Somatic Mutational Analysis in Endoscopic Ultrasound-Guided Biopsy of Pancreatic Adenocarcinoma: Assessing Yield and Impact.

INTRODUCTION: We sought to determine the yield of somatic mutational analysis from endoscopic ultrasound (EUS)-guided biopsies of pancreatic adenocarcinoma compared with that of surgical resection and to assess the impact of these results on oncologic treatment. METHODS: We determined the yield of EUS sampling and surgical resection. We evaluated the potential impact of mutational analysis by identifying actionable mutations and its direct impact by reviewing actual treatment decisions. RESULTS: Yield of EUS sampling was 89.5%, comparable with the 95.8% yield of surgical resection. More than a quarter in the EUS cohort carried actionable mutations, and of these, more than 1 in 6 had treatment impacted by mutational analysis. DISCUSSION: EUS sampling is nearly always adequate for somatic testing and may have substantial potential and real impact on treatment decisions.

Chemotherapy and Immune Checkpoint Blockade for Gastric and Gastroesophageal Junction Adenocarcinoma.

Importance: Combining immune checkpoint blockade (ICB) with chemotherapy improves outcomes in patients with metastatic gastric and gastroesophageal junction (G/GEJ) adenocarcinoma; however, whether this combination has activity in the perioperative setting remains unknown. Objective: To evaluate the safety and preliminary activity of perioperative chemotherapy and ICB followed by maintenance ICB in resectable G/GEJ adenocarcinoma. Design, Setting, and Participants: This investigator-initiated, multicenter, open-label, single-stage, phase 2 nonrandomized controlled trial screened 49 patients and enrolled 36 patients with resectable G/GEJ adenocarcinoma from February 10, 2017, to June 17, 2021, with a median (range) follow-up of 35.2 (17.4-73.0) months. Thirty-four patients were deemed evaluable for efficacy analysis, with 28 (82.4%) undergoing curative resection. This study was performed at 4 referral institutions in the US. Interventions: Patients received 3 cycles of capecitabine, 625 mg/m2, orally twice daily for 21 days; oxaliplatin, 130 mg/m2, intravenously and pembrolizumab, 200 mg, intravenously with optional epirubicin, 50 mg/m2, every 3 weeks before and after surgery with an additional cycle of pembrolizumab before surgery. Patients received 14 additional doses of maintenance pembrolizumab. Main Outcomes and Measures: The primary end point was pathologic complete response (pCR) rate. Secondary end points included overall response rate, disease-free survival (DFS), overall survival (OS), and safety. Results: A total of 34 patients (median [range] age, 65.5 [25-90] years; 23 [67.6%] male) were evaluable for efficacy. Of these patients, 28 (82.4%) underwent curative resection, 7 (20.6%; 95% CI, 10.1%-100%) achieved pCR, and 6 (17.6%) achieved a pathologic near-complete response. Of the 28 patients who underwent resection, 4 (14.3%) experienced disease recurrence. The median DFS and OS were not reached. The 2-year DFS was 67.8% (95% CI, 0.53%-0.87%) and the OS was 80.6% (95% CI, 0.68%-0.96%). Treatment-related grade 3 or higher adverse events for evaluable patients occurred in 20 patients (57.1%), and 12 (34.3%) experienced immune-related grade 3 or higher adverse events. Conclusion and Relevance: In this trial of unselected patients with resectable G/GEJ adenocarcinoma, capecitabine, oxaliplatin, and pembrolizumab resulted in a pCR rate of 20.6% and was well tolerated. This trial met its primary end point and supports the development of checkpoint inhibition in combination with perioperative chemotherapy in locally advanced G/GEJ adenocarcinoma. Trial Registration: ClinicalTrials.gov Identifier: NCT02918162.

Corrigendum: Tumor infiltrating T cell states and checkpoint inhibitor expression in hepatic and pancreatic malignancies.

[This corrects the article DOI: 10.3389/fimmu.2023.1067352.].

Grain, Gluten, and Dietary Fiber Intake Influence Gut Microbial Diversity: Data from the Food and Microbiome Longitudinal Investigation.

