International experts agree: tumor tissue and normal tissue must be examined and compared for tumor diagnostics. The examination of tumor tissue alone leads to misdiagnoses.
CeGaT has been following this approach for 10 years. Tumor and normal tissue are always examined and compared. This is the only way to reliably distinguish somatic (tumor-specific) from germline variants (inherited variants). A medical report is then drawn up on this basis. The CeGaT report not only shows the tumor-specific variants and relevant germline alterations; it also always gives specific scores for TMB, MSI, and HRD. Of course, tumor-specific therapy recommendations are included as well as information on the drugs and a representation of the affected pathways.
Knowledge of tumor-specific variants is essential for a therapy recommendation. If these are not correctly determined because only tumor tissue was examined, therapies may be proposed that also attack healthy cells. In addition, the total number of changes in the tumor (TMB) cannot be correctly determined, which complicates the indication for immunotherapies.
The following recent studies confirm our approach:
TMB (Tumor Mutational Burden) – Overestimation due to lack of comparison with normal tissue.
Nassar et al. 20221 have shown that tumor tissue sequencing alone can lead to overestimated TMB values. This applies in particular, but not exclusively, to non-European patients. The reason for this TMB overestimation is the missing comparison of the sequencing data of the tumor with the data of the patient’s normal tissue. In tumor-only panels, the germline variants are identified by bioinformatic comparison with reference data (e.g., GnomAD), which are predominantly based on data from patients of European descent. Asian or African decendant patient data is underrepresented in the database, so germline variants are incorrectly assigned to somatic and thus tumor-specific variants. In addition, the heritability of a tumor disease is often overlooked.
Consequently, human genetic counseling of patients, and in some cases family members, about the existing familial predisposition is impossible. In addition, necessary screening examinations cannot be offered, which may lead to avoidable late detection of cancer in family members.
Furthermore, Nasser et al. demonstrated that ancestry bias in TMB classification is directly reflected in treatment outcomes. Patients with non-small cell lung cancer and overestimated TMB levels did not show improved survival after immune checkpoint treatment (ICI) compared to patients whose TMB levels were correctly estimated thanks to paired tumor-normal comparison. Tumor-normal matching might avoid ineffective ICI therapies with side effects, and appropriate treatments could be applied directly without losing time.
To avoid TMB misestimation, the authors call for sequencing of tumor and normal tissue.
Tumor-to-normal tissue comparison helps interpret pathogenic germline mutations.
In a study with more than 17,000 patients, Srinivasan et al. 20212 investigated the extent to which cancers that occur in patients with inherited pathogenic and hence, potentially disease-causing variants, are caused by a germline defect. Their comprehensive analyses of tumor and normal tissues indicate that the mere presence of a germline mutation, even in a canonical/common cancer, does not mean that the tumor is caused or primarily driven by the altered germline allele. Instead, the penetrance of a germline alteration in a particular gene, the zygosity of alterations, and the presence of a somatic “second hit” are essential determinants, among others. This misinterpretation can lead to completely inefficient therapeutic proposals. An example would be a BRCA2 mutation that did not cause the tumor development but is present by chance. A tumor-to-normal tissue comparison clearly shows this by the absence of HRD (homologous recombination deficiency). PARP inhibition would not be effective here.
Therefore, to obtain a complete picture of a tumor and to be able to determine the role of germline variants in disease pathogenesis, as well as to identify somatic alterations as drivers in a given tumor, it is necessary to identify the somatic alterations by comparing tumor and normal tissue, and to include them in the interpretation of the germline variants. Only the complete picture makes it possible to say with certainty what really is driving a tumor and, thus, what is a promising therapeutic molecular target.
Conclusion
Both studies underline the necessity for comparing tumor and matched normal tissue to develop promising therapy recommendations for the patient.
Furthermore, it is ensured that family-relevant germline variants are not overlooked and can thus be considered in patient care or genetic counseling. This is the only way to address possible consequences for the affected person and other family members if a hereditary variant is detected.
1Nassar et al. (2022). Cancer Cell 40, 1161–1172.
2Srinivasan, P., Bandlamudi, C., Jonsson, P. et al. (2021). The context-specific role of germline pathogenicity in tumorigenesis. Nat Genet 53, 1577–1585.