Introduction

Brain metastases occur in 20–40% of patients (pts) affected by solid tumours [21, 28]. The optimal therapy of brain metastases is still under investigation. Corticosteroids, radiotherapy, surgical therapy, and radiosurgery all have an established role in the management of brain metastatic disease [3, 17, 26] but all remain strictly palliative approaches. Although chemotherapy has not yet emerged as a standard treatment for pts with brain metastases, the current evidence suggests that cytotoxic drugs may indeed have a role in the treatment of selected pts. Several relatively new anticancer drugs have been developed with pharmacokinetic and bio-distribution properties that allow good accumulation in the brain tissue.

An anticancer drug with a significant accumulation in brain tissue is temozolomide (TMZ), a new orally administered imidazo-tetrazine with a mechanism of action and efficacy similar to dacarbazine (DTIC). It has been already used alone or in combination with radiotherapy in the treatment of primary and secondary brain tumours and is currently used in the therapy of glioblastomas also as single agent [1, 4, 5, 18–20]. It is well tolerated, and is therefore a suitable candidate for combination chemotherapy [8]. A recent study has demonstrated that TMZ adjuvant and concomitant with radiotherapy improves significantly the survival of pts with glioblastoma [24]. In a phase I pharmacokinetic study of TMZ and cisplatin in pts with advanced solid tumours, this combination showed some activity in non-small cell lung cancer (NSCLC) pts. Moreover, objective responses induced by TMZ have been demonstrated also in breast and colon cancers [1, 6]. The good distribution of TMZ in the brain suggests that this drug may be an attractive agent against secondary brain malignancies.

Pegylated liposomal doxorubicin (doxorubicin HCl liposome injection; Doxil or Caelyx, PLD) is a liposomal formulation of doxorubicin, with a low uptake by the reticulo–endothelial system due to the attachment of polyethylene glycol polymers to a lipid anchor. The PLD has demonstrated to be an effective agent against metastatic breast and ovarian cancer and cutaneous T-cell lymphoma [9–12, 16, 22, 29]. It has been recently demonstrated that PLD can also cross the brain–blood barrier with a consequent accumulation in primary and secondary brain lesions; in ten pts with metastatic brain tumours treated with radiolabeled liposomal doxorubicin concurrent with radiotherapy the accumulation of radiolabeled PLD was 7–13 times higher in the metastatic lesions, compared to the normal brain [14]. In a recent case report it is described a patient (pt) affected by a vulvar adenocarcinoma that has metastasized at multiple sites including also brain; in this pt, PLD therapy resulted in a dramatic regression of metastatic lesions and marked improvement in quality-of-life (QoL) [13].

The TMZ and PLD have been already used in combination in phase I clinical studies in the treatment of advanced solid tumours [2, 27]. We have used the TMZ/PLD combination in the treatment of brain metastases derived from solid tumours in order to evaluate the activity and the toxicity profile of the schedule in a prospectively designed phase II study. Moreover, we have evaluated the impact of such combination on the QoL of the pts enrolled in the study.

Patients and methods

Study design

A phase II study was prospectively projected according to the Simon’s two-stage optimal design [23]. The primary endpoint was the objective response rate (CR+PR) at the brain site. According to this design a number (n1) of pts enter the first stage of the trial. The accrual continues to a total of n2 pts only if a specified r1 response rate is achieved in the first series. We have selected as target activity a 10–30% response rate, with a 0.05 α error and a 0.20 β error. In this case the treatment under investigation should be considered non-active if it produced less than two responses out of 10 consecutive pts in the first series and fewer than 5/29 pts in the overall series.

Patient selection

The basic requirement for pt accrual was to be carrier of at least one measurable brain lesion. The minimum size of a target lesion in the brain was 2.0 cm. The pts who had completed Whole Brain Radiotherapy (WBRT) and/or any other kind of therapy at least 8 weeks before study entry (or had refused WBRT) and showed progressive disease (PD) at the brain-site at the time of enrollment were considered eligible. The pts were required to be more than 18-year old, to have a WHO/ECOG performance status of 0–2, without of symptomatic heart disease and with a good heart performance (EF%>50%) evaluated with cardiac scintigraphy and/or ultrasonography, and not pregnant or nursing. Normal liver and kidney biochemistry (total bilirubin<1.5 mg/dl, aspartate aminotransferase and alanine aminotransferase <3 times the normal limit, prothrombin and partial thromboplastin <1.5 times the normal limit and creatinine <1.2 mg/dl) and histological diagnosis of a single cancer histotype were also required. No pts with previous malignancies were allowed except for adequately treated in situ carcinoma of the cervix or squamous carcinoma of the skin. All the extra-brain sites of disease should also be considered not progressive by the investigators at the time of enrollment in the study. The study was approved by the Institutional Bioethical Committee and all pts provided written informed consent before the beginning of the treatment.

