells to survive in metabolically restrictive environments.

3.2 PI3K Regulates Fatty Acid Synthesis (FAS) in Primary Effusion Lymphoma (PEL) and Other B-NHLs

While many lymphomas rely on glucose to produce lactate and energize their metabolism, this is not always the case. Dysregulation of cell metabolism in primary effusion lymphoma (PEL), an aggressive type of B-cell lymphoma, increased not only aerobic glycolysis but also fatty acid synthesis [39, 40]. By using 14C-labeled glucose, Bhatt et al. showed that PEL created more lipids from glucose compared to primary B cells. Furthermore, these cells were sensitive to both an inhibitor of fatty acid synthase, C75, and an inhibitor of glycolysis, 2-DG. Each of these inhibitors affected both glycolysis and fatty acid synthesis (FAS) [39]. Bhatt et al. showed a significant difference in the metabolic profiles of primary B cells and those of human B-cell non-Hodgkin lymphomas (B-NHL), including PEL. Poor-prognosis PEL and other B-NHLs exhibit high levels of aerobic glycolysis and FAS. This suggests that different types of malignant lymphomas can be distinguished by the rate of fatty acid biosynthesis, which may have the potential for targeted therapy against these aggressive lymphomas [39].

Previous work of Bhatt et al. and others showed that PEL cells exhibit high activities of phosphatidylinositol 3-kinase (PI3K), protein kinase B (AKT), and mTOR, genes related to proliferation and survival as well as glycolysis [41,42,43]. In their more recent work, they showed that inhibiting PI3K by LY294002 decreased not only glycolytic flux but also the incorporation of 14C glucose into lipids [39].

In summary, glucose was important to PEL cells for providing not only energy but also acetyl-CoA for lipid synthesis. This illustrates that lymphoma’s metabolism is complex and can exhibit multidimensional alterations.

3.3 AMPK Regulates NADPH Balance for Fatty Acid Oxidation (FAO) as a Means of Supplementing the Tricarboxylic Acid (TCA) Cycle

Jeon et al. showed that AMPK orchestrates NADPH consumption (by FAS) and production (from fatty acid oxidation (FAO)) in lymphoma to support ATP synthesis, redox homeostasis and biosynthesis under low-glucose environments [44]. By doing so, AMPK decreases pentose phosphate pathway activity and increases FAO [44].

Diffuse large B-cell lymphoma (DLBCL), a common lymphoma, utilizes FAO to support energy production and growth [45]. Fatty acids provide fuel for oxidative phosphorylation (OXPHOS), and to increase glutathione levels, and attenuate oxidative stress [45]. The DLBCL with OXPHOS is aggressive and resistant to ibrutinib, an inhibitor of B-cell receptor (BCR) survival signaling [46, 47]. More research into the combination of fatty acid oxidation-targeting drugs and the BCR inhibitor could provide potential therapeutic approaches for patients with DLBCL [45,46,47,48].

3.4 PRPS2 Couples Protein and Nucleotide Biosynthesis to Drive Lymphomagenesis

One of the current mainstay chemotherapeutic strategies involves targeting one-carbon metabolism in malignant cancers. This strategy reduces the production of nucleotides and ATP, as well as alters redox homeostasis. Individual drugs often inhibit the metabolism of folate, nucleotides, and, most notably, thymidine [48, 49]. Key enzyme targets of nucleic acid synthesis include dihydrofolate reductase, thymidylate synthase, adenine/adenosine deaminase, and DNA polymerase/ribonucleotide reductase [48,49,50,51]. As indicated by Cunningham et al., nucleotide biosynthesis is coupled to protein biosynthesis by a critical enzyme, phosphoribosyl-pyrophosphate synthetase 2 (PRPS2), which specifically promotes increased nucleotide biosynthesis in MYC-driven lymphoma. In these lymphomas, PRPS2 may be an effective anticancer target, and other enzymes in this pathway utilized by oncogenes may also exist as potential targets [52].