Although short-term feeding studies demonstrated effects of grains, fiber, and gluten on gut microbiome composition, the impact of habitual intake of these dietary factors is poorly understood. We examined whether habitual intakes of whole and refined grains, fiber, and gluten are associated with gut microbiota in a cross-sectional study. This study included 779 participants from the multi-ethnic Food and Microbiome Longitudinal Investigation study. Bacterial 16SV4 rRNA gene from baseline stool was amplified and sequenced using Illumina MiSeq. Read clustering and taxonomic assignment was performed using QIIME2. Usual dietary intake was assessed by a 137-item food frequency questionnaire. Association of diet with gut microbiota was assessed with respect to overall composition and specific taxon abundances. Whole grain intake was associated with overall composition, as measured by the Jensen-Shannon divergence (multivariable-adjusted P trend for quartiles = 0.03). The highest intake quartile was associated with higher abundance of Bacteroides plebeius, Faecalibacterium prausnitzii, Blautia producta, and Erysipelotrichaceae and lower abundance of Bacteroides uniformis. These bacteria also varied by dietary fiber intake. Higher refined grain and gluten intake was associated with lower Shannon diversity (P trend < 0.05). These findings suggest that whole grain and dietary fiber are associated with overall gut microbiome structure, largely fiber-fermenting microbiota. Higher refined grain and gluten intakes may be associated with lower microbial diversity. Significance: Regular consumption of whole grains and dietary fiber was associated with greater abundance of gut bacteria that may lower risk of colorectal cancer. Further research on the association of refined grains and gluten with gut microbial composition is needed to understand their roles in health and disease.

Single-cell RNA sequencing reveals the effects of chemotherapy on human pancreatic adenocarcinoma and its tumor microenvironment.

The tumor microenvironment (TME) in pancreatic ductal adenocarcinoma (PDAC) is a complex ecosystem that drives tumor progression; however, in-depth single cell characterization of the PDAC TME and its role in response to therapy is lacking. Here, we perform single-cell RNA sequencing on freshly collected human PDAC samples either before or after chemotherapy. Overall, we find a heterogeneous mixture of basal and classical cancer cell subtypes, along with distinct cancer-associated fibroblast and macrophage subpopulations. Strikingly, classical and basal-like cancer cells exhibit similar transcriptional responses to chemotherapy and do not demonstrate a shift towards a basal-like transcriptional program among treated samples. We observe decreased ligand-receptor interactions in treated samples, particularly between TIGIT on CD8 + T cells and its receptor on cancer cells, and identify TIGIT as the major inhibitory checkpoint molecule of CD8 + T cells. Our results suggest that chemotherapy profoundly impacts the PDAC TME and may promote resistance to immunotherapy.

Tumor infiltrating T cell states and checkpoint inhibitor expression in hepatic and pancreatic malignancies.

Hepato-pancreatico-biliary (HPB) malignancies are difficult-to-treat and continue to to have a high mortality and significant therapeutic resistance to standard therapies. Immune oncology (IO) therapies have demonstrated efficacy in several solid malignancies when combined with chemotherapy, whereas response rates in pancreatic ductal adenocarcinoma (PDA) are poor. While promising in hepatocellular carcinoma (HCC) and cholangiocarcinoma (CCA), there remains an unmet need to fully leverage IO therapies to treat HPB tumors. We therefore defined T cell subsets in the tumor microenvironment of HPB patients utilizing a novel, multiparameter flow cytometry and bioinformatics analysis. Our findings quantify the T cell phenotypic states in relation to checkpoint receptor expression. We demonstrate the presence of CD103+ tissue resident memory T cells (TRM), CCR7+ central memory T cells, and CD57+ terminally differentiated effector cells across all HPB cancers, while the anti-tumor function was dampened by expression of multiple co-inhibitory checkpoint receptors. Terminally exhausted T cells lacking co-stimulatory receptors were more prevalent in PDA, whereas partially exhausted T cells expressing both co-inhibitory and co-stimulatory receptors were most prevalent in HCC, especially in early stage. HCC patients had significantly higher TRM with a phenotype that could confer restored activation in response to immune checkpoint therapies. Further, we found a lack of robust alteration in T cell activation state or checkpoint expression in response to chemotherapy in PDA patients. These results support that HCC patients might benefit most from combined checkpoint therapies, whereas efforts other than cytotoxic chemotherapy will likely be necessary to increase overall T cell activation in CCA and PDA for future clinical development.

Impact of comprehensive family history and genetic analysis in the multidisciplinary pancreatic tumor clinic setting.

