Variable body and tissue weight reporting in preclinical cachexia literature may alter study outcomes and interpretation

ABSTRACT Cancer cachexia is a multifactorial syndrome of body weight loss, muscle wasting and progressive functional decline, affecting many advanced cancer patients and leading to worsened clinical outcomes. Despite inherent limitations of many preclinical cachexia models, including large tumor burden, rapid tumor growth and young age of animals, these animal models are widely used and imperative for the study of cachexia mechanisms and experimental therapeutics. However, there are currently no guidelines for the reporting and representation of data in preclinical cachexia literature. We examined the current state of data reporting in publications using the colon-26 adenocarcinoma (C26) model of cachexia and compared statistical differences in reporting mechanisms using animals from our laboratory. We show that data reporting and representation in C26 preclinical cachexia literature are diverse, making comparison of study outcomes difficult. Further, different expression of body and tissue weights in our animals led to differential statistical significance, which could significantly alter data interpretation. This study highlights a need for consistent data reporting in preclinical cancer cachexia literature to effectively compare outcomes between studies and increase translatability to the human condition.


Introduction
Cancer cachexia is a complex syndrome of bodily wasting and progressive functional decline (Baracos et al., 2018). Cachexia occurs in 30% of all cancer patients, and the risk for cachexia development is 70-90% for patients with cancers of the lung, liver, and gastrointestinal tract (Anker et al., 2019). Cancer cachexia is thought to contribute to ~20% of cancer deaths; however, this statistic may underestimate the true impact of cancer cachexia, as prevalence statistics are not currently included in the national cancer records of any country (Baracos et al., 2018;Tisdale, 2002). Patients with cancer cachexia often experience profound fatigue and weakness (Zhou et al., 2017), reduced anticancer therapy tolerance and effectiveness (Aapro et al., 2014;Muscaritoli et al., 2016;Ross et al., 2004;Vaughan & Martin, 2022), decreased quality of life (Baracos et al., 2018;K. C. Fearon et al., 2006;Wallengren et al., 2013), and increased mortality (Baracos et al., 2018;K. Fearon et al., 2011;K. C. Fearon et al., 2006;Penet & Bhujwalla, 2015;Wallengren et al., 2013). Heightened risk of cancer-related medical complications and increased medical costs have also been documented in this patient population (Arthur et al., 2016;Gourin et al., 2014). Limited therapeutic options exist for cachexia patients.
The use of preclinical animal models has greatly enhanced our understanding of cancer cachexia pathophysiology and allows for the avoidance of limitations faced in clinical research. Preclinical cancer cachexia models are valuable for the exploration of mechanistic questions and the assessment of experimental therapeutics.
Multiple categories of preclinical animal models have been developed for the study of cancer cachexia, including ectopic or orthotopic tumor cell implantation, human cancer cell or patient-derived xenografting, and spontaneous tumor growth models in genetically engineered mice (Ballarò et al., 2016).
Ectopic tumor cell implantation models are currently the most commonly used in cachexia literature. These models involve subcutaneous, intramuscular, or intraperitoneal injection of syngeneic rodent cancer cells into the experimental animal, followed by rapid tumor growth and the inflammatory cascade that accompanies it. Commonly used models in this category are colon-26 adenocarcinoma (C26), Lewis lung carcinoma (LLC), MAC 16 adenocarcinoma, B16 melanoma, Walker 256 carcinosarcoma, and Yoshida ascites hepatoma 130 (AH130) (Ballarò et al., 2016;Penna et al., 2016). Rudimentary characteristics of this category of preclinical cancer cachexia models that limit translatability include lack of natural tumor development, heterogeneity of tumor growth amongst experimental animals, ectopic tumor placement, large tumor burden, accelerated cachexia development and progression, decreased age of host, lack of metastasis, and absence of anti-cancer therapies compared with most humans with cancer. Despite limitations, ectopic syngeneic tumor cell implantation models remain widely used due to their relative ease of use, rapid tumor development, and subsequent manifestation of body weight loss and muscle wasting. Additionally, they are well-characterized in the published literature. Nonetheless, these models require careful study conduction and data representation to enhance replication and translatability.
