Higher antitumor activity of trabectedin in germline BRCA2 carriers with advanced breast cancer as compared to BRCA1 carriers: A subset analysis of a dedicated phase II trial
A B S T R A C T
Specific alkylators may allow synthetic lethality among patients with germline BRCA1/2-mutations related cancers. The tetrahydroisoquinolone trabectedin administered at 1.3 mg/m2 as a 3-h intravenous infusion every 3 weeks showed activity in patients with pretreated metastatic breast cancer (MBC) and BRCA germline mutations, but mainly in BRCA2 carriers. Data from a phase II study were retrospectively analyzed to compare the efficacy and safety of this trabectedin dose and schedule in pretreated MBC patients bearing germline BRCA1/2 mutations. The primary efficacy endpoint was the objective response rate (ORR) as per Response Evaluation Criteria In Solid Tumors (RECIST) by independent expert review. Duration of response (DR) and progression-free survival (PFS) were secondary efficacy endpoints. Safety was evaluated using the National Cancer Institute Common Toxicity Criteria (NCI-CTC). Data from 26 BRCA1-mutated and 13 BRCA2-mutated patients were analyzed. 69% of BRCA1-mutated cancers were triple-negative vs. 31% of BRCA2-mutated ones. 77% of BRCA1 and 31% of BRCA2 carriers were platinum- pretreated. The ORR in BRCA2-mutated patients was higher than in BRCA1-mutated patients (33.3% vs. 9.1%). DR ranged for 1.4e6.8 months in BRCA2-mutated patients and for 1.5e1.7 months in BRCA1- mutated patients. More BRCA2-mutated patients had disease stabilization for ≥4 months (25.0% vs. 9.1%) and their median PFS was longer (4.7 vs. 2.5 months). Trabectedin was well tolerated in both patient subtypes. In conclusion, trabectedin showed higher antitumor activity in relapsed MBC patients with germline BRCA2 mutations than in those with BRCA1 mutations.
1. Introduction
Trabectedin is a marine-derived tetrahydroisoquinoline alkaloid currently approved as single agent in the EU, USA, Japan and other countries for treating soft tissue sarcoma, and in the EU and other countries combined with pegylated liposomal doxorubicin for recurrent platinum-sensitive ovarian cancer. Trabectedin binds to the DNA minor groove and also modifies the tumor microenvi- ronment, selectively decreasing the counts of blood monocytes and tumor-associated macrophages. Preclinical studies have shown that the antiproliferative effect of trabectedin is maximal in cells with an efficient transcription-coupled nucleotide excision repair (NER) system and a deficient homologous recombination (HR) repair activity, therefore suggesting cross-talk between the two DNA repair systems [1]. It has been hypothesized that the HR ma- chinery might convert the death-inducing NER-trabectedin-DNA complex into double-strand breaks (DSBs) that would then be repaired, thus allowing cell survival [2]. Hence, a deficient HR machinery would lead to persistence of DNA lesions and a greater number of replication-dependent DSBs that could no longer be repaired [3].
Mutations in BRCA1 and BRCA2 genes have been associated with functionally impaired HR repair mechanisms. Greater antitumor activity has been reported for trabectedin in ovarian cancer or metastatic breast cancer (MBC) patients carrying mutated BRCA genes [4,5]. However, BRCA1 and BRCA2 are structurally and mechanistically different tumor suppressor proteins involved in HR repair. Based on this finding, an exploratory, hypothesis-generating analysis of data from a previous phase II clinical trial was conducted to compare the activity of trabectedin in MBC patients according to their mutation type, BRCA1 or BRCA2.
2. Materials and methods
The design of the study has been described in detail elsewhere [5]. In brief, patients were included if they were adults with pro- gressive MBC and germline BRCA1 or BRCA2 mutation, treated with at least one prior chemotherapy for advanced disease and no limit of previous therapies; at least one measurable lesion with longest diameter 20 mm; recovered from acute toxicities derived from previous therapies; an Eastern Cooperative Oncology Group (ECOG) performance status score of 0e1; and adequate hematological, renal, and hepatic function. Patients were excluded if they had a life expectancy <3 months; had been pretreated with trabectedin; had known hypersensitivity to the trabectedin formulation or to dexamethasone; had received radiation or hormone therapy within 2 weeks, chemotherapy or biological therapy within 3 weeks, or other investigational product within 4 weeks before study onset; or if they had known intracranial disease or other relevant diseases and medical conditions.
