Skeletal Muscle Fibrosis in Pancreatic Cancer Patients with Respect to Survival

Abstract

Background: Cancer cachexia is a catabolic condition characterized by skeletal muscle wasting, consequent to tumor burden, which negatively impacts tolerance to cancer therapies and contributes to increased mortality. Partly because of the limited knowledge of the underlying mechanisms of cancer cachexia derived from human studies, however, the ability to therapeutically intervene remains elusive. The purpose of the current study was therefore to better define the phenotype of skeletal muscle obtained from patients with pancreatic ductal adenocarcinoma (PDAC), which has one of the highest rates of cachexia.

Methods: Morphological analyses were performed on rectus abdominis muscle biopsies obtained from resectable PDAC patients undergoing tumor resection surgery (N = 20) and from weight-stable non-cancer control subjects undergoing benign abdominal surgery (N = 16). PDAC patients with a body weight loss of greater than 5% during the previous 6 months were considered cachectic (N = 15). Statistical tests were two sided.

Results: Skeletal muscle from cachectic PDAC patients had increased collagen content compared with non-cancer control subjects (1.43% vs 9.66%, P = .0004, Dunn test). Across all PDAC patients, collagen content positively correlated with body weight loss (P = .0016, r = 0.672), was increased in patients with lymph node metastasis (P = .007, Mann-Whitney U test), and was associated with survival on univariate (HR = 1.08, 95% confidence interval [CI] = 1.02 to 1.04, P = .008) and multivariable analyses (HR = 1.08, 95% CI = 1.00 to 1.17, P = .038). Cachectic PDAC patients also displayed increased lipid deposition (2.63% vs 5.72%, P = .042), infiltration of CD68+ macrophages (63.6 cells/mm² vs 233.8 cells/mm², P = .0238), calcium deposition (0.21% vs 2.51%, P = .030), and evidence of deficient cellular quality control mechanisms (Mann-Whitney U test). Transcriptional profiling of all patients supported these findings by identifying gene clusters related to wounding, inflammation, and cellular response to TGF-β upregulated in cachectic PDAC patients compared with non-cancer control subjects.

Conclusions: To our knowledge, this work is the first to demonstrate increased collagen content in cachectic PDAC patients that is associated with poor survival.

Figure 1.

Morphological characteristics of skeletal muscle from non-cancer control patients and PDAC patients. Representative hematoxylin and eosin (H&E) stained sections from the rectus abdominis muscle of weight stable non-cancer control patients (A) and PDAC patients (B, C) with varying degrees of BW loss (indicated as a percentage) within the 6 months prior to surgery. PDAC patients are further stratified based on survival time of more than 1 year post-surgery (B) vs less than 1 year post surgery (C). Morphological features consistent with skeletal muscle pathology are observed in muscle from PDAC patients, including mononuclear cell infiltration, muscle fiber fragmentation (white arrowheads), muscle fibers with centralized nuclei (white arrows), and connective tissue deposition (*). Images are representative of n = 16 non-cancer control subjects, n = 5 non-cachectic PDAC patients, and n = 15 cachectic PDAC patients. M = male; F = female; neo = received neoadjuvant therapy prior to tumor resection surgery; naïve = naïve to neoadjuvant therapy; PDAC = pancreatic ductal adenocarcinoma. Scale bar = 100 µm.

Figure 2.

Skeletal muscle collagen content and its relationship with body weight loss in PDAC patients. Representative Masson’s Trichrome stained sections (collagen stains blue) from the rectus abdominis muscle of weight stable non-cancer control patients (A) and PDAC patients (B, C) with varying degrees of BW loss (indicated as a percentage) within the 6 months prior to surgery. PDAC patients are further stratified based on survival time of more than 1 year post surgery (B), vs less than 1 year post surgery (C). Magnified images demonstrate increased staining of collagen (fibrosis) in the endomysium surrounding individual muscle fibers in cachectic PDAC patients. D) The percentage of total muscle area (per 5x field) occupied by collagen was quantified in n = 16 non-cancer control subjects, n = 4 non-cachectic PDAC patients (<5% BW loss) and n = 15 cachectic PDAC patients (>5% BW loss) and is expressed as the mean ± SEM (*P = .0004, Dunn test). E) Relationship between collagen content (% area) and BW loss (%) in PDAC patients (Spearman’s correlation, P = .0016, r = 0.672). BW = body weight; PDAC = pancreatic ductal adenocarcinoma.

Figure 3.

Relationship between muscle collagen content, lymph node metastasis and survival in PDAC patients. A) The percent of total muscle area occupied by collagen in PDAC patients stratified by N stage (N0 vs N1), expressed as mean ± SEM. PDAC patients with positive lymph node metastasis (N1, n = 12) vs PDAC patients with no lymph node metastasis (N0, n = 7) (P = .0099, Mann-Whitney U test). B, C) Kaplan-Meier analyses on PDAC patients dichotomized by either N stage (B, N0 vs N1) or collagen content (C, <10%>). D) The percent of total muscle area occupied by collagen in PDAC patients surviving >1 year post-surgery vs PDAC patients surviving <1 year post-surgery (P = .0040, *Mann-Whitney U test). *Two patients still alive at 10 months and 11 months postsurgery, who had not reached the 1-year mark postsurgery and who had 6.1% and 2.0% collagen content, respectively, were excluded from the analysis. E) Univariate analysis was performed using a Cox proportional hazard model. All variables statistically significant (P < .05) on univariate analyses were included in multivariable Cox regression.

Figure 4.

