A Critical Assessment of Postneoadjuvant Therapy Pancreatic Cancer Regression Grading Schemes With a Proposal for a Novel Approach

Artificial Intelligence-based Segmentation of Residual Pancreatic Cancer in Resection Specimens Following Neoadjuvant Treatment (ISGPP-2): International Improvement and Validation Study

Neoadjuvant therapy (NAT) has become routine in patients with borderline resectable pancreatic cancer. Pathologists examine pancreatic cancer resection specimens to evaluate the effect of NAT. However, an automated scoring system to objectively quantify residual pancreatic cancer (RPC) is currently lacking. Herein, we developed and validated the first automated segmentation model using artificial intelligence techniques to objectively quantify RPC. Digitized histopathological tissue slides were included from resected pancreatic cancer specimens from 14 centers in 7 countries in Europe, North America, Australia, and Asia. Four different scanner types were used: Philips (56%), Hamamatsu (27%), 3DHistech (10%), and Leica (7%). Regions of interest were annotated and classified as cancer, non-neoplastic pancreatic ducts, and others. A U-Net model was trained to detect RPC. Validation consisted of by-scanner internal-external cross-validation. Overall, 528 unique hematoxylin and eosin (H & E) slides from 528 patients were included. In the individual Philips, Hamamatsu, 3DHistech, and Leica scanner cross-validations, mean F1 scores of 0.81 (95% CI, 0.77-0.84), 0.80 (0.78-0.83), 0.76 (0.65-0.78), and 0.71 (0.65-0.78) were achieved, respectively. In the meta-analysis of the cross-validations, the mean F1 score was 0.78 (0.71-0.84). A final model was trained on the entire data set. This ISGPP model is the first segmentation model using artificial intelligence techniques to objectively quantify RPC following NAT. The internally-externally cross-validated model in this study demonstrated robust performance in detecting RPC in specimens. The ISGPP model, now made publically available, enables automated RPC segmentation and forms the basis for objective NAT response evaluation in pancreatic cancer.

Identification of tumour regression in neoadjuvantly treated pancreatic cancer is based on divergent and nonspecific criteria

AIMS

Following the increased use of neoadjuvant therapy for pancreatic cancer, grading of tumour regression (TR) has become part of routine diagnostics. However, it suffers from marked interobserver variation, which is mainly ascribed to the subjectivity of the defining criteria of the categories in TR grading systems. We hypothesized that a further cause for the interobserver variation is the use of divergent and nonspecific morphological criteria to identify tumour regression.

METHODS AND RESULTS

Twenty treatment-naïve pancreatic cancers and 20 pancreatic cancers treated with neoadjuvant chemotherapy were reviewed by three experienced pancreatic pathologists who, blinded for treatment status, categorized each tumour as treatment-naïve or neoadjuvantly treated, and annotated all tissue areas they considered showing tumour regression. Only 50%-65% of the cases were categorized correctly, and the annotated tissue areas were highly discrepant (only 3%-41% overlap). When the prevalence of various morphological features deemed to indicate TR was compared between treatment-naïve and neoadjuvantly treated tumours, only one pattern, characterized by reduced cancer cell density and prominent stroma affecting a large area of the tumour bed, occurred significantly more frequently, but not exclusively, in the neoadjuvantly treated group. Finally, stromal features, both morphological and biological, were investigated as possible markers for tumour regression, but failed to distinguish TR from native tumour stroma.

CONCLUSION

There is considerable divergence in opinion between pathologists when it comes to the identification of tumour regression. Reliable identification of TR is only possible if it is extensive, while lesser degrees of treatment effect cannot be recognized with certainty.

Microscopic tumor mapping of post-neoadjuvant therapy pancreatic cancer specimens to predict post-surgical recurrence: A prospective cohort study

BACKGROUND

Although various pathological grading systems are available for evaluating the response of pancreatic ductal adenocarcinoma (PDAC) to neoadjuvant therapy (NAT), their prognostic value has not been thoroughly validated. This study examined whether microscopic tumor mapping of post-NAT specimens could predict tumor recurrence.