After its genesis in 1975, the C26 model of cancer cachexia became prominent in the literature in the 1990s (Aulino et al., 2010;Corbett et al., 1975). Throughout the past three decades, this classical model of cancer cachexia has been used in over 246 experimental publications, with exponential growth in publication numbers over the past decade. The C26 model utilizes BALB/c or CD2F1 (F1 hybrid of female BALB/c and male DBA/2) mice. Study initiation often occurs at about 6 to 8 weeks of age, when C26 cells are subcutaneously injected. Tumors are visible after 7-14 days and significant body weight loss is achieved after approximately 16-27 days. This cancer model induces cachexia through tumor growth and systemic inflammation, leading to muscle and adipose wasting.
Anorexia likely contributes to negative energy balance and body weight loss towards the end of the study (Asp et al., 2011;Flint et al., 2016), but decreased food intake alone is insufficient to induce lean tissue wasting, as described in pair-feeding experiments (Tian et al., 2010). Body and tissue wasting occurs predominantly from increased lipid and protein catabolism, the latter attributed to increased activity of the ubiquitin-proteasome system (UPS) (Aulino et al., 2010) and autophagy (Penna et al., 2013(Penna et al., , 2019, as well as decreased protein synthesis (Argilés et al., 2014;Smith & Tisdale, 1993). Splenomegaly and elevated circulating levels of pro-inflammatory cytokines, namely high levels of IL-6, are observed (Hetzler et al., 2015;Narsale & Carson, 2014;Strassmann et al., 1992).
Clinically, the diagnostic criteria for cachexia are 1) body weight loss > 5% within six months, or 2) > 2% weight loss in individuals with a body mass index (BMI) < 20 kg/m 2 or with an appendicular skeletal muscle index consistent with sarcopenia (K. Fearon et al., 2011). Although there are certainly other symptoms and biomarkers that may predict cachexia severity in patients (e.g., fatigue, decreased quality of life and mobility/activity, insulin resistance, and circulating inflammatory mediators and tumor-derived factors), body weight and lean tissue loss remain key defining features of patients with cancer cachexia. As such, accurate and thorough reporting of animal body and tissue weight in the preclinical literature is a necessary component for assessing cachexia severity, comparing results across studies, and improving translatability to the human condition. Nonetheless, at present, there are no guidelines for methodological and data reporting procedures in preclinical cancer cachexia research.
Numerous methods of body and tissue weight reporting are currently used, and these methods may provide inconsistent information to readers. Given the wide utilization and necessity of preclinical cancer cachexia models, coupled with the inherent limitations of these models, particularly large tumor burden, rapid tumor growth, and use of young animals that may still be growing, development of guidelines for methodological and data reporting procedures would benefit the cachexia research community by improving transparency and interpretation of research results. As expressed by the Animal Research: Reporting of In Vivo Experiments (ARRIVE) Guidelines 2.0 , appropriate reporting and representation of methodological procedures and transparent data reporting is essential for study interpretation and translatability.
Before recommendations for reporting body and tissue weight can be developed, current practices in data reporting must be assessed. Therefore, the purpose of the present study was to 1) identify and quantify different mechanisms for reporting body and tissue weights in the published literature using the C26 model, and 2) apply these reporting mechanisms to data from our laboratory to identify whether differences exist in statistical significance between data presentations. We show here a wide variation in the types of data reported and the presentation of data in C26 studies. When applied to our animals, different data reporting mechanisms yield differences in statistical significance between mice with and without tumors. These results have important implications for interpretation of preclinical cancer cachexia research and highlight a need for the cancer cachexia research community to develop guidelines for data reporting to enhance replicability and translatability of research results.

Literature Data Extraction
In May of 2021, a PubMed literature search was conducted. The following search terms were used: "cachexia [Mesh] AND (colon-26 OR C26 OR colon 26 OR CT26)", and "(cachexia AND (colon-26 OR C26 OR colon 26 OR CT26)) NOT MEDLINE." A secondary search using the same search terms was completed in May of 2022 to identify new papers published during the past year. From the combination of these searches, 285 publications were identified and analyzed. Of these, 39 publications were eliminated due to lack of fit. Publications not written in English (n = 4), literature reviews (n = 3), those only using C26 cells in culture (n = 10), those that did not include cancer cachexia or mice (n = 8), duplicates (n = 13), and retractions (n = 1) were excluded from the analysis. Publications that did fit the search criteria, a total of 246, were assessed.