BACKGROUND: Genetic testing is recommended for all pancreatic ductal adenocarcinoma (PDAC) patients. Prior research demonstrates that multidisciplinary pancreatic cancer clinics (MDPCs) improve treatment- and survival-related outcomes for PDAC patients. However, limited information exists regarding the utility of integrated genetics in the MDPC setting. We hypothesized that incorporating genetics in an MDPC serving both PDAC patients and high-risk individuals (HRI) could: (1) improve compliance with guideline-based genetic testing for PDAC patients, and (2) optimize HRI identification and PDAC surveillance participation to improve early detection and survival. METHODS: Demographics, genetic testing results, and pedigrees were reviewed for PDAC patients and HRI at one institution over 45 months. Genetic testing analyzed 16 PDAC-associated genes at minimum. RESULTS: Overall, 969 MDPC subjects were evaluated during the study period; another 56 PDAC patients were seen outside the MDPC. Among 425 MDPC PDAC patients, 333 (78.4%) completed genetic testing; 29 (8.7%) carried a PDAC-related pathogenic germline variant (PGV). Additionally, 32 (9.6%) met familial pancreatic cancer (FPC) criteria. These PDAC patients had 191 relatives eligible for surveillance or genetic testing. Only 2/56 (3.6%) non-MDPC PDAC patients completed genetic testing (p < 0.01). Among 544 HRI, 253 (46.5%) had a known PGV or a designation of FPC, and were eligible for surveillance at baseline; of the remainder, 15/291 (5.2%) were eligible following genetic testing and PGV identification. CONCLUSION: Integrating genetics into the multidisciplinary setting significantly improved genetic testing compliance by reducing logistical barriers for PDAC patients, and clarified cancer risks for their relatives while conserving clinical resources. Overall, we identified 206 individuals newly eligible for surveillance or genetic testing (191 relatives of MDPC PDAC patients, and 15 HRI from this cohort), enabling continuity of care for PDAC patients and at-risk relatives in one clinic.

Elevated dietary carbohydrate and glycemic intake associate with an altered oral microbial ecosystem in two large U.S. cohorts.

The human oral microbiome is associated with chronic diseases including cancer. However, our understanding of its relationship with diet is limited. We assessed the associations between carbohydrate and glycemic index (GI) with oral microbiome composition in 834 non-diabetic subjects from the NCI-PLCO and ACS-CPSII cohorts. The oral microbiome was characterized using 16Sv3-4 rRNA-sequencing from oral mouthwash samples. Daily carbohydrate and GI were assessed from food frequency questionnaires. We used linear regression, permutational MANOVA, and negative binomial Generalized Linear Models (GLM) to test associations of diet with α- and ß-diversity and taxon abundance (adjusting for age, sex, cohort, BMI, smoking, caloric intake, and alcohol). A q-value (FDR-adjusted P-value) of <0.05 was considered significant. Oral bacterial α-diversity trended higher in participants in the highest quintiles of carbohydrate intake, with marginally increased richness and Shannon diversity (p-trend=0.06 and 0.07). Greater carbohydrate intake was associated with greater abundance of class Fusobacteriia (q=0.02) and genus Leptotrichia (q=0.01) and with lesser abundance of an Actinomyces OTU (q=4.7E-04). Higher GI was significantly related to greater abundance of genus Gemella (q=0.001). This large, nationwide study provides evidence that diets high in carbohydrates and GI may influence the oral microbiome.

Multiplexed single-cell analysis reveals prognostic and nonprognostic T cell types in human colorectal cancer.

Clinical outcomes in colorectal cancer (CRC) correlate with T cell infiltrates, but the specific contributions of heterogenous T cell types remain unclear. To investigate the diverse function of T cells in CRC, we profiled 37,931 T cells from tumors and adjacent normal colon of 16 patients with CRC with respect to transcriptome, TCR sequence, and cell surface markers. Our analysis identified phenotypically and functionally distinguishable effector T cell types. We employed single-cell gene signatures from these T cell subsets to query the TCGA database to assess their prognostic significance. We found 2 distinct cytotoxic T cell types. GZMK+KLRG1+ cytotoxic T cells were enriched in CRC patients with good outcomes. GNLY+CD103+ cytotoxic T cells with a dysfunctional phenotype were not associated with good outcomes, despite coexpression of CD39 and CD103, markers that denote tumor reactivity. We found 2 distinct Treg subtypes associated with opposite outcomes. While total Tregs were associated with good outcomes, CD38+ Tregs were associated with bad outcomes independently of stage and possessed a highly suppressive phenotype, suggesting that they inhibit antitumor immunity in CRC. These findings highlight the potential utility of these subpopulations in predicting outcomes and support the potential for novel therapies directed at CD38+ Tregs or CD8+CD103+ T cells.

Motixafortide and Pembrolizumab Combined to Nanoliposomal Irinotecan, Fluorouracil, and Folinic Acid in Metastatic Pancreatic Cancer: The COMBAT/KEYNOTE-202 Trial.

