Imaging Cancer-associated Cachexia: Utilizing Clinical Imaging Modalities for Early Diagnosis

Abstract

Cancer-associated cachexia (CAC) is a prevalent condition that accelerates cancer progression and heightens treatment-related adverse effects in patients by affecting multiple organ systems. Despite the profound impact of CAC on clinical management and treatment outcomes of patients with cancer, the current understanding of mechanisms associated with the condition, as well as the tools necessary for early diagnosis, are limited. Currently, the clinical diagnosis of CAC relies on weight change–based assessments, which have limited sensitivity and cannot identify patients at risk for CAC. In this context, noninvasive imaging-based biomarkers, such as the composition and properties of adipose and muscle tissues, may allow for diagnosis of CAC before substantial weight loss occurs. Such early detection can potentially enable more timely and effective interventions. Furthermore, imaging allows for quantitative assessment of CAC, enabling monitoring of prognosis and treatment response. This article reviews current applications and future developments of imaging techniques, particularly those employed in current clinical radiology, that can reveal diagnostic information and facilitate early detection of CAC and quantitative evaluation of associated metabolic alterations.

Keywords: Molecular Imaging, Cancer, MRI, PET/CT, Ultrasound, Muscular, Oncology

© RSNA, 2025

Summary

Early detection and comprehensive assessment of cancer-associated cachexia using diverse imaging modalities, such as dual-energy x-ray absorptiometry, US, CT, MRI, and PET/CT, combined with artificial intelligence–based analysis can improve patient outcomes by enabling timely interventions.

Essentials

• ■ Diverse imaging modalities—dual-energy x-ray absorptiometry, CT, US, MRI, and PET/CT—offer unique advantages for assessing muscle and adipose tissue changes in cancer-associated cachexia.

• ■ MRI and PET are expected to play increasingly important roles in early detection of cachexia given their capabilities to help detect and quantify metabolic changes that precede measurable weight loss during the development and progression of the condition.

• ■ Ongoing research should focus on improving imaging techniques that target imaging biomarkers of cancer-associated cachexia and using artificial intelligence to improve diagnostic accuracy and patient outcomes.

Introduction

Cancer-associated cachexia (CAC) is a multifactorial syndrome characterized by sustained muscle loss, with or without concurrent fat loss, that cannot be fully reversed by nutrition alone (1–3). It accelerates disease progression, exacerbates treatment-related adverse effects, and leads to poor clinical outcomes and reduced survival rates in patients with cancer (4). Approximately 20%–70% of patients with cancer are affected by CAC, with 10%–20% of cancer-related deaths directly attributed to CAC (5). The prevalence of CAC varies across different cancer types, ranging from 12% to 67% (2); incidences are relatively lower in prostate cancer (14%) and breast cancer (21%) (6) but higher in pancreatic cancer (67%), gastroesophageal cancer (60%), and colorectal cancer (48%) (6,7). CAC can develop as early as the initial stages of cancer, but its exact mechanisms remain unclear (2).

Existing evidence indicates that CAC is driven by tumor-derived substances, such as inflammatory cytokines, hormones, and growth factors, which disrupt normal metabolism and adversely affect multiple organ systems (2). These substances trigger a series of biologic responses, including inflammation, appetite suppression, and alterations in muscle and fat metabolism, involving signaling pathways such as JAK/STAT (Janus kinase/signal transducer and activator of transcription), SMAD (small mothers against decapentaplegic), MAPK (mitogen-activated protein kinases), PI3K/AKT/mTOR (phosphoinositide 3-kinases/protein kinase B/mammalian target of rapamycin), and NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) (2,8). Consequently, both muscle and fat mass decrease substantially in patients with CAC (9,10). Beyond its detrimental impact on physical function and quality of life, CAC also compromises the efficacy of anticancer therapies, while exacerbating their toxic adverse effects (8). Therefore, CAC is now recognized as a critical contributor to cancer progression and mortality.

