Beyond the tumor: towards a cachexia-based host phenotype through body composition analysis in patients with resectable lung cancer

Abstract

Cancer cachexia, characterized by involuntary weight loss and extensive muscle and adipose tissue wasting, is a major contributor to morbidity and mortality in cancer patients. To date, no effective medical intervention can completely reverse this multifactorial syndrome, which is driven by different metabolic changes. Identification of cachectic patients is primarily based on alterations in body composition and the assessment of systemic metabolic and inflammatory changes. While these changes have been thoroughly described in patients with more advanced stages of lung cancer, their role in resectable lung cancer remains less explored. In this review, we summarize the different methods to assess body composition metrics such as skeletal muscle (SM) mass, fat distribution and overall body composition. As the dominant driver of cancer cachexia, we also describe the two most widely accepted acute phase proteins. Furthermore, we discuss the short and long-term clinical implications of cancer cachexia and the corresponding body composition and inflammatory changes in resectable lung cancer patients. Finally, we explore the possibility of identifying a specific host phenotype of cachectic lung cancer patients that predisposes to adverse outcomes of lung cancer surgery, which might enhance the predictive value for overall survival and aid in treatment decision-making in lung cancer patients in the future.

Introduction

With the recognition of cancer as a systemic disease that is not solely confined to cellular processes, there is increasing awareness that cancer cachexia is a major contributor to morbidity and mortality in cancer patients. Cancer cachexia has a prevalence of 50% to 80% depending on the cancer type and stage (1). Moreover, it is the primary cause of death in 22% of cancer patients (2). Cancer cachexia is characterized by involuntary weight loss and loss of strength, mainly due to extensive muscle and adipose tissue wasting. It is driven by reduced food intake and metabolic changes such as an ongoing acute phase response, excess catabolism, and insulin resistance (2,3). To date, no effective medical intervention can completely reverse this multifactorial syndrome, and strategies such as conventional nutritional support or exercise have failed so far but might be effective when accompanied by a multimodal approach (4).

The term cachexia is derived from two Greek words, kakós, and hexis, which literally translate to “bad” and “condition”, and has been described as early as Ancient Greece by Hippocrates (5). He wrote “the flesh is consumed and becomes water, the shoulders, clavicles, chest, and thighs melt away, the illness is fatal” (6). Cachexia induced by cancer was first described in 1844, when a 49-year-old patient died of “cancerous cachexia” after a long history of recurring cancer (7). In 1954, Hugh Donovan reported on “malignant cachexia” in kidney and testicle cancer (8). He described it as a “genuine entity” and “generalized disease” that could be recoverable if the primary tumor was removed. He depicted the clinical features of malignant cachexia as “only too familiar” with symptoms like “an unsmiling, wasted face, flaccid skin, complete lack of energy, the loss of appetite and weight, and the infection of the mouth with thrush”. He stated that “malignant cachexia” might also have affected laboratory tests in these patients (2-4).

Nowadays, identification of cachectic patients is primarily based on alterations in body composition and the assessment of systemic metabolic and inflammatory changes (3,4,9,10). Skeletal muscle (SM) and adipose tissue are the two most studied compartments in cancer patients in terms of body composition (11,12). So far, different methods have been proposed to assess body composition and inflammatory status in cancer cachectic patients. Cancer cachexia affects a large proportion of patients with non-small cell lung cancer (NSCLC) and profoundly diminishes their therapeutic options. Much is known about lung cancer patients with cachexia in advanced stages, but knowledge on the impact of cancer cachexia on early-stage lung cancer surgery is rather new (13,14). The goal of this review is to summarize the currently available (imaging) techniques and laboratory tests to objectively assess body composition and inflammatory changes in lung cancer patients with cancer cachexia and resectable disease. Furthermore, we will discuss how cancer cachexia and the corresponding body composition changes impact lung cancer survival after surgery, and how this knowledge could be used in the future to improve treatment decision-making in this population.

