A Subtoxic Microdosing Approach To Identify And Predict Resistance To Carboplatin Chemotherapy And Its Mechanisms

Objectives:Platinum (Pt)-based drugs are widely used in the treatment of a variety of cancers. Carboplatin, the second generation of Pt drug, is among the most commomly used anticancer chemotherapeutic drugs because of approximately the same spectrum of activity as cisplatin but reduced toxicity. Pt-based combination chemotherapy is the standard regimen for the treatment of non-small cell lung cancer (NSCLC) or bladder transitional cell carcinoma (TCC). However, most patients will not benefit from chemotherapy because of chemoresistance. The ability to characterize tumors for drug resistance prior to toxic chemotherapy is an unmet medical need. DNA is considered as the major target of Pt drugs. DNA damage or Pt-DNA adduct formation, is the critical step in cancer cell response to Pt chemotherapy. We hypothesize that low levels of Pt-induced DNA damage are predictive of chemoresistance, that is, cancer cells with no or low DNA damage will survive and be resistant to Pt chemotherapy. Accelerator mass spectrometry (AMS) is an ultrasensitive method for measuring carbon-14. When 14C-labeled carboplatin induces carboplatin-DNA monoadducts, carboplatin, together with 14C, is covalently linked to DNA. By measuring the amount of 14C on genomic DNA with AMS, we are able to determine the level of carboplatin-DNA adducts. Furthermore, some relevant parameters, such as drug uptake/ efflux, intracellular drug inactivation, and DNA repair, can also be measured that helps determine the mechanisms of chemoresistance. Because of the supersensitivity of AMS, all these studies can be performed after patients or cancer cells are treated with non-toxic midcrodose of 14C-labeled carboplatin. It is desirable to establish the microdosing approach to identify and predict the chemoresistance to Pt-based drugs. Predictive tests of how a tumor would react to a particular chemotherapeutic regimen would be extremely useful, since this would permit personalized chemotherapy to maximize the therapeutic effects while minimizing the risk of serious side effects and not delaying effective treatment of the disease. This proposal focuses on measurement of carboplatin-DNA adduct formation and repair in human cell lines and human cancer patients with the ultimate goal of identifying chemoresistance and determining the underlying chemoresistant mechanisms for designing of personalized therapy before patients receive toxic chemotherapy.Methods and results:Six NSCLC cell lines and five bladder TCC cell lines were studied in vitro. The levels of DNA monoadduct formation induced by subtoxic microdosing and therapeutic 14C-labeled carboplatin and its repair rate were detected by AMS at different time points, and the linear regression analysis was calculated. The data showed that microdosing 14C-labeled carboplatin at 50,000 dpm/ml and 1% of therapeutic dose could induce DNA damage, the physiological target of therapeutic carboplatin. The DNA damage induced by microdoses is linearly correlated to that of therapeutic carboplatin (r2=0.95,P<0.0001 for both lung and bladder cancer). Cellular resistance to carboplatin (IC50 value) was determined by the MTT assay. The data showed that DNA monoadduct levels at 4h time point (r2=0.77, p=0.022 for microdose; r2=0.83,p=0.011 for therapeutic dose) and AUC (r2=0.70,p=0.038 for microdose; r2=0.75,p=0.026 for therapeutic dose) induced by both microdosing and therapeutic carboplatin were linearly correlated to the IC50 values of NSCLC cell lines, suggesting the levels of DNA monoadduct induced by microdosing carboplatin can potentially identify and predict the cellular resistance to carboplatin. In addition, DNA repair rate induced by microdosing or therapeutic carboplatin was linearly correlated to the IC50 value of bladder cancer cell lines (r2=0.80, p=0.04 for microdose; r2=0.82, p=0.03 for therapeutic dose). The level of total Pt induced by therapeutic carboplatin and its repair rate in five bladder cancer cell lines were confirmed in parallel by inductively coupled plasma mass spectrometry (ICP-MS) at different time points. The data showed that the kinetics of total Pt level is consistent to that of DNA monoadduct level measured by AMS (r2=0.85,p＜0.0001), suggesting the DNA monoadduct level can substitute total adduct level to potentially identify and predict cellular resistance to carboplatin in bladder cancer cell lines. Quantitative real time PCR was performed to determine the mRNA expression levels of ERCC1 and RRM1 in six lung cancer cell lines and five bladder cancer cell lines. These two genes have been used to predict resistance to chemotherapy in NSCLC patients. The carboplatin-DNA monoadduct levels induced by both microdosing and therapeutic carboplatin treatment were superior in predicting chemoresistance, when