Research progress on the biological function and molecular mechanism of GDF-15 in solid malignant tumors

Abstract

Growth differentiation factor 15 (GDF-15), a stress-responsive member of the transforming growth factor-β (TGF-β) superfamily, is consistently upregulated in multiple solid tumors and closely linked to poor clinical outcomes. This review offers a systematic overview of the pleiotropic functions and principal signaling pathways of GDF-15 in solid malignancies. Within the tumor microenvironment (TME), GDF-15 fuels tumor progression by promoting proliferation, sustaining stemness, remodeling metabolism, and conferring therapy resistance via the TGF-β, Leukemia Inhibitory Factor (LIF)-Signal Transducer and Activator of Transcription 3 (STAT3), and AKT pathways. Notably, GDF-15 orchestrates an immunosuppressive TME by limiting T cell infiltration and expanding regulatory T cells, thereby facilitating immune evasion and resistance to immune checkpoint inhibitors (ICIs). Systemically, GDF-15 contributes to cancer cachexia through activation of the brainstem glial-cell-line-derived neurotrophic factor family receptor α-like (GFRAL)-rearranged during transfection (RET) receptor axis. Accumulating preclinical evidence positions GDF-15 as a promising therapeutic target, particularly for mitigating cachexia and potentiating immunotherapy. However, the context-dependent and dualistic nature of GDF-15 signaling, varying with tumor type, microenvironment, and disease stage, poses substantial hurdles for clinical translation. Future efforts should focus on deciphering the molecular determinants underlying GDF-15’s functional duality, paving the way for precise, context-tailored intervention strategies.

Introduction

Growth differentiation factor 15 (GDF-15) is a distant member of the transforming growth factor-β (TGF-β) superfamily, first identified as macrophage inhibitory cytokine-1 (MIC-1) [1]. Under normal conditions, its expression remains low but rises sharply in response to diverse cellular stresses, such as mitochondrial dysfunction, tissue injury, inflammation, and hypoxia, placing it among the mitochondrial stress-secreted factors mitokines that relay cellular stress signals to systemic levels [2–8].

In metabolic regulation, GDF-15 binds specifically to the glial-cell-line-derived neurotrophic factor family receptor α-like (GFRAL)-rearranged during transfection (RET) receptor complex, which is expressed on neurons in the brainstem area postrema (AP) and nucleus tractus solitarius (NTS). This interaction is independent of classical appetite regulators such as leptin and ghrelin, and it triggers potent anorectic effects while modulating energy expenditure. Through this mechanism, GDF-15 contributes to obesity, chemotherapy-induced anorexia, and cancer cachexia [9–11]. Beyond metabolism, GDF-15 also exerts tissue-protective effects in models of myocardial infarction and neurodegeneration, primarily by limiting reactive oxygen species (ROS) production and preserving mitochondrial dynamics [12–14].

However, GDF-15 takes on a decidedly pro-tumorigenic character within the tumor microenvironment (TME). Solid tumor cells secrete amounts of the cytokine, and elevated circulating levels consistently track with tumor aggressiveness, poor prognosis, and resistance to therapies, most notably immune checkpoint inhibitors (ICIs) [15–19]. Mechanistically, GDF-15 drives proliferation, sustains stem-like phenotypes, and rewires metabolism including the Warburg effect through the TGF-β, Leukemia Inhibitory Factor (LIF)-Signal Transducer and Activator of Transcription 3 (STAT3), and AKT pathways [20]. More critically, it remodels the TME into an immunosuppressive niche by blocking T cell infiltration and expanding regulatory T cells, thereby enabling immune evasion [21].

Preclinical evidence has identified GDF-15 as a promising therapeutic target, especially for alleviating cachexia and enhancing immunotherapy responses. Nevertheless, the effects are highly context-dependent. GDF-15 can either promote or suppress malignancy depending on tumor type, disease stage, and microenvironmental cues, a duality that complicates the design of uniform therapeutic strategies [22–24]. Systemic inhibition of GDF-15 also raises legitimate concerns about interfering with its physiological functions in normal tissues.

In this review, we summarize the biological characteristics of GDF-15, dissect the major signaling pathways through which it operates in solid tumors, and evaluate the emerging clinical significance as both a prognostic biomarker and a therapeutic target. We also discuss unresolved questions and controversies, aiming to inform future precision oncology efforts directed at GDF-15.

Biological characteristics of GDF-15

Structure and classification

GDF-15, also known as macrophage inhibitory cytokine-1 (MIC-1), belongs to the GDF subfamily of the TGF-β superfamily [25]. Unlike classical TGF-β members, GDF-15 exhibits several distinctive structural characterizations. The mature GDF-15 monomer contains the conserved cysteine knot motif characteristic of the TGF-β superfamily, which is essential for dimerization and receptor binding [25]. However, GDF-15 shows low sequence homology (< 30%) with other family members, reflecting its evolutionary divergence [26]. One distinguishing feature is the prolonged prodomain, which remains non-covalently associated with the mature dimer after proteolytic cleavage [27]. This prodomain retention may influence GDF-15 latency, stability, and bioavailability, potentially contributing to its distinct receptor specificity, primarily binding GFRAL-RET rather than canonical TGF-β receptors [27, 28].