3.5 mTOR Activation Promotes Fatty Acid Synthesis (FAS)

mTOR activation during nutrient abundance enhances aerobic glycolysis and lipid synthesis, which is mediated by the sterol regulatory element-binding protein (SREBP) group by inducing the transcription of the fatty acid-synthesizing enzyme (FASN) [53]. FASN is present at elevated levels in the liver and at lower levels in other tissues, but cancerous tissues express excessive FASN, which has been identified as a metabolic oncogene [40, 54,55,56].

In PEL, rapamycin treatment improves survival time in the in vivo model by inhibiting autocrine signaling and vascular endothelial growth factor (VEGF) [43, 57]. On the other hand, Shestov et al. revealed that inhibition of mTOR has an impact on the flux of glycolysis, pentose phosphate pathway, and TCA cycle [58]. Thus, while mTOR activation is linked to FAS, inhibition of mTOR can impact other metabolic pathways critical to cancer growth.

3.6 MYC Regulates Cancer Cell Metabolism under Glucose-Deprived and Hypoxic Conditions

MYC is considered to be a regulator in glycolysis and mitochondrial respiration [33, 34, 59,60,61,62,63]. Using stable isotope-resolved metabolomics [12], Le et al. explored the metabolic alterations that occur in the oncogenic transcription factor c-MYC-inducible human Burkitt lymphoma model P493 cell line under aerobic and hypoxic conditions as well as glucose deprivation. They found the coexistence of oxidative and aerobic glycolysis. They also documented the prominent contribution of glutamine to the TCA cycle of proliferating cells and that hypoxic cancer cells continue to oxidize glutamine for cell growth and survival. Furthermore, this study showed that glutamine metabolism alone could sustain the TCA cycle for cell survival and growth in the absence of glucose [64]. This glucose-independent pathway reflects the dependence of cancer cells on metabolic reprogramming, allowing for the survival and proliferation of cancer cells under the harsh hypoxic and nutrient-deprived conditions of the tumor microenvironment. Their study demonstrated that inhibition of glutaminase, the enzyme that catalyzes the reaction of glutamine to glutamate, by BPTES, impaired MYC-transformed B lymphoma growth in vivo [64]. Other glutaminase inhibitors, such as BPTES analogs, have been developed to target glutamine metabolism in cancers [65].

MYC also regulates proline metabolism, as found by Liu et al. [66]. They found that proline dehydrogenase (POX/PRODH), the first enzyme in proline catabolism, was suppressed by MYC through upregulating miR-23b*. This study provided a deeper understanding of cancer metabolism while enabling the development of novel therapeutic strategies.

3.7 HIF-1 Acts as a Regulator in Hypoxia Adaptation and the Related Metabolic Changes

HIF-1 activity is enhanced by mTOR-altered metabolism and promotes glycolysis as a hypoxia-adaptive transcriptional program. The HIF-1 and HIF-2 heterodimers respond to and are stabilized by hypoxia, resulting in metabolic changes [67]. Of these two heterodimers, HIF-1 is a critical component involved in tumor metabolism that upregulates glucose transporters, glycolytic enzymes, and pyruvate dehydrogenase kinase, isozyme 1 (PDK1), an enzyme which prevents pyruvate from entering the TCA cycle [36]. Qiao et al. demonstrated that malignant lymphomas exhibit constitutive expression of HIF-1α. This expression is mediated by NF-κB, and ionizing radiation treatment of lymphoma showed increased NF-κB activation and elevated HIF-1α levels. This indicates that additional treatment targeting HIF-1α in combination with radiation therapy of lymphoma cells could potentially improve patient outcomes [68].

3.8 Understanding the PI3K/AKT/mTOR Pathway in Lymphoma Can Lead to a Variety of Treatments

Thus far, we have discussed how metabolic analysis has-improved-our-understanding-of NHL metabolism. In some cases, the genetic alterations that lead to unique metabolic phenotypes, such as those discussed in Sect. 3, have been targeted for therapy and have resulted in clinical trials. Additionally, other tools such as [18F]FDG PET/computed tomography (CT). can help gauge the lymphoma metabolism to inform therapy and predict outcomes.