Body composition assessment

Although anthropometric measurements [e.g., body mass index (BMI), calf circumference, and skinfold thickness] are the least expensive methods for objective body composition measurement, these lack accuracy and do not provide a clear and detailed picture of the different tissues of interest to assess cancer cachexia. Over time, several imaging techniques with higher validity have been adopted to measure SM mass, fat distribution and overall body composition, such as dual energy X-ray absorptiometry (DXA), computed tomography (CT), magnetic resonance imaging (MRI), and ultrasonography (Table 1) (9).

Table 1. Summary of different modalities of body composition assessment and inflammation markers, their mechanisms of action, application, merits and dismerits.

Modality	Mechanism	Application	Advantages	Limitations
Dual energy X-ray absorptiometry	Comparison of absorption levels of two X-ray beams on different tissues	Differentiate between bone tissue, adipose soft tissue and lean soft tissue, and quantify these tissues	Easy to perform, widely available, cheap	Ionizing radiation. Relatively limited application and validation due to limited research in lung cancer patients
CT	Measurement of different levels of radiodensity (Hounsfield units) of soft tissues, mostly single slice L3-image	Quantification of the area of SM, VAT, SAT. Tissue triglyceride quantification through SM-RA, VAT-RA, SAT-RA	Easy to perform, standard work-up in lung cancer patients. Free automated segmentation and assessment tools. Single-slice and whole-body applications. Well-validated	Ionizing radiation. Not always available in follow-up
Magnetic resonance imaging	Different levels of energy emission of soft tissues in a magnetic field that is displaced by radiofrequency pulses	Quantification of the area and signal intensity of SM, VAT, SAT	No ionizing radiation. High level of detail	Not standard in work-up. Time-consuming scan/procedure for the patient. Variation between different scanners and acquisition methods
Ultrasonography	Measurement of soft tissue properties through ultrasonic reflections	Quantification of intra-abdominal fat thickness, mesenteric fat thickness, pre-peritoneal fat thickness, muscle thickness, fascicle length, echo-intensity	No ionizing radiation. Widely available, cheap. Bed-side and outpatient availability	Not standard in work-up. Quality and reproducibility are user-dependent
Biolectrical impedance	Measurement of electrical properties (impedance) of soft tissues	Estimation of the amount of fat, and fat-free mass in different patient groups	No ionizing radiation. Portable and easily available. Viable option for body composition estimation if other modalities are not available	Inconsistent correlations between CT-derived body composition metrics. Unreliable in patients with high tumor volume or edema

CT, computed tomography; L3, third lumbar vertebra; RA, radiation attenuation; SAT, subcutaneous adipose tissue; SM, skeletal muscle; VAT, visceral adipose tissue.

Imaging

DXA

DXA has traditionally been regarded as the gold standard for assessing bone mineral density and body composition metrics (15). DXA operates by using the differential radiation attenuation (RA) levels of two X-ray beams at two different energy levels at either the whole-body level, or on regional levels (extremities for example) (16). By comparing the absorption patterns of the two X-ray beams, it can differentiate between bone tissue, lean soft tissue, and adipose tissue, and thus accurately quantify these tissues in the human body (17). DXA does so by quantifying the R-value, which is the ratio between the attenuation coefficients at both energy levels. The R-value is specific for each tissue and is constant for bone and fat in all patients but varies for other soft tissues based on the patients’ body composition. To define SM loss, the SM area is used, which, in general, is then normalized by height to calculate the SM mass index (cm²/m²). In addition to these assessments, measurements of muscle contractility and muscle function, like handgrip strength, are recommended to further diagnose SM wasting (17). The same principles are used for adipose tissue by calculating the fat mass index, which defines male subjects as class I, II, and III obese based on the adipose tissue index (kg/m²).

Although DXA is a relatively easy, widely available, and cheap method to assess body composition, only a few studies can be found that used DXA to assess body composition in lung cancer patients. One study in a cohort of 11 patients assessed appendicular SM and abdominal fat mass in patients before and after systemic therapy and compared them with healthy controls (18). They did not find a difference in appendicular SM or adipose tissue mass before and after lung cancer treatment, nor did they find any differences with healthy controls. However, they did find differences between patients and healthy controls in terms of exercise tests and cellular level muscle contractility, highlighting the importance of additional exercise and muscle function tests. Another study only investigated SM wasting through DXA in stage IV lung cancer patients that were treated with immune check-point inhibitors (19). They found that patients who showed SM wasting had a significantly shorter progression-free survival when compared to patients who did not show SM wasting. However, they did not assess adipose tissue or inflammation markers.

