Safety and preliminary efficacy of pembrolizumab following trans-arterial chemoembolization for hepatocellular carcinoma: the PETAL phase Ib study

BACKGROUND

TACE may prime adaptive immunity and enhance immunotherapy efficacy. PETAL evaluated safety, preliminary activity of TACE plus pembrolizumab and explored mechanisms of efficacy.

METHODS

Patients with liver-confined HCC were planned to receive up to 2 rounds of TACE followed by pembrolizumab 200 mg every 21 days commencing 30-days post-TACE until disease progression or unacceptable toxicity for up to 1 year. Primary endpoint was safety, 21-days dose-limiting toxicities (DLT) from pembrolizumab initiation. Secondary endpoints included progression-free survival (PFS) and evaluation of tumour and host determinants of response.

RESULTS

Fifteen patients were included in the safety and efficacy population: 73% had non-viral cirrhosis, median age was 72 years. Child-Pugh (CP) class was A in 14 patients. Median tumour size was 4 cm. Ten patients (67%) received pembrolizumab after 1 TACE, 5 patients after 2 (33%). Pembrolizumab yielded no synergistic toxicity nor DLTs post-TACE. Treatment-related adverse events occurred in 93% of patients most commonly skin rash (40%), fatigue and diarrhoea (27%). After a median follow-up of 38.5 months, objective response rate (ORR) 12 weeks post-TACE was 53%. PFS rate at 12 weeks was 93% and median PFS was 8.95 months (95%CI 7.30-NA). Median duration of response was 7.3 months (95%CI: 6.3-8.3). Median OS was 33.5 months (95%CI: 11.6-NA). Dynamic changes in peripheral T-cell subsets, circulating tumour DNA, serum metabolites and in stool bacterial profiles highlight potential mechanisms of action of multi-modal therapy.

CONCLUSIONS

TACE plus pembrolizumab was tolerable with no evidence of synergistic toxicity, encouraging further clinical development of immunotherapy alongside TACE.

Are current machine learning applications comparable to radiologist classification of degenerate and herniated discs and Modic change? A systematic review and meta-analysis

INTRODUCTION

Low back pain is the leading contributor to disability burden globally. It is commonly due to degeneration of the lumbar intervertebral discs (LDD). Magnetic resonance imaging (MRI) is the current best tool to visualize and diagnose LDD, but places high time demands on clinical radiologists. Automated reading of spine MRIs could improve speed, accuracy, reliability and cost effectiveness in radiology departments. The aim of this review and meta-analysis was to determine if current machine learning algorithms perform well identifying disc degeneration, herniation, bulge and Modic change compared to radiologists.

METHODS

A PRISMA systematic review protocol was developed and four electronic databases and reference lists were searched. Strict inclusion and exclusion criteria were defined. A PROBAST risk of bias and applicability analysis was performed.

RESULTS

1350 articles were extracted. Duplicates were removed and title and abstract searching identified original research articles that used machine learning (ML) algorithms to identify disc degeneration, herniation, bulge and Modic change from MRIs. 27 studies were included in the review; 25 and 14 studies were included multi-variate and bivariate meta-analysis, respectively. Studies used machine learning algorithms to assess LDD, disc herniation, bulge and Modic change. Models using deep learning, support vector machine, k-nearest neighbors, random forest and naïve Bayes algorithms were included. Meta-analyses found no differences in algorithm or classification performance. When algorithms were tested in replication or external validation studies, they did not perform as well as when assessed in developmental studies. Data augmentation improved algorithm performance when compared to models used with smaller datasets, there were no performance differences between augmented data and large datasets.

DISCUSSION

This review highlights several shortcomings of current approaches, including few validation attempts or use of large sample sizes. To the best of the authors' knowledge, this is the first systematic review to explore this topic. We suggest the utilization of deep learning coupled with semi- or unsupervised learning approaches. Use of all information contained in MRI data will improve accuracy. Clear and complete reporting of study design, statistics and results will improve the reliability and quality of published literature.