PURPOSE: Pancreatic ductal adenocarcinoma (PDAC) is largely unresponsive to checkpoint inhibitors. Blockade of the CXCR4/CXCL12 axis increases intratumoral trafficking of activated T cells while restraining immunosuppressive elements. This study evaluates dual blockade of CXCR4 and PD1 with chemotherapy in PDAC. PATIENTS AND METHODS: Multicenter, single-arm, phase II study to evaluate the safety and efficacy of motixafortide and pembrolizumab combined with chemotherapy in patients with de novo metastatic PDAC and disease progression on front-line gemcitabine-based therapy (NCT02826486). Subjects received a priming phase of motixafortide daily on days 1-5, followed by repeated cycles of motixafortide twice a week; pembrolizumab every 3 weeks; and nanoliposomal irinotecan, fluorouracil, and leucovorin every 2 weeks (NAPOLI-1 regimen). The primary objective was objective response rate (ORR). Secondary objectives included overall survival (OS), progression-free survival (PFS), disease control rate (DCR), safety, and tolerability. RESULTS: A total of 43 patients were enrolled. The ORR according to RECISTv1.1 was 21.1% with confirmed ORR of 13.2%. The DCR was 63.2% with median duration of clinical benefit of 5.7 months. In the intention-to-treat population, median PFS was 3.8 months and median OS was 6.6 months. The triple combination was safe and well tolerated, with toxicity comparable with the NAPOLI-1 regimen. Notably, the incidence of grade 3 or higher neutropenia and infection was 7%, lower than expected for this chemotherapy regimen. CONCLUSIONS: Triple combination of motixafortide, pembrolizumab, and chemotherapy was safe and well tolerated, and showed signs of efficacy in a population with poor prognosis and aggressive disease.

Cancer surveillance awareness and practice among families at increased risk for pancreatic adenocarcinoma.

BACKGROUND: Early detection of pancreatic ductal adenocarcinoma (PDAC) is an important goal for improving survival. Individuals who meet published guidelines for surveillance may be underidentified, and family communication about risk represents a pathway to increasing participation in surveillance. We investigated the uptake of and barriers to surveillance in at-risk relatives of clinic patients. METHODS: We conducted a retrospective record review of patients with personal or family history of PDAC evaluated over 12 months. The first relative presenting to clinic (proband) reported surveillance status and reasons for nonparticipation for at-risk relatives. Descriptive analyses and Fisher's exact tests were conducted to evaluate differences in surveillance participation. RESULTS: Among 193 at-risk relatives, 21% were in surveillance. The primary reasons for nonparticipation were lack of awareness (36%) and lack of interest (24%). Neither the sex nor the cancer status of probands impacted surveillance. At-risk relatives with familial pancreatic cancer (FPC) who also carried relevant pathogenic germline variants (PGVs) were more likely to undergo surveillance than those with FPC or PGVs alone (P = .003). Among families with PGVs, 59% of relatives potentially eligible for surveillance had not completed genetic testing. CONCLUSION: PDAC surveillance is underutilized in high-risk families. Communication interventions to address informational needs and decisional support could improve outcomes.

Moving Beyond the Momentum: Innovative Approaches to Clinical Trial Implementation.

Despite efforts to enhance enrollment and the merger of national cooperative groups, < 5% of patients with cancer will enroll into a clinical trial. Additionally, clinical trials are affected by a lack of diversity inclusive of minority patients, rural residents, or low-income individuals. COVID-19 further exacerbated known barriers of reduced physician-patient interaction, physician availability, trial activation and enrollment, financial resources, and capacity for conducting research. Based on the cumulative insight of academic and community clinical researchers, we have created a white paper identifying existing challenges in clinical trial conduct and have provided specific recommendations of sustainable modifications to improve efficiency in the activation and conduct of clinical trials with an overarching goal of providing improved access and care to our patients with cancer.

Randomized Phase III Trial of Pegvorhyaluronidase Alfa With Nab-Paclitaxel Plus Gemcitabine for Patients With Hyaluronan-High Metastatic Pancreatic Adenocarcinoma.