Structurally, GDF-15 functions as a disulfide-linked homodimer, with each monomer adopting the characteristic TGF-β fold consisting of two pairs of antiparallel β-strands forming a cysteine knot [25]. The dimer interface involves hydrophobic interactions between the two monomers, stabilized by an interchain disulfide bond [26]. Together, these structural characteristics underlie GDF-15’s unique functional pleiotropy and receptor selectivity.

Stress-induced regulation

GDF-15 is expressed at low levels in most healthy tissues but is robustly induced by diverse cellular stresses. As a canonical mitokine, GDF-15 expression is upregulated by mitochondrial dysfunction via the integrated stress response (ISR), particularly the Activating Transcription Factor 4 (ATF4)- C/EBP Homologous Protein (CHOP) axis [29]. Hypoxia induces GDF-15 through Hypoxia-Inducible Factor 1-alpha (HIF-1α) binding to its promoter, a mechanism observed in low-grade gliomas and linked to glycolytic pathway activation [27]. Tissue injury and inflammation also trigger GDF-15 secretion, probably as part of a compensatory anti-inflammatory response [30].

Pathophysiological association

The only well-established receptor for GDF-15 is the GFRAL, which forms a complex with the co-receptor RET [28, 31, 32]. GFRAL expression is largely restricted to neurons in the brainstem AP and NTS, key hubs for appetite and nausea regulation [33].

When GDF-15 binds, GFRAL recruits RET and sets off downstream signals including ERK, AKT, and β-adrenergic pathways. The net effect is reduced food intake and increased energy expenditure [34, 35]. This system works independently of classical appetite regulators like leptin and ghrelin, though cross-regulation may occur under certain metabolic conditions [36, 37].

Dysregulation of this pathway is implicated in multiple disease states. In cancer cachexia, tumor-derived GDF-15 overactivates GFRAL-RET signaling, driving anorexia and wasting [38, 39]. Similarly, chemotherapy-induced anorexia involves hepatic GDF-15 upregulation via the ISR [38, 40], while the anorectic effect of GDF-15 has also sparked interest in targeting this axis for obesity therapy [41–43]. The detailed mechanisms underlying these pathological conditions, particularly the central and peripheral pathways driving cachexia, are discussed in Sect.  2.3.

Tissue-protective functions of GDF-15

Beyond metabolic regulation, GDF-15 exerts tissue-protective effects in cardiovascular and neurodegenerative contexts. In myocardial infarction models, GDF-15 reduces apoptosis by inhibiting ROS production and mitochondrial fission via AMPK activation, and enhances angiogenesis through paracrine VEGF signaling [44–46]. In neurodegenerative settings, GDF-15 upregulation is linked to mitochondrial stress responses and may confer neuroprotection, although mechanistic understanding remains incomplete.

GDF-15 in disease: beyond oncology

Beyond its roles in cancer, GDF-15 is implicated in various non-malignant diseases, where it primarily serves as a biomarker of cellular stress rather than a direct pathological mediator. In cardiovascular disease, elevated GDF-15 correlates with atherosclerosis severity and predicts adverse cardiovascular outcomes, though its direct pathogenic role remains debated; current evidence suggests it functions primarily as an integrative marker of inflammation and oxidative stress [47–51]. Circulating GDF-15 also increases with age and is independently associated with all-cause mortality, reflecting its role as a marker of mitochondrial dysfunction and cellular senescence [52, 53]. As a canonical mitokine, GDF-15 is robustly induced by mitochondrial damage and serves as a clinically useful biomarker for mitochondrial diseases, alongside Fibroblast Growth Factor 21 (FGF-21) [54–56]. Recent studies have further reported elevated GDF-15 levels in patients with chronic migraine, though confounding factors require careful adjustment [57].

Taken together, these observations identify GDF-15 as a universal stress-responsive cytokine across diverse pathological contexts. However, its tissue-specific functions and mechanistic roles in non-oncological diseases remain incompletely understood and further investigation.

Role of GDF-15 in the TME

Within the TME, GDF-15 takes on functions distinct from its metabolic and tissue-protective roles, acting as a key driver of tumor progression across multiple cancer types. In gastric cancer, GDF-15 promotes cisplatin resistance through activation of the ISR pathway and induction of mitochondrial dysfunction. High GDF-15 expression in tumor tissues correlates with increased risk of disease recurrence in patients receiving adjuvant chemotherapy, and elevated serum levels are associated with poor prognosis [58–60]. In multiple myeloma (MM), GDF-15 drives tumor proliferation via the TGF-β-glycolysis axis, enhancing expression of key glycolytic genes while blocking apoptosis. Overexpression GDF-15 significantly accelerates tumor growth in xenograft models, whereas knockdown suppresses it, indicating a direct oncogenic role [59, 61]. In thyroid cancer, GDF-15 functions as a myokine induced by mitochondrial stress, promoting tumor cell invasiveness through STAT3 signaling activation. Clinical data show that circulating GDF-15 levels are significantly elevated in thyroid cancer patients and correlate with tumor malignancy grade [62].