METHODS

This prospective study enrolled 52 patients who underwent pancreaticoduodenectomy after NAT for PDAC between 2019 and 2021. Microscopic mapping was performed to identify residual tumor loci within the tumor bed using 4 mm2 pixels. Patients were divided into small extent (SE; n = 26) and large extent (LE; n = 26) groups using a cutoff value of 226 mm2. The diagnostic performance for predicting tumor recurrence was evaluated using receiver operating characteristic (ROC) curves.

RESULTS

Carbohydrate antigen 19-9 levels were normalised after NAT in more patients in the SE group (SE 21 [80.8%] vs. LE 13 [50.0%]; P = 0.041). Tumor size (P < 0.001), T stage (P < 0.001), positive lymph node yield (P = 0.024), and perineural invasion rate (P = 0.018) were significantly greater in the LE group. The 3-year disease-free survival rate was significantly lower in the LE group (SE 83.3% vs. LE 50.0%, P = 0.004). The area under the ROC curve for mapping extent was 0.743, which was greater than that of the other tumor response scoring systems.

CONCLUSIONS

Microscopic tumor mapping of the residual tumor in post-NAT specimens is a significant predictor of post-surgical recurrence, and offers better prognostic performance than the current grading systems.

Currently Debated Topics on Surgical Treatment of Pancreatic Ductal Adenocarcinoma: A Narrative Review on Surgical Treatment of Borderline Resectable, Locally Advanced, and Synchronous or Metachronous Oligometastatic Tumor

BACKGROUND

Previously considered inoperable patients (borderline resectable, locally advanced, synchronous oligometastatic or metachronous pancreatic adenocarcinoma (PDAC)) are starting to become resectable thanks to advances in chemo/radiotherapy and the reduction in operative mortality.

METHODS

This narrative review presents a chosen literature selection, giving a picture of the current state of treatment of these patients.

RESULTS

Neoadjuvant therapy (NAT) is generally recognized as the treatment of choice before surgery. However, despite the increased efficacy, the best pathological response is still limited to 10.9-27.9% of patients. There are still limited data on the selection of possible NAT responders and how to diagnose non-responders early. Multidetector computed tomography has high sensitivity and low specificity in evaluating resectability after NAT, limiting the resection rate of resectable patients. Ca 19-9 and Positron emission tomography are giving promising results. The prediction of early recurrence after a radical resection of synchronous or metachronous metastatic PDAC, thus identifying patients with poor prognosis and saving them from a resection of little benefit, is still ongoing, although some promising data are available.

CONCLUSION

In conclusion, high-level evidence demonstrating the benefit of the surgical treatment of such patients is still lacking and should not be performed outside of high-volume centers with interdisciplinary teams of surgeons and oncologists.

Pathologic tumor response to neoadjuvant therapy in resected pancreatic cancer: does it affect prognosis?

Neoadjuvant therapy (NAT) + surgical resection for pancreatic cancer (PC) has gained consensus in recent years. Pathological response (PR) is generally assessed according to the College of American Pathologists grading system, ranging from 0 (complete response) to 3 (no response). The aim of our study is to evaluate the PR in a series of resections for PC after NAT and its prognostic implication. 112 patients undergone NAT and resection for PC between 2011 and 2020 were retrospectively evaluated. PR was 0/1, 2 and 3 in 18 (15%), 79 (61%) and 29 (24%) cases, respectively. Chemotherapy regimens different from FOLFIRINOX and gemcitabine + nab-paclitaxel (OR 11.61 (2.53-53.36), p = 0.002) and lymphovascular invasion (OR 11.28 (1.89-67.23), p = 0.008) were associated to PR-3. Median follow-up was 25.8 (3.6-130.5) months. For PR-0/1, PR-2 and PR-3, median DFS was 45.8, 11.5, 4.6 months (p < 0.0001), respectively, while median OS was not reached, 27.1 and 17.5 months (p = 0.0006), respectively. At univariate analysis, PR-0/1 was significantly associated to better DFS and OS (HR 0.33 (0.17-0.67), p = 0.002; HR 0.20 (0.07-0.54), p = 0.002, respectively). At multivariate analysis, pancreaticoduodenectomy (HR 0.50 (0.30-0.84), p = 0.009), LNR (HR 27.14 (1.21-608.9), p = 0.038) and lymphovascular invasion (HR 1.99 (1.06-3.76), p = 0.033) were independently associated to DFS; pre-treatment CA 19.9 value (HR 1.00 (1.00-1.00), p = 0.025), post-treatment resectability status (HR 0.51 (0.28-0.95), p = 0.035), pancreaticoduodenectomy (HR 0.56 (0.32-0.99), p = 0.050), severe morbidity (2.99 (1.22-7.55), p = 0.017), LNR (HR 56.8 (2.08-1548.3), p = 0.017), lymphovascular invasion (HR 2.18 (1.08-4.37), p = 0.029) were independently associated to OS. PR did not reach statistical significance at multivariate analysis. A favorable PR is observed only in a limited number of cases. The prognostic role of PR, despite being promising, remains unclear and further multicentric studies are needed.