CT

CT is currently regarded as the most accurate imaging modality (along with MRI) to assess body composition metrics. CT scans are widely available as they are routinely performed for the diagnostic work-up of (lung) cancer patients (20). CT images are generated based on different levels of tissue radiodensity, or RA. The radiodensity is expressed in Hounsfield units (HU) (21), with more dense tissues having a higher HU value. The HU values of different tissues can be used to identify and quantify them on CT-images. For example, adipose tissue is characterized by a lower radiodensity (−190 to −30 HU) than SM (−29 to 150 HU) (22). By segmenting a single slice at the level of the third lumbar vertebra (L3), precise estimations of the mass and radiodensity of skeletal muscle (SM) tissue, visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT) can be made (Figure 1), which strongly correlate with the total body SM and adipose tissue mass assessed by DXA (22). While L3 is the most optimal level for single-slice segmentation, some studies have validated that segmentations at the first lumbar (L1), or fourth thoracic (T4) level vertebrae, which are both depicted on chest CT scans, are acceptable alternatives when images at the L3 level in a full-body CT scan are not available (23,24). Other vertebral levels have also been investigated such as the fifth, eighth, tenth, and twelfth thoracic vertebra (25-27). While these vertebral levels might correlate slightly less with total-body SM and adipose tissue mass, they do offer a wider scale of applications in body composition analysis, especially when only thoracic vertebra levels are depicted on a scan.

Figure 1.

Automatic segmentation of L3 CT-slice by Mosamatic showing the area of subcutaneous adipose tissue, visceral adipose tissue, and skeletal muscle tissue. Yellow: visceral adipose tissue, red: skeletal muscle tissue, blue: subcutaneous adipose tissue. The muscles that are analyzed are the transversus abdominis, rectus abdominis, psoas muscles, quadratus lumborum, erector spinae, and the internal and external oblique muscles. L3, third lumbar vertebra; CT, computed tomography.

Software packages like SliceOmatic and OsiriX can be used to manually segment and quantify the tissues of interest from a single CT-slice, however this is time-consuming work and prone to human error (28). Automatization of CT-based body composition segmentation is an essential development towards large cohort studies and future clinical implementation. Hence, validated software tools that perform automated segmentation and quantification of the tissues of interest at the L3-level have been developed through deep-learning models, and are freely available, such as Mosamatic^TM (29). Despite the fact that CT-imaging uses ionizing radiation to create images, it is probably the most suitable for longitudinal follow-up of cancer cachexia in lung cancer patients, despite the risk of radiation-related complications. Given that chest CT scans are part of routine follow-up in lung cancer patients, they inherently provide the opportunity to assess body composition at various vertebral levels over time (24). Hence, most studies assessing body composition in lung cancer patients utilized CT-imaging as their primary method.

MRI

Like CT imaging, MRI is a powerful modality to assess body composition on both organ and tissue level, as well as on whole-body level (30). MRI works through the principle of creating a strong magnetic field to align hydrogen protons, which are then temporarily displaced by radiofrequency pulses, after which they return to their equilibrium state, during which they emit energy that is captured and converted into images of tissues and organs by the MRI scanner (31). It has some advantages over CT and DXA imaging. For instance, it does not depend on ionizing radiation, which makes it suitable for imaging of radiosensitive populations. Additionally, the soft-tissue contrast imaging on MRI is superior to CT-imaging, making it easier to delineate the different tissues of interest (30). However, it does have some downsides when compared to CT imaging. First of all, CT imaging is the gold standard in lung cancer work-up (32) and therefore MRI-imaging is not commonly used to assess body composition metrics in this population. Second, standardization is more difficult due to higher inter-scanner variability than seen with CT imaging (33). And third, due to the high level of detail, segmentation might actually be more difficult than on the more fundamental and less detailed CT scan (34). Therefore, MRI is a promising technique to assess body composition metrics and has proven to be valuable to do so in other diseases (34,35), but further development is necessary for lung cancer patients, and thus, no studies on MRI-based body composition assessment in resectable lung cancer patients have been conducted yet. Figure 2 shows a single slice MRI segmentation (35).