PURPOSE: To evaluate the efficacy and safety of pegvorhyaluronidase alfa (PEGPH20) plus nab-paclitaxel/gemcitabine (AG) in patients with hyaluronan-high metastatic pancreatic ductal adenocarcinoma (PDA). PATIENTS AND METHODS: HALO 109-301 was a phase III, randomized, double-blind, placebo-controlled study. Patients ≥ 18 years of age with untreated, metastatic, hyaluronan-high PDA were randomly assigned 2:1 to PEGPH20 plus AG or placebo plus AG. Treatment was administered intravenously in 4-week cycles (3 weeks on, 1 week off) until progression or intolerable adverse events: PEGPH20 3.0 µg/kg twice per week for cycle 1 and once per week thereafter; nab-paclitaxel 125 mg/m2 once per week; and gemcitabine 1,000 mg/m2 once per week. The primary end point was overall survival (OS); secondary end points included progression-free survival (PFS), objective response rate (ORR), and safety. Response was independently assessed per RECIST v1.1. RESULTS: At data cutoff, 494 patients were randomly assigned, with 492 (327 for PEGPH20 and 165 for placebo) included in intention-to-treat analyses. Baseline characteristics were balanced for PEGPH20 plus AG versus placebo plus AG. There were 330 deaths, with a median OS of 11.2 months for PEGPH20 plus AG versus 11.5 months for placebo plus AG (hazard ratio [HR], 1.00; 95% CI, 0.80 to 1.27; P = .97); median PFS was 7.1 months versus 7.1 months (HR, 0.97 [95% CI, 0.75 to 1.26]); ORR was 47% versus 36% (ORR ratio, 1.29 [95% CI, 1.03 to 1.63]). Grade ≥ 3 adverse events with a ≥ 2% higher rate with PEGPH20 plus AG than with placebo plus AG included fatigue (16.0% v 9.6%), muscle spasms (6.5% v 0.6%), and hyponatremia (8.0% v 3.8%). CONCLUSION: The addition of PEGPH20 to AG increased the ORR but did not improve OS or PFS. The safety profile of PEGPH20 plus AG was consistent with that found in previous studies. These results do not support additional development of PEGPH20 in metastatic PDA.

Noninvasive Young's modulus visualization of fibrosis progression and delineation of pancreatic ductal adenocarcinoma (PDAC) tumors using Harmonic Motion Elastography (HME) in vivo.

Background and aims: Poor specificity and predictive values of current cross-sectional radiological imaging methods in evaluation of pancreatic adenocarcinoma (PDAC) limit the clinical capability to accurately stage the tumor pre-operatively and provide optimal surgical treatment and improve patient outcomes. Methods: In this study, we applied Harmonic Motion Elastography (HME), a quantitative ultrasound-based imaging method to calculate Young's modulus (YM) in PDAC mouse models (n = 30) and human pancreatic resection specimens of PDAC (n=32). We compared the YM to the collagen assessment by Picrosirius red (PSR) stain on corresponding histologic sections. Results: HME is capable of differentiating between different levels of fibrosis in transgenic mice. In mice without pancreatic fibrosis, the measured YM was 4.2 ± 1.3 kPa, in fibrotic murine pancreata, YM was 5.5 ± 2.0 kPa and in murine PDAC tumors, YM was 11.3 ± 1.7 kPa. The corresponding PSR values were 2.0 ± 0.8 %, 9.8 ± 3.4 %, and 13.2 ± 1.2%, respectively. In addition, three regions within each human surgical PDAC specimen were assessed: tumor, which had both the highest Young's modulus (YM > 40 kPa) and collagen density (PSR > 40 %); non-neoplastic adjacent pancreas, which had the lowest Young's modulus (YM < 15 kPa) and collagen density (PSR < 10%) and a transitional peri-lesional region between the tumor and non-neoplastic pancreas with an intermediate value of measured Young's modulus (15 kPa < YM < 40 kPa) and collagen density (15% < PSR < 35 %). Conclusion: In conclusion, a non-invasive, quantitative imaging tool for detecting, staging and delineating PDAC tumor margins based on the change in collagen density was developed.

Harmonic Motion Imaging of Pancreatic Tumor Stiffness Indicates Disease State and Treatment Response.

