Influence of chemoresistance and p53 status on Fluoro-2-deoxy-D-glucose incorporation in cancer

Tim A D Smith

Read Also

Resistance of cancers to chemotherapy and the presence of p53 mutations are both associated with poor prognosis. A considerable number of clinical studies have examined the influence of these two factors on FDG incorporation with a view to using FDG-PET as a non-invasive determinant of prognosis. The findings from these studies are summarised below along with the in-vitro cell studies which help to explain the respective associations between FDG incorporation, p53 status and chemoresistance.

Resistance to chemotherapy

Cancer cure is frequently abrogated either by inherent, or acquired drug resistance. An inherently resistant tumor is one that shows little or no sensitivity to therapeutic agents without prior exposure to the drug in question. Acquired resistance is where the tumor is initially sensitive to treatment but becomes increasingly unresponsive to the agent. Even very responsive tumors can develop drug resistance during the course of their treatment [1].

Each cancer cell has thousands of genetic errors [2] and some will induce differing states of drug sensitivity. When the tumor is exposed to a specific chemotherapy agent, cells resistance to the drug will dominate. Residual tumor disease post-chemotherapy is often associated with the presence of subsets of cells resistant to the chemotherapy agent [1,3]. For example patients with ovarian cancer have high response rates to initial chemotherapy after cytoreductive surgery but most develop resistance to chemotherapy during the course of their treatment [3].

Mechanisms of drug resistance

There are a number of mechanisms for drug resistance. These include changes in the rate of drug uptake and efflux, altered drug metabolism, decreased drug-target complex formation, enhanced DNA repair mechanisms [4].

Anticancer drugs enter cells by diffusion or trans-membrane transport proteins. Drug resistance can result from modulation of membrane receptors e.g. the anti-folate agents methotrexate and tomudex enter the cell via the reduced folate carrier (RFC) and resistance to methotrexate has been shown to be associated with decreased expression of RFC [5]. Some chemotherapeutics directly affect target molecules such as the platinum drugs that interact directly with DNA. Resistance to platinum drugs can be due to increased tumour cell levels of molecules that inactivate these compounds such as glutathione [6]. Drugs such as 5fluorouracil (5FU) require chemical modification. 5FU is anabolised to a thymidylate synthase inhibitor although most 5FU is catabolised by dihydropyrimidine dehydrogenase (DPD) and 5FU-resistant cells have been shown to express higher than normal levels of DPD and increased levels of thymidylate synthase [5]. A further mechanism of resistance is induced by changes in the protein target of some drugs e.g. resistance to paclitaxel and docetaxel include acquired mutations at the drug binding site of tubulin and differential expression of tubulin isoforms [7,8] Increased efflux of drugs including taxanes and anthracyclines can result from expression of the ABC transporter proteins such as P-glycoprotein. Multi drug resistance (MDR) is the state in which tumor cells are resistant to a wide range of different chemotherapy drugs including doxorubicin (an anthracycline), alkaloids (e.g colchicine) due to the presence of the plasma membrane protein P-glycoprotein (P-gp). Changes in the apoptotic signal induction pathways are also involved in chemotherapy-resistance [6].

P53 status, drug resistance and prognosis

To prevent neoplastic formation, cellular P53 protein responds to DNA damaging cellular insults by either orchestrating DNA repair mechanisms or, where DNA damage is irreparable, by induction of apoptosis pathways. In view of its crucial role in maintaining DNA integrity p53 protein is referred to as the ‘Guardian of the genome’. In the absence of cell stress p53 protein levels are low but X-rays, UV, DNA-damaging chemotherapy drugs, DNA synthesis inhibitors, disruptors of microtubule components, hypoxia, myc introduction into cell, depletion of intracellular nucleotide precursor pools  all cause increased p53 within minutes. Loss of p53 function through mutation is associated with many cancer types [9]. Since most anti-cancer chemotherapy works through induction of apoptosis, loss of p53 function through mutation is associated with resistance to some drugs including doxorubicin [10] and cisplatin [11]. Further therapy response and survival have been shown to be detrimentally influenced by p53 abnormality in patients with leukemia [12], lymphoma [13] and epithelial ovarian cancer [14].