Clinically, CAC can be staged into three phases: (a) precachexia, characterized by weight loss of 5% or less, along with anorexia and metabolic changes; (b) cachexia, defined by weight loss greater than 5% within 6 months, weight loss greater than 2% in individuals with body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) less than 20 kg/m², or weight loss greater than 2% with sarcopenia; and (c) refractory cachexia, marked by continuous tumor progression, no response to antitumor therapy, active catabolism, and irreversible continuous weight loss (1). In the cachexia phase, sarcopenia can be determined based on the appendicular skeletal muscle index (SMI) measured by dual-energy x-ray absorptiometry (DXA) (cutoff: <7.00 kg/m² for men; <5.50 kg/m² for women) (11), or by the lumbar SMI measured using CT imaging (cutoff: <52.40 cm²/m² for men; <38.50 cm²/m² for women) (12). It is important to note that there are numerous cutoff values for diagnosing sarcopenia, which vary depending on the measurement modality, sex, and consensus (13,14). In the refractory cachexia phase, patients typically exhibit low performance status (scores 3 or 4) based on the World Health Organization criteria, with a prognosis of less than 3 months of survival (2). Currently, there are no approved drugs to treat CAC (8). Adequate nutritional support remains the primary intervention strategy for this condition (15), though its effectiveness hinges on the early diagnosis of CAC. Therefore, detecting CAC or predicting the risk for early intervention is crucial for improving patient survival. In current clinical management of patients with cancer, weight loss and BMI alterations are the main indicators for diagnosing CAC (1). However, these metrics are not specific to disease biology and may be confounded by other abnormal physiologic and metabolic conditions of patients, such as obesity, edema, and pleural or peritoneal effusion. To overcome these limitations and enable a more precise assessment, imaging modalities have become essential tools in diagnosing and evaluating CAC.

With the increasing use of imaging in cancer management and advances in imaging technology, various clinically available cancer imaging modalities, including DXA, US, CT, MRI, and PET, have been investigated for early detection and intervention in CAC (Fig 1). These imaging techniques play a crucial role in improving the accuracy of CAC diagnosis and guiding timely interventions. Among these, CT stands out for its ability to accurately estimate the volume of fat and skeletal muscle, calculate the SMI, and depict the presence of sarcopenia, thereby providing a fairly comprehensive assessment of whether a patient is in a cachectic state (16). The development of biomarker-based imaging methods is motivated by the potential to enable early and precise diagnosis before substantial weight loss occurs (17). For instance, previous studies have consistently demonstrated that low SMI and low skeletal muscle attenuation in patients with cancer, as determined by using imaging-based body composition analysis, are strongly correlated with systemic inflammatory markers, including C-reactive protein, albumin, modified Glasgow prognostic score, and the neutrophil to lymphocyte ratio (18). These findings are pivotal for defining and managing CAC (18). Furthermore, sarcopenia, characterized by a reduced SMI, has been shown to complement these inflammatory markers in the pathogenesis of CAC, thereby providing additional prognostic value (19). Therefore, incorporating these parameters from the imaging-based body composition analysis into routine clinical practice offers a promising approach to refine the characterization, management, and prognosis of CAC in patients with cancer. This review introduces the emerging roles of various clinical imaging modalities in the diagnosis and evaluation of CAC, providing examples that highlight their capabilities and potential. It also discusses the current limitations of imaging technologies and methods in their applications, as well as approaches to overcome these challenges. Additionally, future research opportunities aimed at advancing imaging-based approaches for the earlier and more accurate diagnosis and characterization of CAC are outlined.

Figure 1:

Diagram illustrates the use of different imaging modalities to assess cancer-associated cachexia. AI = artificial intelligence, CRP = C-reactive protein, DXA = dual-energy x-ray absorptiometry.

Imaging of Skeletal Muscle with DXA

DXA is a widely used imaging technique that differentiates various tissue types based on their attenuation of x-ray photons at two distinct energy levels (20). This capability allows DXA to assess body composition by producing image contrast based on differences in the densities of muscle, fat, and bone (20). While DXA is valuable for evaluating body composition and bone health, it is limited in visualizing internal adipose and muscular tissues compared with three-dimensional imaging modalities like CT and MRI.