Sections 3.2 and 3.5 of this chapter have highlighted how PI3K and mTOR play key roles in controlling lymphoma metabolism. Due to the importance of the PI3K/AKT/mTOR pathway, targeting this pathway has already begun to show promising results in the treatment of different lymphoma types. For instance, in a 220-patient study of chronic lymphocytic leukemia (CLL), dual therapy with both idelalisib, a PI3Kδ inhibitor, and rituximab, an established CD20 antibody commonly used against many lymphomas, improved the rate of progression-free survival (PFS) from 46% to 93% compared to monotherapy with rituximab alone [69].

However, in clinical studies, sometimes a less straightforward approach may be required for treatment due to unforeseen pathway changes. For instance, the PI3K/AKT/mTOR pathway is thought to play a role in refractory mantle cell lymphoma (MCL) resistance to ibrutinib [70,71,72]. As such, one may logically reason that PI3K/AKT/mTOR inhibitors would be a potential therapy for refractory MCL. However, the use of the PI3K inhibitor idelalisib and the rapamycin analog temsirolimus has provided underwhelming clinical outcomes [70, 73,74,75]. To explain this ineffective treatment, consider a study by Garcia et al. focusing on multidrug resistance (MDR), which coincided with an upregulated PI3K/AKT pathway. In this study, inhibition of PI3K via wortmannin and LY294002 induced cell death, but also coincided with the activation of NF-κB, which can promote cell survival, thus mitigating the effect of PI3K-targeting treatment [76]. Zhang et al. postulate that understanding the metabolic reprogramming coincident with genetic alterations will be critical to a successful treatment approach. They identified a strong reliance on OXPHOS in refractory MCL, and subsequent treatment with IACS-010759, a clinically relevant inhibitor of complex 1 of the electron transport chain (ETC), yielded more promising results [70]. Additionally, glutamine uptake was also upregulated, and treatment with amino-oxyacetate, which inhibits glutaminolysis, induced ROS formation and oxidative stress [70]. Thus, knowledge about both the genetic alterations and the metabolic shifts is key to developing effective treatment strategies.

4 Metabolic Profiling for Monitoring Tumor Progression and Guiding Treatment

4.1 [18F]FDG PET/CT

Regions of the body exhibiting elevated levels of glycolysis, such as many tumors, will show accumulation of the [18F]FDG as assessed by PET/CT. Various metabolic parameters can be determined and used to diagnose a particular lymphoma case. The most frequently measured parameter is the standardized uptake value (SUV). The mean SUV and the max SUV are also useful parameters [77]. From the SUV, the metabolic tumor volume (MTV) and total lesion glycolysis (TLG) can be determined. MTV is the volume of the tumor that is metabolically active and is also sometimes referred to as total metabolic tumor volume (TMTV) when distinct regions are totaled. TLG is a dimensionless index rating the average glycolysis activity for the entire tumor and calculated by multiplying the mean SUV by the MTV [77, 78]. Additionally, metabolic heterogeneity can be approximated by the “area under the curve of the cumulative SUV histogram” (AUC CSH) method, where a lower AUC implies greater variability in metabolism [79]. Figure 3 shows a simplified visualization of the process for determining these key parameters and terms.

Fig. 3
figure 3

Simplified [18F]FDG PET/CT process. Injection of the isotopic glucose analog [18F]FDG and subsequent analysis by PET/CT can determine key metabolic parameters (SUV, MTV, TLG, and MH) for research or clinical applications

Full size image

In DLBCL, patients with high TMTV and TLG exhibited a more advanced stage of lymphoma, and these parameters were strong inverse predictors of progression-free survival and overall survival (OS) outcomes [80, 81]. For instance, high TLG indicated a higher progression rate (41% PFS) and worse overall survival (45% OS) compared to low TLG (72% PFS and 73% OS) [80]. Additionally, patients with overexpression of MYC were also at a greater risk for relapse and progression. In this study, Cottereau et al. combined molecular evaluation (MYC expression) and metabolic imaging (TMTV measurements) characteristics to more accurately diagnose and monitor patients and concluded that this approach could lead to patient-tailored therapies [80].

