Research
Curing cancer remains a formidable challenge, largely because tumors are highly heterogeneous. At the time of diagnosis, a patient may already harbor billions of tumor cells comprising a vast number of genetically distinct subclones. While most subclonal mutations are biologically neutral, some confer a selective advantage under specific therapies. As a result, even when an initial treatment is effective, resistant subclones can survive, expand,
and ultimately drive relapse.
Despite broad recognition of tumor heterogeneity as a central obstacle, few therapeutic strategies explicitly address it. Until recently, this was primarily a technical limitation: it was not possible to track more than a small number of tumor clones simultaneously. Advances in cell barcoding technologies now enable the parallel tracking of thousands of subclones. By combining cell barcoding with high-throughput drug and CRISPR screens, single-cell sequencing, and biologically relevant tumor organoid models, we aim to develop rational therapeutic strategies that overcome tumor heterogeneity and prevent the emergence of resistant disease.
Therapeutic strategies
Precision Lethality​
​Most cancers require combination therapies to achieve durable responses, reflecting the underlying heterogeneity of tumor cell populations. Different drugs often eliminate different tumor subclones, but without lineage resolution it is difficult to determine which subclones are killed and which survive a given treatment. Instead, drug combinations are most commonly evaluated using drug synergy, defined as a greater-than-expected reduction in overall tumor viability compared to the effects of the individual drugs. While useful, synergy is a bulk measure and does not account for clonal heterogeneity or the survival of rare, resistant subpopulations.
We are developing a new strategy, termed Precision Lethality, that explicitly targets tumor heterogeneity. This approach relies on cell barcoding, where individual tumor cells are labeled with short, randomized DNA sequences that act as unique clonal identifiers. By tracking thousands of barcoded tumor subclones in parallel during drug treatment, we can directly quantify clonal survival and identify drugs with non-overlapping resistance profiles. This enables the rational selection of drug combinations that collectively eliminate resistant subclones, even in the absence of classical synergy. We apply Precision Lethality to identify treatments that complement standard-of-care chemotherapy in neuroblastoma and rhabdomyosarcoma.
Radiopharmaceutical Therapy​​
Radiopharmaceutical Therapy (RPT) provides a strategy to overcome tumor
heterogeneity. While tumor cells vary in intrinsic radiosensitivity, most are susceptible to lethal DNA damage at sufficiently high radiation doses. Importantly, effective treatment does not require uniform target expression: radiation emitted from receptor-positive cells can kill neighboring tumor cells through cross-fire irradiation, enabling eradication of small tumor clusters despite heterogeneous receptor expression.
As shown in the figure, this is illustrated by a neuroblastoma patient treated with 177 Lu-DOTATATE targeting the SSTR2 receptor, where immunohistochemistry and PET/CT imaging reveal variable target expression across primary and metastatic lesions. Through our involvement in the analysis of this clinical trial led by our collaborator Jakob Stenman, we identified substantial inter- and intra-tumoral heterogeneity, motivating our efforts to develop next-generation RPT targets with more stable expression. In parallel, we use high-throughput screening to identify neuroblastoma-specific radiosensitizers that further enhance therapeutic efficacy in the presence of heterogeneity
Differentiation Therapy​​
Differentiation therapy is an established component of neuroblastoma treatment, where retinoic acid is administered after cytotoxic therapy to suppress residual tumor cells by driving them toward a differentiated, less malignant state. However, differentiation responses are often incomplete, reflecting phenotypic heterogeneity within the tumor cell population and allowing subsets of tumor cells to persist after treatment.
As illustrated in the figure, retinoic acid induces pronounced morphological changes consistent with neuronal differentiation, yet many tumor cells remain refractory. In collaboration with Malin Wickström, we previously performed large-scale drug screens using morphology-based readouts to identify compounds that induce neuronal differentiation beyond current standards of care. We are now following up these findings with validation experiments and combinatorial CRISPR and drug screens to identify clinically relevant drug combinations and genetic dependencies that enhance differentiation across heterogeneous tumor cell populations. Promising strategies are evaluated in tumoroid models and in vivo systems, with the goal of improving differentiation-based therapy and reducing relapse driven by phenotypic heterogeneity.
Tumor types
The lab primarily focuses on pediatric cancers. Despite the huge clinical importance, development in pediatric cancer lags substantially behind adult cancer, for example in the development of targeted therapies. The lab primarily works with neuroblastoma, which is the deadliest solid pediatric cancer apart from brain cancers, and rhabdomyosarcoma, where the high-risk subtype has very low survival.