Spatial transcriptomics (ST) revolutionized the study of gene expression within tissue sections, yet its application to clinically relevant material has remained limited by technical constraints related to sample characteristics and RNA quality. This thesis advances the use of spatial transcriptomics by establishing frameworks for profiling archival clinical specimens, ensuring tissue quality prior to ST, and combining spatial data and other complementary modalities to reveal the cellular, molecular and immune-clonal architecture of human cancer.

In Article I, we developed a method that adapted genome-wide spatial transcriptomics to formalinfixed paraffin embedded (FFPE) samples, which constitute the vast majority of clinical specimens, unlocking their potential. The workflow was validated on mouse brain tissue, showing high correlation with matched fresh-frozen data, and then applied on other various tissues to show its robustness across sample types. This development expanded spatial transcriptomics to previously inaccessible FFPE material. We also developed a TSO-based QC assay to assess spatial RNA accessibility directly in FFPE tissue sections, minimizing the risk of failed or biased spatial transcriptomics analysis.

In Article II, we developed an assay that enables evaluation of RNA integrity directly within tissue sections, the spatial RNA integrity number (sRIN) assay. Traditional RIN values are obtained from bulk tissue and therefore cannot reveal local variability in RNA quality. sRIN was developed as a practical quality-control step allowing users to identify well-preserved versus degraded regions before committing to costly downstream experiments.

In Article III, we analyzed genomic and clinical data from phase III melanoma trials and used NMF to define seven melanoma subtypes reflecting distinct differentiation states. By integrating bulk, single-cell, and spatial transcriptomics, we showed that these states coexist within tumors. Our analyses revealed that only differentiated melanoma becomes sensitized to immune checkpoint blockade therapy when combined with BRAF/MEK inhibition through enhanced antigen presentation, whereas undifferentiated states remain resistant, colocalize with CAF-rich niches, and could potentially be targeted by CDK7 inhibitors.

In Article IV, we combined single-cell RNA-seq, spatial transcriptomics and spatial V(D)J profiling to characterize the spatial and clonal organization of the immune landscape of sinonasal squamous cell carcinoma (SNSCC). Although immune infiltration was extensive, T- and B- cell clones were found preferentially within antigen presenting cell -rich stromal and peritumoral niches rather than undifferentiated tumor areas. We identified diverse CD8+ T-cell activation states that followed a bifurcated differentiation trajectory, and we also found an unusually large population of FOXP3+ regulatory T cells (Tregs), many expressing CXCR3 and TBX21 (T-bet). Integration of spatial gene expression with sV(D)J associated immune diversification and clonal expansion with CAF-associated matrix remodeling programs and interferon-activated antigen presenting cell programs. CXCL9+/CXCL10+ macrophages were found as drivers of lymphocyte recruitment and IDO1+ regulatory dendritic cells as immunosuppressive within the same niches.

Spatial transkriptomik (ST) har revolutionerat studiet av genuttryck i vävnadssnitt, men dess till-lämpning på kliniskt relevant material har länge begränsats av tekniska hinder kopplade till provets egenskaper och RNA-kvalitet. Denna avhandling utvecklar användningen av spatial transkriptomik genom att etablera ramverk för profilering av arkiverade kliniska prov, säkerställa vävnadskvalitet före ST och kombinera spatiala data med andra kompletterande modaliteter för att belysa den cellulära, molekylära och immun-klonala arkitekturen i human cancer.

I Artikel I utvecklade vi en metod som anpassade genomomfattande spatial transkriptomik till formalinfixerade paraffininkapslade (FFPE) prover, vilka utgör majoriteten av kliniska vävnader, och därmed frigjorde deras potential. Arbetsflödet validerades på mus-hjärnvävnad, där vi observerade hög korrelation med motsvarande färskfrusna data, och tillämpades därefter på olika andra vävnadstyper för att visa metodens robusthet. Denna utveckling utökade ST till tidigare otillgängligt FFPE-material. Vi utvecklade även ett TSO-baserat QC-test för att bedöma rumslig RNA-tillgänglighet direkt i FFPE-vävnadssnitt, vilket minimerar risken för misslyckad eller snedvriden ST-analys.

I Artikel II utvecklade vi ett test som möjliggör bedömning av RNA-integritet direkt i vävnadssnitt – det spatiala RNA-integritetsnumret (sRIN). Traditionella RIN-värden erhålls från bulkvävnad och kan därför inte avslöja lokal variation i RNA-kvalitet. sRIN utvecklades som ett praktiskt kvalitetskontrollsteg som gör det möjligt att identifiera välbevarade respektive degraderade regioner innan man investerar i kostsamma efterföljande experiment.

