DOCTORAL THESIS 2012 POKORNA Abstract We have used a well characterized transplantable transgenic mouse model which mimics human acute promyelocytic leukemia (APL), both in its biological characteristics and its response to conventional therapeutic drugs. The aim of our study was to better characterize the efficacy of the combined treatment and to determine molecular markers of clinical outcome. We established a minimal residual disease monitoring based on the high sensitivity of detection of PML-RAR transcripts by polymerase chain reaction (PCR) technology in APL mice. We showed that oncogene-specific PCR-based assays allow, like in patients, the diagnosis, follow-up and prediction of disease evolution. Furthermore, PCR assay was used to assess various tissues and organs for the presence of PML-RAR-positive cells in minimal residual disease free long-term survivors. As expected, majority of mice had no measurable tissue level of PML-RAR demonstrating the efficacy of immunotherapy. However, tracking the oncogene-positive cells reveals for the first time that extramedullary PML-RAR-positive cell reservoirs such as the brain may persist and be involved in the leukemia relapse. We aimed at investigating the immune responses involved in the anti-leukemic effect of the combined immutherapy. To evaluate the...

Background: In childhood acute lymphoblastic leukemia (ALL), the cure rate has now reached 80% in the western world. Even so, 15¬–20% will die from the disease or treatment-related causes, among them children who did not present any known unfavorable features at diagnosis. Treatment of childhood ALL is risk-adapted, meaning that certain factors that are related to the child or the leukemic blasts stratifies to more or less intensive treatment. In this thesis, characteristics of the bone marrow (BM) stroma, reflecting the interaction between the leukemic cells and their microenvironment, were evaluated. The aims were to investigate these factors in relation to other known data in order to further understand the biology of leukemia, and to suggest additional risk factors that would further improve decision making for the treatment of individual children diagnosed with ALL. Methods: We retrospectively investigated microvessel density (MVD), blast-congested vessel fraction (BCVF), and degree of fibrosis – reticulin fiber density (RFD) – in sections from diagnostic BM biopsies from children diagnosed in Umeå, Uppsala, and Stockholm. RFD was also studied in BM sections from treatment day 29. Results: RFD had prognostic impact in patients with high-hyperdiploid (HeH) leukemia. Moreover, rapid reduction of RFD during induction treatment was associated with a favorable prognosis compared to slow reduction, in B-cell precursor (BCP) ALL patients. There was also a correlation between RFD at diagnosis and minimal residual disease (MRD) measured by flow cytometry on treatment day 29 in BCP patients. BCP patients with high RFD and high MVD had an unfavorable outcome compared to all other BCP patients. In addition, MVD and RFD were both associated with immunophenotype, and MVD with cytogenetic aberrations. There was a correlation between MVD and WBC count in BCP high-risk patients. There was also a strong correlation between BCVF and WBC count in all BCP patients, but not between BCVF and MVD or RFD. There was a negative correlation between MVD and in vitro cellular resistance to several drugs in BCP patients. A drug-resistance score combining the drugs most strongly correlated to MVD – cytarabine, doxorubicin, and dexametasone (ADD score) – identified the prognostic potential of ADD score in HeH patients with no unfavorable features. Conclusions: Taken together, these studies indicate that stroma factors in leukemia are related to both phenotypic and genotypic features of acute leukemia. Stroma factors also seem to influence the response to induction treatment, in vitro drug resistance, and outcome in certain subgroups of childhood ALL patients. The results emphasize the importance of BM stroma in leukemia and the need for greater use of BM biopsy at diagnosis.

The diagnosis and monitoring of AML relies predominantly on the identification of blast cells in the bone marrow and peripheral blood. While at the time of diagnosis the identification of leukemic cells is relatively easy, during remission the identification of small numbers of blasts is problematic. This is most evident by the fact that patients who achieve complete remission frequently relapse, despite pathologic examination indicating a marked reduction in leukemic cell burden. In this thesis I have explored the potential of using serum proteins secreted by leukemic cells as a means of monitoring disease in patients. To identify proteins that might be useful for monitoring, I took advantage of published gene expression arrays and looked into online bioinformatics databases. Using specific characteristics, I was able to identify approximately 107 candidate proteins secreted by AML cells. RT-PCR analysis and ELISA assays were performed to evaluate the variability of expressions and serum level differences of twelve different proteins in the list.

