Therapeutic implications of intratumor heterogeneity for TP53 mutational status in Burkitt lymphoma

Genomic instability, defined as the tendency to acquire DNA damage determining accumulation of genomic alterations over time, is a hallmark of cancer conferring evolutionary advantages, and resulting in resistance to chemotherapy and increased metastatic potential [1]. A common mechanism determining genomic instability in tumors is oncogene-induced replication stress, leading to DNA damage accumulation during the S phase of the cell cycle [2]. Our group and others recently reported that MYC-driven cancers such as Burkitt lymphoma (BL) and Diffuse large B-cell lymphoma (DLBCL) overexpress active components of the DNA damage response pathway (DDR) such as checkpoint kinases (CHK1/2), in order to cope with the high levels of replication stress deriving from MYC overexpression, and are sensitive to pharmacologic DDR inhibition [3‐5]. BL is characterized by a high level of MYC expression due to the occurrence of chromosomal translocations which are hallmarks of the disease, and G1/S checkpoint dysfunction with frequent TP53 mutations (30 % of cases) [6, 7]. TP53 mutations drive chemoresistance in many different cancers including aggressive B-cell lymphomas [8, 9], and cooperate with MYC by preventing its intrinsic proapoptotic effects and by further increasing genomic instability [10]. Intra-tumor heterogeneity, intended as the occurrence of genomic diversities within the same tumor over space and time, is intimately related to genomic instability, and has been recently unraveled by next generation sequencing (NGS) studies [11]. Nevertheless, its clinical significance and therapeutic implications in aggressive B-cell lymphomas are yet to be elucidated.

In the current study we report a genetic and functional analysis aimed at defining the mechanisms of chemoresistance in a 43-year old woman affected by stage IVB Burkitt lymphoma with bulky abdominal masses and peritoneal effusion. The patient, despite a transient initial response to chemotherapy with reduction of the bulky masses, rapidly progressed and died of her disease. Targeted TP53 sequencing found that the bulky mass was wild-type (WT) whereas peritoneal fluid cells harbored a R282W mutation, depicting a paradigmatic example of intra-tumor heterogeneity for the TP53 mutational status at disease onset in BL. Functional studies on the TP53 mutant clone confirmed an impaired p53-mediated response and resistance to ex vivo doxorubicin administration. Finally, we demonstrated that these cells were characterized by overexpression of markers of genomic instability and DDR pathway activation, and were sensitive to pharmacologic inhibition of CHK kinases.