I Artikel III analyserade vi genomiska och kliniska data från fas III-studier på melanom och använde NMF för att definiera sju melanomsubtyper som återspeglar distinkta differentieringsstadier. Genom att integrera bulk-, enkelcells- och spatial transkriptomik visade vi att dessa tillstånd samexisterar inom tumörer. Våra analyser visade att endast differentierat melanom blir känsligt för immuncheckpoint-blockad när behandlingen kombineras med BRAF/MEK-hämning via ökad antigenpresentation, medan odifferentierade tillstånd förblir resistenta, samlokaliserar med CAFrika nischer och potentiellt kan riktas med CDK7-hämmare.

I Artikel IV kombinerade vi enkelcells-RNA-sekvensering, spatial transkriptomik och långläsande spatial V(D)J-sekvensering för att karaktärisera immunlandskapet i sinonasalt skivepitelcarcinom (SNSCC). Vi identifierade olika CD8⁺ T-cellsaktiveringsstadier och en ovanligt stor population av FOXP3⁺ regulatoriska T-celler, många med uttryck av CXCR3 och TBX21 (T-bet). Integrering av spatialt genuttryck med sV(D)J kopplade immun diversifiering och klonal expansion till CAF-associerade matrix-remodelleringsprogram och interferonaktiverade antigenpresenterande cellprogram. CXCL9⁺/CXCL10⁺ makrofager identifierades som drivande för lymfocytrekrytering, medan IDO1⁺ regulatoriska dendritiska celler utgjorde immunsuppressiva komponenter inom samma nischer.

While triple-negative breast cancer (TNBC) is known to be heterogeneous at the genomic and transcriptomic levels, spatial information on tumor organization and cell composition is still lacking. Here, we investigate TNBC tumor architecture including its microenvironment using spatial transcriptomics on a series of 92 patients. We perform an in-depth characterization of tumor and stroma organization and composition using an integrative approach combining histomorphological and spatial transcriptomics. Furthermore, a detailed molecular characterization of tertiary lymphoid structures leads to identify a gene signature strongly associated to disease outcome and response to immunotherapy in several tumor types beyond TNBC. A stepwise clustering analysis identifies nine TNBC spatial archetypes, further validated in external datasets. Several spatial archetypes are associated with disease outcome and characterized by potentially actionable features. In this work, we provide a comprehensive insight into the complexity of TNBC ecosystem with potential clinical relevance, opening avenues for treatment tailoring including immunotherapy.

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer. Although standard-of-care chemotherapeutics are sufficient for most ALL cases, there are subsets of patients with poor response who relapse in disease. The biology underlying differences between subtypes and their response to therapy has only partially been explained by genetic and transcriptomic profiling. Here, we perform comprehensive multi-omic analyses of 49 readily available childhood ALL cell lines, using proteomics, transcriptomics, and pharmacoproteomic characterization. We connect the molecular phenotypes with drug responses to 528 oncology drugs, identifying drug correlations as well as lineage-dependent correlations. We also identify the diacylglycerol-analog bryostatin-1 as a therapeutic candidate in the MEF2D-HNRNPUL1 fusion high-risk subtype, for which this drug activates pro-apoptotic ERK signaling associated with molecular mediators of pre-B cell negative selection. Our data is the foundation for the interactive online Functional Omics Resource of ALL (FORALL) with navigable proteomics, transcriptomics, and drug sensitivity profiles at https://proteomics.se/fora.

Formalin-fixed paraffin embedding (FFPE) is the most widespread long-term tissue preservation approach. Here, we report a procedure to perform genome-wide spatial analysis of mRNA in FFPE-fixed tissue sections, using well-established, commercially available methods for imaging and spatial barcoding using slides spotted with barcoded oligo(dT) probes to capture the 3′ end of mRNA molecules in tissue sections. We applied this method for expression profiling and cell type mapping in coronal sections from the mouse brain to demonstrate the method's capability to delineate anatomical regions from a molecular perspective. We also profiled the spatial composition of transcriptomic signatures in two ovarian carcinosarcoma samples, exemplifying the method's potential to elucidate molecular mechanisms in heterogeneous clinical samples. Finally, we demonstrate the applicability of the assay to characterize human lung and kidney organoids and a human lung biopsy specimen infected with SARS-CoV-2. We anticipate that genome-wide spatial gene expression profiling in FFPE biospecimens will be used for retrospective analysis of biobank samples, which will facilitate longitudinal studies of biological processes and biomarker discovery.