The diagnosis and monitoring of AML relies predominantly on the identification of blast cells in the bone marrow and peripheral blood. While at the time of diagnosis the identification of leukemic cells is relatively easy, during remission the identification of small numbers of blasts is problematic. This is most evident by the fact that patients who achieve complete remission frequently relapse, despite pathologic examination indicating a marked reduction in leukemic cell burden. In this thesis I have explored the potential of using serum proteins secreted by leukemic cells as a means of monitoring disease in patients. To identify proteins that might be useful for monitoring, I took advantage of published gene expression arrays and looked into online bioinformatics databases. Using specific characteristics, I was able to identify approximately 107 candidate proteins secreted by AML cells. RT-PCR analysis and ELISA assays were performed to evaluate the variability of expressions and serum level differences of twelve different proteins in the list.

Le myélome multiple (MM) est une hémopathie maligne caractérisée par la prolifération d’un clone plasmocytaire tumoral dans la moelle osseuse et d’une accumulation d’une immunoglobuline monoclonale dans le sérum et/ou les urines. La diversité des anomalies cytogénétiques, rendant la maladie plus ou moins agressive, et la variabilité de la réponse au traitement font du MM une maladie hétérogène. Le MM reste incurable dans la majorité des cas avec une médiane de survie de 5 à 7 ans. La rechute après traitement est due à la persistance de cellules tumorales au sein de la moelle osseuse, appelée maladie résiduelle ou en anglais « Minimal Residual Disease » (MRD). La cytométrie en flux multiparamétrique (CFM) est une technique sensible, simple et rapide qui permet d’identifier et de caractériser des cellules d’intérêt dans un échantillon biologique. C’est dans le but de simplifier le suivi de la MRD du MM que nous avons développé une solution complète basée sur la CFM. Cette solution comprend (i) la conception d’un panel à 5 couleurs, composés des anticorps (Ac) anti-CD38, anti-kappa et anti-lambda pour identifier les plasmocytes totaux et de deux pools d’Ac couplés au même fluorochrome (anti-CD19/anti-CD27, pool négatif et anti-CD56/anti-CD117/anti-CD200, pool positif) pour détecter les plasmocytes tumoraux ; (ii) le développement d’un préparateur afin d’automatiser l’ensemble des étapes de préparation de l’échantillon ; et (iii) l’automatisation de l’analyse des résultats de CFM grâce à un logiciel que nous avons créé. Cette solution simple et entièrement automatisée permet d’augmenter la reproductibilité et la productivité, de diminuer le coût du test, sans altérer la sensibilité ou la spécificité. En parallèle de ces travaux, nous avons construit un score de risque simple basé sur l’expression de gènes codant pour des protéines de surface (CD24, CD27, CD36 et CD302) permettant de prédire la survie des patients atteints de MM au diagnostic ainsi qu’à la rechute. / Multiple myeloma (MM) is a hematological malignancy characterized by clonal plasma cell proliferation in bone marrow and abnormal monoclonal immunoglobulin accumulation in the serum and/or urine. The heterogeneity of the disease is partly due to the cytogenetic abnormalities diversity making the disease more or less aggressive. MM is incurable in the majority of cases with a median survival between 5 and 7 years. The persistence of abnormal plasma cells in bone marrow after treatment is called minimal residual disease (MRD) and leads to the patient relapse. Multiparametric flow cytometry (MFC) is a sensitive, simple and fast technique to identify and characterize cells of interest in biological samples. In order to simplify MRD follow-up we have developed a complete solution based on MFC. This solution includes (i) the 5-color panel design, composed of anti-CD38, anti-kappa and anti-lambda antibodies (Ab) to identify total plasma cell population and two pools of Ab paired to the same fluorophore (anti-CD19/anti-CD27, negative pool and anti-CD56/anti-CD117/anti-CD200, positive pool) to detect abnormal plasma cells; (ii) the development of a device used to automatically prepare biological samples before MFC; and (iii) the analysis automation of MFC results using a homemade software. This fully automated solution increases reproducibility and productivity, decreases processing and analyzing time as well as test cost, without affecting sensibility and specificity. In parallel, we have built a simple risk score based on gene expression encoding surface proteins (CD24, CD27, CD36 and CD302) providing MM patient outcome at diagnostic and MRD follow-up.