These tumor-promoting effects, coupled with GDF-15’s immunosuppressive functions (detailed in Sect.  2.5), including inhibition of T cell infiltration, induction of regulatory T cells, and modulation of myeloid cells, establishing GDF-15 as a central node in cancer progression. The following sections dissect the key signaling pathways through which GDF-15 exerts these effects within the TME.

Signaling pathways and functional outcomes

GDF-15 in tumor proliferation and survival

GDF-15 promotes tumor cell proliferation and survival through convergent activation of TGF-β, STAT3, and AKT signaling, with the dominant pathway varying by tumor type. In multiple myeloma, GDF-15 activates the TGF-β-glycolysis axis, leading to upregulation of glycolytic genes and inhibition of apoptosis. Pharmacological inhibition of TGF-β signaling blocks this pro-proliferative effect, and in vivo xenograft models confirm that GDF-15 overexpression accelerates tumor growth whereas knockdown suppresses it [61]. In glioma stem-like cells (GSCLCs), GDF-15 upregulates LIF expression via ERK1/2-c-Fos signaling, subsequently activating STAT3 to maintain stem cell phenotype and enhance self-renewal [63]. The GDF-15-LIF-STAT3 axis also promotes proliferation in thyroid cancer, where STAT3 activation regulates invasiveness [29]. In pancreatic cancer, GDF-15 activates AKT signaling in a strictly context-dependent manner-pro-tumorigenesis is observed in the aged microenvironment, with no significant effect in young hosts [22]. This senescence-dependent AKT activation sensitizes tumors to AKT inhibitors, suggesting therapeutic vulnerability in specific patient subsets [22]. In silico docking experiments suggested that GDF-15 may directly bind to TGF-βR2, though this interaction requires further in vivo validation [63, 64]. In inflammatory microenvironments, GDF-15 regulates PI3K/AKT to modulate pro-cancer factor expression, while absence of GDF-15 inhibits PI3K/AKT activation, and PI3K inhibition enhances GDF-15’s anti-inflammatory effects [65].

The convergence of the aforementioned pathways on common downstream effectors-glycolytic enzymes, anti-apoptotic B-cell lymphoma 2 (Bcl-2) family members, and cell cycle regulators, amplifying GDF-15’s net pro-tumorigenic effect. However, pathway dominance is tumor-type specific, with TGF-β signaling predominates in multiple myeloma, STAT3 in glioma and thyroid cancer, and AKT in age-dependent pancreatic cancer. This specificity has important therapeutic implications, suggesting that pathway inhibitors should be matched to tumor context rather than applied uniformly.

GDF-15 in immune evasion

GDF-15 orchestrates a multifaceted immunosuppressive TME through parallel and intersecting mechanisms affecting T cells, regulatory T cells, and myeloid populations. A primary mechanism involves impaired T cell trafficking. GDF-15 disrupts LFA-1/β2 integrin-mediated adhesion of T cells to activated endothelium, blocking their extravasation into tumors [18]. This effect has been shown to influence PD-1 ICB efficacy in preclinical models. Clinically, high serum GDF-15 levels correlate significantly with treatment failure in melanoma patients, while neutralizing antibodies restore T cell infiltration and improve therapeutic outcomes in mouse models.

Concurrently, GDF-15 promotes the expansion of immunosuppressive regulatory T cells. In hepatocellular carcinoma, it drives regulatory T cell (Treg) generation and functional enhancement via CD48 receptor signaling, and blocking GDF-15 markedly enhances anti-tumor immunity, identifying this axis as a distinct immunosuppressive mechanism [66].

Beyond lymphoid cells, GDF-15 also molds the myeloid compartment through two related mechanisms. It pushes macrophages toward an M2-like phenotype, inducing CD14 + monocytes to adopt this immunosuppressive state while restraining M1 polarization [67]. At the same time, GDF-15 helps recruit myeloid-derived suppressor cells (MDSCs) into tumors [68, 69]. Although these myeloid effects are supported by emerging data, they need confirmation across more tumor types. Metabolic reprogramming adds another layer. Through TGF-β signaling, GDF-15 drives glycolysis in tumor cells, creating a metabolically competitive environment that leaves T cells short of nutrients [63, 70]. This metabolic-immune crosstalk links tumor cell metabolism directly to immune suppression—an indirect but important piece of how GDF-15 evades immunity.

Notably, the immunomodulatory functions of GDF-15 are not uniform across all contexts. Emerging evidence points to context-dependent effects on myeloid cells and inflammatory gene expression [71]. Despite the prevailing view of GDF-15 as an immunosuppressive factor, emerging evidence suggests that its immunomodulatory effects may be more nuanced. While multiple studies demonstrate that GDF-15 impairs T cell trafficking and promotes Treg expansion [66], a recent report in adrenocortical carcinoma found that high GDF-15 levels were associated with decreased expression of pro-inflammatory immune genes, but the functional consequences for anti-tumor immunity remain unclear. Patients with low GDF-15 levels show better immunotherapy response and longer progression-free survival [6]. Furthermore, the degree to which GDF-15-mediated immunosuppression is tumor cell-autonomous versus microenvironment-driven is poorly understood. Most existing studies rely on GDF-15 overexpression or neutralization in established tumors, which may not capture its role during early immune editing [72]. Future investigations should employ temporal and cell-type-specific manipulation of GDF-15 to dissect its context-dependent immune functions. It is worth noting that the role of GDF-15 is tumor type and microenvironment-specific, which provides a theoretical basis for the development of precise immunotherapy strategies.

