The purpose of this study was to 1) assess the expression of XIAP in a quantitative fashion on a large cohort of melanoma specimens in an objective, automated fashion, and to compare expression of XIAP in malignant specimens to benign nevi and 2) assess the association between XIAP expression and resistance to Carboplatin, and the effect of Phenoxodiol on XIAP. In our large cohort of melanomas and nevi we show significantly higher expression in tumors than in nevi, and XIAP expression was significantly higher in metastatic specimens than in primary melanomas. Among primary tumors we found an association between high XIAP expression and deep lesions (>2 mm). Furthermore, we demonstrate that degradation of XIAP by Phenoxodiol confers Carboplatin sensitization in melanoma cells.
One of the major problems in the treatment of unresectable melanoma is the inherent resistance to chemotherapy; response rates to any single chemotherapeutic agent or combination of agents are in the order of 25% at best, and the responses are typically not durable . Chemotherapy resistance in melanoma cells is multifactorial. Mechanisms implicated in chemoresistance in melanoma include (but are not limited to) multi-drug resistance transporter proteins [47, 48], oxidative agents such as nitric oxide , DNA repair pathways  and intrinsic resistance to apoptosis, as reviewed by Soengas et al .
Chemotherapy resistance associated with defects in the apoptotic cascade is due to the over-expression of factors that inhibit the apoptotic signal initiated by cytotoxic agents. XIAP is one of the major inhibitors of apoptosis, due to its ability to block both the intrinsic and extrinsic pathways. We did not have information on chemotherapy treatment and response to therapy on the patients included in this cohort. We are currently collecting specimens from a randomized clinical trial in which metastatic melanoma patients on the control arm are being treated with chemotherapy alone, and we will assess XIAP levels and the association with chemotherapy response.
A number of drugs that directly target XIAP are in development, including antisense oligonucleotide molecules (AEG 35156, Aegra pharmaceuticals, Montreal, Canada) and small molecule inhibitors, as reviewed by Schimmer et al , and these inhibitors might enhance chemosensitivity in XIAP over-expressing tumors. Finding significantly higher expression in melanoma than in nevi was encouraging for development of XIAP-targeting therapies in melanoma. However, there was some overlap in the range of expression between benign nevi and melanoma specimens, particularly for primary melanomas. It is unclear why some nevi had relatively high XIAP levels; although these nevi were not thought to be "atypical" on pathologic evaluation, some benign melanocytes might be undergoing early malignant transformation. Given that these benign nevi were excised, we cannot assess the association between XIAP expression and clinical outcome in nevi. Moreover, other researchers have shown that somatic mutations are already evident in nevi, but not in normal skin, such as the presence of activating B-raf and N-ras mutations [52, 53]. Because level of expression may be an important determinant of sensitivity to XIAP targeted therapies, we sought to determine the percent of samples that displayed XIAP over-expression as compared to nevi. For this purpose, we defined high expression as exceeding the AQUA score in 95% of benign nevi, as done in prior studies . We found that 61% of the primary and 77% of the metastatic specimens met our definition for high expression. While it is unknown whether there is a threshold level above which XIAP inhibitors might be more effective as therapeutics, our data show that there is variability in XIAP expression in melanoma, which should, at the very least, be assessed when XIAP targeting therapies are used in clinical trials in melanoma patients.
Extensive research has been done on the structure and function of XIAP, as reviewed in detail by Schimmer et al . In addition to caspase inhibition, XIAP has a role in activation of the JNK pathway, with resultant MAP kinase pathway activation, leading to NF-κB activation . XIAP also activates NF-κB by promoting its nuclear translocation . XIAP mediates cell cycle arrest, via regulation of cyclins and cyclin-dependent kinase inhibitors (CDKIs), and is involved in receptor-mediated signaling .
Targeting XIAP to sensitize cancer cells to chemotherapy is supported by a number of studies showing that over-expression of XIAP in cell lines confers chemotherapy resistance, and that down-regulation of XIAP results in activation of both the mitochondrial and death receptor pathways and sensitization to chemotherapy [17, 19, 57–59]. Targeting XIAP is unlikely to result in significant toxicity as XIAP knock-out mice exhibit normal organ function . Although clinical toxicity data using the XIAP targeting drug AEG 35156 are pending, we believe that the approach of targeting XIAP in order to sensitize melanoma cells to chemotherapy is worthy of further investigation.