The key advantages of DXA include its low cost, portability, and accessibility, making it especially useful in resource-limited settings. With a short examination time (typically 3 to 8 minutes), DXA provides a rapid assessment of body composition (21). DXA produces detailed images of various anatomic regions, including the head, trunk, upper and lower limbs, gynoid, and abdominal areas (22). These images categorize tissues into fat, lean soft tissue, and bone mineral content (Fig 2) (23), providing quantitative metrics such as the whole-body lean mass index and appendicular lean mass index, as recommended by the International Society for Clinical Densitometry (24). In the context of CAC, radiologists using DXA can assess both systemic and regional distributions of skeletal muscle and adipose tissue (25), enabling the evaluation of sarcopenia and cachexia, particularly in patients with muscle loss. Moreover, the low radiation exposure (<1 mrem) at DXA scans makes it especially suitable for more frail patients.

Figure 2:

Body composition components measured with DXA. Body mass (M) is the sum of fat, LST, and BMC. BMC = bone mineral content, DXA = dual-energy x-ray absorptiometry, LST = lean soft tissue. (Adapted and reprinted, under a CC BY 3.0 license, from reference 23.)

Despite its advantages, DXA has several limitations. It is less effective at discriminating specific muscle tissues, particularly in the thoracic and abdominal regions (26). Additionally, variability in patients’ height and weight can affect scan accuracy, as some patients may require multiple scans if they do not fit entirely within the scanning field (27). Furthermore, DXA does not provide information on metabolic changes associated with CAC, nor can it depict structural alterations in muscle and adipose tissues, limiting its application in the early diagnosis of cachexia.

In summary, DXA serves as a cost-effective and accessible option for identifying patients with CAC with substantial muscle and adipose tissue loss. However, it lacks utility in the early diagnosis of CAC due to its inability to depict metabolic changes and structural alterations in muscle and adipose tissue.

Measurement of Muscle Properties with US

US imaging is a low-cost, portable modality widely available in most clinical settings, including nonradiology specialties. Its durability for bedside use makes it particularly beneficial for critically ill patients (28–30). Unlike high-energy x-ray, US uses low-energy, high-frequency sound waves that are sensitive to soft tissues, such as muscles and internal organs, to generate tissue contrast and assess tissue properties. Given its widespread availability and robustness in implementation, there is a growing trend to use US imaging for body composition analysis (Fig 3A) (31,32).

Figure 3:

(A) Axial US imaging shows echogenicity measurement; yellow circle, region of interest; white arrowhead, the mean echogenicity value from the histogram; red rectangle, a schematic representation of the transducer’s position. (Reprinted, under a CC BY 4.0 license, from reference 32.) (B) Exemplary US image of a cross-section of the quadriceps. All quadriceps muscles (vastus medialis [VM], rectus femoris [RF], vastus lateralis [VL], vastus intermedius [VI]) were captured in a single motion. The femur (F) is also discernible within the thigh. (Adapted and reprinted, under a CCBY NC ND 4.0 license, from reference 39.)

Early changes in muscle and body composition often precede the clinical signs of cachexia, such as substantial weight loss. These changes can serve as crucial biomarkers for identifying the risk of CAC before it becomes clinically apparent.

By measuring muscle thickness, cross-sectional area, stiffness, and echogenicity (33,34), radiologists using US can detect these early changes in muscle properties, often before patients experience weight loss. This ability to identify muscle wasting before the onset of noticeable weight loss makes US a valuable tool in the early diagnosis and management of CAC. The absence of ionizing radiation and enhanced portability of US imaging is particularly advantageous compared with DXA and CT (35,36). Unlike MRI and nuclear imaging, which require dedicated facilities and personnel with specialized training, US imaging is readily accessible in health care settings beyond radiology departments (36).