Figure 2.

Segmentation of L3 MRI-slice showing the dorsal portion of skeletal muscle at this level. The area highlighted in red is the dorsal area of muscle that is analyzed, which include the erector spinae and psoas muscles, quadratus lumborum, and internal and external oblique muscles. The area highlighted in green represents the anterior part of the muscles that are analyzed, which also include part of the oblique muscles, and the rectus abdominis muscle. L3, third lumbar vertebra; MRI, magnetic resonance imaging.

Ultrasonography

While CT and MRI are not always accessible and affordable, ultrasonography has gained interest as a radiation-free alternative, that allows bedside use and is associated with lower costs. It works through a transducer emitting ultrasonic sound and measuring its reflections but is less commonly used to assess body composition. However, it can be applied to measure visceral adiposity by measuring intra-abdominal fat thickness, mesenteric fat thickness, pre-peritoneal fat thickness, and the abdominal wall fat index for example (36). SM can also be measured in the form of muscle thickness, fascicle length and echo intensity (36). Correlation of quantitative measurements of these tissues correlates well with the other body composition assessment modalities like DXA, CT, and MRI (37), however, no uniform cut-off values for tissue measurement have been determined, and it is less accurate than the total body composition analysis provided by DXA, CT, or MRI. Furthermore, its accuracy and reproducibility are user-dependent, opposed to the other imaging modalities. So far, there are no known reports that assessed body composition through ultrasonography in lung cancer patients, and its widespread adoption is challenged by the lack of standardized measurements, warranting further research in the future.

Bioelectrical impedance analysis (BIA)

Another portable and radiation-free method to measure body composition is BIA (Table 1). BIA assesses hydration and cell mass, independently of body size, by measuring the electrical properties of different body tissues. It works by quantifying impedance, which results from the resistance and reactance of specific tissues (38,39). As such, it enables an estimation of a two-compartment model, including fat mass and fat-free mass, of body composition (38). It must be noted, though, that, especially in patients suffering from cancer, factors such as edema and tumor load could affect BIA readings (40,41). Edema, for example, can alter BIA readings by affecting tissue water content and muscle mass measurements.

Multiple studies have examined the correlation between BIA measurements and CT-derived assessments of skeletal and adipose tissues, unfortunately with inconsistent findings (38,39,42). One study (43) found that BIA measurements correlated well with CT-derived measurements of skeletal and adipose tissues in head and neck cancer patients (43), while another study showed the opposite in lung cancer patients (42), suggesting that the agreement between body composition measured with BIA and software analysis of CT-scans depends on cancer type (42). Therefore, BIA might be an inferior modality to assess body composition in cancer patients, especially in an in-hospital setting, in which CT and MRI scans are widely available and part of standard work-up in lung cancer patients.

Phase angle

BIA can, however, also be used to calculate phase angle, which reflects the relative contributions of body fluid and cellular membrane integrity. Phase angle is positively associated with cellular membrane capacitance and negatively associated with tissue resistance, which reflects fluid status (44). Lower phase angles are indicative for cell death, impaired cellular function, or diminished membrane integrity (45).

Calculating Z-scores

Z-scores are increasingly used in body composition assessment to standardize individual measurements against population-based norms, accounting for age, sex, and occasionally for ethnicity. They provide a dimensionless index of how many standard deviations an actual value deviates from the mean of a reference group and thus facilitate comparisons across individuals and time points (46). This standardization is particularly valuable in clinical and research settings, where detecting deviations from normative ranges—such as identifying cachexia—can have diagnostic and prognostic implications.

A Z-score represents the number of standard deviations that each patient differs from a mean value, which, in body composition analysis is mostly age- and sex-specific (12). These Z-scores are calculated by the following formula: Z = (X − µ)/σ, in which Z is the Z-score, X the actual value, µ the mean and σ the standard deviation (12,46). DXA employs Z-scores to contextualize lean and fat mass, as well as bone mineral content (47). In CT-imaging, Z-scores are applied to muscle, and adipose tissue, cross-sectional area and radiodensity (12,48). Although less common, the use of Z-scores in BIA and ultrasound is gaining recognition. Nonetheless, their clinical utility remains limited by the availability of comprehensive, modality-specific reference datasets, many of which have only recently begun to be systematically developed.