Treatment and monitoring

The TMZ was administered orally at the dose of 200 mg/m2 daily for five consecutive days under fasting conditions. The PLD was intravenously administered at the dose of 35 mg/m2 (infusion time 60 min) at the day 1. The treatment was repeated every 28 days. The pts underwent clinical and biochemical examination before entry into the study and then monthly during the study and follow-up; blood counts were assessed weekly during treatment. Each evaluation included complete physical examination, a routine biochemical profile, the assessment of side effects and the identification of any complications. All adverse events were recorded and graded according to CTC-NCI criteria (Version 2.0). Treatment was continued until disease progression, unacceptable toxicity, pt refusal or for a maximum of eight cycles.

Treatment evaluation

Target lesions were assessed by computed tomography (CT) and/or gadolinium-enhanced magnetic resonance imaging (Gd-MRI). The initial examinations had to be performed within 14 days prior the inclusion in the study. Evaluation of target lesions was performed at every third cycles and, if necessary, based on the clinical situation, according to the WHO/ECOG [25] as described below.

Complete response (CR): disappearance of all known brain metastases. Partial response (PR): 50% or greater decrease in measurable brain lesions or an objective improvement in evaluable, but non-measurable brain lesions. It is not necessary for every brain lesion to have regressed, but no brain lesion should have progressed. Stable disease (SD): brain lesions unchanged (<50% decrease or <25% increase in the size of measurable lesions). PD: progression of some or all brain lesions and/or the appearance of new brain lesions.

Disease staging of extra-brain metastases was performed before the beginning of the treatment and then modulated on the basis of different tumour diagnosis or disease sites. A complete reassessment by appropriate imaging was however required every 6 months in all cases.

In order to assess the effect of treatment on the QoL, pts completed the Functional Assessment of Cancer Therapy-General questionnaire (FACT-G, Version 4) [7]. The FACT-G questionnaire was administered before starting treatment and after 3–6–9 months.

Statistical analysis

The QoL data were analysed by ANOVA test. Survival plots were constructed by the Kaplan Meyer method and survival data were analysed by the GraphPad Instat 3.2 statistic software.

Results

Patient characteristics

According to the study design at least two brain responses out of ten consecutive pts were required in order to proceed to the phase II step. Three pts achieved major responses in the first ten pt cohort. Therefore, the study proceeded to the second step and it was completed when a total of seven responses clearly exceeded the six response target in the global pt series. We have treated 19 consecutive pts with different solid tumours. Eighty-two cycles were performed (range: 1–8 cycles). Thirteen out of 19 pts (68.4%) had multiple brain lesions (cortical and/or subcortical and/or cerebellar) and six (31.5%) had a single brain localization (Table 1).

Table 1 Patient characteristics
Full size table

Activity of TMZ/PLD combination

Three CRs (15.8%) were achieved in pts with metastases from breast cancer and lasted 23+, 5+ and 8 months. All of them progressed after WBRT, while two pts had undergone previous systemic chemotherapy. All the three pts had multiple brain lesions and extra-brain disease at different sites. Four PRs (21.0%) were recorded in two pts affected by colo-rectal and two by breast cancer and lasted 10+, 6, 13+ and 4 months, respectively. All the PRs have been defined on the basis of measurable index brain lesions. Eight out of 19 (42.1%) pts remained in SD up to 9+ months and four (21.0%) pts showed a PD. The overall response rate (CR+PR) obtained on brain metastasis was 36.8% (95% CI: 19.1–59.2), while disease control rate was 78.9% (95% CI: 56.3–91.3) (Table 2). Median Progression free survival (PFS) and median overall survival (OS) were 5.5 (95% CI: 2.7–8.2) and 10.0 (95% CI: 6.3–13.7) months, respectively (Fig. 1a, b).

Table 2 Clinical response
Full size table
Fig. 1
figure 1

The PFS and OS have been constructed by the Kaplan–Meyer method. The median PFS was 5.5 months (95% CI: 2.7–8.2) and the median OS was 10.0 months (95% CI: 6.3–13.7)

Full size image

Toxicity of TMZ/PLD combination

Eighty-two cycles of TMZ/PLD were administered from April 2001 to December 2003. Grade III neutropenia, controlled with the administration of G-CSF and not requiring dose-reduction, and thrombocytopenia were recorded in one pt. Grades II and I (CTC) neutropenia were recorded in other eight and six pts, respectively. Moreover, grades II and I thrombocytopenia developed in five and ten pts, respectively. Grade III erythrodisesthesia occurred in two pts and was controlled with the administration of flavopirimidines and corticosteroids. Grades II and I erythrodisesthesia was also found in four and in six pts, respectively. Grade II nausea/vomiting was recorded in 12 pts and grade II headache in only one pt. Liver or renal or cardiac or severe neurological toxicity were never observed in our series (Table 3).

Table 3 Major toxicities (grades I–III according to Common Toxicity Criteria extended by National Cancer Institute, NCI CTC) in the 19 enrolled pts
Full size table

Effects of TMZ/PLD combination on the QoL

We also measured the quality of life through the questionnaire FACT-G that was well accepted by all the pts. A pt recall is shown in Fig. 2. We observed a highly statistically significant improvement in the QoL as measured by FACT-G, suggesting a clinically relevant impact of combination therapy in pts with brain metastases after 3 and 6 months from beginning of TMZ/PLD treatment (p≤0.0002 and ≤0.0004). The maximal percent increase in mean FACT-G values from baseline occurred after 9 months of therapy (Fig. 2). Matched t-test between baseline and 9 months value in pts who where still alive was also significant for FACT-G (p≤0.0001) indicating that the benefit derived from the treatment was still present after a long-term treatment. The FACT-G improvement involved all the domains and was more pronounced on the Physical and Functional Well Being domains.