Indiana University-Purdue University Indianapolis (IUPUI) / Triple-negative breast cancers (TNBCs) are pathologically defined by the absence of estrogen, progesterone, and HER2 receptors. Compared to other breast cancers, TNBC has a relatively high mortality. In addition, TNBC patients are more likely to relapse in the first few years after treatment, and experiencing a shorter median time from recurrence to death. Detecting the presence of tumor in patients who are technically “disease-free” after neoadjuvant chemotherapy and surgery as early as possible might be able to predict recurrence of patients, and then provide timely intervention for additional therapy. To this end, I applied the analysis of “liquid biopsies” for early detection of minimal residual disease (MRD) on early-stage TNBC patients using next-generation sequencing. For the first part of this study, I focused on detecting circulating tumor DNA (ctDNA) from TNBC patients after neoadjuvant chemotherapy and surgery. First, patient-specific somatic mutations were identified by sequencing primary tumors. From these data, 82% of the patients had at least one TP53 mutation, followed by 16% of the patients having at least one PIK3CA mutation. Next, I sequenced matched plasma samples collected after surgery to identify ctDNA with the same mutations. I observed that by detecting corresponding ctDNA I was able to predict rapid recurrence, but not distant recurrence. To increase the sensitivity of MRD detection, in the second part I developed a strategy to co-detect ctDNA along with circulating tumor RNA (ctRNA). An advantage of ctRNA is its active release into the circulation from living cancer cells. Preliminary data showed that more mutant molecules were identified after incorporating ctRNA with ctDNA detection in a metastatic breast cancer setting. A validation study in early-stage TNBC is in progress. In summary, my study suggests that co-detection of ctDNA and ctRNA could be a potential solution for the early detection of disease recurrence. / 2021-08-05

The ability to detect residual levels of leukemic blasts (measurable residual disease, MRD) has already been integrated in the daily routine for treatment of patients with chronic myeloid and acute lymphoblastic leukemia. In acute myeloid leukemia (AML), a variety of mostly retrospective studies have shown that individuals in AML remission who tested positive for MRD at specific time-points or had increasing MRD levels are at significantly higher risk of relapse and death compared to MRD-negative patients. However, these studies differ with respect to the “MRD-target”, time-point of MRD determination, material analyzed, and method applied. How this probably very valuable MRD information in individual patients may be adapted in the daily clinical routine, e.g., to separate patients who need more aggressive therapies from those who may be spared additional—potentially toxic—therapies is still a work-in-progress. With the exception of MRD assessment in acute promyelocytic leukemia (APL), the lack of randomized, prospective trials renders MRD-based decisions and clinical implications in AML a difficult task. As of today, we still do not have proof that early intervention in MRD-positive AML patients would improve outcomes, although this is very likely. In this article, we review the current knowledge on non-APL AML MRD assessment and possible clinical consequences.

Analysis of donor chimerism has become a routine method for the documentation of engraftment after allogeneic hematopoietic stem cell transplantation (HSCT). In recent years several groups have also focused on the application of this technique for the detection of relapsing disease after allogeneic HSCT. This review addresses technical issues (sensitivity, specificity) and discusses the advantages and limitations of methods currently used for chimerism analysis and their usefulness for the detection of MRD. In addition, the potential impact of novel procedures, e.g. subset chimerism or real-time PCR-based procedures, is discussed. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Traditionally, response to treatment in hematological malignancies is evaluated by light microscopy of bone marrow (BM) smears, but due to more effective therapies more sensitive methods are needed. Today, detection of minimal residual disease (MRD) using immunological and molecular techniques can be 100 times more sensitive than morphology. The main aim of this thesis was to compare and evaluate three currently available MRD methods in childhood acute lymphoblastic leukemia (ALL): (i) real-time quantitative PCR (RQ-PCR) of rearranged antigen receptor genes, (ii) multicolor flow cytometry (FCM) of leukemia-associated immunophenotypes and (iii) real-time quantitative PCR of fusion gene transcripts (RT-PCR). In paper I, we assessed the applicability of RQ-PCR in a population-based cohort of childhood ALL diagnosed in Sweden between 2002-2006. Clonal IG/TCR rearrangements were identified in the 96% of the 279 ALL cases. Using RQ-PCR, the quantitative range of 10-3 was reached in 93% of B-cell precursor (BCP) ALL and 86% of T-cell ALL (T-ALL) by at least one target gene. In paper II, we compared MRD detection using both RQ-PCR and FCM in the context of NOPHO ALL-2000 protocol. By applying the stratification threshold of ≥0.1% MRD late during induction therapy (day 29), we could demonstrate that both methods can predict the risk of BM relapse but not extramedullary relapse. However, the threshold of ≥0.2% MRD appears to be more optimal using RQ-PCR in BCP ALL, whilst in T-ALL, the results indicate that RQ-PCR is preferable for MRD assessment. The stability of RNA in vitro is a critical factor when using sensitive molecular techniques such as MRD detection. In paper III, we evaluated the influence on MRD detection when blood is collected in tubes with RNA stabilization reagents (PAX gene Vacutatiner®) compared to collection in EDTA-tubes (non-stabilized). We analyzed 68 matched samples from chronic myeloid leukemia patients and the results indicated that non-stabilized blood processed within 30 hours is preferable for MRD detection. In paper IV, follow-up samples from eight children with Philadelphia positive (Ph+) ALL were evaluated with the three available MRD methods. MRD measured by the fusion gene transcripts (BCR-ABL1) appeared to be the most sensitive method, however, precise quantification can be difficult and the other methods are thus complementary. In conclusion, all three applied MRD methods are useful and correlate to each other, although not necessary exchangeable in individual patients. We also conclude that MRD assessment by RQ-PCR, based on rearranged IG/TCR genes and multicolor FCM are predictive for identification of high risk childhood ALL patients.