Antibiotics promote intestinal growth of carbapenem-resistant Enterobacteriaceae by enriching nutrients and depleting microbial metabolites

The intestine is the primary colonisation site for carbapenem-resistant Enterobacteriaceae (CRE) and serves as a reservoir of CRE that cause invasive infections (e.g. bloodstream infections). Broad-spectrum antibiotics disrupt colonisation resistance mediated by the gut microbiota, promoting the expansion of CRE within the intestine. Here, we show that antibiotic-induced reduction of gut microbial populations leads to an enrichment of nutrients and depletion of inhibitory metabolites, which enhances CRE growth. Antibiotics decrease the abundance of gut commensals (including Bifidobacteriaceae and Bacteroidales) in ex vivo cultures of human faecal microbiota; this is accompanied by depletion of microbial metabolites and enrichment of nutrients. We measure the nutrient utilisation abilities, nutrient preferences, and metabolite inhibition susceptibilities of several CRE strains. We find that CRE can use the nutrients (enriched after antibiotic treatment) as carbon and nitrogen sources for growth. These nutrients also increase in faeces from antibiotic-treated mice and decrease following intestinal colonisation with carbapenem-resistant Escherichia coli. Furthermore, certain microbial metabolites (depleted upon antibiotic treatment) inhibit CRE growth. Our results show that killing gut commensals with antibiotics facilitates CRE colonisation by enriching nutrients and depleting inhibitory microbial metabolites.

Abstract 3056: Gut microbial co-metabolism influences the tumor microenvironment

Right sided colorectal cancers (RCRCs) have distinct molecular phenotypes and clinical presentations to those on the left. Gut microbial co-metabolic interactions with the RCRC microenvironment are poorly defined and present a novel therapeutic target. Here, we present a multi-omics analysis of RCRCs.

A prospective observational pilot study was performed at Imperial College London NHS Trust, UK. Treatment naive patients undergoing elective resectional surgery for primary RCRCs were recruited. Fresh intra-luminal fecal samples were obtained at three anatomical sites: 5 cm pre-, tumor, and 5 cm post-tumor. Microsatellite instability (MSI) and CDX2 expression were immunohistochemically evaluated. Metataxonomic analysis was carried out on a MiSeq Platform (Illumina, USA). Metabolomics analysis was performed using liquid chromatography mass spectrometry (Acquity UPLC, Xevo G2-S Q-TOF, Waters). After pre-processing, univariate and multivariate statistics were carried out in MetaboAnalyst and RStudio.

20 patients (8 females, mean age 69.9(53-83)) were included. Of the n=20 RCRCs, n=2 were ileo-cecal valve, n=4 were cecal, n=4 were cecum/proximal ascending colon, n=4 were mid-ascending colon, n=2 were ascending/hepatic flexure, n=2 were hepatic flexure, and n=2 were transverse colon cancers. Tumor classification according to TNM 8 indicated n=4 T1, n=1 T2, n=11 T3, n=4 T4; n=12 as N0, n=6 as N1, and n=2 as N2; n=2 as M0, n=1 as M1c, and n=17 as MX. All n=20 were CDX2 positive and n=16 had normal expression of mismatch repair proteins, n=4 showed MSI. Triglyceride (TG) species were significantly associated with T3-4 rather than T1-2 (p<0.01). Ceramide and diacylglycerol species were significantly associated with T1-2 rather than T3-4 (p<0.01). Multiple TG, an acyl carnitine, and a lysophosphatidic acid (LPA) species were associated with MSI status rather than no MSI status (p<0.01). An integrative analysis pipeline was developed to investigate statistically significant covariance between specific metabolites and taxa (p<0.01). Specifically, Procrustes similarity analysis of the multi-omics dataset revealed a negative correlation at pre- (p=0.606) and tumor level (p=0.738), but significant similarity at post-tumor level (p=0.045). This pattern in post-tumor may indicate cancer specific co-metabolism. At the post-tumor site, the LPA species increased in MSI, showed a negative correlation with members of the Oscillospiraceae family. Ceramide species, significantly increased in early-stage CRC, demonstrate negative correlation with the Lachnospiraceae family. Ongoing work is examining site-matched tissue samples utilizing imaging mass cytometry to define immune interactions.

Mapping the intra-luminal fecal metabolome and metataxonome in RCRC reveals potential bacterial-host co-metabolic targets that could be leveraged for stratifying immunotherapy treatment.

Citation Format: Petra Paizs, Duncan Roberts, Nathan Danckert, James Lewis, Maria Sani, Marta Jamroziak, Robert Goldin, Julian Marchesi, Lauren Ford, James Alexander, Zoltan Takats, James Kinross. Gut microbial co-metabolism influences the tumor microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3056.