p53 therapy in a patient with Li-Fraumeni syndrome

Li-Fraumeni syndrome is an inherited genetic disorder characterized by familial clustering of multiple malignancies predominantly including sarcomas, breast cancers, brain tumors, and other diverse neoplasms (1–3). In this syndrome, patients often develop multiple primary cancers typically with initial occurrence at a young age. The genetic basis of this syndrome is a germ-line mutation in the p53 gene (2, 4). In addition to pathogenesis, defects in p53-mediated apoptotic pathways contribute to the resistance of these tumors to conventional treatment. These tumors invariably become refractory to standard therapies and result in early death (2, 3). Hence, treatment of Li-Fraumeni tumors with p53 gene transfer represents a novel, prototypical targeted cancer therapy for these neoplasms. Somatic mutations in p53 occur in the majority of sporadic, nonfamilial tumors (5). The following report describes the treatment of a Li-Fraumeni patient with Advexin adenoviral p53 gene therapy and the relationships between treatment response, radiographic findings, and molecular markers associated with p53 mechanisms of action.

The patient is a 25-year-old woman from a Li-Fraumeni family (Fig. 1). The paternal grandfather died of cancer at the age of 38 years. The patient's father had two siblings (of four) who died of sarcoma and uterine/cervical cancer at ages 24 and 34 years, respectively. The patient's father died of breast cancer at age 52 years. Of the patient's three siblings, one died at age 12 years of sarcoma and another was diagnosed with breast cancer at age 31 years. The patient presented at age 21 years with abdominal pain, anorexia, and vomiting leading to the diagnosis of granulosa cell cancer (inhibin positive). Analyses of the BRCA1/BRCA2 genes and immunohistochemistry for the mismatch repair genes were uninformative, indicating that these genes are unlikely to be involved. p53 sequencing of DNA obtained from peripheral blood revealed the presence of a heterozygous germ-line p53 abnormality with a nucleotide 151ΔG frameshift mutation (c.151delG; p.Glu51AsnX71) in exon 4 resulting in a truncated protein containing only 51 amino acids of wild-type p53 protein.

The patient was initially treated with oophorectomy followed by five cycles of bleomycin, cisplatin, and vinblastine. She subsequently developed embryonal carcinoma of the vagina (inhibin negative) 2 years later and was treated with cisplatin, etoposide, and ifosfamide initially and then switched from ifosfamide to Taxol after development of encephalopathy. She experienced progressive disease with bone invasion after three cycles of treatment and received pelvic radiotherapy (45 Gy) and whole brain radiation on development of central nervous system metastases.

The patient's pelvic disease continued to progress resulting in development of lower extremity edema and pain. Other disease sites (adrenal, brain, and mediastinum) were nonsymptomatic. Because she had exhausted all conventional chemotherapy, radiation, and surgical options, she was referred for experimental treatment with Advexin p53 therapy under a single-patient (compassionate) use protocol approved by the Food and Drug Administration and the local Institutional Review Board.

Immunohistochemistry was done to evaluate pretreatment and posttreatment expression of molecular markers associated with p53 function using an automated immunostainer (TechMate 500, Ventana Medical Systems) as described (6). Cell cycle arrest and proliferation were evaluated by antibodies specific for p21 (MS-230, Lab Vision) and Ki-67 (Mib-1, Dako). Markers relevant to the induction of apoptosis were detected by antibodies to cleaved caspase-3 (Cell Signaling) and terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling staining using the Frag EL kit (QIA33) from EMD Biosciences. Negative regulators of p53 activity, HDM2 (clone 1F2, EMD Biosciences), and bcl-2 (clone 124, Dako) were also evaluated. Expression of the coxsackie adenovirus receptor was also determined (CAR, Santa Cruz Biotechnology). Endogenous mutated p53 expression was detected with an antibody specific for the p53 NH₂ terminus (DO7, Lab Vision) whereas expression of wild-type p53 delivered by Advexin was determined with an antibody specific for the COOH terminus (ICA.9, Lab Vision) that was not present in the mutated germ-line p53 gene sequence. Photo images were taken using a Dage video camera on an Olympus BX60 microscope using a 40× or 60× objective lens.