Inflammation assessment

Acute phase response

Inflammatory markers have been investigated to identify patients with cachexia and those who are at risk, as the origin of cachexia is rooted in systemic inflammation. Catabolic changes in cancer cachexia are signaled by tumor-secreted pro-inflammatory cytokines (particularly interleukin 1 and 6, and tumor necrosis factor-α) through tissue receptors in SM, adipose tissue, and hypothalamic receptors (3). However, serum levels of these cytokines are not suitable to diagnose cancer cachexia, as a direct correlation between the particular cytokine and degree of cachexia or mechanism for controlling tissue wasting has been difficult to prove (49). Furthermore, some cytokines only showed effect in murine models, and no association has been shown with the human form of cachexia.

Pro-inflammatory cytokines are the primary signaling molecules that activate the acute phase response, which is increasingly being recognized as the dominant driver of cancer cachexia (3). The acute phase response involves up- and downregulation of acute phase proteins, which can be measured long before the development of cachexia can be observed. In the next section, we describe the two most widely accepted acute phase proteins, C-reactive protein (CRP) and albumin, and another early marker of disease progression, the neutrophil-to-lymphocyte ratio (NLR).

CRP and albumin

The most widely assessed acute phase proteins are CRP, which is positively associated with systemic inflammation, and albumin, which is negatively associated with systemic inflammation. Typical values that are associated with cancer cachexia are albumin levels of <35 g/L and CRP levels >10 mg/L, however, there is considerable variation in the thresholds that are reported (3,50). By combining both these markers, the Glasgow Prognostic Score can be calculated, which ranges from 0–2, and has no variability in threshold determination (51). The Glasgow Prognostic Score correlates strongly with weight-loss and is shown to be a powerful predictor for tumor progression, survival, and quality of life in patients with lung cancer (3,51).

CRP alone can also serve as a marker of systemic inflammation and has potential prognostic value for body composition analysis and overall survival in cancer patients (52). It correlates well with muscle RA, intramyocellular lipid content, and cachexia in cancer patients, and can therefore be used as a prognostic marker by itself, but also in combination with other body composition metrics (53). One could therefore argue that CRP is a marker that can be well used for longitudinal follow-up of cachectic lung cancer patients, who do not undergo routine total body scans.

NLR

NLR is an alternative marker of the acute phase response when acute phase protein testing is not available (3,50). Neutrophils promote tumor growth as they secrete pro-angiogenic and anti-apoptotic factors (54). Lymphocytes (T-lymphocytes in particular), on the other hand, suppress tumor growth and are crucial for anti-tumor immune responses (55). Therefore, an increased NLR (more neutrophils than lymphocytes) is associated with worse survival outcomes in cancer patients. The NLR offers similar prognostic information and has less variable reported thresholds than CRP or albumin thresholds alone (3,50). The NLR threshold ranges from higher than three to higher than five, based on a systematic review (50) and all NLR ratio thresholds (>3 to >5) investigated were significantly associated with poor survival in (lung) cancer patients. Nevertheless, acute phase response proteins such as CRP seem to correlate strongly to overall and cancer-specific survival when compared to the NLR (56) and combining both seems to yield a more accurate survival prognosis than NLR alone (57).

Clinical implications for lung cancer surgery

A substantial proportion of patients with resectable lung cancer suffer from cancer cachexia (20). Cancer cachexia might not only affect the quality of life of these patients, but also their short-term surgical outcomes and long-term overall survival. In the next part, we delve deeper into the impact of cachexia on surgery outcomes and the other way around, and how cachexia-related changes in body composition and inflammatory markers can predict these outcomes.