PURPOSE: Pancreatic ductal adenocarcinoma (PDA) is a common, deadly cancer that is challenging both to diagnose and to manage. Its hallmark is an expansive, desmoplastic stroma characterized by high mechanical stiffness. In this study, we sought to leverage this feature of PDA for two purposes: differential diagnosis and monitoring of response to treatment. EXPERIMENTAL DESIGN: Harmonic motion imaging (HMI) is a functional ultrasound technique that yields a quantitative relative measurement of stiffness suitable for comparisons between individuals and over time. We used HMI to quantify pancreatic stiffness in mouse models of pancreatitis and PDA as well as in a series of freshly resected human pancreatic cancer specimens. RESULTS: In mice, we learned that stiffness increased during progression from preneoplasia to adenocarcinoma and also effectively distinguished PDA from several forms of pancreatitis. In human specimens, the distinction of tumors versus adjacent pancreatitis or normal pancreas tissue was even more stark. Moreover, in both mice and humans, stiffness increased in proportion to tumor size, indicating that tuning of mechanical stiffness is an ongoing process during tumor progression. Finally, using a brca2-mutant mouse model of PDA that is sensitive to cisplatin, we found that tissue stiffness decreases when tumors respond successfully to chemotherapy. Consistent with this observation, we found that tumor tissues from patients who had undergone neoadjuvant therapy were less stiff than those of untreated patients. CONCLUSIONS: These findings support further development of HMI for clinical applications in disease staging and treatment response assessment in PDA.

Results from a Phase IIb, Randomized, Multicenter Study of GVAX Pancreas and CRS-207 Compared with Chemotherapy in Adults with Previously Treated Metastatic Pancreatic Adenocarcinoma (ECLIPSE Study).

PURPOSE: Limited options exist for patients with advanced pancreatic cancer progressing after 1 or more lines of therapy. A phase II study in patients with previously treated metastatic pancreatic cancer showed that combining GVAX pancreas (granulocyte-macrophage colony-stimulating factor-secreting allogeneic pancreatic tumor cells) with cyclophosphamide (Cy) and CRS-207 (live, attenuated Listeria monocytogenes expressing mesothelin) resulted in median overall survival (OS) of 6.1 months, which compares favorably with historical OS achieved with chemotherapy. In the current study, we compared Cy/GVAX + CRS-207, CRS-207 alone, and standard chemotherapy in a three-arm, randomized, controlled phase IIb trial. PATIENTS AND METHODS: Patients with previously treated metastatic pancreatic adenocarcinoma were randomized 1:1:1 to receive Cy/GVAX + CRS-207 (arm A), CRS-207 (arm B), or physician's choice of single-agent chemotherapy (arm C). The primary cohort included patients who had failed ≥2 prior lines of therapy, including gemcitabine. The primary objective compared OS between arms A and C in the primary cohort. The second-line cohort included patients who had received 1 prior line of therapy. Additional objectives included OS between all treatment arms, safety, and tumor responses. RESULTS: The study did not meet its primary efficacy endpoint. At the final study analysis, median OS [95% confidence interval (CI)] in the primary cohort (N = 213) was 3.7 (2.9-5.3), 5.4 (4.2-6.4), and 4.6 (4.2-5.7) months in arms A, B, and C, respectively, showing no significant difference between arm A and arm C [P = not significant (NS), HR = 1.17; 95% CI, 0.84-1.64]. The most frequently reported adverse events in all treatment groups were chills, pyrexia, fatigue, and nausea. No treatment-related deaths occurred. CONCLUSIONS: The combination of Cy/GVAX + CRS-207 did not improve survival over chemotherapy. (ClinicalTrials.gov ID: NCT02004262)See related commentary by Salas-Benito et al., p. 5435.

Experimental microdissection enables functional harmonisation of pancreatic cancer subtypes.

OBJECTIVE: Pancreatic ductal adenocarcinoma (PDA) has among the highest stromal fractions of any cancer and this has complicated attempts at expression-based molecular classification. The goal of this work is to profile purified samples of human PDA epithelium and stroma and examine their respective contributions to gene expression in bulk PDA samples. DESIGN: We used laser capture microdissection (LCM) and RNA sequencing to profile the expression of 60 matched pairs of human PDA malignant epithelium and stroma samples. We then used these data to train a computational model that allowed us to infer tissue composition and generate virtual compartment-specific expression profiles from bulk gene expression cohorts. RESULTS: Our analysis found significant variation in the tissue composition of pancreatic tumours from different public cohorts. Computational removal of stromal gene expression resulted in the reclassification of some tumours, reconciling functional differences between different cohorts. Furthermore, we established a novel classification signature from a total of 110 purified human PDA stroma samples, finding two groups that differ in the extracellular matrix-associated and immune-associated processes. Lastly, a systematic evaluation of cross-compartment subtypes spanning four patient cohorts indicated partial dependence between epithelial and stromal molecular subtypes. CONCLUSION: Our findings add clarity to the nature and number of molecular subtypes in PDA, expand our understanding of global transcriptional programmes in the stroma and harmonise the results of molecular subtyping efforts across independent cohorts.