High activity of sequential low dose chemo-modulating Temozolomide in combination with Fotemustine in metastatic melanoma. A feasibility study

Metastatic melanoma (MM) is an incurable chemoresistant cancer with poor prognosis. Until now, only few drugs have shown some activity. So this tumor represents an opportunity to verify new and more effective treatment strategies.

Presently, Dacarbazine (DTIC) remains the standard chemotherapy for MM with an overall response rate of approximately 10-15% with complete response in less than 5% of patients and a survival about 8-10 months [1, 2]. No other agents have demonstrated better results than DTIC in phase III studies also when utilized as polichemotherapy or in association to immunotherapy [3–6].

Temozolomide (TMZ) has been recently utilized in MM. It is a novel oral alkylating agent having a high oral bioavailability and extensive tissue distribution, including penetration through the blood-brain barrier. Patients with MM achieved overall response rates of nearly 20% with single-agent TMZ as similar as DTIC [7–9]. Also nitrosureas are considered drugs of any activity in MM including patients with brain metastatic. Among nitrosurea analogs, fotemustine (FM) has been more extensively studied in MM, especially in Europe. It is a third generation chloroethylnitrosourea that has demonstrated significant antitumoral effects in MM with a response rate averaging 20%. However, its use is somewhat limited by its myelotoxic side effect, especially when old schedules are utilized [10–12].

The activity of alkylating agents depends on their capacity to form alkyl adducts with DNA, in some cases causing cross-linking of DNA strands. However, the antineoplastic activity of these agents is limited by cellular resistance principally induced by the DNA repair enzyme O(6)-methylguanine DNA-methyltransferase (MGMT), a DNA suicide enzyme which removes alkyl groups from alkylated DNA strands [13–16]. In tumor cell lines and xerografts an inverse correlation between the level of this protein and the sensibility to the cytotoxic effects of nitrosureas including FM has been demonstrated [17, 18]. Moreover, studies evaluating the tumor MGMT levels in patients with brain tumors receiving nitrosureas reported a positive correlation between low level content of MGMT and a better survival [19, 20].

Preclinical studies and recent clinical experiences also support the concept that continuous exposure to alkylating agent TMZ, streptozocin, procarbazine, and DTIC, can effectively deplete cells of MGMT, which is the primary mechanism of tumor resistance to nitrosureas, thus reversing the resistance to these chemotherapeutic agents [21–23]. In particular, sequential administration of TMZ and FM is able to induce depletion of MGMT both in blood lymphocytes and in tumoral tissue [24].

Recent clinical experiences have confirmed that continuous exposure to alkilating agent procarbazine in association with FM is an active treatment in patients with recurrent malignant gliomas [25]. At present, in spite of numerous experimental experience, very few data exist regarding the clinical use of TMZ as chemomodulating agent in MM patients. In particular, no established doses, timing and schedules are known.

Thus, we planned this study in MM patients to verify the hypothesis that depletion of MGMT induced by low dose TMZ could render melanoma cells more susceptible to FM. We used two different schedules of sequential combination of TMZ and FM in to assess their profile of toxicity and efficacy.

This is the first clinical experience in MM using sequential non-therapeutic low dose TMZ previous full dose FM. We demonstrated that this is an active regimen in MM patients with an acceptable profile of toxicity. In fact, our preliminary data showed that as compared to TMZ or FM single agent, the sequential regimen of the two drugs together significantly enhances their antitumoral activity inducing high response rate and regression also in visceral sites as bowel and liver.

We used this sequential regimen to verify the hypothesis that continuous exposure to alkilating agent TMZ could effectively deplete tumoral cells of MGMT which is the primary mechanism of tumor resistance to nitrosureas. This hypothesis is supported by preclinical studies and clinical experiences [21–24]. Also recent experimental data in human melanoma cell lines have confirmed the presence of a close correlation between MGMT activity and the level of resistance to TMZ and FM, although a wide variability in MGMT activity among different cell lines was noted [26]. The Authors also reported that the MGMT inactivation by O(6)-benzylguanine sensitized all melanoma cell lines expressing MGMT to TMZ and FM-induced apoptosis. Moreover, the MGMT transfection attenuated the apoptotic response, supporting the hypothesis that O(6)-alkylguanines are critical lesions involved in the initiation of programmed melanoma cell death [26].

Further clinical experiences carried out in patients with recurrent cerebral tumons confirmed that continuous exposure to alkilating agent procarbazine in association to FM is an active therapeutic option for patients with glioblastoma heavily pretreated [25].