Lipid content in skeletal muscle of non-cachectic and cachectic PDAC patients versus non-cancer control patients. A, B) Representative serial sections from the rectus abdominis muscle of non-cancer control patients and PDAC patients with varying levels of BW loss (indicated as a percentage) stained with Oil Red O, which stains lipid orange, or Masson’s Trichrome, which stains collagen blue. Scale Bar = 200 µm. C) The percent of total muscle area occupied by fat in non-cancer control patients (n = 16), non-cachectic PDAC patients (n = 4) and cachectic PDAC patients (n = 15), expressed as mean ± SEM (*P = .0418, Dunn’s test). D) Representative skeletal muscle sections from a non-cancer control and cachectic PDAC patients immunostained with antibodies against platelet-derived growth factor receptor alpha (to label fibroadipogenic progenitor [FAP] cells, green) and laminin (to label basement membranes, red), and counterstained with DAPI (to label cell nuclei, blue). Scale bar = 50 µm. White arrows = FAP cells. BW = body weight; M = male, F = female, neo = neoadjuvant therapy; naïve = naïve to neoadjuvant therapy; PDAC = pancreatic ductal adenocarcinoma.

Figure 5.

Figure 5. Ultrastructural damage, calcifications and macrophages present in skeletal muscle from cachectic PDAC patients. A–C) Representative electron micrographs of skeletal muscle from non-cancer control patients (n = 2) and cachectic PDAC patients (n = 2) following transmission electron microscopy. Areas of ECM deposition are indicated by black arrows. D–K) Representative skeletal muscle sections from non-cancer control patients (D, H) and cachectic PDAC patients (E–G, I–K) stained with Alizarin Red S to label calcium deposition (stains red). Calcium deposits localized inside (*) and at the periphery (white arrows) of muscle fibers, in the extracellular matrix (white arrowheads) and in blood vessels walls (v) are noted. Scale bar: 50 µm. L) The percent of total muscle area positive for calcium deposition in non-cancer controls (n = 3) versus cachectic PDAC patients (n = 6), expressed as the mean ± the SEM (P = .0238, Mann-Whitney U test). M–O) Representative skeletal muscle sections from a non-cancer control patient (M) and cachectic PDAC patients (M, O) immunostained with a CD68 antibody to label macrophages (brown staining, black arrows). Scale bar: 200 µm. P) Staining of a serial muscle section with H&E to demonstrate the localization of CD68+ macrophages in cachectic muscle. Areas of collagen (light pink staining, white arrows), lipid (white arrowheads) and muscle fibers (black astericks) are indicated. Q) The average number of CD68+ macrophages in muscle of cachectic PDAC patients compared to non-cancer control subjects (P = .0303, Mann-Whitney U test). Data represent mean ± SEM, from n = 5 non-cancer controls and n = 7 cachectic PDAC patients. CON = controls; PDAC = pancreatic ductal adenocarcinoma.

Figure 6.

Accumulation of protein aggregates, giant vesicles in skeletal muscle from cachectic PDAC patients. A–F) Representative electron micrographs of skeletal muscle from non-cancer control subjects (A, D) and cachectic PDAC patients (B, C, E, F). The morphology of these vesicles was consistent with that of autolysosomes containing lipofuscin granules (ie, aggregates of incompletely degraded proteins and lipids). Many of the light-density lipid inclusions (white arrowheads) present in the vesicles were partially or entirely replaced by darkly stained calcifications (long white arrows), which can be observed more clearly in magnified images. White scale bars = 500 nm. G–R) Representative muscle sections from non-cancer control subjects or cachectic PDAC patients stained with antibodies against UBIQUITIN (G–I), p62 (J–L), LAMP1 (M–O), or LC3 (P–R) (red staining) plus antibodies against either dystrophin or laminin to label fiber membranes (green staining) and counterstained with DAPI to label nuclei (blue). Scale bar = 100 µm. Magnified images depict the size of p62+ aggregates and LAMP1+ lysosomes in muscle of cachectic PDAC patients, their localization near the myofiber membrane, and their proximity to myonuclei (white arrows). S–Y) The average staining intensity of ubiquitin and the average size and number of p62+, LAMP1+, or LC3+ puncta in non-cancer control subjects versus cachectic PDAC patients. All data represent the mean ± SEM, from n = 4-5 non-cancer control patients and n = 5-6 cachectic PDAC patients. *P < .05, Mann-Whitney U test. PDAC = pancreatic ductal adenocarcinoma.

Figure 7.

Gene expression of pro-fibrotic factors in skeletal muscle from control and cachectic PDAC patients. Rectus abdominis muscle biopsies from non-cancer control subjects and cachectic PDAC patients were processed for qRT-PCR analyses to validate select genes identified via microarray as increased in cachectic PDAC patients. Compared with non-cancer control subjects, cachectic PDAC patients showed increased gene expression of thrombospondin-1 (THBS1/TSP-1, 6.9-fold, P = .0491) (A), TGFBR2 (1.9-fold, P = .0378) (B), TGFBR3 (2.0-fold, P = .0116) (C), connective tissue growth factor (CTGF, 4.3-fold, P = .0134) (D), and plasminogen activator inhibitor-1 (PAI-1/SERPINE1, 4.9-fold, P = .0062) (E, F) The housekeeping gene, 18S, was not statistically significantly different between non-cancer control subjects and cachectic PDAC patients. All samples were run in triplicate on the same PCR plate and were normalized to the mean of the non-cancer control group. Data are expressed as mean ± SEM, from n = 8-10 non-cancer control subjects and n = 12-13 cachectic PDAC patients. *P < .05, **P < .01, Mann-Whitney U test. PDAC = pancreatic ductal adenocarcinoma; qRT-PCR = quantitative reverse transcriptase polymerase chain reaction.