Interestingly, estimations of intertumoral metabolic heterogeneity (MH) were shown to be an accurate predictor of poor outcomes in patients with higher MTV, coinciding with shorter PFS and OS. Models taking into account both MTV and MH improved upon MTV alone as a predictor for PFS and OS [81, 82]. Similar results were found in primary mediastinal B-cell lymphoma (PMBCL), where a model combining both MH and TLG was found to be a more effective predictive tool than models using only one of these parameters [83]. Thus, [18F]FDG PET/CT can be used to determine key parameters relevant to the diagnosis and predicted outcomes of patients.

Evaluation of metabolism is not only helpful in evaluating DLBCL but also useful for other lymphomas as well. In FL, baseline TMTV was a strong predictor of outcome, where a TMTV of 510 cm3 cutoff is related to a less than 3-year PFS [84]. In primary brain lymphoma (PBL), MTV and TLG were the only [18F]FDG-measured parameters shown to be correlated with PFS and OS [85]. Additionally, [18F]FDG PET/CT analysis is also a beneficial tool to monitor disease progression during treatment. In relapsed or refractory B-cell lymphoma treated with yttrium 90 (90Y) ibritumomab tiuxetan radioimmunotherapy, a drop in SUV of 49% or higher after either 2 or 6 months was shown to be an indicator of successful treatment, and lesser changes indicated a need for a different strategy [86]. [18F]FDG PET/CT is a flexible tool that can be applied to different forms of lymphoma and various therapeutic strategies.

[18F]FDG PET/CT is also useful to predict disease transformation over time, such as Richter’s transformation (RT), which is the conversion of CLL to DLBCL. In multiple studies, RT was shown to coincide with a max SUV greater than 5.0 with a relatively high predictive potential [87,88,89]. Of particular note, a heterogeneous distribution of the [18F]FDG also coincided with lower survival, suggesting greater proliferation for these lymphomas [89].

[18F]FDG PET/CT has become a powerful metabolic tool for medical professionals for diagnosis and progression monitoring in various types of lymphoma. It can be used to visualize the glucose uptake and glycolytic activity of a tumor, as well as to characterize metabolic heterogeneity. Furthermore, it has been used as a methodology to direct patient treatment, where it can be used to confirm successful approaches or provide suggestions for alternative strategies for personalized medicine.

4.2 Systemic NAAG Concentrations for Tumor Growth Monitoring

One recent study found that concentrations of N-acetyl-aspartyl-glutamate (NAAG) in brain tumors positively correlated with patient tumor grades [90]. The authors of this study then investigated the role of plasma NAAG concentrations in tumor growth monitoring using the MYC-transformed human P493-6 model in vivo. They found that systemic NAAG concentrations in plasma of the mice bearing MYC-transformed P493-6 tumors strongly mirrored tumor growth where increases in NAAG concentrations preceded the rise of tumor sizes when MYC was expressed and decreases in NAAG concentrations preceded the decrease of tumor sizes when MYC was suppressed. These findings suggest that plasma NAAG concentrations are linked to tumor growth rates and that changes in systemic NAAG concentrations are detectable before the corresponding changes in tumor sizes. Measurement of NAAG concentrations in peripheral blood is thus a promising non-invasive strategy for timely tumor growth monitoring during cancer treatment [90].

5 Conclusion

The therapeutic challenges in malignant lymphomas include chemoresistance, radiation tolerance, and multidrug resistance. Novel therapeutic strategies, which are based on the metabolic phenotypes of aggressive lymphomas are being pursued [51]. The malignant cells exhibit different pathways for altering catabolism and enhancing anabolism for rapid cell proliferation in order to adapt to the tumor microenvironment [9]. The metabolic differences can occur in many lymphomas; therefore, understanding and learning about these specific differences can lead to new targets for therapy, both individually and in combination with other treatments. Furthermore, metabolomics technologies can be critical to predict outcomes and to elucidate appropriate and novel treatment strategies going forward.