Each publication was evaluated independently by two reviewers (AGB and MLL), and discrepancies between reviewers' data were discussed to reach consensus. Weight reporting mechanisms used to report body and tissue weights were recorded. Body weight was categorized as either baseline weight, tumor-free or tumor-bearing final body weight, or tumor-free or tumor-bearing time course weight. Publications were counted as recording time course weight if they included any additional body weight measurements beyond baseline and final. Within these categories, body weight was reported as actual weight in grams, change from baseline weight (grams or as a percentage), or as the difference from tumor-free control animals (grams or as a percentage). Tissues included adipose, various skeletal muscles, heart, liver, and spleen. Tissue weight reporting mechanisms captured included weight in grams or milligrams, percent of final body weight (including or not including tumor), percent of baseline body weight, percent of control group tissue weight, or weight normalized to tibia length. Tumor weight reported as actual or estimated was also captured. Actual tumor weight refers to the weight of the tumor as measured via dissection on necropsy day, while estimated tumor weight refers to estimated tumor volume measured by calipers. Sex, strain, and age of mice, as well as the date of publication, were also captured. A schematic to illustrate the data collected is shown in Fig. 1.

Live-Animal Portion of Study
To compare data reporting methods for body and tissue weights found in our literature search, a study was conducted using the C26 cachexia model in our laboratory. All methods were approved by the Institutional Animal Care and Use Committee at Baylor University.
Cell culture: Colon-26 adenocarcinoma cells were obtained from the Division of Cancer Treatment and Diagnosis Tumor Repository, National Cancer Institute (Frederick, MD, USA), where they had been tested for purity and contamination. Cells were cultured in RPMI 1640 medium (Gibco, Waltham, MA, USA) supplemented with 5% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37°C in a 5% CO 2 humidified atmosphere. Subconfluent (~75%), low-passage (≤ 6 passages from attainment) cells were trypsinized, centrifuged, counted, and resuspended in sterile phosphate-buffered saline (PBS) at a concentration of 1x10 7 cells/ml immediately before implantation.
Animals: Eight to ten-week-old male (n = 27) and female (n = 27) CD2F1 (F1 hybrid of female BALB/c and male DBA/2) mice were included in this study. Sample size was determined based on previous literature (Asp et al., 2011). The mice were kept on a 12:12 h light-dark cycle with ad libitum access to standard rodent chow and water. Mice were randomly assigned (by cage) to tumor or control groups. Male (n = 14) and female (n = 14) mice were inoculated with 1x10 6 colon-26 adenocarcinoma (C26) cells in 100 µl PBS, while the remaining mice received an equal volume of sterile PBS via subcutaneous injection in the left flank. Injections were performed under isoflurane anesthesia. Body weight, body condition, food intake, and tumor dimensions were assessed regularly.
Mice that met humane endpoint criteria prior to the end of the study (n = 2 females with tumors) were excluded from analysis. Mice were sacrificed 25-27 days after tumor cell injection. Each mouse was placed under isoflurane general anesthesia (5% in oxygen), and after non-response to a firm toe pinch, blood was drawn through cardiac puncture. Mice were then euthanized via terminal thoracotomy. Tissue excision and organ collection commenced.
Tissues were collected, weighed, and snap-frozen in liquid nitrogen. Tissues were then stored at -80 °C.

Statistical Analysis
Two-tailed t-tests were used to compare tumor-bearing male animals to control male animals, and tumorbearing female animals to control female animals. A two-tailed t-test was also used to compare male tumor weight to female tumor weight. Assumptions of independence, normality, homogeneity of variance, and random sampling were all met for each t-test. Significance was set at p < 0.05. All data were analyzed using the IBM Statistical Package for the Social Sciences 28 (IBM SPSS Statistics, Cary, NC, USA), and figures were compiled using GraphPad Prism 9 (La Jolla, CA, USA). Data are expressed as mean ± standard error of the mean (SEM).

