A Rational Therapeutic to Enhance Apoptosis in Therapy of Lung Cancer
Bcl-2 protein inhibits apoptosis and confers resistance to treatment with traditional cytotoxic chemotherapy, radiotherapy, and monoclonal antibodies. Oblimersen sodium is an antisense oligonucleotide compound designed to specifically bind to human bcl-2 mRNA, resulting in catalytic degradation of bcl-2 mRNA and subsequent decrease in bcl-2 protein translation. Both small cell and non-small cell lung cancer show baseline and inducible expression of bcl-2, which may contribute to resistance to therapy. Preclinical studies have shown that combining bcl-2 antisense with chemotherapy improves antitumor response, increases apoptosis of tumor cells, and increases survival. Preliminary data from a large international randomized trial in melanoma show a trend toward increased survival and significantly improved response rates and response duration when oblimersen is added to dacarbazine. Phase I studies in small cell lung cancer patients demonstrate that oblimersen can be combined with paclitaxel or carboplatin and etoposide. The combination of docetaxel and oblimersen has been shown to be feasible in Phase I studies and is currently undergoing evaluation in comparison with docetaxel alone as first-line salvage therapy in patients refractory or relapsed after one prior chemotherapy regimen. Enhancement of the efficacy of anticancer treatments with oblimersen bcl-2 antisense therapy represents a promising new apoptosis-modulating strategy.
BCL-2: A CRITICAL ANTIAPOPTOTIC PROTEIN
The development of cancer requires dysregulated proliferation combined with dysregulation of cell death. The components of the apoptotic program are therefore targets for anticancer therapy (1, 2, 3, 4) . The apoptotic control mechanisms of cells can be simplified to proapoptotic stimuli and antiapoptotic forces. Three families of apoptotic control proteins have been described based on their sequence homology, three-dimensional structure, and function: (a) proapoptotic proteins including Bax and Bak; (b) BH3 only proteins that are proapoptotic; and (c) antiapoptotic proteins including bcl-2 (2 , 3 , 5, 6, 7, 8) . The mechanisms of interaction between these classes of proteins continue to be described.
Located on the inner mitochondrial membrane, bcl-2 serves as a key inhibitor of apoptosis that blocks release of cytochrome c and maintains mitochondrial integrity (5 , 8, 9, 10) . By inhibiting apoptosis, bcl-2 confers resistance to treatment with traditional cytotoxic chemotherapy, radiotherapy, and monoclonal antibodies (11 , 12) . In studies of human cancer, most evidence suggests that bcl-2 contributes to a more malignant tumor phenotype. Elevated bcl-2 protein correlates with poor response to chemotherapy and/or hormonal therapy in non-Hodgkin’s lymphoma, acute myelogenous leukemia, multiple myeloma, and prostate cancer (13, 14, 15, 16, 17, 18) . In xenograft models, nontumorigenic cell lines can be made highly tumorigenic by transfection with the bcl-2 gene (14 , 19 , 20) .
BCL-2 EXPRESSION IN NON-SMALL CELL LUNG CANCER (NSCLC)
Observations made in NSCLC suggest that the relationship between bcl-2 expression and tumor phenotype is complex and multifactorial. Most prior studies of bcl-2 expression in NSCLC have been conducted using immunohistochemistry on formalin-fixed, paraffin-embedded archival tissue. Contrary to the experience in hematological neoplasms and prostate cancer, early studies suggested that detection of bcl-2 in NSCLC by immunohistochemistry was associated with a lower risk of metastatic disease and possibly improved overall survival (21 , 22) . However, none of these early studies were correlated with the type of chemotherapy administered to patients. More recently, one study examined 34 tumor samples taken from patients with advanced NSCLC who were treated with the combination docetaxel and vinorelbine. In that study, 16% of cases were positive for bcl-2 expression, but there was no apparent correlation with response to therapy (23) . In part, the small sample size of prior studies has limited proper statistical analysis. Unexamined variables of potential importance include the relative expression of other genes, stage of disease, source of biopsy specimen from a primary tumor or metastatic site, and whether the biopsy is from a previously treated or chemotherapy-naive patient. In lung cancer cell lines, bcl-2 antisense reduced bcl-2 protein expression, enhancing apoptotic activity of standard anticancer drugs (24, 25, 26) .