Regarding the association of TMZ and nitrosureas in MM patients, only two studies have been published in which TMZ was used in association to lomustine [27] and FM [28], respectively. In both studies full doses of both drugs were utilized with an therapeutic additive/synergistic intent. Nevertheless, despite of a high response rate, an unacceptable toxicity was reported with myelotoxicity being the principal dose-limiting toxicity. In particular, the study of Tas et al [28] reported a response rate of 35%, but the median survival was only 6.7 months with a dose reduction in the 45% of patients, a dose delay in 32,5%, and an early treatment discontinuing in 27,5%. Notably, in our study we report a response rate of 35.7% and a stable disease in 35.7% of patients with a survival over 13 months.

At present, very few data regarding the use of low dose TMZ as chemomodulating agent are available and no established doses and schedules exist [22, 24]. Also the interval between the two drugs administration is not clear. Some Authors have reported that the administration of TMZ divided over two consecutive day at the dose of 100-200 mg/m² per day, seems to induce a substantial MGMT depletion at the time of FM administration given in the second day about 4 hours after TMZ [24, 25]. Nevertheless, a wide inter-individual variation and no definitive data are available.

So, we used two schedules of TMZ and FM (day 1,8-21 vs day 1-21) in two well balanced cohorts of 7 patients each, to identify the regimen that better conciliates antitumor activity with an acceptable toxicity. In according to the data previously reported [24, 25], we administered TMZ at 100 mg/m² per two days and FM at 100 mg/m² in the second day 4 hours after TMZ. We reported high response rate with this regimen in both cohorts of patients and a disease regression also in visceral sites and in patients with multiple metastatic localizations. Globally, we obtained an overall response of 35,7% with 1 CR and 2 PR in cohort A (regimen d1,8-28) and 2 PR in cohort B (regimen d 1-21). Five SD were also reported (35.7%), 2 in cohort A and 3 in cohort B. The median overall survival of the entire group was over 13 months. At this time, 7 of 14 patients are alive yet.

The unique CR occurred in a Cohort A patient with mediastinal lymphopaty and bowel localizations lasting about 8 months (Figure 1). The 4 PR occurred in patients with multiple and diffuse disease including lung, liver, bone, adrenal gland, lymph nodes and soft tissue. The unique patient with brain metastases died after 4 months of SD because of a cerebral hemorrhagic accident arising in the metastatic lesion without evidence of disease progression.

Interestingly, one PR in cohort B, occurred in a female with a disseminated disease involving axillaries lymph nodes, diffuse subcutaneous lesions, and multiple liver metastases. After 2 cycles of therapy the patient showed a dramatic response in all metastatic sites (Figure 2). At present, after 13 months from starting therapy, this patient is alive in PR. The biopsy of a subcutaneous lesion performed after the third cycle of therapy confirmed the diagnosis of metastatic melanoma and evidenced a diffuse regression of the neoplastic cells with abundant melanocitic pigment. Immunohistochemistry staining showed a strong positivity for melanoma associated antigen S100 protein, HMB 45 and MART 1. Surprisingly, an impressive lymphocytic (CD3+, CD4+, CD8+) and macrophage cells (CD68+) infiltration was also present (Figure 3, 4) meaning that immunomediated mechanisms have been also burst after TMZ-FM treatment, probably due to the massive disruption of neoplastic cell and consequent deliverance of tumoral associated antigens.

Regarding the toxic profile, there was a significant difference between the two cohorts of patients, principally in terms of myelotoxicity. In particular, the d1,8-28 schedule was characterized by a heavier G3-4 thrombocytopenia (3 of 7 patients) with respect to the d1-21 regimen (1 case of G3 thrombocytopenia) with a dose reduction in 4 patients in cohort A and in only 2 patients in cohort B, and chemotherapy delayed in 4 patients in cohort A and in 2 patients in cohort B. In summary, the d1-21 schedule resulted similar to the 1,8-28d schedule in term of activity, but it was superior in terms of tolerability and manageability guarantying the dose and timing planned.

Of course, an attempt to correlate the basal level of MGMT as well as the measurement of its depletion during therapy could permit to distinguish responder from non-responder patients. Nevertheless, this was not an objective of present study. In fact, our purpose was to evaluate the feasibility, tolerability and the activity of this new treatment. The study of the correlation between MGMT level and clinical outcomes has been planned in our ongoing phase II study.

In the current study we demonstrated that the sequential combination of low dose TMZ and FM has a high activity in MM patients with an acceptable toxicity. The 1-21d schedule showed similar activity and a better toxic profile with respect to the 1,8,28d schedule; thus, we are using the 1,21d schedule in our phase II ongoing study aiming to confirm the high activity of this association in MM patients.