GDF-15 in cancer cachexia and metabolism

Systemically, GDF-15 drives cancer cachexia primarily through the brainstem GFRAL-RET receptor axis, as introduced in Sect.  1.3. Tumors secrete excessive GDF-15 (typically 10–100 fold above physiological levels), chronically activating GFRAL-RET signaling in the AP and NTS. This sustained activation suppresses appetite and increases energy expenditure, ultimately resulting in weight loss and muscle atrophy [28, 38]. Notably, this anorectic effect operates independently of leptin and ghrelin [36, 37] and can be reversed by neutralizing antibodies such as ponsegromab, which is under clinical evaluation [39].

Beyond central appetite regulation, GDF-15 exerts direct peripheral effects on muscle tissue. In C2C12 myotubes and C26 tumor-bearing mice, GDF-15 downregulates anti-apoptotic Bcl-2 and activates caspase-3, promoting muscle fiber degradation independently of GFRAL [73]. This mechanism has been verified in tumor tissue, serum exosomes, and gastrocnemius muscle, confirming the relevance in vivo. The coexistence of central GFRAL-RET and peripheral Bcl-2/caspase-3 pathways highlights a dual mechanism by which GDF-15 contributes to cachexia.

In addition to its role in wasting, GDF-15 influences tumor cell metabolism. Through TGF-β and AKT signaling, GDF-15 enhances aerobic glycolysis (the “Warburg effect”) while suppressing oxidative phosphorylation, thereby supporting tumor cell proliferation and chemotherapy resistance [61]. In gastric cancer, GDF-15 upregulation drives cisplatin resistance through the ATF4-CHOP pathway and oxidative stress, with Bcl-2/caspase-3 dysregulation serving as a downstream mechanism [59].

The convergence of central GFRAL-RET and peripheral Bcl-2/caspase-3, metabolic mechanisms position GDF-15 as a key mediator of the cancer cachexia syndrome. Notably, the same molecule that suppresses appetite centrally also promotes tumor growth peripherally, a duality that complicates therapeutic targeting but also offers opportunities for integrated intervention.

Pathway cross-talk and tumor-type specificity

Several nodes of pathway cross-talk have been identified. First, a STAT3-AKT interaction has been proposed, wherein STAT3 and AKT may form a positive feedback loop, with STAT3 activation enhancing PI3K/AKT signaling and vice versa. However, direct biochemical evidence of this interaction in cancer contexts remains limited, with current support coming primarily from thyroid cancer studies relying on in vitro assays [29]. Second, a TGF-β-STAT3 axis operates in glioma stem-like cells, where GDF-15 upregulates LIF via ERK1/2-c-Fos, subsequently activating STAT3 [63]. Third, TGF-β-AKT metabolic integration occurs in multiple myeloma, where TGF-β signaling drives glycolysis while AKT supports cell survival, creating coordinated metabolic and anti-apoptotic effects [61]. Additionally, emerging evidence points to GFRAL-RET peripheral signaling, suggesting that although GFRAL is primarily associated with central metabolic regulation, it may activate survival signals such as AKT through non-classical pathways in tumors [11].

Despite these collective findings implicating AKT signaling in GDF-15-mediated tumor progression, the strength of evidence varies considerably across studies. Specifically, the pancreatic cancer study has utilized both GDF-15 knockout mice and pharmacological AKT inhibition, providing relatively robust genetic and pharmacologic validation [22]. In contrast, evidence from thyroid cancer relies primarily on in vitro STAT3 activation assays with limited in vivo confirmation. Additionally, the proposed cross-talk between STAT3 and PI3K/AKT in thyroid cancer remains hypothetical, as direct biochemical evidence of pathway interaction is lacking [29]. It should be noted that the described limitations will affect the translational impact of targeted targeting GDF-15/AKT signaling as well.

Adding another layer of complexity, tumor-type specificity further modulates GDF-15’s effects. The pro-tumorigenic role of GDF-15 via TGF-β signaling is not universally observed. For instance, while studies in multiple myeloma and glioma demonstrate clear tumor-promoting effects [61], evidence from pancreatic cancer models suggests that GDF-15’s oncogenic activity is strictly dependent on the aged microenvironment, with no significant effect observed in young hosts. This discrepancy highlights the critical influence of host factors—particularly senescence-associated stromal changes on GDF-15 function. Methodologically, these contrasting findings may also reflect differences in experimental systems: xenograft models versus genetically engineered mouse models, or variations in GDF-15 source (tumor-derived vs. host-derived). Future studies should carefully control for such variables to delineate the true context-dependent nature of GDF-15/TGF-β signaling [74, 75].