Weight Reporting Mechanisms in Published Literature
Body Weight: Baseline body weight was reported in 52.0% of publications (Fig. 3A). Final body weight was reported by some method in 86.5% of publications. Body weight including tumor weight was reported by 63.4% of publications and tumor-free final body weight by 49.2% of publications (Fig. 3B). Time course weight, body weight reported throughout the study period, was reported by some method in 61.4% of publications. Time course weight including tumor weight was reported by 48.8% of publications, and time course weight with tumor weight removed via tumor volume calculation was reported in 16.7% of publications (Fig. 3C).
Beyond weight reporting category (i.e., baseline body weight, final body weight, muscle and organ weights, etc.), the manner in which data was reported within each category varied. Within body weight reporting categories, body weight was reported as grams, change from baseline weight (grams or as a percentage), or as the difference from tumor-free control (grams or as a percentage) (Table 1). Tissue weight categories were reported as weight in grams, percent of final body weight (including or not including tumor mass), percent of baseline weight, percent of tumor-free control animal tissue weight, or weight normalized to tibia length (Table 2). For body and tissue weight, actual weight was the most commonly reported. Secondarily, body weight expressed as a percent change from baseline weight and tissue weights as a percentage of the control animal tissue weights were most prevalent.

Examination of Weight Reporting Mechanisms from Published Literature using Animals from our Laboratory
Body Weight: Upon sacrifice, whole body weight for male tumor-bearing mice (23.0 g) was 16.2% less than male control mice (27.3 g), while tumor-free final body weight of male tumor-bearing mice (20.9 g) was 23.5% less. Whole body weight for female tumor-bearing mice (21.5 g) was 103.2% of the female control mice (20.9 g), but when tumor mass was subtracted, female tumor-bearing mice (18.9 g) weighed 9% less than control female mice ( Fig. 4A). All male tumor-bearing final body weight values with and without tumor (actual, change from baseline, change from max weight, difference from control) were significantly different from male control animal values (p < 0.001) ( Table 3). When comparing female tumor-bearing final body weight values with tumor to female control animals (actual, change from baseline, change from max weight, difference from control), the only statistically significant value was the change from max body weight (control = -0.4 g, tumor = -1.4 g, p = 0.004). However, tumor-free final body weight represented as actual, change from baseline, change from max weight, and difference from control were all significantly different from female control animal values (p = 0.01 or p < 0.001) (Table 3).
Tissue Weights: Actual tumor weight was non-significantly different in female tumor-bearing mice compared to male tumor-bearing mice (Fig. 4B). Tumor weight expressed as a percent of tumor-free final body weight was 13.6% for female mice and 9.9% for male mice, and this was statistically significant (p < 0.05) ( Table   4). Heart weight of male tumor-bearing mice (105.5 mg) was significantly smaller than that of control male mice (132.9 mg) (p < 0.001) (Fig. 4C). Male tumor-bearing heart weight expressed as a percent of final body weight (0.46%) was significantly different from that of control males (0.49%) (p = 0.016), but when tumor weight was subtracted, no significant difference existed (control = 0.49%, tumor = 0.51%, p = 0.070) ( Table 4). Heart weight of female tumor-bearing mice (94.5 mg) was significantly smaller than that of control female mice (101.3 mg) (p < 0.001) (Fig. 4C). Female tumor-bearing heart weight expressed as a percent of final body weight (0.44%) was significantly different from that of control females (0.49%) (p < 0.001), but when tumor weight was subtracted, no significant difference existed (control = 0.49%, tumor = 0.50%, p = 0.581) ( Table 4). Gastrocnemius weight of male tumor-bearing mice (99.3 mg) was significantly smaller than that of control male mice (122.8 mg) (p < 0.001) (Fig.   4D). Male tumor-bearing gastrocnemius weight expressed as a percent of final body weight (0.43%) was not significantly different from that of control males (0.45%) (p = 0.115), but when tumor weight was subtracted, a significant difference existed with tumor-bearing animals having a larger percentage (control = 0.45%, tumor = 0.48%, p = 0.018) ( Table 4). Gastrocnemius weight of female tumor-bearing mice (72.3 mg) was significantly smaller than that of control female mice (86.9 mg) (p < 0.001) (Fig. 4D). Female tumor-bearing gastrocnemius weight expressed as a percent of final body weight (0.34%) was also significantly smaller than that of control females (0.42%) (p < 0.001). Likewise, when tumor weight was subtracted, a significant difference remained, with tumor-bearing animals having a smaller percentage (control = 0.42%, tumor = 0.38%, p = .018) ( Table 4).
