Our cross-sectional study demonstrates that there is a high rate of co-trimoxazole induced drug rash, in patients treated for immune related adverse events, with those developing rash appearing to have a reduced survival.

Background: Genetic evolution of clear cell renal cell carcinoma (ccRCC) follows distinct trajectories, with varying levels of intratumor heterogeneity (ITH) and chromosomal complexity (WGII). While these patterns associate with clinical outcomes, it remains unknown whether they fully reconcile tumor behavior and how genetic and transcriptional features co-evolve in relation to the tumor microenvironment (TME). Methods: To analyze the patterns of transcriptional and TME heterogeneity, we performed bulk whole-transcriptome sequencing on 244 samples, including 22 metastatic and 12 tumor-adjacent normal samples, from 79 ccRCC patients recruited to the TRACERx Renal study. We integrated transcriptional data with previously published genetic, phylogenetic, spatial and clinical information. Results: Transcriptional distances between paired samples from the same primary tumor mirrored but were not fully determined by genetic distance (p-value < 0.001); and increased from primary-primary to primary-metastasis and primary-normal pairs. Within primary-metastasis pairs, metastasis-seeding primary tumor regions were transcriptionally closest to their matched metastasis (p-value < 0.001), suggesting that an important fraction of metastatic transcriptional traits were acquired in the primary tumor. Regarding the tumor clonal structure, transcriptional evolution followed a conserved path through increasing cell proliferation and oxidative phosphorylation and downregulating DNA repair from earlier to later clones. Further, within tumors with increasing WGII we observed upregulation and downregulation of repressors and downstream effectors, respectively, of the canonical cGAS-STING pathway. Combining the presence of this transcriptional pattern with WGII predicted shorter PFS in TRACERx Renal (p-value < 0.001) and in TCGA-KIRC (p-value < 0.001). Clonal evolution was also linked to changes in TME, with each of the previously defined genetic evolutionary trajectories associated to a specific TME (p-value < 0.001). For example, ccRCCs on a PBRM1-SETD2 trajectory demonstrated increased infiltration of cytotoxic immune cells. TME ITH was pervasive and associated with shorter PFS (p-value = 0.03). A recurrent trend from earlier to later clones was progressive T cell depletion (p-value < 0.001). The evolution of the TCR repertoire mirrored the tumor clonal structure (p-value = 0.002), suggesting the thus far elusive antigenic source in ccRCC is heritable. Accordingly, the TCR repertoire in metastasis-seeding primary tumor regions resembled the closest the TCR repertoire of matched metastasis (p-value = 0.06). Conclusion: Integrated analysis of genetic and transcriptional data in TRACERx Renal showed i) transcriptional and TME ITH not fully recapitulated by genetic ITH, ii) conserved paths of transcriptional and TME evolution and iii) a heritable nature of part of the ccRCC antigen source. Citation Format: Ángel Fernández Sanromán, Lewis Au, Benjy Jek Yang Tan, Charlotte Spencer, Anne-Laure Catin, Irene Lobon, Husayn Pallikonda, Kevin Litchfield, Fiona Byrne, James Larkin, Annika Fendler, Samra Turajlic. Integrated analysis of genetic, transcriptional and TME evolution of ccRCC: TRACERx Renal [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1621.

Diverse clinical presentations of clear cell renal cell carcinoma (ccRCC) confound clinical decision making, leading to over and undertreatment. Clonal evolution of ccRCC proceeds through distinct trajectories characterised by differing levels of genomic intratumoral heterogeneity (gITH) and chromosomal complexity (weighted genomic instability index, wGII). However, accurate evaluation of these indices requires multiregional profiling of fresh tumour; cost prohibitive and logistically challenging in the clinical setting. Clinical histopathology workflows routinely capture multiple tumour areas enabling the use artificial intelligence (AI) to predict tumour evolutionary features directly from clinical grade H&E whole slide image (WSIs). ccRCC displays profound genetic and histological ITH but the link between these entities remains unclear. We leverage the TRACERx Renal cohort, comprising 1485 WSIs from 81 tumours to predict WGII and gITH and to gain insights into the relationship between genetic and histological ITH. Critically, each WSI is associated with a wGII and gITH label derived from a closely linked fresh tumour sample. For both prediction tasks, we extracted meaningful features for each WSI using self-supervised representation learning “MoCo”. Since high wGII confers poor prognosis we focussed on predicting binary stratification label of high wGII or low wGII (relative to the cohort median). First we predicted wGII as a continuous variable using a supervised multiple instance learning regression model trained on the MoCo features, and then classified the predicted wGII into “high” or “low” achieving 0.80 AUROC. To predict gITH we postulated that the degree of gITH would correlate with histological ITH. Using an unsupervised clustering of refined MoCo features we defined 24 histological clusters. The number of computationally derived histological clusters within a single tumour positively correlated with gITH (pearson’s 0.56). We used the number of clusters to classify WSIs into prognostic binary groups of high or low gITH (relative to the cohort median) achieving an AUROC of 0.80. To understand the biological relationship between histological and genetic ITH we pathologically characterised the histological clusters: a pathologist annotated WSIs with tumour architecture and cytomorphology. Image tiles were associated with the annotations using spatial coordinates, illuminating phenotypic traits of different evolutionary trajectories and providing an interpretability framework for our AI pipelines. Since the tumour evolutionary course dictates disease progression tempo, applying evolutionary classification in clinic can fundamentally improve patient care. Here, for the first time, we provide a framework to translate fundamental evolutionary principles underpinning tumour biology and clinical progression into a prognostic computational pathology biomarker possible to clinically implement. Citation Format: Charlotte E. Spencer, Axel Camara, Auriane Riou, Lewis Au, Jose I. Lopez, Zayd Tippu, Charles Maussion, Kenneth Ho, Amy Strange, Emma Nye, Veronique Birault, Lydwine Van-praet, Kim Edmonds, Eleanor Carlyle, Steve Hazell, Sarah Rudman, James Larkin, Samra Turajlic. Predicting tumor evolution from digital histology using AI [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4298.