Fig. 2
figure 2

Variations (mean ± SD) in QoL in TMZ/PLD treated pts. Statistical analysis has been performed by ANOVA comparing different time pts to the baseline values

Full size image

Discussion

Radiation therapy is the standard treatment for the majority of pts with brain metastases, while surgical resection should be considered in pts with isolated brain metastasis and no extracranial disease. The role of chemotherapy in the treatment of brain metastases from solid tumours is not clearly defined. However, several relatively new anticancer drugs have been developed with pharmacokinetic and bio-distribution properties that allow the selective accumulation in the brain tissue. The TMZ and PLD are two suitable agents for the treatment of brain metastases for their capacity to accumulate in brain tissue and in tumour tissue within brain [8, 15].

In this study, we have evaluated the activity and the toxicity profile of the combination between TMZ and PLD in the treatment of brain metastases derived from several solid tumours in a prospective phase II trial which has been performed according to the two stage optimal design [23]. The recommended phase II dose for this combination, from a previous phase I study, is TMZ 1000 mg/m2 and PLD 40 mg/m2 every 4 weeks [27]. We have selected the 35 mg/m2 PLD dose taking in account that the study accrual might include heavily pretreated pts.

In the present study the overall number of responders to the TMZ/PLD combination exceeded the target, which has been prospectively established and the trial could be therefore defined as positive on the primary outcome of the objective response rate at the brain site. Three CRs (15.8%) were achieved in pts with metastases from breast cancer and four PRs (21.0%) were recorded in two pts affected by colo-rectal and two by breast cancer. Moreover eight (42.1%) pts showed a SD. The overall response rate (CR+PR) of 36.8% (95% CI: 19.1–59.2) obtained on brain metastasis, and the brain disease control rate of 78.9% (95% CI: 56.3–91.3) appears encouraging, although the definition of SD is somewhat difficult in this specific setting. It has to be considered in fact that several pts, who achieved a response after TMZ/PLD treatment, at the time of study accrual were in PD after WBRT

This is the first report, to the best of our knowledge, of the use of TMZ/PLD combination in the treatment of brain metastases even if the two drugs have been already shown to be active as single agents in this setting [1, 4, 5, 15, 18–20]. We have also demonstrated that this combination is able to rescue some of those pts who were refractory to the conventional radiotherapy and to induce in some cases long-lasting responses in a peculiar subset of pts with a bad prognosis where the median OS of 10.0 months can be considered extremely promising.

The majority of the ORs were obtained in pts affected by breast and colo-rectal cancers. While breast tumours are considered highly sensitive to antracyclin-containing regimens, the good activity on colo-rectal cancer suggests a specific sensitivity profile of brain metastasis as compared to other disease sites. This compelling hypothesis needs however to be explored in larger series.

The TMZ/PLD regimen was well-tolerated being scarcely myelotoxic and not inducing mucositis or relevant nausea/vomiting effects. In fact, grade III neutropenia, controlled with the administration of G-CSF and not requiring dose-reduction, and thrombocytopenia were recorded in only one heavily pre-treated pt. Grade III erythrodisesthesia was found in two pts, but it did not require treatment discontinuation and/or dose reduction and was controlled by the administration of flavopirimidines and corticosteroids. The good safety profile of the TMZ/PLD combination has to be considered also taking in account that our pt population has a relatively high mean age (60.8±15 years) and includes also several pts with more than 70 years (6/19 pts). Our findings indicate that the TMZ/PLD combination is well tolerated in elderly pts and can be administered to subjects that could not be treated by conventional approaches both for the presence of brain metastases and old age factors. It has to be considered that an increase in the TMZ dose/intensity should be achieved by a 7 days on, 7 days off schedule at the daily dose of 125 mg/m2 as recently suggested in a recent phase I study of TMZ/PLD combination [2].

It is now widely accepted that relief from symptoms and protection against brain complications should not be overlooked as therapeutic goals to be achieved in cancer pts with brain metastases. Therefore, the improvement in the QoL may be as important as the increase in survival rates in these pts. We indeed observed a remarkable improvement in the QoL, assessed by FACT-G, during the therapy. It must be noted that the effects of TMZ/PLD combination on QoL persisted after 9 months of therapy. Interestingly, a statistically significant amelioration in the QoL was recorded also in pts who did not achieve an objective response. A suggestive hypothesis is that imaging alone might not fully demonstrate the benefit of chemotherapy on brain metastatic disease.

In conclusions, we have demostrated in a prospectively designed phase II trial that the combination of TMZ and PLD has a definite activity in brain metastases from solid tumours and is a well tolerated treatment also in elder pts. Moreover, the TMZ/PLD combination had a relevant impact on the amelioration of the quality of life also in pts who did not obtain an OR; an encouraging 10.0 median OS has been also recorded.