Pathway integration

In summary, GDF-15 signaling in solid tumors is characterized by functional convergence, multiple pathways (TGF-β, STAT3, AKT) converge on common outcomes including proliferation, immune evasion, and cachexia—and extensive pathway cross-talk that creates signaling networks rather than linear cascades (Fig. 1). Critically, context dependence determines which pathway dominates: tumor type, microenvironment (particularly aging), and disease stage collectively shape the net biological effect. These features have important translational implications: effective therapeutic targeting requires matching pathway inhibitors to specific tumor contexts and considering both local and systemic GDF-15 functions. This inherent complexity positions GDF-15 as both a challenging target and a promising opportunity for context-specific intervention. Building on these mechanistic insights, the following section examines the clinical implications of GDF-15 as a prognostic biomarker and therapeutic target, highlighting its correlation with disease progression, treatment response, and patient outcomes across multiple tumor types.

Fig. 1.

GDF-15 signaling in solid tumors. a Local tumor microenvironment effects. GDF-15 secreted by tumor cells activates multiple signaling pathways: TGF-β signaling promotes glycolysis, LIF-JAK-STAT3 signaling enhances stemness, PI3K-AKT signaling supports survival, and the Bcl-2/caspase-3 pathway regulates apoptosis. Emerging evidence suggests that GDF-15 may also modulate ferroptosis, though this mechanism requires further investigation. In addition, GDF-15 shapes the immune landscape within the tumor microenvironment. It disrupts T cell adhesion by interfering with LFA-1/β2 integrin binding, thereby impairing T cell infiltration. It also promotes regulatory T cell induction through CD48 signaling, drives M2 macrophage polarization, and facilitates the recruitment of myeloid-derived suppressor cells. Solid arrows denote well-established mechanisms, whereas dotted arrows indicate pathways that await further confirmation. b Systemic effects – cancer cachexia. GDF-15 released from tumors enters the circulation and binds to GFRAL-RET receptor complexes on neurons in the brainstem area postrema and nucleus tractus solitarius. This interaction suppresses appetite and increases energy expenditure, ultimately leading to weight loss and muscle atrophy. Direct peripheral effects also contribute to muscle wasting, mediated through downregulation of Bcl-2 and activation of caspase-3, which promotes myofiber degradation. c Pathway cross-talk and context dependence. Several points of interaction among signaling pathways have been identified. TGF-β and AKT signaling are metabolically integrated; TGF-β activates STAT3 via LIF; and a potential feedback loop between STAT3 and AKT has been proposed but requires experimental validation. The dominant signaling pathway varies by tumor type: TGF-β predominates in multiple myeloma, STAT3 in glioma, AKT in pancreatic cancer where senescence-associated stromal changes are required, and T cell exclusion characterizes melanoma. The net biological outcome of GDF-15 signaling is therefore shaped by tumor type, microenvironmental context, and disease stage

Clinical significance and therapeutic targets

Prognostic value

GDF-15 correlates significantly with disease progression and poor prognosis across multiple solid tumors. In colorectal cancer (CRC), preoperative serum GDF-15 levels above 1,200 pg/mL independently associate with poor disease-free survival (DFS) and overall survival (OS) in a retrospective cohort of 214 patients (HR = 2.31 and 2.89, respectively) [76]. In adrenocortical carcinoma (ACC), patients with serum GDF-15 below the median (2,450 pg/mL) have significantly longer progression-free survival (PFS) (HR = 2.64) [6]. In non-small cell lung cancer (NSCLC), high GDF-15 levels associate with poor treatment response and shorter OS (HR = 2.43) and PFS (HR = 2.18) [77]. In low-grade gliomas (LGG), high GDF-15 expression independently predicts poor OS (HR = 2.17) [27].

Immunotherapy response

GDF-15 impairs anti-tumor immune responses by disrupting LFA-1/β2 integrin-mediated adhesion of T cells, preventing their infiltration into the TME. In melanoma patients, high serum levels of GDF-15 were significantly associated with failure to respond to therapy with PD-1 inhibitors [15]. Moreover, a similar picture was found in ACC patients where higher GDF-15 levels were associated to poorer response to ICIs and significantly with shortened pfa [6]. Recently, a large prospective cohort study from Vienna confirmed these findings in 807 patients with different solid tumors. In the cohort receiving ICIs, non- responders showed significantly higher baseline levels of GDF-15 (3,387 vs. 1,956 pg/mL) whereas above-median GDF-15 was associated with a 2-fold higher risk of dying (HR = 2.1) [Viena cohort 2025].

Cachexia and metabolic disorders

GDF-15 is an important driver of cancer-related cachexia causing body weight loss and muscle atrophy by regulating metabolic pathways. Its underlying mechanisms involved both mitochondrial stress and inflammation: GDF-15 levels strongly correlate with host factors such a low body mass index, muscle loss and inflammatory markers [73]. Preclinical studies have also shown that neutralizing antibodies against GDF-15 can improve chemotherapy-induced anorexia, vomiting and body weight loss in animal models [78]. Additionally, GDF-15 participates to the metabolic regulation under both acute or chronic energy deprivation conditions through inhibition of leptin signals [78].

GDF-15 as a biomarker

Beyond oncology, GDF-15 serves as a non-specific biomarker for disease severity and treatment response. In idiopathic pulmonary fibrosis, increased serum GDF-15 associates with acute exacerbation and reduced survival [79]. Antibodies targeting GDF-15 (e.g., GDFATHER-1/2a trial) are being explored for synergistic effects with immunotherapy [24].