Comparative Analyses of the Clinicopathologic Features of Short-Term and Long-Term Survivors of Patients with Pancreatic Ductal Adenocarcinoma Who Received Neoadjuvant Therapy and Pancreatoduodenectomy

Neoadjuvant therapy (NAT) is increasingly used to treat patients with pancreatic ductal adenocarcinoma (PDAC). Patients with PDAC often show heterogenous responses to NAT with variable clinical outcomes, and the clinicopathologic parameters associated with these variable outcomes remain unclear. In this study, we systematically examined the clinicopathologic characteristics of 60 short-term survivors (overall survival < 15 months) and 149 long-term survivors (overall survival > 60 months) and compared them to 352 intermediate-term survivors (overall survival: 15-60 months) of PDAC who received NAT and pancreatoduodenectomy. We found that the short-term survivor group was associated with male gender (p = 0.03), tumor resectability prior to NAT (p = 0.04), poorly differentiated tumor histology (p = 0.006), more positive lymph nodes (p = 0.04), higher ypN stage (p = 0.002), and higher positive lymph node ratio (p = 0.03). The long-term survivor group had smaller tumor size (p = 0.001), lower ypT stage (p = 0.001), fewer positive lymph nodes (p < 0.001), lower ypN stage (p < 0.001), lower positive lymph node ratio (p < 0.001), lower rate of lymphovascular invasion (p = 0.001) and perineural invasion (p < 0.001), better tumor response grading (p < 0.001), and less frequent recurrence/metastasis (p < 0.001). The ypN stage is an independent predictor of both short-term and long-term survivors by multivariate logistic regression analyses. In addition, tumor differentiation was also an independent predictor for short-term survivors, and tumor response grading and perineural invasion were independent predictors for long-term survivors. Our results may help to plan and select post-operative adjuvant therapy for patients with PDAC who received NAT and pancreatoduodenectomy based on the pathologic data.

Prognosticators for Patients with Pancreatic Ductal Adenocarcinoma Who Received Neoadjuvant FOLFIRINOX or Gemcitabine/Nab-Paclitaxel Therapy and Pancreatectomy

Neoadjuvant FOLFIRINOX and gemcitabine/nab-paclitaxel (GemNP) therapies are increasingly used to treat patients with pancreatic ductal adenocarcinoma (PDAC). However, limited data are available on their clinicopathologic prognosticators. We examined the clinicopathologic factors and survival of 213 PDAC patients who received FOLFIRINOX with 71 patients who received GemNP. The FOLFIRINOX group was younger (p < 0.01) and had a higher rate of radiation (p = 0.049), higher rate of borderline resectable and locally advanced disease (p < 0.001), higher rate of Group 1 response (p = 0.045) and lower ypN stage (p = 0.03) than the GemNP group. Within FOLFIRINOX group, radiation was associated with decreased lymph node metastasis (p = 0.01) and lower ypN stage (p = 0.01). The tumor response group, ypT, ypN, LVI and PNI, correlated significantly with both DFS and OS (p < 0.05). Patients with the ypT0/T1a/T1b tumor had better DFS (p = 0.04) and OS (p = 0.03) than those with ypT1c tumor. In multivariate analysis, the tumor response group and ypN were independently prognostic factors for DFS and OS (p < 0.05). Our study demonstrated that the FOLFIRINOX group was younger and had a better pathologic response than the GemNP group and that the tumor response group, ypN, ypT, LVI and PNI, are significant prognostic factors for survival in these patients. Our results also suggest that the tumor size of 1.0 cm is a better cut off for ypT2. Our study highlights the importance of systemic pathologic examination and the reporting of post-treatment pancreatectomies.