Background: Cancer research autopsy genomic studies offer insight into the metastatic cancer landscape but come with complexities that relate to the sampling and processing of post-mortem tissue. Clarifying the effect of autopsy variables on pre- and post-sequencing quality control (QC) is an unmet need that may inform tissue collection strategies. Methods: The effect of age, sex, post-mortem interval (PMI), and sample type (primary, metastatic, or normal) on pre-sequencing QC (nucleic acid concentration and integrity) was examined in 2678 samples (88.6% metastatic, 8.0% primary, 3.4% normal) from 83 patients with melanoma, lung, renal, or prostate cancer in the PEACE study. In the lung cohort, 160 surgical samples were also included through the TRACERx study, allowing surgery-autopsy tissue comparisons. Post-sequencing QC metrics were evaluated for lung samples that underwent DNA (n=522) or RNA (n=366) sequencing. Results: RNA concentration and RIN were greater in surgical samples than those collected at autopsy. Across cohorts, metastatic autopsy samples had greater nucleic acid concentrations than primary or normal autopsy samples, but not integrity. DNA and RNA concentration and integrity differed significantly between primary tumor types. When comparing samples of different metastatic sites from the whole cohort, concentration was lowest in bone (DNA) or the digestive tract (RNA), while integrity was greatest in the brain and lowest in the digestive tract (DIN, RIN). Although autopsy variables like age, sex and PMI correlated with pre-sequencing QC metrics in univariate analysis, they were not found to significantly correlate with these metrics in multivariate analysis, which identified that only primary cancer type and metastatic site were independent determinants of concentration and integrity. Similarly, for post-DNA (whole exome) sequencing QC, only the metastatic site was found to independently influence sequencing QC metrics like total number of sequences, average sequence length, and FastQC score. For RNA sequencing, only the metastatic site was found to influence sequencing QC metrics like total number of sequences, percentage of non-duplicated sequences, one hit-one genome percentage, and the alignment percentage on the human genome. Discussion: The lack of influence of PMI on QC in the largest QC-focused autopsy cancer study to date suggests that quality tissue can be obtained from non-rapid autopsy programs, which are more feasible and less resource-intensive than rapid programs. Citation Format: Petros Fessas, Sonya Hessey, Corentin Richard, Cristina Naceur-Lombardelli, Sophia Ward, David A. Moore, Karolina Nowakowska, Blanca Trujillo, Irene Lobon, Scott T. Shepherd, Fiona Byrne, Samra Turajlic, Gerhardt Attard, Charles Swanton, Mariam Jamal-Hanjani. The effect of cancer research autopsy parameters on DNA and RNA sequencing quality [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2926.