Despite well-documented prognostic value across multiple tumor types, several factors limit the immediate clinical utility of circulating GDF-15. First, optimal cutoff values vary widely between studies (e.g., ranging from 1,200 to 2,500 pg/mL across cohorts) [77], reflecting heterogeneity in assay platforms, patient populations, and sample processing methods. Second, GDF-15 levels are influenced by non-cancer factors including age, renal function, and inflammatory status [48, 53], complicating individual patient interpretation. Third, most prognostic studies to date have been retrospective with limited independent cohort validation. Prospective studies with standardized assays and pre-specified cutoffs are urgently needed before GDF-15 can be incorporated into routine clinical risk stratification.

Therapeutic strategies targeting GDF-15

GDF-15 promotes pancreatic cancer growth specifically in the aged microenvironment via AKT activation, with knockout studies demonstrating slowed tumor growth in aged but not young mice [22]. In multiple myeloma, GDF-15 drives tumor proliferation through the TGF-β-glycolysis axis, with overexpression significantly accelerating tumor growth in xenograft models [61].

Regarding immunomodulation, preclinical studies have shown that blocking GDF-15 may enhance PD-1 inhibitor efficacy. The ongoing GDFATHER-1/2a trial (NCT04725474) is evaluating a GDF-15 antibody combined with ICIs in advanced cancer patients, with preliminary results suggesting potential clinical activity [74].

Current intervention strategies targeting GDF-15 include neutralizing antibodies (e.g., ponsegromab for cachexia), receptor targeting via GFRAL to regulate metabolic pathways, and combination therapy with chemotherapy or immunotherapy to overcome drug resistance [80]. Additional agents under investigation include AZD8853, evaluated in a Phase 1/2a trial in advanced MSS-CRC and urothelial carcinoma (NCT05397171), though no objective responses were observed [NCT05397171]. The ongoing GDFATHER-NSCLC-01 (NCT07098988) and GDFATHER-NSCLC-02 (NCT07246863) trials are evaluating visugromab combined with chemoimmunotherapy in metastatic NSCLC. To provide a comprehensive overview of clinical studies directly targeting GDF-15 or evaluating its biomarker potential in cancer patients, we summarize relevant interventional trials and biomarker studies in Table 1. This table complements the mechanistic discussions in Chaps. 2 and 3 by offering a concise reference for the clinical evidence discussed throughout this review.

Table 1.

Summary of GDF-15-targeted interventions and biomarker studies in cancer patients

Study/Identifier	Phase	Population	Intervention	Objectives	Status	Source/Registry
GDFATHER-1/2a (NCT04725474)	Phase 1/2a	Anti-PD-1/PD-L1 refractory advanced solid tumors (nsq NSCLC, UC, HCC)	Visugromab (anti-GDF-15) + nivolumab	ORR: 18.2% (NSCLC), 18.5% (UC), 14.3% (HCC); 61.5% of responders achieved CR/CMR; median DoR > 28.4 mo (NSCLC/UC)	Completed	[81]
GDFATHER-NSCLC-01 (NCT07098988)	Phase 2b	First-line metastatic nsq NSCLC	Visugromab + chemoimmunotherapy vs. placebo + chemoimmunotherapy	Primary: ORR; Secondary: DoR, PFS, OS, body weight trends	Recruiting	[81]
GDFATHER-NSCLC-02 (NCT07246863)	Phase 2b	Second-line metastatic nsq NSCLC (progressed on prior ICI)	Visugromab + anti-PD-1 + chemotherapy (multiple arms)	Primary: ORR; Secondary: DoR, PFS, OS, body weight trends	Recruiting	82
PROACC-1 (NCT05546449)	Phase 2	Cancer cachexia with elevated GDF-15 (≥ 1,500 pg/mL)	Ponsegromab (anti-GDF-15)	Dose-responsive weight gain (placebo-adjusted: 1.33–3.00 kg at 12 weeks); improved appetite and activity	Completed	[83]
NCT05397171	Phase 1/2a	Advanced MSS-CRC, urothelial carcinoma	AZD8853 (anti-GDF-15) monotherapy	Well tolerated; no objective responses; transient GDF-15 suppression	Terminated	84
NCT07112196	Phase 2b	Cancer cachexia	Visugromab	Evaluating cachexia outcomes	Recruiting	85
Yamashita et al. 2025	Retrospective	214 CRC patients	Serum/tissue GDF-15 analysis	Preoperative serum GDF-15 independently associated with DFS (HR = 2.31, 95% CI: 1.45–3.68) and OS (HR = 2.89, 95% CI: 1.72–4.86); cutoff 1,200 pg/mL (sensitivity 78%, specificity 71%)	Completed	[86]
Vienna cohort 2025	Prospective	807 solid tumor patients	Serum GDF-15 analysis	GDF-15 correlated with age, inflammation, metastatic disease; in ICI-treated patients (n = 187), non-responders had higher baseline GDF-15 (3,387 vs. 1,956 pg/mL); above-median GDF-15 associated with 2-fold increased risk of death (HR = 2.1, 95% CI: 1.7–2.5)	Completed	[87]