Integrated Pathologic Score Effectively Stratifies Patients With Pancreatic Ductal Adenocarcinoma Who Received Neoadjuvant Therapy and Pancreaticoduodenectomy

Neoadjuvant therapy is increasingly used to treat patients with pancreatic ductal adenocarcinoma (PDAC). Pathologic parameters of treated PDAC, including tumor (ypT) and lymph node (ypN) stage, and tumor response grading (TRG) are important prognostic factors in this group of patients. To our knowledge, a multifactorial prognostic score combining pathologic features including ypT, ypN, and TRG in treated PDAC patients has not been reported. Our cohort consisted of 398 PDAC patients who received neoadjuvant therapy and underwent pancreaticoduodenectomy at our institution. All pancreaticoduodenectomy specimens were grossly and microscopically evaluated using a standard protocol. The integrated pathologic score (IPS) was calculated as the sum of the scores for ypT, ypN, and TRG according to either the MD Anderson grading system (IPSMDA) or the College of American Pathologists (CAP) grading system (IPSCAP). The IPSMDA and IPSCAP were correlated with clinicopathologic parameters and patient survival. Using either IPSMDA or IPSCAP, PDAC patients were stratified into 3 distinct prognostic groups for both disease-free survival (DFS) ( P <0.001) and overall survival (OS) ( P <0.001). The IPSMDA and IPSCAP correlated with tumor differentiation, margin status, the American Joint Committee on Cancer (AJCC) stage, and tumor recurrence ( P <0.05). In multivariate analysis, IPSMDA, IPSCAP, margin status, and tumor differentiation were independent prognostic factors for both DFS ( P <0.05) and OS ( P <0.05). However, patients with AJCC stage IB, IIA, or IIB disease had no significant difference in either DFS or OS ( P >0.05). The IPS appears to provide improved prognostic information compared with AJCC staging for preoperatively treated patients with PDAC.

The Prognostic and Predictive Role of the Neutrophil-to-Lymphocyte Ratio (NLR), Platelet-to-Lymphocyte Ratio (PLR), and Lymphocyte-to-Monocyte Ratio (LMR) as Biomarkers in Resected Pancreatic Cancer

Pancreatic cancer has poor survival despite modern-day advances in its management. At present, there are no available biomarkers that can predict chemotherapy response or help inform prognosis. In more recent years, there has been increased interest in potential inflammatory biomarkers, with studies revealing a worse prognosis of patients with a higher neutrophil-to-lymphocyte ratio in a range of tumour types. Our aim was to assess the role of three inflammatory biomarkers in peripheral blood in predicting chemotherapy response in patients with earlier disease treated with neoadjuvant chemotherapy and as a prognostic marker in all patients that underwent surgery for pancreatic cancer. Using retrospective records, we discovered that patients with a higher neutrophil-to-lymphocyte ratio (>5) at the time of diagnosis had worse median overall survival than those with ratios ≤5 at 13 and 32.4 months (p = 0.001, HR 2.43), respectively. We were able to appreciate a correlation between a higher platelet-to-lymphocyte ratio and increased residual tumour in the histopathological specimen in patients receiving neoadjuvant chemotherapy; however, the association was weak (p = 0.03, coefficient 0.21). Due to the dynamic relationship between the immune system and pancreatic cancer, it is unsurprising that immune markers may be useful as potential biomarkers; however, larger prospective studies are needed to validate these findings.