[¹⁸F]Fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET)

Positron emission tomography (PET) using the glucose analogue fluoro-2-deoxy-D-glucose (FDG-PET) is becoming a routine tool for probing tumour metabolism non-invasively and exploits the enhanced glucose utilization characterized by tumour cells. The fluorinated glucose analogue, ¹⁸F labelled glucose analogue, fluoro-2-deoxy-d-glucose FDG is transported into tumour cells via a family of glucose transporter proteins (Gluts) then phosphorylated by the enzyme hexokinase (HK) to FDG-6-phosphate after which it undergoes little further metabolism (except dephosphorylation by glucose-6-phosphatase (G-6-Pase)). Higher levels of HK and Glut and low levels of G-6-Pase have been reported [15] in tumour tissue compared with corresponding normal tissue. The use of serial FDG-PET scans of patients with tumours during the course of chemotherapy has been demonstrated by many studies to be a useful tool in cancer management [16]. Generally, compared with pre-treatment, responding tumours show decreased FDG uptake within a few days of starting chemotherapy.

 

FDG incorporation and resistance to chemotherapy

Sestabimi is a substrate for the P-gp pump and so the rate of efflux of ^99mTc-sestabimi can be used to diagnose MDR-related P glycoprotein expression in solid tumours [17]. However the utility of FDG incorporation as an indicator of the MDR phenotype has been investigated in a number of studies [18-20]. In each of these studies, consisting respectively of 47 patients with untreated lung cancer [18], 35 patients with intrahepatic cholangiocarcinoma (ICC) [19] and 70 patients with hepatocellular carcinoma (HCC) [20] expression of P-gp, determined immunohistochemically and found to negatively correlate with tumour SUV (standardised uptake value).

In-vitro studies [21-23] also showed that cells exhibiting the MDR phenotype incorporate FDG at lower levels than MDR-negative control cells.  Thus Lorke et al [21] found that FDG incorporation by multi-drug resistant HT-29 colon carcinoma cells grown in-vitro or as xenografts in SCID mice was lower compared with incorporation by sensitive HT-29 cells whilst incubation of cells with inhibitors of P-gp showed that efflux of FDG by the P-gp is a mechanism responsible for decreased FDG incorporation by cells expressing MDR. Recently Seo et al [22] compared  FDG incorporation by PLC/PRF/5 cells and doxorubicin-resistant P-gp expressing PLC/DOR cells and found FDG incorporation to be lower in PLC/DOR cells compared with PLC/PRF/5 cells. Treatment with the P-gp inhibitors Verapamil and cepharanthine restored FDG uptake in PLC/DOR cells, but not in PLC/PRF/5 cells. Yamada et al [23] have also shown by treating the multi-drug resistant melanoma cell line SK-MEL 24 with inhibitors of Pgp that FDG is a substrate for the Pgp pump.

Another possible mechanism responsible for the lower FDG incorporation by MDR positive tumour cells is decreased expression of glucose transporters. In a series of colchicine-selected multidrug-resistant (MDR) human KB carcinoma cell lines MDR resistant cells exhibit decreased glucose transport which correlated with the sensitivity to the toxic effect of 2-deoxy-D-glucose [24].

Tumor resistance to 5FU has been shown to be directly associated with changes in FDG incorporation [25] and to induce modifications in pathways associated with FDG incorporation [26].  Cells can be selected for drug resistance by exposure to increasing concentrations of the drug over a period of months or years. We previously reported [25] that although glucose transport was increased in cells selected for resistance to 5FU compared with sensitive control MCF-7 cells FDG incorporation by the resistant cells was actually lower. Using the glucose transport inhibitor phloretin we demonstrated that the decreased FDG incorporation by the resistant compared with the sensitive cells was due to increased efflux of FDG via the glucose transporters. Using a clone of T47D breast tumour cells selected for 5FU-resistance it has also been shown that ATP-synthase activity is diminished in the resistant cells when compared with WT cells [26].