The medical literature is inconsistent regarding the importance of bcl-2 alone as a clinical prognostic factor. Various methods (e.g., immunohistochemistry or Northern or Western analysis) have yielded nonuniform results. Other problems have included contamination of solid tumor biopsies with epithelial, blood, and stromal cells and a generally low standard for quality control, particularly with regard to basal levels of expression and diurnal variation (27) . More recent studies have assessed relative imbalances in anti- versus proapoptotic protein pools, some of which suggest that a high ratio of bcl-2:Bax may be clinically more informative. Overall, few firm conclusions can be drawn from these studies.
ANTISENSE THERAPY TO TARGET BCL-2
Oblimersen sodium (G3139, Genasense; Genta Inc., Berkeley Heights, NJ) is an antisense phosphorothioate oligonucleotide compound designed to specifically bind to the first six codons of the human bcl-2 mRNA sequence, resulting in degradation of bcl-2 mRNA and subsequent decrease in bcl-2 protein translation and intracellular concentration (Fig. 1<$REFLINK> ; Refs. 28 and 29 ). Oblimersen is the first oligonucleotide to demonstrate proof of principle of an antisense effect in human tumors by the documented down-regulation of the target bcl-2 protein (30) .
PRECLINICAL ACTIVITY OF OBLIMERSEN
A growing body of preclinical and clinical evidence suggests that oblimersen synergizes with many cytotoxic and biological/immunotherapeutic agents against a variety of hematological malignancies and solid tumors. Studies of oblimersen in xenograft models have shown marked enhancement of the efficacy of standard cytotoxic chemotherapy and of rituximab in several cancers including non-Hodgkin’s lymphoma, melanoma, breast cancer, gastric cancer, and NSCLC (Table 1<$REFLINK> ; Ref. 31 ). Durable regressions of aggressive human breast cancer xenografts have been observed after combination therapy with docetaxel and oblimersen (32) . Oblimersen down-regulates bcl-2 mRNA within 48 h and protein levels within 96 h in ex vivo-treated myeloma cells (33) . This down-regulation is associated in a sequence-specific manner with sensitization of myeloma cells to cytotoxic activity of dexamethasone and doxorubicin (33 , 34) .
CLINICAL STUDIES OF OBLIMERSEN
Phase I/II trials indicate that oblimersen provides biologically relevant plasma levels, down-regulates target bcl-2 protein within 3–5 days of initiating treatment, and yields an acceptable safety profile. The most common toxicities are low-grade fever that is usually self-limiting and fatigue, particularly with longer durations of infusion (35 , 36) .
In a small pilot study, 12 patients with chemorefractory small cell lung cancer (SCLC) received paclitaxel (150 mg/m2 on day 6 of a 21-day cycle) plus oblimersen [3 mg/kg on days 1–8 of the cycle (36)] . All had previous treatment with etoposide plus a platinum; five patients had been treated with 1–3 additional chemotherapy regimens, and four patients had progressed after prior paclitaxel treatment. No objective responses occurred on the combination paclitaxel/oblimersen regimen; four patients had stable disease after 2 treatment cycles, with two of the four patients progressing within 1 month and terminating therapy with cycle 3. Of the two remaining patients, one maintained stable disease until cycle 6, and one remained stable over 10 cycles of therapy and free of progression for over 1 year. It is of interest that only the patient with prolonged stable disease had consistently high plasma oblimersen levels.
Given the very poor prognosis for chemorefractory SCLC, these findings were sufficiently promising to prompt a dose-finding study in 16 previously untreated SCLC patients (37) . Patients were divided into three cohorts to receive either of the following regimens: (a) 5 mg/kg oblimersen on days 1–8 of a 21-day cycle, carboplatin at a dose of area under the curve 6 on day 6, and 80 mg/m2 etoposide on days 1–8 of cycle; (b) 5 mg/kg oblimersen on days 1–8 of a 21-day cycle, carboplatin at a dose of area under the curve 5 on day 6, and 80 mg/m2 etoposide on days 1–8 of cycle; or (c) 7 mg/kg oblimersen on days 1–8 of a 21-day cycle, carboplatin at a dose of area under the curve 5 on day 6, and 80 mg/m2 etoposide on days 1–8 of cycle. Of 14 evaluable patients, partial response was seen in 12 patients, and stable disease was seen in 2 patients. Dose-limiting toxicity (grade 4 neutropenia) occurred in two of three evaluable patients in cohort 1, but these patients were able to continue with a dose reduction of the carboplatin to area under the curve 5; one grade 3 neutropenia occurred in cohort 2; in cohort 3, five of six evaluable patients experienced at least one episode of neutropenia (grade 3 or 4), and four of six patients experienced thrombocytopenia (grade 3), with toxicity observed primarily in later cycles. The majority of patients in all three cohorts were able to complete all six planned cycles of therapy.