The patient had three pelvic tumors of which only one was accessible for intralesional injection. The patient's pelvic neoplasm, contiguous with the vaginal primary, was treated by intratumoral injection of Advexin. Advexin is a replication-defective adenoviral vector containing the wild-type human p53 cDNA under the control of the cytomegalovirus immediate-early promoter (5). The patient received intratumoral Advexin at a dose of 2 × 10¹² viral particles per injection twice weekly on days 2 and 4 of week 1, then every 28 days for two additional treatments. Consistent with previous reports, Advexin p53 therapy was well tolerated (5, 7, 8) and side effects were limited to grade 1 pain at the injection site and transient fever. There were no significant differences related to Advexin treatment between pretreatment and posttreatment complete blood counts, serum chemistries, or urinalyses. Needle penetration resistance changed from firm to soft after two injections, concomitant with reduction in edema and pain. A fusion fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computed tomography (CT) scan done at 2 months, following four intratumoral injections of Advexin therapy, showed complete resolution of 2,3-FDG uptake in the treated tumor compared with the pretreatment evaluation (Fig. 2). The standardized uptake value of the injected lesion 2 months after Advexin administration was 1.1 compared with a baseline value of 5.3, whereas CT scans of the lesion showed stable disease for >4 months. FDG-PET/CT of a nontreated pelvic lesion in the same transverse plane showed progression with an increased standardized uptake value from 5.0 to 11.5 (Fig. 2). A repeat fusion FDG-PET/CT done 4 months later revealed continued complete remission by FDG uptake in the treated tumor with further progression in the untreated lesions. These observations are consistent with the clinical trial results of other agents that mediate antitumor effects through cell cycle arrest and apoptosis induction wherein PET scans indicated significant tumor responses when CT scans showed stable disease (9, 10). In multiple studies wherein the two methods were compared to evaluate treatment response, FDG-PET/CT scans were shown to be superior and provided better correlation with clinical outcomes than CT scans (10, 11). An important implication of these observations is that p53 therapy may be best evaluated by fusion FDG-PET/CT scans and that CT findings of durable stable disease may reflect meaningful therapeutic activity. In this regard, a CT angiogram done 5 months after Advexin treatment showed that the treated lesion was avascular, consistent with the complete metabolic response by fusion PET/CT scan and confirmed progression in the nontreated lesion that increased in size and vascularity.

The adenoviral antibody titer was 1:32 before treatment and increased to 1:2,048 on day 28 after four adenoviral p53 treatments. Consistent with previous findings, the antibody response did not correlate with any pathologic sequelae and did not seem to inhibit Advexin activity (8).

Successful treatment of patients with the Li-Fraumeni syndrome is problematic due to the frequency of multiple malignancies that share a common p53 genetic defect fundamental to cancer progression and the development of treatment resistance. A major paradox in the treatment of Li-Fraumeni syndrome tumors is that conventional cytotoxic therapies (chemotherapy and radiotherapy) that induce DNA damage also contribute to the high incidence of treatment related secondary malignancies in these patients (3, 12–14). It is likely that the p53 abnormalities in Li-Fraumeni syndrome tissues mediate defects in repair of DNA damage following conventional cancer treatments, thus promoting mutations that result in secondary tumor development (12–14). Hence, novel, nongenotoxic therapies for Li-Fraumeni syndrome tumors are clearly needed. Adenoviral p53 is a targeted molecular therapy that addresses the fundamental genetic defect of Li-Fraumeni syndrome and does not induce DNA damage in normal tissues (5, 15).

As a prototypical tumor suppressor gene, p53 is a transcription factor that controls several biological processes important for tumor growth control including regulation of the cell cycle, angiogenesis, and apoptosis (16, 17). This investigation identified the correlation of molecular markers of p53 tumor suppression with radiographic findings and treatment response and extended the results of previous clinical studies of p53 tumor therapy.

Molecular markers of p53 tumor suppression were evaluated by immunohistochemistry using samples from the Li-Fraumeni syndrome embryonal tumor taken before treatment and 7 days after Advexin injection. Immunostaining with an antibody directed against the NH₂ terminus of p53 (amino acids 37–45) present in the mutated p53 germ-line sequence showed constitutive mutant p53 protein expression in the nuclei of the patient's normal skin and Li-Fraumeni syndrome tumor cells (Fig. 3A). These observations are consistent with previous immunohistochemical studies showing overexpression of abnormal p53 compared with normal tissues wherein p53 is tightly regulated and undetectable by standard immunohistochemical assays (18). As expected, tumor levels of wild-type p53 increased after adenoviral p53 treatment, reflecting exogenous delivery of high levels of therapeutic wild-type p53 (Fig. 3A). The presence of increased Advexin-derived p53 was confirmed with an antibody directed to the COOH terminus of p53. This region (amino acids 388–393) is not present in the truncated mutant Li-Fraumeni syndrome p53.