The RNA integrity number (RIN) is a frequently used quality metric to assess the completeness of rRNA, as a proxy for the corresponding mRNA in a tissue. Current methods operate at bulk resolution and provide a single average estimate for the whole sample. Spatial transcriptomics technologies have emerged and shown their value by placing gene expression into a tissue context, resulting in transcriptional information from all tissue regions. Thus, the ability to estimate RNA quality in situ has become of utmost importance to overcome the limitation with a bulk rRNA measurement. Here we show a new tool, the spatial RNA integrity number (sRIN) assay, to assess the rRNA completeness in a tissue wide manner at cellular resolution. We demonstrate the use of sRIN to identify spatial variation in tissue quality prior to more comprehensive spatial transcriptomics workflows. Kvastad et al. develop the spatial RNA Integrity Number (sRIN) assay that evaluates the RNA integrity at cellular resolution. This method improves the resolution of a similar method called the RNA Integrity Number (RIN), demonstrating spatial variation in the quality of RNA samples.

Sinonasal squamous cell carcinoma (SNSCC) is frequently immune infiltrated yet responds variably to immunotherapy, suggesting that immune effectiveness may depend on tissue organization rather than abundance. Here, we integrated single-cell RNA-seq with paired V(D)J profiling across seven SNSCC tumors and performed deep spatial transcriptomics with spatial V(D)J sequencing in three tumors with distinct histopathology.

Single-cell analyses revealed diverse CD8⁺ states spanning stem-like, cytotoxic, tissue-resident, and exhausted phenotypes, alongside an unusually dominant and clonally expanded regulatory CD4⁺ compartment enriched for T-bet⁺CXCR3⁺ Tregs. Trajectory, clonal overlap, and phylogenetic inference supported bifurcated differentiation pathways, separating a CX3CR1⁺TBX21⁺ cytotoxic endpoint from a resident–exhausted lineage.

Spatial V(D)J mapping uncovered colocalization of TCR and BCR activity within discrete stromal and peritumoral immune aggregates that lacked canonical tertiary lymphoid structure features. Across patients, TRB and IGH signals showed a positive spatial association. Linking spatial clonal diversity to transcriptome-wide programs identified convergent associations with IFN-activated antigen-presenting cell (APC) and matrix-remodeling programs driven by cancer-associated fibroblasts (CAFs). Single-cell resolution implicated CXCL9/10⁺ macrophages as key chemokine sources and IDO1⁺ regulatory dendritic cells as local immunoregulatory elements within the same niches. Together, these data define APC-centered stromal microenvironments as focal points of adaptive clonal expansion that simultaneously recruit and constrain immunity in SNSCC.

Formalin-fixed paraffin embedding (FFPE) is the most widespread long-term tissue preservation approach. Here we present a procedure to perform genome-wide spatial analysis of mRNA in FFPE fixed tissue sections. The procedure takes advantage of well-established, commercially available methods for imaging and spatial barcoding using slides spotted with barcoded oligo(dT) probes to capture the 3’ end of mRNA molecules in tissue sections. First, we conducted expression profiling and cell type mapping in coronal sections from the mouse brain to demonstrate the method’s capability to delineate anatomical regions from a molecular perspective. Second, we explored the spatial composition of transcriptomic signatures in ovarian carcinosarcoma samples using data-driven analysis methods, exemplifying the method’s potential to elucidate molecular mechanisms in heterogeneous clinical samples. Finally, we demonstrate the applicability of the assay to characterize organoids and a human lung biopsy specimen infected with SARS-CoV-2.

In BRAF-mutant melanoma, BRAF/MEK inhibitors induce high initial response rates, but resistance limits durability, whereas immune checkpoint blockade (ICB) provides long-term benefit only to a subset of patients. Combining BRAF/MEK inhibitors (BRAF/MEKi) with ICB has the potential to extend the durability of response to a broader patient population. However, clinical trials testing this approach have yielded mixed outcomes.

To identify patients most likely to benefit from BRAF/MEKi-ICB combination therapy, we integrated clinical and multi-omics data—including transcriptomics, genomics, and spatial RNA sequencing—from IMspire150, BRIM, AVAST-M, and published ICB monotherapy datasets. We identified two ICB-resistant tumor states—differentiated and undifferentiated—with distinct microenvironments. Only differentiated tumors could be sensitized to ICB through MAPK inhibition via the MITF/NLRC5 axis, whereas undifferentiated tumors remained refractory but were sensitive to CDK7 inhibition.

Spatial transcriptomics revealed the coexistence of ICB-sensitive and resistant subtypes within tumors occupying distinct spatial neighborhoods, with cancer-associated fibroblasts (CAFs) localizing uniquely to undifferentiated regions. These findings establish tumor differentiation, microenvironment, and spatial architecture as critical determinants of BRAF/MEKi-ICB response, providing a framework for optimizing patient selection.