US also provides both quantitative and qualitative muscle information (37). Muscle quantity is typically assessed through measurements of muscle thickness and cross-sectional area (38), while muscle quality is evaluated based on muscle echogenicity (33). The updated consensus from the European Working Group on Sarcopenia in Older People recommends US for assessing muscle quantity (11). A recent study comparing panoramic US and MRI measurements of skeletal muscle cross-sectional areas found excellent concordance between the two methods in monitoring quadriceps muscle atrophy and hypertrophy (Fig 3B) (39).

US-based muscle echogenicity is associated with fat infiltration, fibrous tissue, and reduced muscle function (40). These early changes in muscle quality often precede substantial weight loss and serve as important markers for early diagnosis of CAC. Combining measurements of the skeletal muscle cross-sectional area and echogenicity can improve diagnostic accuracy for muscle wasting compared with using these parameters individually (31,41). Muscle fat infiltration, as indicated by US measurements of echogenicity and shear wave velocity, can effectively monitor the biologic and physiologic processes in the tissue, providing a more detailed assessment compared with traditional gross weight measurements such as BMI (29). Studies have demonstrated that US assessment of calf muscles can effectively depict sarcopenia based on muscle thickness and echogenicity of the anterior tibial muscle (41).

US elastography enables the assessment of stiffness and elastic properties of soft tissue (42), offering additional insights into muscle quality and functional loss in patients with CAC. Increased intramuscular fat and muscle fibrosis detected with elastography are associated with decreased muscle function, reduced strength, and higher mortality rates (43,44). Shear wave US elastography measures tissue stiffness as a parameter of muscle quality and functional loss, independent of muscle quantity (45).

A challenge in using US imaging for the quantitative assessment of tissue properties is the inconsistency in US imaging protocols for monitoring longitudinal muscle loss. This variability arises primarily from differences in patient muscle compression, site selection, and fluid status (23,46). To address this limitation, standardized scanning protocols have been developed to enhance the reproducibility of US systems for grayscale analysis (47).

Comprehensive Measurement of Skeletal Muscles Using CT

CT is an essential imaging modality for evaluating body composition in patients with cancer due to its high resolution (0.25–0.50 mm) and ability to differentiate skeletal muscle and adipose tissue, including visceral and subcutaneous fat, based on distinct Hounsfield unit values (48,49). This capability to quantify muscle and fat mass has proven beneficial for assessing CAC.

CT imaging at the third lumbar vertebra (L3) is commonly used to assess muscle and fat distribution in patients with CAC, as it correlates well with whole-body muscle and fat mass (50). Several studies have demonstrated that reductions in subcutaneous adipose tissue (SAT) and subsequent muscle atrophy are early indicators of CAC. In patients with pancreatic cancer, CT images at the L3 level revealed that SAT depletion occurred first, followed by atrophy in visceral adipose tissue (VAT) and skeletal muscle (51). One study found that reductions in adipose tissue were detectable 6 months before diagnosis, while muscle mass decline was observed 18 months prior (52). These findings highlight the importance of early detection of changes in skeletal muscle and adipose tissue for cancer diagnosis.

CAC can be diagnosed using the SMI at the L3 level (cutoff: <52.40 cm²/m² for men; <38.50 cm²/m² for women) as a diagnostic criterion for sarcopenia, in combination with a weight loss exceeding 2% of BMI (1,12). In patients with metastatic lung cancer, sarcopenia due to CAC, identified using SMI thresholds, has been found to be effective in independently predicting response rate and prognosis (53). Furthermore, preexisting sarcopenia increases the risk of perioperative complications and worsens long-term prognosis in patients with lung cancer (54). SMI at L1 is sometimes used as an alternative to L3 for diagnosing sarcopenia (55). Additionally, CT measurement of the pectoralis muscle area at the fourth thoracic vertebra has been shown to predict survival in non–small cell lung cancer (56). The role of SMI in predicting CAC prognosis has also been investigated in other cancers, including esophageal (57), colorectal (58), and pancreatic cancers (59). The combination of sarcopenia and adipopenia at the L3 level has been integrated into a cachexia scoring system, providing prognostic markers for diffuse large B-cell lymphoma (60).