Additional statistical differences in tibialis anterior and liver weights can be found in Table 4 and Fig. 4. Due to massively increased spleen weight and decreased adipose weight, the reporting methodology did not yield differences in statistical significance for these tissues (Table 4 and Fig. 4).

Discussion
In this study, we extracted data from all publications in the PubMed database using the C26 cachexia model from 1990-2022 to quantify prevalence and types of reporting for body and tissue weights. We followed this analysis with data from our own laboratory to examine differences in statistical outcomes using the reporting methodologies identified in our data extraction. We found variability in both what body and tissue weight data was reported, and the presentation of these data. We found differential effects on statistical significance using various reporting methodologies with our own data, particularly related to whole and tumor-free body weight and its use in the presentation of tissue weights. This is the first, to our knowledge, comprehensive analysis of body and tissue weight reporting mechanisms in the cachexia literature. Our findings have important implications for interpretation and replicability of preclinical cancer cachexia studies.
At present, the cancer cachexia literature is inconsistent in terms of both what data is reported and how that data is reported in preclinical studies. The present study reveals that 48.0% and 13.4% of studies did not report any type of baseline and final body weight of animals, respectively. These data points are imperative for confirming the existence and severity of cachexia in experimental animals. Fifty-two percent of publications reported actual tumor weight upon study conclusion. This is also an important finding to note, as lack of tumor weight reporting does not allow for the comparison of tumor burden across studies. Further, less than half of publications reported any form of adipose weight, only 71.5% of publications reported any form of skeletal muscle weight, and 10% of publications did not report any type of body or muscle weight data. Body weight loss and wasting of skeletal muscle and adipose tissue are cornerstones of cachexia pathology, and thus, thorough reporting of these measurements is needed for an accurate assessment of cachexia severity.
Beyond reporting body and tissue weights, much variability was observed regarding the manner in which this information was represented. Body weight was reported as grams, change from baseline (grams or as a percentage), or as the difference from tumor-free control (grams or as a percentage). Tissue weights were reported as weight in grams or milligrams, percent of final body weight (including or not including tumor), percent of baseline body weight, percent of tumor-free control tissue weight, or weight normalized to tibia length. Tumor weight was reported as actual or estimated tumor volume. Within these two categories, tumor weight was reported as weight in grams, percent of final body weight (including or not including tumor), or percent of baseline body weight. The present study using data from our laboratory demonstrates that these various mechanisms of data reporting can result in varied statistical significance within the same cohort of animals. For example, comparing our male tumor-bearing mice to male control mice resulted in statistical significance in all categories of final body weight including tumor (actual, change from baseline, change from max weight, difference from control), while data from female tumorbearing mice compared to female control mice only demonstrated statistical significance for final body weight including tumor when it was expressed as change from max weight. However, when tumor weight was removed from final weight, the comparison of both male and female tumor-bearing animals to controls displayed statistical significance in all tumor-free final body weight measures. Similar trends were discovered when exploring tissue weight measures. For example, actual gastrocnemius weight was statistically different when comparing tumor and control animals of both sexes but was no longer significant in male animals when expressed as a percent of final body weight. Thus, while the inclusion of these various reporting mechanisms and data presentations can be helpful in the context of individual studies, the selection of one over the other without foundational reporting (such as actual weight in grams or milligrams) may introduce selection bias, thus misleading readers, and future compounding work.