The induction of cell cycle arrest and apoptosis is an important mechanism of p53-mediated tumor suppression in model systems, and activation of these pathways was shown by the expression of molecular markers in the patient's tumor. p21 is a critical mediator of cell cycle arrest (19) and expression of this marker was increased from <1% positively stained cells in the pretreatment tumor to 20% positive staining after Advexin treatment (Fig. 3B). Activation of apoptosis was shown by the detection of increased cleaved caspase-3 in posttreatment tumor samples with 20% positivity after Advexin treatment compared with <5% staining before therapy (Fig. 3B). These results were confirmed by terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling staining that detects fragmented DNA, which is another marker of apoptosis. Staining of Advexin-derived p53 corresponded to regions of positive staining for p21 and apoptosis markers in the posttreatment samples. Consistent with the induction of cell cycle arrest and increased apoptosis, staining for the cellular proliferation marker Ki-67 was decreased from 55% in pretreatment tumor samples to 35% after Advexin treatment. Staining of the antiapoptotic marker bcl-2 was low in both pretreatment and posttreatment samples and decreased from 20% to 10% after Advexin treatment. Overall, these molecular analyses indicate that the p53 signaling pathway was intact in the Li-Fraumeni syndrome tumor and that p53 gene transfer resulted in the induction of cell cycle arrest and apoptosis.

Another mechanism of p53 tumor suppression shown in tumor models is antiangiogenesis. p53 has been shown to down-regulate the expression of vascular endothelial growth factor and to activate the transcription of secreted inhibitors of angiogenesis (20, 21). The results of the posttreatment CT angiogram showing avascularity of the treated tumor and increased vascularity of a nontreated lesion are consistent with this mechanism of p53 action.

Additional molecular markers associated with the regulation of p53 activity were also evaluated. HDM2 is a negative regulator of p53 activity whereas bcl-2 is an inhibitor of p53-mediated apoptosis (17). The Li-Fraumeni syndrome tumor expressed low levels of HDM2 and bcl-2 (10–20% positivity), representing a marker profile conducive to p53 activity. Additional studies will be required to determine whether these markers are predictive of p53 treatment response in other patients.

Another interesting finding was the increased expression of coxsackie adenovirus receptors after p53 treatment. This observation suggests that adenoviral p53 may enhance its own activity by up-regulating the receptor that mediates its cellular binding and entry. Corroboration of this phenomenon and further evaluation in a larger number of patients may optimize future adenoviral p53 treatment schedules to coincide with maximum coxsackie adenovirus receptor expression.

In summary, our findings show important relationships between treatment response, radiographic findings, and molecular markers of p53 tumor suppression. Improvement of the patient's tumor-related symptoms after adenoviral p53 treatment was associated with a durable complete response by PET scan and durable stable disease by CT scan. This important result implies that, like other agents which induce cell cycle arrest and apoptosis, p53 treatment responses may be better monitored by fusion PET/CT scans rather than by CT scans alone (9–11). The absence of tumor vascularity in treated versus nontreated tumors by CT angiography was consistent with the induction of p53-mediated antiangiogenic mechanisms. Molecular markers of p53 tumor suppression documented the induction of cell cycle arrest and apoptosis by p21 and cleaved caspase-3 detection, respectively, in posttreatment tumor samples. In regard to regulators of p53 activity, the patient's tumor expressed a low HDM2 and bcl-2 marker profile predictive of p53 activity. In conclusion, our findings indicate that evaluation of additional Li-Fraumeni syndrome tumors with adenoviral p53 treatment is warranted. Corroboration of the identified relationships between treatment response, radiology findings, and molecular markers identified in this study should be assessed in a larger number of patients to guide further development of p53 therapy for both familial and nonfamilial cancers with p53 abnormalities.

We thank Drs. Guillermina Lozano, Louise Strong, and Martha French for their review of the manuscript and Wesley Gage and Dr. Hector Battifora for assistance with the immunohistochemistry studies.