CT-derived muscle quality measurements, such as mean muscle attenuation, negatively correlate with muscular fat content, indicating latent sarcopenia (61). As CAC progresses, adipose tissue degradation leads to intramuscular fat infiltration, known as myosteatosis, which is associated with decreased CT attenuation values (62). Elevated myosteatosis has been linked to poorer survival and prolonged hospital stays in patients with cancer (63,64). Skeletal muscle attenuation, as quantified with CT, has proven to be a prognostic marker for outcomes in cancers, including colorectal cancer (65) and diffuse large B-cell lymphoma (66).

The major advantages of CT lie in its ability to assess metabolic changes in CAC, such as hepatic fat content. Recent studies suggest that CT may be useful for monitoring hepatic fat accumulation, which is often associated with increased energy expenditure in patients with CAC (5,67). These findings emphasize CT’s potential as an early diagnostic tool for depicting metabolic shifts in CAC.

CT offers rapid, high-resolution, multiplanar three-dimensional volumetric imaging, making it an effective tool for assessing muscle, adipose tissue, and liver in CAC prognosis. Despite its strengths, CT has limited ability to distinctly visualize muscular layers due to the inherent nature of x-ray–based imaging. Furthermore, concerns over radiation exposure during repeated scans may restrict its use in CAC (68).

Functional and Metabolic Imaging with MRI

Cancer-associated skeletal muscle atrophy involves changes such as loss of muscle fibers and increased fat infiltration (69). These functional and metabolic changes occur earlier than structural or morphologic changes, making MRI a promising tool for depicting early-stage CAC. MRI provides high contrast and resolution, making it ideal for assessing muscle and adipose tissue changes associated with CAC (25). Unlike CT, MRI can also reveal functional and metabolic changes, offering insights beyond anatomic imaging (25,70). For instance, hyperpolarized MRI is valuable for monitoring tumor metabolism and treatment response (71,72), while chemical exchange saturation transfer MRI provides insights into the metabolic pathways of creatine and phosphocreatine in muscle tissue (73,74), among other applications.

Advanced MRI techniques, such as diffusion-weighted imaging and diffusion-tensor imaging, enable early detection of muscle abnormalities (75). Diffusion-weighted imaging can assess skeletal muscle integrity, while diffusion-tensor imaging evaluates muscle microstructure and fiber composition, offering valuable biomarkers for CAC (75). Additionally, MRI elastography can help assess tissue stiffness, aiding in the evaluation of muscle changes in CAC (Fig 4A) (76). Moreover, T1 and T2 mapping sequences, which measure relaxation times, provide insights into both pathologic and physiologic changes in muscle (77). Proton MR spectroscopy combined with MRI can quantify intracellular lipids in muscle, offering a direct measure of fat infiltration, a key feature in CAC progression (78). Together with ³¹P-MR spectroscopy, which depicts muscle energy metabolism, and ¹³C-MR spectroscopy, which depicts muscle triglycerides and glycogen, these methods enable a comprehensive assessment of metabolic changes associated with CAC (78).

Figure 4:

(A) MR elastography assessment of biceps brachii stiffness in the right arm under varying loads in a healthy male individual lying in the right lateral decubitus position. The images show muscle orientation and driver placement, with wave patterns for 0-, 4-, and 8-kg loads, indicating muscle stiffness with higher loads. (Reprinted, under a CC BY 4.0 license, from reference 76.) (B) The proton density fat fraction (PDFF) maps of the psoas and erector spinae muscles (highlighted in red) in a 56-year-old male individual with esophagogastric junction adenocarcinoma show a decrease in PDFF over 218 days. (Reprinted, under a CC BY 4.0 license, from reference 80.) (C) MRI employs chemical shift-selective imaging in both healthy control mice and mice with pancreatic cancer, demonstrating metabolic changes in white adipose tissue (WAT). CSSI = chemical shift-selective imaging, eWAT = epididymal WAT, PC = pancreatic cancer, sWAT = subcutaneous WAT. (Adapted and reprinted, under a CC BY 4.0 license, from reference 83.)

