The cachexia hematologic severity score (CHeSS): Use of a laboratory-based tool to predict survival and mutation-associated wasting in Acute Myeloid Leukemia and myelodysplastic syndrome Free

Introduction: Cachexia, marked by inflammation-driven losses in nutritional status, muscle mass, and fat tissue is known to lead to significant functional impairment and poor outcomes in solid tumor malignancies. In Acute Myeloid Leukemia (AML) and Myelodysplastic syndrome (MDS), however, cachexia remains under-recognized, with no consensus guidelines on utilizing it for disease diagnosis, management, or prognosis. This is due to typical definitions requiring expensive and time-consuming CT or MRI-based assessments of skeletal and adipose tissue. In our study, we aimed to create a more practical, scalable method to diagnose cachexia using routine laboratory markers that reflect systemic catabolism and malnutrition and rates of weight loss over time.

Methods: We retrospectively analyzed 203 adult patients (≥18 years old) treated for myelodysplastic syndrome (MDS), primary (pAML), and secondary AML (sAML), at Karmanos Cancer Institute from 2013 to 2024. To classify cachexia, collected total protein, albumin, creatine, weight, and BMI at time of diagnosis, and one, three, and six-months post-diagnosis. Weight and BMI were combined to calculate weight loss grading scale (WLGS), a validated malnutrition tool that compares BMI and weight change across specific time points, and weight loss velocity (WLV).

These parameters were combined to obtain a Cachexia Hematologic Severity Score (CHeSS). Using established survival-based cutoffs, we assigned our laboratory biomarkers and WLV a score of 0-2 and WLGS a score of 0-4. Greater weight was placed on the WLGS, given its previous validation in solid tumor malignancies. Cachexia severity was assessed using a composite hematologic cachexia scoring index and categorized into three strata: none-cachectic (score 0–3), pre-cachectic (4–7), and cachectic (≥8). Primary endpoints included overall survival (OS) using Kaplan–Meier methodology and compared between groups via log-rank testing. Mutation associations (e.g., TP53, TET2) with cachexia severity were assessed using Fisher's exact test to compare distribution across cachexia groups.

Results: 103 (51%) had MDS and 100 (49%) had AML (including pAML and sAML). The median age at diagnosis was 66 yrs for MDS and 65 yrs for AML. 64 (62%) MDS and 63 (63%) AML patients were male. Amongst MDS patients, 66 (64%), 32 (31%), and six (6%) met criteria for non-cachectic, pre-cachectic and cachectic at one-to-six-months post-diagnosis, respectively. OS greater than 12 months was sequentially lower in non-cachectic, pre-cachectic, and cachectic patients at 76%, 64%, and 50% (OR=0.56). Median OS was 102 and 89 months for non-cachectic and pre-cachectic patients, compared to 13 months for cachectic patients (Log Rank Test; p-value = 0.065, NS).

The most common mutations at diagnosis in MDS patients who met pre-cachectic and/or cachectic criteria were CUX1 (6, 16%), DNMT3A (5, 13%), TP53 (5, 13%), SRSF2 (4, 11%), TET2 (4, 11%). There was a statistically significant association with progression to pre-cachectic/cachectic status from baseline with an ASXL1 mutation at diagnosis (chi-square; p = 0.002). In AML patients who progressed to pre-cachectic or cachectic status, TP53 (9, 26%), BCOR (7, 21%), DNMT3A (7, 21%), and TET2 (6, 18%) were the most common. No individual mutations reached statistical significance; however, TP53 showed higher rates of pre-cachexia and cachexia progression from time of diagnosis.

Conclusion: Our study determined that progression to pre-cachectic or cachectic status was associated with reduced survival outcomes in both AML and MDS. While not significant in the latter, this still reflects a large, clinically impactful survival interval difference. Furthermore, mutations such as ASXL1, DNMT3A, TP53, and RUNX1, have long been associated with in vitro cellular inflammation and clinically categorized as high risk. Our analysis reveals a potential explanation for this through development of cachexia.

With a readily calculable cachexia scoring system, our aim is to identify at-risk patients early in their disease course. This provides clinicians with opportunities for early intervention, through supportive care services, nutritionists, and exercise therapy programs. Ongoing work is still needed to compare CHeSS with current gold-standard prognostic scores and investigate the metabolomic impact of high-risk mutations in promoting cachexia development.