Short-term outcomes

While some studies focus on body composition metrics that influence short-term postoperative outcomes such as 30- and 90-day mortality, and complications (26,58-61), most studies assess long-term overall and disease-free survival after lung cancer surgery. In terms of postoperative complications after lung cancer surgery, evidence on the influence of body composition remains ambiguous. Most studies report increased complications and length of hospital stay in patients with CT-derived sarcopenia, measured at single-slice lumbar and thoracic vertebral levels (26,59-62), while one study reports similar complication rates in sarcopenic and non-sarcopenic patients (58). Overall complication rates varied from 22.7% to 39.2% in the non-sarcopenic group, and from 53.2% to 62.5% in the sarcopenic group in the studies that found a significant difference in complication rates (60-62). The study that did not find a difference in complication rate showed an overall complication rate of 29.5% in sarcopenic patients and 20.9% in non-sarcopenic patients (58). All these studies assessed body composition by calculating the SM (or psoas muscle) index by measuring the cross-sectional area of SM at a single level vertebra CT image, which was then normalized for height (cm²/m²). Cut-off values of these cross-sectional areas have been reported in existing literature. However, opposed to using set cut-off values for men and women of all ages, it might be more valuable to determine Z-scores of the values of the population based on their age and sex. This principle has been used in a study that examined short-term outcomes after gastrectomy for gastric cancer (63). This study found that patients with a low calculated Z-score of SM mass, had more postoperative complications. Until now, no studies have been conducted that assess the impact of cancer cachexia as a whole on short-term outcomes in patients undergoing surgery for lung cancer. Short-term outcomes are also mitigated by the current enhanced recovery after surgery (ERAS) guidelines that were published in 2019 (64). These guidelines advocate for progressive drain management, optimal intra-operative fluid management, optimized pain management, thrombo-prophylaxis, and early mobilization after surgery, which all contribute to improved short-term outcomes. In terms of body composition metrics, these guidelines also recommend that patients undergoing surgery, should preferably have a BMI >18.5 kg/m². Furthermore, they even warrant postponement of the surgery to optimize nutritional status and BMI, as a low BMI is correlated to worse short-term postoperative outcomes like 90-day mortality, air leakage, and atelectasis (65,66).

Long-term prognosis

Body composition metrics and cancer cachexia are more likely to be of aid in treatment decision-making in multidisciplinary team meetings and further specify long-term prognosis in lung cancer patients. Several studies have investigated the impact of body composition and cancer cachexia on overall and disease-free survival, taking into account combinations of different parameters such as SM-, VAT-, and SAT-mass and RA, and inflammatory markers (27,67-73). SM quality is usually measured through the radiodensity of skeletal muscle tissue (SM-RA), as radiodensity is an indicator of tissue fat content (27,67,69,72). As such, a lower radiodensity of SM tissue is indicative of myosteatosis. All studies investigating this found that myosteatosis was associated with poor long-term survival (67,69,72).

In addition to SM mass and SM-RA, VAT and SAT have also been investigated as prognostic markers in patients undergoing surgery for lung cancer (69,74). One of these studies found that patients with a lower VAT volume, and higher subcutaneous adipose RA (lower fat content of SAT), and lower total fat volume showed worse disease-free survival (69). Additionally, another study found that a higher VAT/SAT ratio (relatively more VAT mass) was associated with poor survival prognosis in patients undergoing surgery for lung cancer (74).

Another study investigated pericardial fat volumes in patients that underwent lung cancer surgery and found that lower pericardial fat mass was associated with worse 5-year overall survival (73). This could be useful in patients that did not, or could not, undergo a full body CT-scan and only a chest CT-scan. This could also be helpful in longitudinal follow-up, as patients then, in general, only undergo chest CT-scans. However, it might be more practical to use the already validated segmentation at the level of the L1 vertebra or T4 vertebra in those cases.

Phase angle measurements through BIA have also been used to study long-term prognosis in advanced stage lung cancer patients, and showed that patients with a lower phase angle exhibited significantly shorter overall survival (75). While this was not investigated in lung cancer patients undergoing surgery, the aforementioned study suggests that measuring body composition and cell quality through BIA and phase angles could be valuable in (resectable) lung cancer patients in which CT-imaging is not available.

Lastly, some studies have investigated inflammatory markers and their role in predicting survival in patients undergoing surgery for lung cancer (68,71,76). One study found that decreased muscle mass was strongly correlated to higher CRP levels, which underscores the fact that cancer cachexia is not only a syndrome of (skeletal) tissue wasting, but also chronic inflammation, and advocates for host phenotype development (76). Another study showed that decreased albumin concentrations in combination with elevated NLR were associated with significantly worse overall survival in patients undergoing surgery for lung cancer (68), while the third study showed decreased overall survival rates in patients with high CRP levels compared to low CRP levels (71). This underscores the importance of systemic inflammation in patients burdened with cancer cachexia that are planned to undergo pulmonary resection.