MRI is also sensitive to fat changes, which is crucial for monitoring CAC. Techniques like chemical shift imaging can quantify fat infiltration (79), with studies showing significant correlations between fat content in muscles and BMI loss in patients with CAC (Fig 4B) (80). CAC is often accompanied by ectopic fat deposition, particularly in the liver (81). Advanced techniques, such as the three-point Dixon method, enhance the accuracy of fat quantification, especially in the liver and vertebrae, allowing for organ- and tissue-specific assessment of fat accumulation in CAC (82).

Moreover, MRI can capture dynamic metabolic changes associated with CAC progression, such as the browning of white adipose tissue, which contributes to lipid breakdown—a key feature of CAC (9,10). A study in mice with pancreatic cancer and CAC showed that chemical shift imaging measurement of fat fraction in SAT and VAT revealed a significant reduction due to white adipose browning (Fig 4C) (83). Functional MRI techniques, such as blood oxygen level–dependent imaging, can further help confirm changes in brown adipose tissue, which are related to the metabolic shifts in CAC (84).

With its nonionizing radiation and high accuracy, MRI shows great promise for the early assessment and diagnosis of CAC. However, clinical applications face challenges such as long scan times, high costs, and complex postprocessing. Advances in rapid imaging sequences, improved image reconstruction algorithms, and optimized clinical workflows should overcome these limitations, making MRI a critical tool for early detection of CAC in oncology and radiology clinics.

Molecular Imaging with PET/CT

PET, often used in combination with CT, offers a multimodal approach to assessing and diagnosing CAC by integrating both morphologic and metabolic imaging modalities. Fluorodeoxyglucose (FDG), a glucose analog, is a widely used tracer in oncology for imaging of cancer metabolism. FDG is taken up by cells and trapped intracellularly after phosphorylation, making it suitable for evaluating cancer metabolism and associated metabolic changes, including those observed in CAC (85).

Recent studies have explored the use of FDG PET/CT for the early detection of muscle changes in CAC. For instance, a study involving 56 patients with stage III non–small cell lung cancer found increased FDG uptake in the psoas muscle of patients with CAC, indicating its potential as a prognostic indicator (86). Additionally, elevated muscle standardized uptake value (SUV) of FDG was associated with decreased overall survival and local recurrence-free survival time in soft tissue sarcomas (87). Lower muscle SUV correlated with reduced serum hemoglobin and albumin levels, highlighting its potential in assessing nutritional status and disease severity (87).

FDG PET/CT is the currently accepted method for quantifying brown adipose tissue, which exhibits high glucose uptake. A preclinical investigation showed that FDG uptake serves as a biomarker for the CAC-associated process of white adipose browning (88). Furthermore, a study of 390 patients with cancer found that reduced hepatic FDG uptake and increased uptake in VAT and SAT were linked to CAC (Fig 5) (89). Additionally, a nomogram incorporating age, hemoglobin levels, maximum SUV of the liver, and minimum SUV of the subcutaneous fat has shown strong predictive ability for the onset of cachexia in patients with cancer, facilitating risk stratification for overall survival in these patients (90).

Figure 5:

Representative PET/CT images of two patients with lung cancer. (A–D) A 60-year-old female individual with cachexia shows lower peak liver SUV (2.22), higher minimum visceral fat SUL (0.50), and subcutaneous fat SUL (0.26). (E–H) A 59-year-old female individual without cancer-associated cachexia, with higher peak liver SUV (2.98), lower minimum visceral fat SUL (0.16), and subcutaneous fat SUL (0.07). Both patients were female individuals under 65 years old, with a BMI more than 20 and WBC less than 9.5×10⁹/L. However, the patient with cachexia showed lower peak liver SUV and higher minimum visceral and subcutaneous fat SUL compared with the patient without cachexia. BMI = body mass index, SUL = standardized uptake value normalized by lean body mass, SUL_min = minimum standardized uptake value normalized by lean body mass, SUV = standardized uptake value, SUV_peak = peak standardized uptake value, WBC = white blood cell. (Adapted and reprinted, with permission, from reference 89.)