Inclusion criteria: Studies listed in this table meet one or both of the following criteria: (1) interventional trials directly targeting GDF-15 or its receptor GFRAL in cancer patients; (2) biomarker studies evaluating circulating or tissue GDF-15 levels in relation to cancer prognosis, treatment response, or clinicopathological features

Abbreviations: ORR Objective response rate, CR Complete response, CMR Complete metabolic response, DoR Duration of response, PFS Progression-free survival, OS Overall survival, ICI Immune checkpoint inhibitor, nsq NSCLC non-squamous non-small cell lung cancer, UC Urothelial cancer, HCC Hepatocellular carcinoma, CRC Colorectal cancer; MSS Microsatellite stable, HR Hazard ratio, CI Confidence interval

Controversies and challenges

Context-dependent duality of GDF-15

A recurring observation in the GDF-15 literature is its seemingly contradictory behavior, promoting tumors in some settings while protecting tissues in others. Several factors help explain these discrepancies.

Tumor type represents one axis of functional variability. GDF-15 consistently drives progression in multiple myeloma and glioma [63], yet its role in pancreatic cancer is conditional: it promotes tumor growth only within an aged microenvironment, with no detectable effect in young hosts [22]. This pattern suggests that GDF-15’s oncogenic activity depends on specific stromal cues rather than being cell-autonomous.

Superimposed on these tissue-specific effects is the influence of disease stage. Most studies have focused on established tumors, leaving early disease largely unexplored. Given that GDF-15 responds to cellular stress, it may initially function as a tumor-suppressive sentinel—eliminating damaged cells—before being co-opted by malignant cells to fuel progression. Longitudinal studies with temporal control of GDF-15 will be needed to test this hypothesis.

Experimental design introduces further complexity. The field relies heavily on xenograft models using human tumor cell lines, which may not fully recapitulate the complex tumor-stroma interactions present in autochthonous tumors. Some discrepancies between studies likely reflect differences in model systems as much as genuine biological variation.

Cross-study comparisons are further complicated by measurement issues stemming from the H202D polymorphism. The H202D polymorphism in GDF-15 affects detection by some antibody-based assays, potentially confounding cross-study comparisons [88]. Studies that do not account for this genetic variant may yield inconsistent results, particularly in ethnically diverse populations. Detailed assay recommendations and correction strategies are discussed in Sect.  4.2. Recognizing these sources of variability is essential for designing interpretable experiments and for developing therapeutic strategies that selectively target GDF-15’s pathogenic functions while sparing its protective roles.

H202D variant and measurement considerations

The common H202D polymorphism significantly affects GDF-15 detection by certain immunoassays. Population frequency varies by ancestry, ranging from approximately 5% in East Asian populations to 25–30% in European populations, with approximately 5–10% of individuals being homozygous D allele [89]. The Roche Elecsys^® platform detects H- and D-containing GDF-15 peptides similarly and is recommended for total GDF-15 measurement regardless of genotype [88]. In contrast, assays based on R&D Systems reagents such as Quantikine^® and DuoSet^® ELISAs consistently underestimate GDF-15 in the presence of the D variant, with recovery approximately 45% for DD homodimers and 62–78% for HD heterodimers compared to HH homodimers [88]. For studies using such assays with known genotype, correction equations can be applied: HH ×1.04, HD ×1.56, DD ×2.56.

Beyond genetic variation, several factors consistently influence circulating GDF-15 levels and should be controlled in clinical studies [48]. These include age approximately 15% increase per decade in healthy adults, renal function (GDF-15 levels correlate inversely with estimated glomerular filtration rate), inflammatory status (GDF-15 correlates with C-reactive protein), body mass index, smoking status, and certain medications (e.g., metformin, cisplatin).

Peripheral receptor controversies and functional complexity

Despite GFRAL currently identified as the primary receptor for GDF − 15 in central nervous system, receptor systems in peripheral tissues remain unknown. In fact, evidence indicates that GDF-15 may also operate independent of GFRAL in peripheral tissues, necessitating investigation of these non-canonical pathways [90]. The same molecule can both suppress appetite through central GFRAL-RET signaling and promote tumor cell proliferation by enhancing Warburg-like metabolism, a duality that remains unexplained [36, 61]. Baseline GDF-15 positively correlates with triglyceride-rich particles and lipoproteins, but during acute metabolic stress, the association pattern in H202D variant carriers shows differences, supporting GDF-15’s function as a metabolic stress biomarker rather than a direct weight-loss mediator [91].

GDF-15 exerts effects through multiple downstream pathways beyond those discussed in Chap. 2, including AMPK (maintaining AMPK activity, improving insulin resistance and hepatic steatosis), TGF-β1/SMAD3 (reducing gluconeogenesis and fibrosis), PPARγ (ketogenic diet increases transcription), and non-canonical pathways in skeletal muscle and heart [92–94]. Additionally, tissue-specific differences have been reported, for example, knockout of macrophage GDF-15 in adipose tissue exacerbates obesity, while during the development of MASH hepatic GDF-15 expression increased via TFEB and DNA Damage Inducible Transcript 3 (DDIT3) signaling [95, 96]. During infection, GDF-15 coordinates metabolic changes of the host to support the immune responses, while a time-dependent-switch of GDF- 15 may have detrimental long-term action. In fact, this time-dependent functional switch remains unexplained.