Clear cell Renal Cell Carcinoma (ccRCC) is profoundly angiogenic, characterised by complex yet heterogenous vascular networks. Blood vessels are an important constituent part of the tumour microenvironment (TME) and, in addition to immune cells, are the target of drug therapies in advanced disease. The TME plays an important role in determining disease progression and response to therapy, acting as a selective pressure on the tumour cells thus influencing evolutionary trajectory. This selective pressure is sculpted by cross-talk between blood vessels, immune cells and the tumour cells themselves. Whilst the details of these carefully orchestrated cellular interactions is not understood their final read-out is reflected in tissue morphology, which can be assessed using an H&E-stained slide, a fundamental component of clinical diagnostic histopathological workflows. A computational pathology approach to assess vascular networks from digital H&E whole slide images (H&E WSIs) would present a powerful tool to understand disease biology. It would permit high-throughput analysis of large cohorts where routine multi-regional sampling captures disease heterogeneity. Such work would lay the foundations for developing a computational pathology biomarker to predict survival outcomes that could be easily implemented into existing clinical workflows. Intricacy of vascular network structures makes reproducible analysis challenging, which can be approached either using morphology, a qualitative evaluation of a shape, or using topography to quantify feature dimensions. Here we reconcile the two methods to develop an interpretable computational pathology solution to study the blood vessels in ccRCC. Further, we have built a deep-learning attention UNET model to segment blood vessels from H&E WSIs. By combining these tools we have developed a computational pathology pipeline able to robustly characterise vascular networks directly from H&E WSIs. We leverage 1064 tumour regions from 82 ccRCC tumours of the TRACERx Renal dataset where ex-vivo multi-regional sampling with closely linked specimens for histological and genomic analysis permits interrogation of the histo:genomic relationship contextualised within the evolutionary dynamics of each tumour. We demonstrate that vascular intratumoral heterogeneity is pervasive and we link different vascular topologies to genetic alterations associated with opposing evolutionary trajectories (PBRM1 and BAP1 mutations) and the acquisition of metastatic competence (loss of 9p). Finally, we show that progressive accumulation of genetic alterations alters vascular network structure, suggesting that vascular topology could be used to assess tumour evolution. Our pipeline is a powerful tool to study ccRCC vasculature in large cohorts with multi-regional sampling to capture intratumoral heterogeneity and ultimately could form the basis of a computational pathology biomarker to predict outcome to therapy. Citation Format: Charlotte E. Spencer, Graham Ross, Thomas Mead, Amy Strange, Anna Song, Katie Bentley, Samra Turajlic. Interpretable computational pathology reveals that vascular networks reflect evolutionary dynamics in kidney cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr PR015.

Background: Although a range of therapeutic options are available in the management of invasive breast carcinomas, spatial segregation of tumour subclones has hampered biomarker identification in single-region samples. Representative sampling (RS) overcomes spatial bias by sampling from a homogenized and well-mixed cancer specimen. The Homogenization of Leftover Surgical Tissue Feasibility Study (HoLST-F - NCT03832062) is a prospective trial aiming to assess the feasibility of RS in tumor tissue leftover after pathology sampling. An interim analysis of the breast cancer cohort is presented. Methods: In the context of the HoLST-F study and pre-specified end-points, representative samples derived from 75 leftover invasive breast carcinoma specimens underwent flow cytometric analysis of CK8/18, CD3, Ki67 and DAPI. Tumor cell enrichment by CK8/18 positivity and ploidy (in aneuploid carcinomas) was performed for downstream DNA extraction and whole exome sequencing. Somatic variants were determined using a bespoke pipeline to remove artefacts associated with fixation. Variant oncogenicity and therapeutic evidence levels were assigned by OncoKB variant annotation. Quantitative image analysis was applied to tissue sections from diagnostic FFPE blocks stained with Ki67 immunohistochemistry (n=78) and H&E for lymphocytic infiltration scores (n=175) to correlate with flow cytometry. Results: In enriched representative samples, oncogenic mutations were commonly identified at high variant allele frequency (VAF) in known breast carcinoma driver genes including PIK3CA, CDH1, TP53, KMT2C and GATA3. Other clinically relevant oncogenic mutations were identified in ESR1, PTEN, ERBB2, RB1, AKT1, BRCA1, NF1 and FOXA1 which have been associated with resistance to various anti-cancer therapies. These mutations occurred at low and high VAFs indicative of both of clonal and sub-clonal resistance mechanisms. Recurrent variants in ESR1 (e.g. D538G) and PIK3CA (e.g. H1047R/L) were associated with Level 1 evidence for use as predictive biomarkers, while variants in twelve other genes across the cohort were associated with Level 2-4 evidence. Lymphocyte infiltration scores and Ki67 expression varied by tumor region, however RS by flow cytometry showed strong correlation with a weighted average across multiregional quantification of Ki67 expression (R=0.81, p=2.8 × 10-8) and lymphocyte infiltration (R=0.61, p<2.2 × 10-16). Conclusions: RS of invasive breast carcinoma in the HoLST-F trial has recapitulated the expected genomic driver landscape in breast cancer, in addition to identifying both clonal and subclonal genomic mechanisms of therapy resistance. Many of these mutations are either current or emerging therapeutic targets. Flow quantification of cells expressing phenotypic biomarkers (Ki67 and CD3) is feasible through RS and early analysis indicates that RS correlates with a weighted average across multiple regions. Leftover surgical tissue is an underutilized resource for biomarker assessment in breast carcinomas and can be examined by RS. Citation Format: Brian Hanley, Lisa Gallegos, Lavinia Spain, Husayn Pallikonda, Zayd Tippu, Samantha Hill, Aoune Barhoumi, Fiona Byrne, Yulia Dogva, Ashley Gilchrist, Glenn Noel-Storr, Hannah Veloz, Stacey Stanislaw, Harold Sansano, Kim Edmonds, Eleanor Carlyle, Nicholas Turner, James Larkin, Nelson Alexander, Samra Turajlic. Representative sampling of invasive breast carcinomas: Interim report from a prospective study (HOLST-F) [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr A007.