Two studies have examined FDG incorporation by tumors resistant to tyrosine kinase inhibitors [27, 28].  In one case two patients with gastrointestinal stromal tumors that were resistant to imatinib were associated with low FDG incorporation [27]. Su et al [28] determined FDG uptake by 4 NSCLC tumor lines with varying levels of sensitivity to the EGFR kinase inhibitor gefitinib in-vitro or grown as xenografts in nude mice during treatment with this compound. FDG incorporation was only found to change in cells sensitive to the gefitinib. Western blot studies carried out on cytosolic and membrane fractions indicated that this was due to translocation of glucose transporters from the membrane to the cytosol in the sensitive tumor cells.

Not all studies have found differences in FDG incorporation between resistant and non-resistant tumor cells. Treatment with paclitaxel increased FDG uptake to a similar extent by WT human adenocarcinoma-derived ovarian cells (A2780) and cells that had been selected for resistance to doxorubicin [29].

P53 and FDG incorporation

Patients with Li-Fraumeni syndrome (LFS) have an underlying germline mutation in their p53 gene resulting in an inherited predisposition to a range of cancer types [30]. FDG-PET/CT has recently been used to screen for cancers in such patients [31]. Of 15 asymptomatic patients with LFS cancers were detected in 3 suggesting that FDG-PET/CT could be utilised in the surveillance of such patients although the risks of radiation exposure to at rick patients needs to be examined.

Loss of p53 has been shown to be associated with enhanced rates of glucose utilisation by tumor cells [32] and many clinical studies have compared tumor FDG incorporation (usually maximal SUV (standardised uptake value)) between patients with wild type p53 tumours and patients with mutant P53. P53 protein content is determined using immunohistochemistry and increased p53 expression is assumed to be due to overexpression of mutant p53. Positive correlations between p53 expression and FDG incorporation were observed in studies of 90 patients with colorectal cancer [33], two studies of primary breast cancer consisting of 275 [34] and 86 patients [35],  two studies of non-small cell lung carcinoma (NSCLC) consisting of 149 [36] and 82 [37] patients respectively, 38 patients with cervical cancer [38], two studies of patients with bone soft tissue sarcomas consisting of 63 [39] and 89 [40] patients respectively. Further two small studies of patients with hepatocellular carcinoma  [41,42] showed that the lesions with the highest FDG incorporation expressed mutant p53. Studies that have shown p53 expression not to be related to FDG incorporation are in the minority and include a small study (19 patients) with ovarian cancer [43], or in a study using a contrast ratio of lung tumor vs non-involved lung of 71 patients with c-stage IA lung adenocarcinomas [44] and a study of 75 patients with breast cancer [45]. Other studies [46] have compared FDG incorporation in tumors of patients patients according to expression of tumor suppressor genes including p53 and showed that the groups with aberrant tumor suppressor genes had a higher FDG incorporation than did tumors without mutations in these genes

Several of these studies [33, 36, 38 and 39] also compared the protein expression of glucose transporters and hexokinases with p53 status. Some found glut1 [33, 36, 39] but not glut-3 [33] expression to be positively associated with p53 expression.  However tumur cells can express several different glucose transporters and their expression does not necessarily correspond with glucose transport at the functional level [47].

To determine the effect of abrogation of functional p53 on FDG incorporation by tumor cells we transfected MCF-7 breast tumor cells with a dominant negative p53 construct [47]. FDG incorporation was found to be increased in the p53 abrogated cells compared with wild-type MCF-7 cells. Using microarray we found that glucose transporters 1, 8, and 10 were expressed in MCF-7 cells and HK I was the principal HK in MCF-7 cells but was not differentially expressed at the messenger RNA level in the dominant negative p53 clones, compared with WT cells. However, increased HK activity was observed in both dominant negative p53 clones, compared with WT MCF-7.

A further mechanism by which FDG incorporation could be increased in cells with attenuated p53 levels is through a switch from aerobic to anaerobic respiration. Pathways associated with p53 Matoba et al 2006 [48] showed that O₂ consumption was lower and lactate production higher in p53^-/- compared with p53^+/+ HCT116 cells and that this was due to p53 controlling the synthesis of cytochrome c oxidase required for the terminal step of the respiratory chain catalysing transfer of e^- to O₂.