Phenoxodiol is an isoflavone derivative that has been shown to cause proteasomal degradation of XIAP, and reverse chemoresistance in ovarian cancer cells . The precise target and mechanism of action of Phenoxodiol remain unknown, and XIAP degradation might be a downstream effect of Phenoxodiol, rather than a direct effect of the drug. In ovarian cancer we showed that both knock down of XIAP by RNA interference and pretreatment with Phenoxodiol result in similar sensitization to chemotherapy . In this study we demonstrated that Phenoxodiol treatment of melanoma cells also caused XIAP degradation, cleavage of pro-caspase-2, and activation of caspases -3, -8 and -9 in Carboplatin resistant melanoma cell lines. The pro-apoptotic effect of Phenoxodiol on melanoma cells potentiated the cytotoxic effects of Carboplatin, and reduced the IC50 of Carboplatin. We showed that in our Carboplatin resistant cells, caspase-3 (the main effector caspase) was not activated by Carboplatin alone, yet pre-treatment of YUMAC cells with Phenoxodiol for 2 or 4 hours resulted in decreased levels of XIAP, sensitization to Carboplatin, and remarkable activation of caspase-3. We note that in the YUSAC2 cells there was much less XIAP degradation and no meaningful sensitization to Carboplatin. These results confirm previous studies by our group and others suggesting that intracellular inhibitors of caspases, such as XIAP, play a major role in drug-induced apoptosis [17, 19, 58], and by triggering caspase activation, Phenoxodiol enhances the cytotoxic effect of chemotherapy. Although the molecular differences between these two cell lines have yet to be determined, these results elucidate potential new targets for chemoresistance. We are presently screening additional potential target(s) mediating this resistance.
Single agent Carboplatin is not commonly used for melanoma, whereas Cisplatin is widely used. The chemical structure of these drugs is very similar, and in most (but not all) diseases these drugs have demonstrated equal efficacy. Carboplatin has been used as a singe agent in Phase II studies for metastatic melanoma, with response rates of 11–16% [61, 62]. Use of single agent Cisplatin in Phase II trials has resulted in response rates of 10–16% [7, 63]. While these two platinum compounds have not been compared head-to-head for melanoma, the toxicity associated with Cisplatin often prohibits its use in patients, and further development of sensitizers to Carboplatin is a logical approach. Phenoxodiol has been assessed in Phase I studies, and is extremely well tolerated in both the oral and the intravenous forms [19, 31, 64]. In our in vitro studies we used a concentration of 10 μg/ml of Phenoxodiol in order to sensitize cells to Carboplatin. In clinical trials with oral Phenoxodiol given at a dose of 50 mg/kg, the serum concentration was in the order of 250 μM after 30 minutes of administration . Thus the dose used in this study (10 μg/ml = 41.62 μM) is readily attainable in vivo, and is likely to be well tolerated. Of the three cell lines we assessed for sensitization to Carboplatin, one remained resistant after Phenoxodiol pre-treatment. Studies are underway in our laboratory to assess additional isoflavone analogues as chemotherapy sensitizers, using a larger panel of cell lines. However, in select patients whose tumors demonstrate XIAP degradation with Phenoxodiol treatment, use of Phenoxodiol as a chemotherapy sensitizer is a reasonable approach that warrants further investigation. These clinical trials should incorporate assessment of XIAP degradation in short term cultures after treatment with Phenoxodiol as a biomarker of response to this combination of therapy. This is currently being assessed in platinum resistant ovarian cancer patients treated with Phenoxodiol and Cisplatin.
In summary, our studies show that XIAP is up-regulated in melanoma specimens compared to nevi, and expression of XIAP is higher in metastatic than in primary melanomas. This is the first study of XIAP expression in a large cohort of melanoma specimens. Furthermore, we demonstrate that Phenoxodiol sensitization to Carboplatin in melanoma is associated with XIAP degradation, although the precise target of Phenoxodiol is still being studied in our laboratory. Further studies are needed to determine whether isoflavone analogues and direct XIAP targeting therapies can be used as chemotherapy sensitizers in melanoma patients, and whether XIAP expression levels are associated with response to therapy.