Impact of surgery on cancer cachexia

Some studies have investigated how SM mass (70,77) and psoas muscle mass (78) changed over time after surgical removal of the tumor and how this impacted survival. Both studies assessing total SM mass found that a decrease of SM mass over time (6 months and 12 months) was associated with poor overall (70) and disease-free survival (77). The study investigating psoas muscle mass found that patients with high psoas muscle loss had a significantly shorter 5-year overall survival rate and a higher risk of recurrence (78). This indicates that follow-up of body composition metrics over time, after surgery, might aid in improving survival estimates, possibly altering postoperative treatment strategies. Up until now, no studies have been performed that investigate how cancer cachexia-associated symptoms change after tumor resection in lung cancer patients, and how this impacts survival. Additionally, no thresholds for minimal detectable changes in body composition metrics after cancer treatment have been determined. As such, it is difficult to correlate body composition changes to either the effect of cancer treatment, or measurement errors of different CT-scans or slightly different L3-levels. This should be studied in the future.

Cancer cachexia host phenotypes

Body composition assessment based on imaging (e.g., the CT-derived body composition metrics SM-RA, SAT-RA, VAT-RA and the SM-, SAT-, and VAT index) and inflammation marker testing (e.g., CRP) has proven to be prognostic for patients suffering from cancer. Most studies until now have focused on individual body composition metrics or chronic inflammation, as opposed to a combination of these factors. But, as cancer cachexia is defined as a syndrome that is characterized by the combination of SM and adipose tissue loss and chronic inflammation, one could hypothesize that combining body composition metrics and systemic inflammation to define so-called cachectic host phenotypes might even be more prognostic, and thus, relevant. Studies focusing on other types of cancer have attempted to define such cachectic host phenotypes (10,50). By correlating all possible body composition metrics and inflammation markers through a multivariable model, a combination of most predictive factors could be determined. In patients with colorectal cancer, for example, these host phenotypes were based on the combination of various body composition variables at a single slice L3 CT-image with CRP (10). In these patients, a combination of elevated CRP and reduced SM mass or elevated CRP and reduced VAT mass was much more strongly correlated to worse survival than these prognostic factors did by themselves.

Currently, prognosis and treatment decision-making in lung cancer care is majorly driven by tumor-based scoring systems such as the tumor, node, metastasis (TNM)-classification system and thus the host response to tumor (i.e., changes in body composition and inflammation) has been left out in the prognosis and treatment decision-making process of lung cancer patients. While it has not been investigated in lung cancer patients yet, it could be argued that a prognostic scoring system, such as the TNM classification, could be developed to determine different host phenotypes. This staging system would then be based on SM-, VAT-, and SAT mass and quality, in combination with systemic inflammation markers, such as the acute phase protein CRP or albumin. Such a host phenotype staging system could then be used in combination with the lung cancer TNM staging system to provide more accurate prognosis and better patient-tailored treatment strategies for both early-stage resectable lung cancer patients as well as the more advanced lung cancer patients. To even further quantify SM strength and function in these host phenotypes, functional tests such as handgrip strength could be added, which is an easy and cheap test to perform with proven prognostic value in cancer patients (79).

Perspectives and conclusion

Cancer cachexia is an important driver of morbidity and mortality in patients with resectable lung cancer. Several simple and well validated assessments of body composition and systemic inflammation are readily available for implementation of standard clinical care. Assessment of (degrees of) cancer cachexia could play a major role in improving survival predictions, and therefore also pre-operative treatment decision-making. Thorough phenotyping of the host (and tumor-associated interactions with the host) by combining body composition characteristics with markers of systemic inflammation shows great promise. Several automatic CT-based body composition assessment tools using deep learning models are already available and are essential for clinical implementation of routine host phenotyping. Prospective implementation studies and possibly intervention studies will be the next step forward in patient-centered lung cancer surgery.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.