Since PET/CT allows for the simultaneous assessment of primary tumors and distant organs, it provides a comprehensive evaluation of metabolic changes in CAC. For example, in esophageal and gastroesophageal junction cancer, pretreatment FDG PET/CT scans showed that higher maximum SUV in the primary tumor was significantly associated with CAC and patient survival (91). Similarly, a large retrospective study found markedly reduced hepatic FDG uptake in patients with CAC, which correlated with poorer survival rates (92).

Recently, efforts have been focused on developing novel radionuclides with high specificity and sensitivity to CAC-specific biomarkers. Preclinical studies have shown that translocator protein tracers, such as fluorine 18 (¹⁸F)-FEPPA and ¹⁸F-DPA-714, can effectively image brown adipose tissue (93,94). While these novel tracers have shown promise in preclinical investigations, they require further development for clinical applications (93,94).

Future Perspectives

As precision and individualized cancer care continue to advance, the impact of CAC on patient outcomes and survival is increasingly recognized. However, a substantial gap remains in the understanding of CAC mechanisms and the effective medical interventions available. Emerging evidence also suggests that CAC patterns vary significantly across different demographic and socioeconomic groups, highlighting the role of health disparities in disease progression. For instance, a study comparing non-Hispanic Black and non-Hispanic White patients with pancreatic cancer found that CAC-related muscle loss was more pronounced in Non-Hispanic Black patients, with greater decreases in core and psoas musculature, lower baseline serum albumin levels, and higher platelet counts (95). These findings underscore the need for a more comprehensive approach to CAC assessment that accounts for racial and socioeconomic factors.

These unmet needs can be addressed through new imaging technologies and approaches that are already available within clinical imaging modalities. The imaging techniques discussed in this review provide examples of noninvasive methods to quantitatively analyze systemic metabolic changes in patients with CAC, including changes in muscle tissue, adipose tissue, and liver function. These approaches and technologies provide valuable information for the potential early diagnosis of CAC, which can inform clinical decisions regarding early intervention to slow its progression.

Importantly, using clinically available imaging modalities is cost-effective and does not impose additional burdens on resource-limited communities. We can utilize the resources and technologies developed for imaging primary cancer problems to also image CAC, thereby accelerating the clinical translation of these applications. This review discusses the use of DXA, US, CT, MRI, and PET to evaluate CAC-related changes in tissue properties and metabolic functions. Other potential applications are discussed in the literature. For instance, functional MRI techniques can depict alterations in brain activity related to appetite regulation and food intake, which are frequently dysregulated in cachectic patients (96). High-resolution proton MR spectroscopy has been used to detect increased levels of free choline and phosphocholine, contributing to higher total choline levels in cachectic tumors (97). Additionally, DXA, CT, and MRI can help identify sarcopenic obesity by evaluating muscle and fat tissue composition (98). In patients with CAC, sarcopenic obesity is prevalent but often overlooked due to minimal weight loss (98). This condition is associated with poor prognosis, including worsened overall outcomes and reduced chemotherapy efficacy (98). Furthermore, research in animal models of CAC has demonstrated progressive cardiac atrophy, accompanied by impaired cardiac function, as shown with echocardiography (99). Cachexia-induced heart failure further exacerbates the progression of CAC (99). The bidirectional relationship between CAC and subsequent heart failure, as observed in preclinical studies, calls for further investigation in humans. Advanced imaging modalities, such as CT, MRI, and echocardiography, can help elucidate these interconnected pathophysiologic processes (99).