Pancreatic Ductal Adenocarcinoma: Molecular Pathology and Predictive Biomarkers

Pancreatic ductal adenocarcinoma (PDAC) has an extremely poor prognosis due to the lack of methods or biomarkers for early diagnosis and its resistance to conventional treatment modalities, targeted therapies, and immunotherapies. PDACs are a heterogenous group of malignant epithelial neoplasms with various histomorphological patterns and complex, heterogenous genetic/molecular landscapes. The newly proposed molecular classifications of PDAC based on extensive genomic, transcriptomic, proteomic and epigenetic data have provided significant insights into the molecular heterogeneity and aggressive biology of this deadly disease. Recent studies characterizing the tumor microenvironment (TME) have shed light on the dynamic interplays between the tumor cells and the immunosuppressive TME of PDAC, which is essential to disease progression, as well as its resistance to chemotherapy, newly developed targeted therapy and immunotherapy. There is a critical need for the development of predictive markers that can be clinically utilized to select effective personalized therapies for PDAC patients. In this review, we provide an overview of the histological and molecular heterogeneity and subtypes of PDAC, as well as its precursor lesions, immunosuppressive TME, and currently available predictive molecular markers for patients.

The Histopathology of Neoadjuvant-Treated (NAT) Pancreatic Ductal Adenocarcinoma

Examination of pancreatic ductal adenocarcinoma after NAT with the intent of diagnosis and outcome prediction remains a challenging task. The lack of a uniform approach to macroscopically assess these cases along with variations in sampling adds to the complexity. Several TRG systems have been proposed to correlate with an overall survival. In clinical practice, most of these TRG schemes have shown low level of interobserver agreement arguing for a need of larger studies and more innovative ways to assess outcome in this population.

Pathologic Examination of Pancreatic Specimens Resected for Treated Pancreatic Ductal Adenocarcinoma: Recommendations From the Pancreatobiliary Pathology Society

Currently, there are no internationally accepted consensus guidelines for pathologic evaluation of posttherapy pancreatectomy specimens. The Neoadjuvant Therapy Working Group of Pancreatobiliary Pathology Society was formed in 2018 to review grossing protocols, literature, and major issues and to develop recommendations for pathologic evaluation of posttherapy pancreatectomy specimens. The working group generated the following recommendations: (1) Systematic and standardized grossing and sampling protocols should be adopted for pancreatectomy specimens for treated pancreatic ductal adenocarcinoma (PDAC). (2) Consecutive mapping sections along the largest gross tumor dimension are recommended to validate tumor size by histology as required by the College of American Pathologists (CAP) cancer protocol. (3) Tumor size of treated PDACs should be measured microscopically as the largest dimension of tumor outer limits that is bound by viable tumor cells, including intervening stroma. (4) The MD Anderson grading system for tumor response has a better correlation with prognosis and better interobserver concordance among pathologists than does the CAP system. (5) A case should not be classified as a complete response unless the entire pancreas, peripancreatic tissues, ampulla of Vater, common bile duct, and duodenum adjacent to the pancreas are submitted for microscopic examination. (6) Future studies on tumor response of lymph node metastases, molecular and/or immunohistochemical markers, as well as application of artificial intelligence in grading tumor response of treated PDAC are needed. In summary, systematic, standardized pathologic evaluation, accurate tumor size measurement, and reproducible tumor response grading to neoadjuvant therapy are needed for optimal patient care. The criteria and discussions provided here may provide guidance towards these goals.

Artificial Intelligence-Based Segmentation of Residual Tumor in Histopathology of Pancreatic Cancer after Neoadjuvant Treatment

Simple Summary

The use of neoadjuvant therapy (NAT) in patients with pancreatic ductal adenocarcinoma (PDAC) is increasing. Objective quantification of the histopathological response to NAT may be used to guide adjuvant treatment and compare the efficacy of neoadjuvant regimens. However, current tumor response scoring (TRS) systems suffer from interobserver variability, originating from subjective definitions, the sometimes challenging histology, and response heterogeneity throughout the tumor bed. This study investigates if artificial intelligence-based segmentation of residual tumor burden in histopathology of PDAC after NAT may offer a more objective and reproducible TRS solution.