The dose of 7 mg/kg/day oblimersen is currently being studied in combination with carboplatin and etoposide as front-line therapy in 50 patients with SCLC in a Phase II trial by the Cancer and Leukemia Group B (CALGB 30103 trial). Oblimersen is also being evaluated in conjunction with docetaxel as second-line therapy in relapsed or refractory stage IIIB-IV NSCLC. This is a multicenter randomized trial in which patients will be randomized to docetaxel (75 mg/m2 on day 5) alone or docetaxel plus oblimersen (7 mg/kg on days 1–8) for up to eight 21-day cycles. The primary end point will be survival, with tumor response and time to progression as secondary end points (Fig. 2)<$REFLINK> .
Phase I/II studies of oblimersen have also been undertaken in melanoma (33 , 38) , prostate cancer (39, 40, 41) , and refractory acute leukemias (42) . Dose-limiting toxicity of thrombocytopenia was reached at 12 mg/kg/day for 5 days when combined with dacarbazine 1000 mg/m2. Samples from tumors in these studies have shown decreases in bcl-2 protein after oblimersen therapy. Randomized clinical trials are currently under way to evaluate the efficacy and tolerability of oblimersen in combination with cytotoxic chemotherapy in lung cancer, chronic lymphocytic leukemia, multiple myeloma, and malignant melanoma. In addition, nonrandomized trials are under way to evaluate oblimersen in combination with different classes of chemotherapy agents and monoclonal antibodies in gastric, colon, breast, hepatocellular, prostate, and Merkel cell cancers; non-Hodgkin’s lymphoma; acute myeloid leukemia; chronic myelogenous leukemia; and multiple myeloma.
Dr. Mark Socinski: In the second-line trials of docetaxel plus or minus oblimersen, what are the statistical assumptions now with that number of patients? Was survival the primary end point?
Dr. Roy S. Herbst: Yes, it was based on survival.
Dr. Raymond DuBois: With this particular pathway, there are a lot of other antiapoptotic members such as MCL1. Have you seen any change in their expression when you knocked down bcl-2?
Dr. Herbst: We have not looked at that because we have been doing only a clinical study. Preclinically, that’s something that needs to be investigated: there are a number of different mechanisms and different members of this pathway. Some of the small molecules that are available now might also come into play.
Dr. Ramaswamy Govindan: That was going to be my question. If other signals can compensate, then whatever you knock out may not make a big difference. Given that this inhibitor is very specific, the initial signals are not that stellar.
Dr. Herbst: My take would be the following: we have a drug that we know is pharmacologically deliverable. We know that it’s safe. We have some evidence in a large trial that it actually works. I think this is now our chance in lung cancer to do one of two things. We can take this to a big Phase III trial, and maybe we need to do that, but we should also consider a small 30-patient biological trial and actually collect tissue and blood monocytes and ask if we are hitting the target and if we are seeing activity as a single agent. Perhaps now is the time to ask more scientific questions.
Dr. Geoffrey Shapiro: In SCLC, is bcl-2 there de novo, or is it only there when resistance develops? Because if it’s there de novo, then one might argue that it is not really preventing cell death because we have very good response rates upfront.
Dr. Herbst: So perhaps we can then improve on an already good response rate or help with resistant disease? Again, there are all these unknowns that need to be addressed.
Presented at the First International Conference on Novel Agents in the Treatment of Lung Cancer, October 17–18, 2003, Cambridge, Massachusetts.
Requests for reprints: Roy S. Herbst, Thoracic Medical Oncology, M. D. Anderson Cancer Center, The University of Texas, 1515 Holcombe Boulevard-432, Houston, TX 77030-4009. Phone: (713) 792-6363; Fax: (713) 796-8655; E-mail: