Prognostic nomogram for patients with unresectable pancreatic cancer treated with gemcitabine plus nab-paclitaxel or FOLFIRINOX: A post-hoc analysis of a multicenter retrospective study in Japan (NAPOLEON study)

The relationship between LDH and GLIM criteria for cancer cachexia: Systematic review and meta-analysis

INTRODUCTION

Cancer cachexia is a clinical condition characterized by recognizable "sickness behaviors" accompanied by loss of lean body tissue. The Global Leadership on Malnutrition (GLIM) has proposed phenotypic (unintentional weight loss, low body mass index and low muscle mass) and aetiologic (reduced food intake and inflammation or disease burden) diagnostic criteria. Recent work has suggested serum lactate dehydrogenase (LDH) might represent a 3rd aetiologic criteria. Little is known of its relationship with GLIM. A systematic review and meta-analysis of their comparative prognostic value and association was performed.

METHODS

A search of electronic databases (PubMed, Medline, Ovid, Cochrane) up to February 2023 was used to identify studies that compared the prognostic value of LDH and components of the GLIM criteria in cancer. An analysis of the relationship between LDH and the components of GLIM was undertaken where this data was available. RevMan 5.4.1 was used to perform a meta-analysis for each diagnostic criteria that had 3 or more studies which reported hazard ratios with a 95 per cent confidence interval for overall survival (OS).

RESULTS

A total of 119 studies were reviewed. Advanced lung cancer was the most studied population. Included in the meta-analysis were 6 studies (n=2165) on LDH and weight loss, 17 studies (n=7540) on LDH and low BMI, 5 studies (n=758) on LDH and low muscle mass, 0 studies on LDH and food intake and 93 studies (n=32,190) on LDH and inflammation. There was a significant association between elevated serum LDH and each of low BMI (OR 1.39, 1.09 - 1.77; p=0.008), elevated NLR (OR 2.04, 1.57 - 2.65; p<0.00001) and elevated CRP (OR 2.58, 1.81 - 3.67; p<0.00001). There was no association between elevated serum LDH and low muscle mass. Only one study presented data on the association between LDH and unintentional weight loss. Elevated LDH showed a comparative OS (HR 1.86, 1.57 - 2.07; p<0.00001) to unintentional weight loss (HR 1.57, 1.23 - 1.99; p=0.0002) and had a similar OS (HR 2.00, 1.70 - 2.34; p<0.00001) to low BMI (HR 1.57, 1.29-2.90; p<0.0001). LDH also showed an OS (HR 2.25, 1.76 - 2.87; p<0.00001) congruous with low muscle mass (HR 1.93, 1.14 - 3.27; p=0.01) and again, LDH conferred as poor an OS (HR 1.77, 1.64-1.90; p<0.00001) as elevated NLR (HR 1.61, 1.48 - 1.77; p<0.00001) or CRP (HR 1.55, 1.43 - 1.69; p<0.00001).

CONCLUSION

Current literature suggests elevated serum LDH is associated with inflammation in cancer (an aetiologic GLIM criterion), however more work is required to establish the relationship between LDH and the phenotypic components of GLIM. Additionally, elevated serum LDH appears to be a comparative prognosticator of overall survival in cancer when compared to the GLIM criteria.

A multicenter retrospective observational NAPOLEON2 study of nanoliposomal irinotecan with fluorouracil and folinic acid in patients with unresectable pancreatic cancer

Nanoliposomal irinotecan with fluorouracil and folinic acid (NFF) is a standard regimen after gemcitabine-based therapy for patients with unresectable or recurrent pancreatic cancer. However, there are limited clinical data on its efficacy and safety in the real-world. We therefore initiated a retrospective and prospective observational study (NAPOLEON-2). The results of the retrospective part were reported herein. In this retrospective study, we evaluated 161 consecutive patients who received NFF as second-or-later-line regimen. The main endpoint was overall survival (OS), and the other endpoints were response rate, disease control rate, progression-free survival (PFS), dose intensity, and adverse events (AEs). The median age was 67 years (range, 38-85 years). The median OS and PFS were 8.1 and 3.4 months, respectively. The objective response and disease control rates were 5% and 52%, respectively. The median relative dose intensity was 81.6% for nanoliposomal irinotecan and 82.9% for fluorouracil. Grade 3 or 4 hematological and nonhematological AEs occurred in 47 and 42 patients, respectively. Common grade 3 or 4 AEs included neutropenia (24%), anorexia (12%), and leukocytopenia (12%). Subanalysis of patients treated with second-line and third-or-later-line demonstrated no statistical significant difference in OS (7.6 months vs. 9.1 months, respectively; hazard ratio, 0.92; 95% confidence interval, 0.64-1.35; p = 0.68). In conclusion, NFF has acceptable efficacy and safety profile even in real-world clinical settings. The prospective study is in progress to validate these findings.

Minimum and Optimal CA19-9 Response After Two Months Induction Chemotherapy in Patients With Locally Advanced Pancreatic Cancer: A Nationwide Multicenter Study

OBJECTIVE

This nationwide multicenter study aimed to define clinically relevant thresholds of relative serum CA19-9 response after 2 months of induction chemotherapy in patients with locally advanced pancreatic cancer (LAPC).

BACKGROUND

CA19-9 is seen as leading biomarker for response evaluation in patients with LAPC, but early clinically useful cut-offs are lacking.

METHODS

All consecutive patients with LAPC after 4 cycles (m)FOLFIRINOX or 2 cycles gemcitabine-nab-paclitaxel induction chemotherapy (±radiotherapy) with CA19-9 ≥5 U/mL at baseline were analyzed (2015-2019). The association of CA19-9 response with median OS (mOS) was evaluated for different CA19-9 cut-off points. Minimum and optimal CA19-9 response were established via log-rank test. Predictors for OS were analyzed using COX regression analysis.

RESULTS

Overall, 212 patients were included, of whom 42 (19.8%) underwent resection. Minimum CA19-9 response demonstrating a clinically significant median OS difference (12.7 vs. 19.6 months) was seen at ≥40% CA19-9 decrease. The optimal cutoff for CA19-9 response was ≥60% decrease (21.7 vs. 14.0 mo, P =0.021). Only for patients with elevated CA19-9 levels at baseline (n=184), CA19-9 decrease ≥60% [hazard ratio (HR)=0.59, 95% CI, 0.36-0.98, P =0.042] was independently associated with prolonged OS, as were SBRT (HR=0.42, 95% CI, 0.25-0.70; P =0.001), and resection (HR=0.25, 95% CI, 0.14-0.46, P <0.001), and duration of chemotherapy (HR=0.75, 95% CI, 0.69-0.82, P <0.001).

CONCLUSIONS

CA19-9 decrease of ≥60% following induction chemotherapy as optimal response cut-off in patients with LAPC is an independent predictor for OS when CA19-9 is increased at baseline. Furthermore, ≥40% is the minimum cut-off demonstrating survival benefit. These cut-offs may be used when discussing treatment strategies during early response evaluation.

Preoperative chemotherapy, radiotherapy and surgical decision-making in patients with borderline resectable and locally advanced pancreatic cancer

Surgical resection combined with systemic chemotherapy is the cornerstone of treatment for patients with localized pancreatic cancer. Upfront surgery is considered suboptimal in cases with extensive vascular involvement, which can be classified as either borderline resectable pancreatic cancer or locally advanced pancreatic cancer. In these patients, FOLFIRINOX or gemcitabine plus nab-paclitaxel chemotherapy is currently used as preoperative chemotherapy and is eventually combined with radiotherapy. Thus, more patients might reach 5-year overall survival. Patient selection for chemotherapy, radiotherapy and subsequent surgery is based on anatomical, biological and conditional parameters. Current guidelines and clinical practices vary considerably regarding preoperative chemotherapy and radiotherapy, response evaluation, and indications for surgery. In this Review, we provide an overview of the clinical evidence regarding disease staging, preoperative therapy, response evaluation and surgery in patients with borderline resectable pancreatic cancer or locally advanced pancreatic cancer. In addition, a clinical work-up is proposed based on the available evidence and guidelines. We identify knowledge gaps and outline a proposed research agenda.

Short-term serial circulating tumor DNA assessment predicts therapeutic efficacy for patients with advanced pancreatic cancer

PURPOSE

We investigated the potential clinical utility of short-term serial KRAS-mutated circulating cell-free tumor DNA (ctDNA) assessment for predicting therapeutic response in patients undergoing first-line chemotherapy for advanced pancreatic cancer.

METHODS

We collected 144 blood samples from 18 patients with locally advanced or metastatic cancer that were undergoing initial first-line chemotherapy of gemcitabine plus nab-paclitaxel (GEM plus nab-PTX). Analysis of KRAS-mutated ctDNA was quantified by digital droplet polymerase chain reaction (ddPCR) as mutant allele frequency (MAF). This study investigated pretreatment KRAS-mutated ctDNA status and ctDNA kinetics every few days (days 1, 3, 5 and 7) after initiation of chemotherapy and their potential as predictive indicators.

RESULTS

Of the 18 enrolled patients, an increase in KRAS-mutated ctDNA MAF values from day 0-7 after initiation of chemotherapy was significantly associated with disease progression (P < 0.001). Meanwhile, positive pretreatment ctDNA status (MAF ≥ 0.02%) (P = 0.585) and carbohydrate antigen 19-9 (CA19-9) values above the median (P = 0.266) were not associated with disease progression. In univariate analysis, this short-term increase in ctDNA MAF values (day 0-7) was found to be associated with significantly shorter progression free survival (PFS) (hazard ration [HR], 24.234; range, (2.761-212.686); P = 0.0002).