La cryoconservation de tissu ovarien peut être proposée, avant traitements hautement gonadotoxiques, à des patientes afin de préserver leur fertilité. L'autogreffe de tissu ovarien est actuellement la seule méthode de réutilisation du tissu ovarien disponible, mais en cas de pathologie néoplasique, elle présente un risque de réintroduction d'éventuelles cellules malignes via le greffon. L'objectif de ce travail a été de développer une méthode pour détecter la maladie résiduelle (MRD) dans le tissu ovarien par cytométrie en flux multicouleurs (CMF) en cas de leucémie aiguë. Une technique de dissociation de cortex ovarien a été mise au point à partir de tissu ovarien de référence provenant de résections percoelioscopiques. Un modèle expérimental de détection de la MRD a été validé et consistait à ajouter des cellules de leucémie aiguë lymphoblastique (LAL) ou myéloïde (LAM) à une suspension de cellules ovariennes isolées de référence. La méthode a ensuite été utilisée pour rechercher la MRD dans le tissu ovarien cryoconservé de 11 patientes atteintes de leucémie aiguë (7 LAL et 4 LAM).Le modèle expérimental a permis de valider une sensibilité de 10"4 et une spécificité élevée tant pour les LAL que pour les LAM. Lorsqu'un marqueur moléculaire était disponible pour la recherche de la MRD, nous avons observé une bonne corrélation entre la CMF et la PCR quantitative. La détection par CMF de la MRD dans le tissu ovarien des patientes leucémiques était positive chez 3 des 11 patientes étudiées.Cette technique est essentielle pour évaluer le risque carcinologique avant de proposer la réutilisation du tissu ovarien cryoconservé par technique d'autogreffe. / Ovarian cryopreservation together with autograft of frozen/thawed ovarian tissue is a real option to preserve and restorefertility in cancer patients. However in cases of leukemia, there is a real concern regarding the presence of metastaticcells in the ovarian tissue, which could lead to the recurrence of the primary disease. The aim was to validate multicolorflow cytometry (MFC) as an original technique for minimal residual disease (MRD) detection in ovarian cortex fromacute leukemia patients.We developed an experimental model which consisted in adding serial dilutions of leukemic cells into isolated ovariancell suspensions obtained from healthy cortex. The modelization was validated for acute lymphoblastic leukemia (ALL)and acute myeloid leukemia (AML). Then the method was applied to MRD detection of leukemic cells in cryopreservedovarian cortex from 11 leukemia patients (7 ALL and 4 AML).This experimental model made it possible to obtain a high specificity and a robust sensitivity of 10~4 for MRD detectionby MFC for both types of acute leukemic cells. When a molecular marker was available, we observed a good correlationbetween CMF and quantitative PCR. Ovarian MRD was positive by MFC in one T-ALL and 2 AML patients.MRD detection in the ovarian cortex is essential to evaluate the risk of cancer reseeding before ovarian tissue autograft.