The patient was hospitalized in August 2011 in critical conditions with two bulky abdominal masses originating from both ovaries, a massive abdominal effusion and small bowel obstruction. Surgical biopsy of the bulky mass (left ovary), cytology of the malignant cells from ascitic fluid, and immunophenotype (CD20+, CD19+, CD10+, BCL6+, CD38+, c-MYC) led to the diagnosis of BL (Fig. 1a–d). Fluorescence in situ hybridization (FISH) on malignant cells from both bulky mass and ascitic fluid showed a t(8;22)(q24;q11) translocation involving the MYC oncogene and the lambda light chain locus (IGL) (Fig. 1e). Detailed description of Immunohistochemistry and FISH studies is available in Additional file 1. The principal comorbidity was a severe bipolar disorder and anorexia nervosa that was still active at the time of disease onset, so that the patient was severely underweight (body mass index <17 kg/m²) and deemed initially unfit for intensive chemotherapy. Initial treatment consisted in 5 days of debulking cyclophosphamide (200 mg/m²/die) followed by 1 CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) cycle, which was complicated by severe tumor lysis syndrome, and bowel perforation requiring surgical intervention (Fig. 1f–i). A computed tomography (CT) scan performed after the first cycle showed marked reduction of the bulky lesions, with persistence of the abdominal effusion (Fig. 1g). After recovering from surgery, she received two additional Rituximab-CHOP-14 cycles but a CT scan performed right after showed marked disease progression (Fig. 1h). At this point the patient underwent therapy intensification according to the B-NHL-2002 regimen [12] but was unresponsive and ultimately died of rapidly progressing disease. Since recent studies confirmed that TP53 mutations occur in about 30 % of BL cases [6, 7], in order to investigate the mechanisms underlying resistance to standard and intensive chemotherapy in this patient, we performed TP53 targeted DNA sequencing of the tumor tissue available from the initial biopsy (left ovary), of tumor cells initially collected from the ascitic fluid, and of matched normal saliva using the 454 GS Junior platform (Roche diagnostics) (Additional file 1). The patient gave informed consent for the use of surplus tissue in research, and the protocol was approved by the Institutional Review Board (Study n 12/2009/U/Tess, protocol 148/2009). We found that the tumor tissue from the initial bulky mass was entirely TP53 wild type, whereas lymphoma cells from the abdominal effusion harbored an heterozygous R282W mutation (Fig. 1j, k), which resulted in a 844C>T aminoacidic change, known to negatively affect p53 function and being associated to shorter survival in different cancer models (IARC database http://​p53.​iarc.​fr/​TP53GeneVariatio​ns.​aspx ) [13, 14]. These findings were confirmed by Sanger sequencing (Fig. 1l). Notably, both samples were chemonaive being collected before the start of chemotherapy. In order to define the impact of the R282W mutation on response to therapy in this specific case, we treated cultured TP53 mutant primary BL cells from the ascitic fluid with either DMSO 0.01 % or doxorubicin 500 nM (Fig. 2a). The TP53 wild type Hodgkin lymphoma cell line KM-H2 was used as a control. According to the TP53 status, primary mutant BL cells were resistant to doxorubicin, whereas KM-H2 cells were sensitive. Consistent with these data, doxorubicin induced p21 expression in KM-H2 cells but not in primary BL cells (Fig. 2b). Interestingly, as shown in Fig. 1, while the TP53 WT bulky masses rapidly responded to chemotherapy, the malignant TP53 mutant ascites was still present at the time of second CT scan despite multiple repeated paracenteses (Fig. 1g), indicating a similar chemoresistant behavior also in vivo. Cell viability assays were performed by using WST-1 reagent (Roche). Detailed description of western blot protocols, antibodies and reagents is available in Additional file 1.

Since we recently reported constitutive DDR activation and high efficacy of CHK inhibitors in TP53 mutant aggressive B-cell lymphomas (DLBCL and BL) [5], we evaluated the expression levels of genomic instability and DDR activation markers [5, 15] in peritoneal fluid cells and in the bulky mass by western blotting (Fig. 2c) and immunohistochemistry (Fig. 2d–i) confirming that peritoneal fluid cells demonstrated constitutive γH2AX (H2AX S139) and p-CHK1 S345 expression (Fig. 2c–e). Notably, although to a lesser extent, we observed positivity for these markers also in the TP53 WT bulky mass (Fig. 2f, g), suggesting that the acquisition of genomic instability and of a DDR+ phenotype was an intrinsic feature of this neoplasm that preceded the development of the TP53 mutation. The TP53 WT KM-H2 cells, used as negative control of DDR activation [5], were negative for both p-CHK1 and γH2AX (Fig. 2h, i). Next, in order to assess whether the TP53 mutant subclone was sensitive to DDR inhibition, we treated primary ascitic fluid BL cells (DDR+) and KM-H2 cells (DDR-) with the CHK inhibitor PF-0477736, finding that peritoneal fluid cells were exquisitely sensitive to CHK inhibition whereas KM-H2 cells were resistant (Fig. 2j). Following CHK inhibition, γH2AX levels increased in primary peritoneal fluid cells, indicating that in these cells the blockade of DDR leads to accumulation of endogenous DNA damage (Fig. 2k). These findings are consistent with a model in which constitutive activation of CHK kinases cooperates with MYC and is crucial to prevent untolerable levels of genomic instability deriving from MYC-induced replication stress and G1/S checkpoint dysfunction.

Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.