CONCLUSION

This short-term ctDNA kinetics assessment may provide predictive information to reflect real-time therapeutic response and lead to effective refinement of regimen in patients with advanced pancreatic cancer undergoing systemic chemotherapy.

Is lung involvement a favorable prognostic factor for pancreatic ductal adenocarcinoma with synchronous liver metastases?-A propensity score analysis

BACKGROUND

For advanced pancreatic cancer, pulmonary metastases (PM) have been considered favorable factors compared to metastases of other sites, but it remains unknown whether the prognosis of patients with synchronous liver and lung metastases is better than that of non-PM.

METHODS

Data was derived from a two-decade cohort and included 932 cases of pancreatic adenocarcinoma with synchronous liver metastases (PACLM). Propensity score matching (PSM) was applied to balance 360 selected cases, grouped into PM (n = 90) and non-PM (n = 270). Overall survival (OS) and survival-related factors were analyzed.

RESULTS

In PSM-adjusted data, the median OS was 7.3 and 5.8 months, for PM and non-PM, respectively (p = 0.16). Multivariate analysis revealed that male gender, poor performance status, higher hepatic tumor burden, ascites, elevated carbohydrate antigen 19-9, and lactate dehydrogenase were factors of poor survival (p < 0.05). Chemotherapy was the only independent significant factor of favorable prognosis (p < 0.05).

CONCLUSION

Although lung involvement was indicated to be a favorable prognostic factor for patients with PACLM in the whole cohort, PM were not associated with better survivals in the subset of cases subjected to PSM adjustment.

Preoperative lung immune prognostic index predicts survival in patients with pancreatic cancer undergoing radical resection

Background

Lung immune prognostic index (LIPI), a combination of derived neutrophil-to-lymphocyte ratio (dNLR) and lactate dehydrogenase (LDH), is currently attracting considerable interest as a potential prognostic indicator in many malignancies. Our study aimed to investigate the prognostic value of preoperative LIPI in patients with pancreatic ductal adenocarcinoma (PDAC) undergoing radical resection.

Methods

We retrospectively reviewed PDAC patients treated with radical resection from February 2019 to April 2021 at Chinese People's Liberation Army (PLA) general hospital. Based on the cut-off value of dNLR and LDH identified by X-tile, patients were divided into LIPI good and LIPI intermediate/poor group. Kaplan-Meier curve and log-rank test were used to compare the recurrence-free survival (RFS) and overall survival (OS) of the two groups. Univariate and multivariate Cox regression was used to identify the independent prognostic value of LIPI. Subgroup analysis was performed to identify specific population benefited from radical resection.

Results

A total of 205 patients were included and the median RFS and OS was 10.8 and 24.3 months, respectively. Preoperative LIPI intermediate/poor was related to worse RFS and OS (p < 0.05). Preoperative LIPI intermediate/poor, vascular invasion and no adjuvant chemotherapy were indicators of poor OS. Patients with LIPI intermediate/poor had worse OS especially among females and those with adjuvant chemotherapy (p < 0.05). Adjuvant chemotherapy related to better RFS and OS in patients with LIPI good (p < 0.05).

Conclusions

Preoperative LIPI intermediate/poor can be an indicator of poor prognosis in patients with PDAC undergoing radical resection. LIPI good could be an effective marker of benefit from adjuvant chemotherapy. Larger studies are warranted for further validation.

Multimodal survival prediction in advanced pancreatic cancer using machine learning

Background

Existing risk scores appear insufficient to assess the individual survival risk of patients with advanced pancreatic ductal adenocarcinoma (PDAC) and do not take advantage of the variety of parameters that are collected during clinical care.

Methods

In this retrospective study, we built a random survival forest model from clinical data of 203 patients with advanced PDAC. The parameters were assessed before initiation of systemic treatment and included age, CA19-9, C-reactive protein, metastatic status, neutrophil-to-lymphocyte ratio and total serum protein level. Separate models including imaging and molecular parameters were built for subgroups.

Results

Over the entire cohort, a model based on clinical parameters achieved a c-index of 0.71. Our approach outperformed the American Joint Committee on Cancer (AJCC) staging system and the modified Glasgow Prognostic Score (mGPS) in the identification of high- and low-risk subgroups. Inclusion of the KRAS p.G12D mutational status could further improve the prediction, whereas radiomics data of the primary tumor only showed little benefit. In an external validation cohort of PDAC patients with liver metastases, our model achieved a c-index of 0.67 (mGPS: 0.59).

Conclusions

The combination of multimodal data and machine-learning algorithms holds potential for personalized prognostication in advanced PDAC already at diagnosis.

•

We developed a machine-learning-based prediction model that outperforms the AJCC staging system and mGPS.

•

Applying our model to an external validation cohort demonstrates generalizability.

•

Explainable machine learning enables to understand the decision making of our model and identifies relevant parameters.

•

Combining clinical, imaging and genetic data holds potential for personalized prognostication in advanced PDAC.