Abstract

Background: Histologic examination of resected pancreatic cancer after neoadjuvant therapy (NAT) is used to assess the effect of NAT and may guide the choice for adjuvant treatment. However, evaluating residual tumor burden in pancreatic cancer is challenging given tumor response heterogeneity and challenging histomorphology. Artificial intelligence techniques may offer a more reproducible approach. Methods: From 64 patients, one H&E-stained slide of resected pancreatic cancer after NAT was digitized. Three separate classes were manually outlined in each slide (i.e., tumor, normal ducts, and remaining epithelium). Corresponding segmentation masks and patches were generated and distributed over training, validation, and test sets. Modified U-nets with varying encoders were trained, and F1 scores were obtained to express segmentation accuracy. Results: The highest mean segmentation accuracy was obtained using modified U-nets with a DenseNet161 encoder. Tumor tissue was segmented with a high mean F1 score of 0.86, while the overall multiclass average F1 score was 0.82. Conclusions: This study shows that artificial intelligence-based assessment of residual tumor burden is feasible given the promising obtained F1 scores for tumor segmentation. This model could be developed into a tool for the objective evaluation of the response to NAT and may potentially guide the choice for adjuvant treatment.

Dataset for the reporting of carcinoma of the exocrine pancreas: recommendations from the International Collaboration on Cancer Reporting (ICCR)

BACKGROUND AND OBJECTIVES

Current guidelines for the pathology reporting on pancreatic cancer differ in certain aspects, resulting in divergent reporting practice and a lack of comparability of data. Here we report on a new international dataset for the pathology reporting of resection specimens with cancer of the exocrine pancreas (ductal adenocarcinoma and acinar cell carcinoma). The dataset was produced under the auspices of the International Collaboration on Cancer Reporting (ICCR), a global alliance of major (inter-)national pathology and cancer organisations.

METHODS AND RESULTS

According to the ICCR's rigorous process for dataset development, an international expert panel consisting of pancreatic pathologists, a pancreatic surgeon and an oncologist produced a set of core and non-core data items based on a critical review and discussion of current evidence. Commentary was provided for each data item to explain the rationale for selecting it as a core or non-core element, its clinical relevance, and to highlight potential areas of disagreement or lack of evidence, in which case a consensus position was formulated. Following international public consultation, the document was finalised and ratified, and the dataset, which includes a synoptic reporting guide, was published on the ICCR website.

CONCLUSIONS

This first international dataset for cancer of the exocrine pancreas is intended to promote high quality, standardised pathology reporting. Its widespread adoption will improve consistency of reporting, facilitate multidisciplinary communication and enhance comparability of data, all of which will help to improve the management of pancreatic cancer patients.

Survival in borderline resectable and locally advanced pancreatic cancer is determined by the duration and response of neoadjuvant therapy

BACKGROUND

Pancreatic cancer is the 8th commonest cancer and the 5th commonest cause of cancer-related death in Australia, with a 9% average 5-year survival. This study aims to investigate the effects of neoadjuvant treatment on overall survival (OS) and recurrence-free survival (RFS) in borderline resectable (BRPC) and locally advanced (LAPC) pancreatic adenocarcinoma followed by curative resection.

MATERIALS AND METHODS

Prospectively-collected demographic, medical, surgical and pathological data of patients with BRPC and LAPC treated with both neoadjuvant therapy (NAT) and surgery at a single tertiary referral centre in Australia were reviewed and analysed.

RESULTS

Between 2012 and 2018, 60 patients, 34 with BRPC and 26 with LAPC, were treated with NAT followed by curative resection. The commonest neoadjuvant chemotherapy regimens were Gemcitabine + Abraxane (51.7%) and FOLFIRINOX (35.0%), with 48.3% of patients additionally receiving neoadjuvant radiotherapy. Median RFS was 30 months and median OS was 35 months. On multivariable analysis, inferior OS was predicted by enlarged loco-regional lymph nodes on initial computed tomography (p = 0.032), larger tumour size post-NAT (p = 0.006) and Common Terminology Criteria for Adverse Events post-NAT toxicity greater than grade 2 (p = 0.015). LAPC patients received more neoadjuvant chemotherapy (p = 0.008) and radiotherapy (p = 0.021) than BRPC and achieved a superior pathological response (p = 0.010).

CONCLUSION

Patients who respond to NAT likely have a favourable disease biology and will progress well following resection. It is these patients who should be selected for more aggressive upfront management, and those with resistant disease should be spared from high-risk surgery.