Each imaging modality discussed in this review has its strengths and limitations, as summarized in the Table. Thus, proper diagnosis and assessment of CAC necessitates the selection of appropriate imaging tools and capabilities based on specific circumstances and resources to maximize diagnostic information from routine clinical imaging examinations. Since metabolic changes often precede cachexia, ongoing research should aim to detect and identify specific biomarkers using molecular and functional imaging to enable highly sensitive and specific imaging of the early stages of cachexia.

Advantages and Disadvantages of Different Imaging Modalities for Assessing Cancer-associated Cachexia

Given the complexity of CAC, multimodal imaging assessments are essential for its accurate diagnosis. In this context, radiomics is emerging as a promising tool for deriving imaging biomarkers for early detection and prediction of CAC. Radiomics uses large datasets and advanced computational capabilities, including artificial intelligence (AI) (100), to extract image features from medical images, enabling the identification of subtle patterns indicative of early structural changes associated with CAC. AI-driven radiomics may allow clinicians to detect CAC at its early stages, facilitating timely intervention and potentially improving patient outcomes. For example, AI-assisted analysis of US images enables assessment of muscle quality through the evaluation of intermuscular adipose tissue (101). This method has been shown to effectively screen and diagnose sarcopenia, an early manifestation of CAC (101). A recent study demonstrated that radiomics can predict CAC in patients with stage IV non–small cell lung cancer, with gray-level co-occurrence matrix features of skeletal muscle serving as strong indicators of muscle loss, a key sign of CAC (102). Additionally, a radiomics signature derived from PET/CT, combined with BMI, M staging, and Eastern Cooperative Oncology Group score, can help predict cachexia onset before immune checkpoint inhibitor therapy in non–small cell lung cancer, enabling proactive monitoring of high-risk cases (103). In addition to advancements in radiomics, platforms like Storyline Health (https://storylinehealth.com), which integrates telemedicine with AI tools, may support early CAC detection and clinical decision-making. By incorporating imaging data, clinical information, and personalized care plans, this platform could assist in assessing CAC risk. While still in its early stages, it holds potential to complement existing diagnostic methods and contribute to more proactive and personalized care for patients with CAC.

Recent advances in deep learning, particularly in automatic segmentation, have further enhanced the utility of radiomics in CAC diagnosis. AI-assisted approaches, such as the V-Net network (https://github.com/faustomilletari/VNet), have demonstrated high accuracy in segmenting adipose and muscle tissue, with Dice indexes of 0.96 for SAT and 0.98 for VAT (104). Additionally, an automatic segmentation method based on a fully convolutional network has been successfully applied to process and analyze abdominal CT images (Fig 6) (105).

Figure 6:

Fully convolutional neural network–based adipose and muscle tissue segmentations on abdominal CT images at the third lumbar vertebra. (A) Integrated depiction of all segmented areas. (B–D) Segmentation maps distinguish subcutaneous fat (B, highlighted in red), skeletal muscle (C, highlighted in purple), and visceral fat (D, highlighted in green). The Dice similarity coefficients stand at 0.98, 0.99, and 0.98 for subcutaneous fat, skeletal muscle, and visceral fat, respectively. (Adapted and reprinted, under a CC BY 4.0 license, from reference 105.)

Despite the promising potential, the clinical application of AI and radiomics in diagnosing CAC remains limited, highlighting the need for further research in this area. Addressing this gap is expected to yield innovative tools for the early detection and prediction of CAC, ultimately contributing to improved clinical management of CAC and patient outcomes.

Conclusion

Clinical imaging technologies, including DXA, US, CT, MRI, and molecular imaging, can complement the quantitative analysis of systemic metabolic changes in CAC, which is crucial for early diagnosis and intervention planning. Implementing cost-effective imaging tools like DXA and US imaging to diagnose and assess CAC not only benefits resource-limited settings but also utilizes existing capabilities from primary cancer imaging, thus expediting applications for CAC. Looking ahead, the integration of AI technologies with molecular and functional imaging to identify specific imaging biomarkers holds great promise for early detection and risk prediction of CAC. This approach has the potential to enhance detection rates, support informed decision-making, and improve clinical outcomes, ultimately enabling more effective management of CAC.