In the last decade, immune checkpoint therapies (ICTs) have revolutionised cancer treatment. ICTs can boost a patients’ immune response against cancer by circumventing immune evasion mechanisms deployed by cancer cells (Esfahani et al., 2020, Pilard et al., 2021, Spiliopoulou et al., 2023). Standard-of-care immunotherapy drugs such as nivolumab and pembrolizumab, block programmed cell death protein 1 (PD-1) releasing the brakes on T cell proliferation and effector functions (Alsaab et al., 2017). Along with anti-PD-1 blocking drugs, anti-programmed death-ligand1 (PD-L1) blocking drugs such as Atezolizumab and Durvalumab can also block PD-1/PD-L1 interactions to enhance T cell activity against cancer (Twomey & Zhang, 2021). Recently, other immune-checkpoint-blocking molecules that have failed as monotherapy, have shown good efficacy in combination with the standard-of-care. For example, Relatlimab in combination with nivolumab got approved by the Food and Drug Administration (FDA) for unresectable or metastatic melanoma (Chocarro et al., 2022). As such, these advancements in ICT strategies resulted for the first time in durable responses in many cancer types.

However, ICTs also have some limitations. One of the main challenges is their limited efficacy, with only around 30% of patients achieving a durable response to single-agent ICTs, a percentage that increases to around 60% with the newly introduced combinational therapies (Jenkins et al., 2018, Kooshkaki et al., 2020). In addition, immune-related side effects can occur due to the unleashing of a patient’s own immune system. (Fasano et al., 2022, Li et al., 2020, Rahman et al., 2022). Furthermore, these therapies are often associated with high economic costs for patients and the national health system (Shi et al., 2022, Zhou et al., 2019). To address these issues, the scientific community has put much effort into discovering predictive biomarkers for ICTs to more precisely match patients with the right type of drug(s). However, current clinical practices deciding whether to administer ICTs or not are still based on traditional features such as: patient history, physical examination, PD-L1 expression status assessed by conventional immunohistochemistry, tumour mutational burden (TMB) and microsatellite instability (MSI) status (Davis & Patel, 2019), which alone are not precise enough to select patients for immunotherapy. Therefore, in most countries, cancers treated with ICTs are not accepted for therapy reimbursement, unless an appropriate complementary diagnostic can be implemented (Heffernan, Nikitas, Shukla, Camejo, & Knott, 2022).

One of the main reasons for a lack of appropriate biomarkers to identify immunotherapy responders, is that current biomarkers do not take into account the patient-to-patient variability and the intratumoral heterogeneity that is characteristic of the tumour microenvironment (TME), a complex, multicellular ecosystem whose functions cannot be properly described using unidimensional parameters. For instance, patients with similar clinical parameters can have completely different response rates to ICTs because of differences in the TME composition (Allam et al., 2022, Sadeghi Rad et al., 2021, Shelton et al., 2021, Upadhrasta and Zheng, 2019).

Initially, methods to dissect the composition of the TME were based on conventional one-marker-at-the-time immunohistochemistry which only allowed for a semiquantitative evaluation of cells present in a tissue section (Giraldo et al., 2015, Kuczkiewicz-Siemion et al., 2022, Tan et al., 2020). For example, in metastatic melanoma, it was suggested that the CD 8+ lymphocyte density correlated with response to anti-PD-1 therapy using conventional immunohistochemistry (Rizk et al., 2019). However, for the accurate identification of T cell subtypes, including CD8+ T cells, the simultaneous detection of multiple markers, such as CD3, CD4, and CD8, turned out to be imperative (Kumar, Connors, & Farber, 2018), which cannot be achieved using conventional IHC. Furthermore, the same cell type can be present in different functional states, which can completely alter the way in which cells interact with their environment. For instance, CD8+ T cells have the capacity to inhabit distinct functional states, including activated, exhausted, and terminally exhausted, thus necessitating the assessment of additional markers, such as OX40/CD69 for activation and TIM3 as an exhaustion indicator (Bosisio et al., 2020, Chow et al., 2022, Zhang et al., 2020). Finally, recent studies have indicated that spatial interactions between immune/tumour cells within specific locations in the tissue (intratumoral/peritumoral/perivascular) can significantly affect how patients respond to immunotherapy (Antoranz et al., 2022, Golesworthy et al., 2022, Zhang et al., 2019).

In order to measure cell functionality and spatial relationships between different cell types, it has become essential to measure the expression of multiple markers in a single tissue section at single-cell resolution. Even if single-cell technologies such as single-cell RNA sequencing (scRNA-seq) are powerful tools to identify cell phenotypes and cell states (Abdelaal et al., 2019, Plass et al., 2018), they require tissue dissociation and therefore do not retain spatial information (Li & Wang, 2021). Moreover, RNA-based techniques cannot capture if there is effective translation to proteins nor can they measure post-translational modifications such as cleavage, phosphorylation and translocation to produce a functional protein (Marguerat & Bähler, 2010). Since proteins are the macromolecules ultimately producing the biological effect targeted by drugs, spatial protein-based multiplexed immunohistochemistry (mIHC) is currently the state-of-the-art for a clinically meaningful investigation of the immune-related tumour microenvironment. The mIHC techniques enable the identification of several proteins on a single tissue section with single-cell resolution (Berry et al., 2021, Bosisio et al., 2022, Schapiro et al., 2022). Since the recent breakthrough of mIHC techniques as highly demanded scientific technologies, a number of innovative studies demonstrating their potential have been published. In this review, we are summarising how mIHC has been used in the field of immunotherapy. In particular, we will focus on how these technologies contributed to (i) characterise the tumour microenvironment, (ii) understand the role of tumour heterogeneity, (iii) study the interplay of the immune microenvironment and tumour progression, (iv) discover biomarkers for immune checkpoint therapies and (v) suggest novel therapeutic strategies (Fig. 1). The characteristics of the various studies using mIHC presented in this review are summarised in Table 1.