compared to the ERCC1 and RRM1 expression levels in these 11 cell lines.Other major steps, like drug metabolism, cell uptake and efflux, and repair of DNA damage, can affect DNA damage and, therefore, cellular sensitivity to Pt chemotherapy. Measuring drug uptake/efflux and intracellular inactivation allows insights into resistance chanisms. Pt-resistant A549 cells had lower DNA damage and higher IC50 when compared to sensitive H23 cells (p<0.001). The uptake/efflux did not contribute significantly to the resistance differences between these two specific cell lines. Intracellular glutathione (GSH) is involved in intracellular inactivation of Pt in A549. Depletion of GSH with buthionine sulphoximine (BSO) increased the carboplatin-DNA monoadduct levels (p<0.05) and sensitized A549 to carboplatin (p<0.01). We also developed a Pt-resistant 5637 sub-cell line by culturing 5637 cells with carboplatin for a long term. Compared with the parental chemosensitive cells, total GSH level increased in the 5637 resistant cells (p<0.05).To determine if DNA repair is involved in chemoresistance, we used siRNA to knock down the expression of ERCC1 in the chemoresistant A549 cells. ERCC1 is one of the key proteins in the multi-subunit nucleotide excision repair complex. Downregulation of ERCC1 expression increased the DNA monoadduct levels (p<0.01), and sensitized A549 cells to carboplatin (p＜0.001). Compared with chemosensitive 5637 parental cells, the levels of MGMT mRNA and protein expression increased in 5637 resistant cells by PCR array and Western blot (p<0.05), indicating that multiple underlying mechanisms contribute to the chemoresistance.We also performed in vivo studies to determine if the microdosing approach could be translated into clinical applications. We titrated and determined the dose of 14C-labeled carboplatin needed for the in vivo studies in mice. Balb/c mice and nude mice carrying tumor xenografts were treated, through intravenous injection, with one microdose (2 mg/m2) or one therapeutic dose (200 mg/m2) of carboplatin, each containing 50,000 dpm per gram of body weight of 14C-labeled carboplatin. Blood samples were taken up after 14C-labeled carboplatin injection. Liquid scintillation counter (LSC) was used to determine the drug metabolism (pharmacokinetics) of 14C-carboplatin in plasma. Peripheral blood mononuclear cells (PBMC) were isolated for DNA extraction. Tumor xenografts were excised. The kinetics of Pt-DNA monoadduct formation in PBMC and tumor xenografts treated with microdosing or therapeutic carboplatin was linear (r2=0.94, p<0.001), with the essentially identical pharmacokinetics. This suggests that the levels of DNA monoadducts induced by microdose can potentially predict the levels of monoadducts induced by therapeutic carboplatin, and microdoses are reasonable surrogates for therapeutic doses. Based on the above results, we have opened a Phase 0 microdosing trial in clinical cancer patients. In this clinical trial, four patients received one subtoxic microdose of 14C-labeled carboplatin at 10×106 dpm/kg and the total unlabeled carboplatin of 1% the therapeutic dose. The half lives of microdosing carboplatin were 1.5 to 2 hours, identical to that of therapeutic dose. The 14C signals in PBMC, representing carboplatin-DNA monoadduct levels, were 10～100 times the background levels, the desired level for AMS analysis. This data suggests that this dose of 14C-carboplatin (10×106 dpm/kg) and total carboplatin dose (1% of therapeutic dose) is the recommended dose for future clinical trials. We did not observe any toxicity in these four patients. The radiation exposure from 14C is less than 0.2% of that from an abdominal CT scan.Conclusions:1. The identical pharmacokinetics was observed for the subtoxic microdosing and therapeutic carboplatin.2. The subtoxic microdosing carboplatin can induce genomic DNA damage in cancer cell lines in vitro, tumor xenografts in vivo, and PBMC from mice or human patients.3. The level of carboplatin-DNA monoadducts in cancer cell lines and tumor xenografts in nude mice induced by microdosing carboplatin is linearly proportional to those induced by therapeutic doses. The subtoxic microdosing approach can potentially be used to identify and predict DNA monoadduct levels induced by therapeutic carboplatin.4. The low level of DNA monoadducts induced by subtoxic microdosing carboplatin correlates with the cellular resistance to carboplatin. The subtoxic microdosing approach can potentially be used to identify and predict cellular chemoresistance to carboplatin.5. DNA monoadduct levels measured by AMS is linearly correlated to the total Pt level induced by therapeutic carboplatin measured by ICP-MS.6. The underlying mechanism, like drug metabolism, intracellular drug inactivation and DNA repair pathways can affect DNA monoadduct formation and, therefore, cellular sensitivity to Pt chemotherapy.