First developed nearly 50 years ago, the C26 model remains one of the most common and wellcharacterized models to study cachexia. Despite this, notable variability exists between studies related to study timeline, tumor size, and severity of body weight loss and tissue atrophy. Identifying contributing factors to this variability is beyond the scope of this study, but may include the source of C26 cells, cell culturing techniques, age and strain of animals, sex, tumor size, and additional procedures and handling of animals done over the duration of the study. It is challenging to evaluate and compare studies with different experimental timelines, variable tumor sizes, or that use mice at a variety of growth stages. To give context to this point, suppose that one study had an 18day experimental timeline with 3-gram tumors and a -4.0-gram change from baseline in tumor-free body weight, while another study had a 26-day experimental timeline with 1.5-gram tumors and -1.0-gram change from baseline in tumor-free body weight. Without the inclusion of this foundational weight data, the reader may incorrectly assume that the study with more experimental days would result in a greater tumor burden and increased weight loss. If actual body and tumor weights are not expressed in grams, the severity of cachexia in these two studies may be difficult to assess, especially in the context of variable ages and baseline weights.
It is also important to consider data representation in studies with less severe cachexia phenotypes, or earlystage studies, where weight changes are more nuanced. The female tumor-bearing animals from our laboratory used in this study, though not sacrificed at an early-stage cachexia timepoint, demonstrated milder wasting compared to their male tumor-bearing counterparts. While female tumor-bearing animals (21.5 g) weighed more on necropsy day than female control animals (20.9 g), their tumor-free final body weight (18.9 g) was 2 grams less than that of female controls (20.9 g). Thus, there are large differences in data presentation and statistical significance in this case when representing final body weight as whole (including tumor) or as tumor-free. Likewise, this then also impacts the presentation and statistical significance of tissue weights if they are reported as a percent of final body weight (whole or tumor-free). It is logical to assume that this is not only true of male vs. female mouse cancer cachexia data but would also be observed in early-stage studies where the cachexia phenotype is less severe.
The present study also revealed, somewhat unsurprisingly, that a sex bias still exists in preclinical cancer cachexia research. Most studies included male animals alone. Fourteen studies failed to report the sex of animals used. Of these 14 publications, one was from 1990-2000, two from 2001-2010, and 11 from 2011-2022. To our knowledge, only 5 studies performed within the dates of our literature search using the C26 model of cancer cachexia used both male and female animals. Of these studies, only 2 performed an actual head-to-head comparison (other works combined male and female animals for analysis). Sex differences in cancer cachexia susceptibility, mechanistic proregression, and outcomes are known to occur (Hetzler et al., 2015;Montalvo et al., 2018;Rosa-Caldwell et al., 2021;Zhong et al., 2022;Zhong & Zimmers, 2020). Thus, appropriate inclusion, analysis, and transparent reporting of sex in preclinical studies is imperative. Strain of mice and animal age upon study initiation were also found to vary amongst publications, potentially further limiting study comparison and translatability.
Apart from a few publications, strain of animals used was divided between CD2F1 (44.3%) and BALB/c (52.8%).
Of all publications, 13.4% did not include age of animals used. Of these 33 studies, 3 were from 1990-2000, 6 from 2001-2010, and 24 from 2011-2022. Age is an important piece of information, particularly as animals younger than 8 weeks are still growing (Fox, 2007;Jackson et al., 2017). In our study, including animals between 8 and 10 weeks of age at baseline, weight gain was seen throughout the course of the study evidenced by differences in baseline and maximum weight. Growth and weight gain are important considerations when studying mechanisms of weight loss and muscle wasting. Significant weight gain of control animals over the course of the study can further complicate interpretation of cachexia severity. For example, expressing weight loss in tumor-bearing animals as a percentage of control animals at study termination, baseline body weight, or maximum body weight may yield differential results and complicate data interpretation.
Not all preclinical cancer cachexia publications are the same (i.e., not every experiment is testing drug 'x's' ability to decrease skeletal muscle wasting), but consistent body and tissue weight data presentation across preclinical works is necessary. This is especially important because hallmark features of cachexia in both preclinical and clinical settings include loss of body weight and lean tissue. Accurate and thorough reporting of animal weight in the preclinical literature is crucial not only for interpretation of study results within the preclinical research community, but translatability to the human condition, which must always be the bedrock of preclinical work. It seems appropriate, then, to suggest all preclinical cancer cachexia publications (beyond just those that use the C26 cell line) clearly represent data that confirms if and to what degree the animals are cachectic. At a minimum, the reporting of a few fundamental data points would provide consistent information to give readership insight into the cachectic state of the animals in that study. We suggest that these include the following measurements of both the control and experimental groupsbaseline body weight in grams, final body weight in grams as both whole and tumor-free weight, actual tumor weight in grams weighed at time of necropsy, spleen weight in grams/milligrams, the weight of one adipose depot in grams/milligrams, and the weight of at least one skeletal muscle in grams/milligrams. Although normalized body and tissue weights can provide additional insight to readers, these representations should not replace reporting of actual weights, particularly due to challenges with statistical analysis of ratios (Allison et al., 1995). We also suggest, at a minimum, animal sex, strain, and age at time of inoculation, as well as number of study days completed prior to necropsy, be included. Studies utilizing cell lines that generate increased levels of IL-6 or other inflammatory markers should also consider including these measures.
This certainly is not an exhaustive list, but one that should serve as a reporting baseline. This information is important not only for the benefit of the reader but also for reporting consistency across the preclinical cancer cachexia research community. Betancourt et al. recently developed an animal cachexia score (ACASCO) that can be used to determine an animal's stage along the cachexia continuum and as a primary endpoint in preclinical cancer cachexia therapeutic research (Betancourt et al., 2019). Though not practical for all preclinical cancer cachexia studies, the use of the ACASCO should be considered for those studies testing experimental therapeutics.
This study was not without limitation. First, in terms of data extraction from publications, we included a reporting mechanism as "yes" if it was reported once. However, there were occasions when multiple experiments were included in one publication, and sometimes different results and reporting mechanisms were included in the different experiments. This was not captured in our analysis and thus our study may overrepresent data reported in individual experiments within publications. Second, our data extraction was limited to the C26 model of cachexia.
However, the C26 model has many similarities to other ectopic tumor models, thus making our recommendations relevant to a large proportion of the preclinical cachexia literature. Third, we did not include analysis of how well other factors such as age, sex, strain, cell line clone, single nucleotide polymorphisms (SNPs) in cell line, source of animals and cells, protocol for cell preparation, number of cells injected, mouse housing characteristics, inflammation, or other biomarkers impact disease progression, predict for cachexia severity, and/or increase variability between studies. Fourth, we do not propose "diagnostic" criteria or defining features of cachexia that should be evident in all preclinical models. Identifying predictive biomarkers for improving the clinical diagnosis of cachexia and developing models with improved clinical translatability are major topics of discussion and research in the cachexia field and are beyond the scope of the present work. The goal of this study was to assess variability in reporting of an important feature of cachexia (body and tissue wasting) and how this variability can affect data interpretation in the C26 model. We acknowledge that biochemical markers, muscle strength, and physical functioning, among other factors are indeed important components of assessing cachexia development and progression and should be included in publications whenever possible. Finally, the second PubMed data search took place in the month of May 2022, so any publications that were produced after this time were not included in our analysis.
The present study shows that body and tissue weight reporting and data representation in C26 preclinical cancer cachexia research is currently chaotic and diverse, making the interpretation and comparison of study outcomes difficult. Despite inherent limitations with C26 and other ectopic tumor-induced cachexia models, these models remain the most predominant and thoroughly characterized for the study of cachexia, further necessitating the standardization of data representation. The inclusion of key methodological details, as well as the transparent reporting of results, influences the work's meaningfulness and clinical translatability. With the omission of foundational details, readers may incorrectly interpret findings and take next steps that are not prudent. Even worse, discoveries may not be applicable to the human condition. This study highlights a need to develop guidelines for data reporting in preclinical cachexia literature to effectively compare outcomes between studies and increase clinical translatability.   (N = 12); data are mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 measured by two-tailed ttest compared to control.   Control (N = 13), Female Control (N = 13), Male Tumor (N = 14), Female Tumor (N = 12); data are mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 using two-tailed t-tests compared to control. BW = body weight.

Control Tumor Control Tumor
Tumor