Classimed logo

Individualized Outcome Prediction for High Risk DLBCL Patients
(Discrimination versus Correlation )


G.Valet1), H.G.Höffkes2)

2) Max-Planck-Institut für Biochemie, Martinsried,
1) Medizinische Klinik III, Klinikum Fulda, Germany



1. Background: Diffuse large-B-cell lymphoma (DLBCL) represents the most frequent lymphoma in adults. Between 35 to 40 % of patients are cured by anthracyclin therapy. The relatively high therapeutic failure may be explained by the existence of lymphoma subgroups with different responsivenes to chemotherapeutic agents.
- The analysis of gene-expression profiles from RNA-expression chip arrays by correlation statistics leads to the distinction of several patient groups with different outcome prognosis (group future) (L1,L2).

2. Goal:

The calculated gene signatures are useful for patient stratification (Kaplan-Meier) and potential identification of malignancy associated metabolic pathways. They concern, however, inhomogeneus patient groups because they consist of survivors and non survivors. Consequently the analysis is only of limited value for individual patient therapy planning.

It was investigated, whether, discriminatory data pattern classification by a data sieving algorithm permits the individualized pretherapeutic identification (>95% correct) and outcome prediction for high risk DLBCL patients.

3. Results: The predictive identification of individual high risk patients is possible for the data of both studies (fig.1) (L3, L4). Predictive (fig.2) and prognostic (fig.3) classification patterns are different.

4. Conclusion: The metaanalysis of reported gene expression data by discriminating instead of correlating analysis permits the individualized pretherapeutic identification of the majority of high risk patients.

Literature References:
external links L1. Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fisher RI, Gascoyne RD, Müller-Hermelink HK, Smeland EB, Staudt LM. The use of molecular profiling to predict survival after chemotherapy for diffuse large B-cell lymphoma. NEJM 346:1937-47(2002), (external links chip data, external links patient data)
external links L2. Grau M, Lenz G, Lenz P. Dissection of gene expressiondata sets into clinically relevant interaction signatures via high-dimensional correlation maximation. Nat Comm 10:5417(2019) (external links chip data, external links patient data)
L3. Valet G, Höffkes HG. Data pattern analysis for the individualised pretherapeutic identification of high-risk diffues large B-cell lymphoma (DLBCL) patients by cytomics. Cytometry Part A 59A:232-236(2004)
L4. Valet G. Individualized outcome prediction for high risk diffuse large B-cell lymphoma (DLBCL) patients (2021).


© 2021 G.Valet
INTERNET: https://www.classimed.de
last update: Apr 24,2021
first display: Apr 02,2003