In summary, GDF-15 complexity and controversy stem from its pleiotropy, microenvironment dependence, and pathway cross-regulation. Future research should focus on clarifying specific action conditions (tumor stage, receptor expression profile, host factors) to develop precise targeting strategies that selectively block pathogenic functions while preserving protective roles. To offer a systematic view of the context-dependency of GDF-15 roles in different tumor types and TME, here we summarize the bidirectional effects and underlying mechanisms in Table 2, providing a straightforward reference to all situations in which the results associated with GDF-15 are different. Table 2 complements our mechanistic summaries in Chaps. 2 and 3 and Fig. 1 to avoid the need to search through those figures to identify what we believe to be the most relevant situations.

Table 2.

Context-dependent dual roles of GDF-15 in solid tumors

Tumor type / Context	GDF-15 effect	Proposed mechanisms	Key references
Multiple myeloma	Pro-tumorigenic	TGF-β-glycolysis axis activation; enhanced glycolysis, inhibited apoptosis	[97]
Glioma (including GSCs)	Pro-tumorigenic	ERK1/2-c-Fos-LIF-STAT3 signaling; maintenance of stem-like phenotype	[63]
Thyroid cancer	Pro-tumorigenic	STAT3 activation; mitokine-induced invasiveness	[29]
Pancreatic cancer (aged microenvironment)	Pro-tumorigenic	AKT activation; senescence-associated stromal changes required	[22]
Pancreatic cancer (young microenvironment)	No significant effect	Lack of age-related stromal factors	[22]
Colorectal cancer (serum GDF-15)	Prognostic marker (poor prognosis)	Correlated with advanced stage, inflammation, cachexia; unclear direct mechanism	[86]
Colorectal cancer (tissue GDF-15)	No significant correlation	Not associated with clinicopathological features	[86]
Adrenocortical carcinoma (high GDF-15)	Poor immunotherapy response	Decreased pro-inflammatory immune gene expression; unclear functional consequences	[6]
Adrenocortical carcinoma (low GDF-15)	Better immunotherapy response	Preserved anti-tumor immunity	[6]
Melanoma	Immunosuppressive	LFA-1/β2 integrin disruption; T cell exclusion	[18]
Hepatocellular carcinoma	Immunosuppressive	CD48-mediated Treg induction	[66]
Gastric cancer	Chemoresistance	ISR activation, mitochondrial dysfunction; ferroptosis resistance	[97]
Myocardial infarction / Neurodegeneration	Protective	ROS inhibition, mitochondrial fission regulation via AMPK	[12]
General context (early disease stage)	Hypothesized tumor-suppressive	Stress-induced elimination of damaged cells; requires validation	-

This table summarizes evidence discussed in the main text. Mechanisms in italics require further validation

Summary and outlook

In conclusion, GDF-15 is increasingly recognized as a core stress-responsive cytokine and a key mitochondrial factor involved in the pathophysiology of solid malignancies. Functioning at the intersection of mitochondrial stress, inflammatory signaling, and systemic metabolic regulation, GDF-15 exerts complex and context-dependent effects on tumor progression. It promotes cancer cell proliferation, invasion, metabolic reprogramming, and immune evasion through the activation of multiple signaling cascades, including TGF-β, LIF-STAT3, and AKT pathways. Clinically, GDF-15 serves as a poor prognosis biomarker and are closely related to cancer cachexia and immune escape. These findings have established GDF-15 as a promising therapeutic target. Currently, strategies targeting GDF-15 to alleviate cachexia or combining it with ICIs to overcome treatment resistance have shown encouraging prospects in preclinical and early clinical studies.

However, there are still some critical scientific issues and challenges encountered in the evolution of GDF-15 hindering the further clinical translations. The contradiction between the pro-cancer and anti-cancer functions of GDF-15 in different tumor types, microenvironments such as aging, and disease stages highlights the significant context dependence of its action. The unknown peripheral receptor system, the selection mechanism of downstream signaling pathways, and the influence of gene polymorphism in humans are still scientific puzzles to be clarified.

Therefore, future research should focus on in-depth analysis of the molecular switch for the functional transformation of GDF-15, clarification of its dominant mechanism of action in specific TME, and exploration of targeting strategies that can accurately distinguish its pathological and physiological functions. Only in this way can the therapeutic potential of targeting GDF-15 be maximally explored, and it can be safely and effectively transformed from a complex biomarker into a precise therapeutic tool for combating cancer, ultimately achieving its successful translation into clinical applications.

Authors’ contributions

DA: Writing – original draft, Data curation, Supervision, Writing – review & editing; YA: data analysis and interpretation, manuscript preparation and editing; RS: Funding acquisition, Supervision; KA: Project administration, Supervision ; KM: study concept and design, quality control of data, manuscript review.

Funding

People’s Hospital of Kizilsu Kirgiz Autonomous Prefecture work station of State Key Laboratory of Pathogenesis, Prevention, Treatment of Central Asian High Incidence Diseases Fund (No: SKL-HIDCA-2024-KZ3).

Data availability

All the data obtained and analyzed during the current study were available from the corresponding authors on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors agree and give consent for the publication of this manuscript.

Competing interests

The authors declare no competing interests.