Trabectedin(Yondelis®; Pharma Mar, S.A., Colmenar Viejo, Madrid, Spain) was administered via a central venous catheter as a 3-h intravenous (i.v.) infusion of 1.3 mg/m2 every 3 weeks (q3wk). Pa- tients also received prophylactic antiemetic premedication, thera- peutic use of granulocyte colony-stimulating factor (G-CSF) during Cycle 1 if needed, and prophylaxis with G-CSF from Cycle 2 onwards. Efficacy was assessed using the Response Evaluation Criteria In Solid Tumors (RECIST) v.1.0 [6]. Patients were evaluable for efficacy if they received at least 2 trabectedin infusions and had at least one disease assessment performed at least 6 weeks after treatment onset. The primary efficacy endpoint was the confirmed objective response rate (ORR), defined as the combined rate of complete (CR) and partial response (PR) confirmed by assessments repeated at least every 4 weeks. Duration of response (DR) and progression- free survival (PFS) were secondary efficacy endpoints. Safety was evaluated using the National Cancer Institute-Common Toxicity Criteria (NCI-CTC), v.3.0. All patients who received at least part of one trabectedin infusion were evaluable for safety. Descriptive statistics (95% confidence interval [CI] and range) were used to characterize the ORR and the safety profile of the study schedule. DR and PFS were analyzed using the Kaplan-Meier method.
3. Results
3.1. Patient characteristics
Data from 39 of the 40 patients with mutated BRCA genes enrolled in this phase II study between June 2007 and December 2010 were used in this analysis: 26 with mutated BRCA1 and 13 with mutated BRCA2 (Fig. 1). One patient was excluded from this analysis because she had mutations in both BRCA1 and BRCA2. The demographic and baseline characteristics of 39 included patients are shown in Table 1. BRCA1-mutated patients were aged 40 years (range, 30e59 years) and most had ductal carcinoma (n 22, 85%). Five (20%) had tumors positive for estrogen receptor expression and 3 (12%) for progesterone receptor expression; 69% were triple negative. Twenty-two patients (85%) had visceral disease and 24 (92%) had non-visceral disease. Median number of sites of disease at baseline was 3 (range, 1e6 sites). Most common sites were lymph nodes (n 19, 73%) and lung (n 14, 54%). All patients had undergone prior surgery and received prior systemic therapy (median: 4 lines per patient [range, 1e10 lines]); 21 (81%) had also received prior radiotherapy.
Fig. 1. Disposition of metastatic breast cancer patients with germline BRCA mutations treated with trabectedin. The patient with mutations in both BRCA genes was excluded from the analysis.
BRCA2-mutated patients were aged 52 years (range, 38e58 years) and all had ductal carcinoma. Positive estrogen and pro- gesterone receptor expression was found in 9 patients (69%); 23% had triple negative tumors. Eleven patients (85%) had visceral disease and 12 (92%) had non-visceral disease. Median number of sites of disease at baseline was 2 (range, 1e5 sites). Most common sites were liver (n 9, 69%), lymph nodes and bone (n 6, 46% each). All patients had undergone prior surgery and had been given prior systemic therapy (median: 4 lines per patient [range, 2e8 lines]), and 12 (92%) had received prior radiotherapy.
3.2. Treatment and dosing
A total of 102 trabectedin cycles were administered to BRCA1- mutated patients, for a median of 4 cycles per patient (range, 1e9 1.3e11.5 mg/m2), median dose intensity was 0.4 mg/m2/week (range, 0.3e0.4 mg/m2/week), and median relative dose intensity was 96% (range, 77e102%). BRCA2-mutated patients received 78 trabectedin cycles, for a median of 6 cycles per patient (range, 1e14 cycles). Median cu- mulative dose was 7.9 mg/m2 (range, 1.3e16.1 mg/m2), median dose intensity was 0.4 mg/m2/week (range, 0.3e0.4 mg/m2/week), and median relative dose intensity was 95% (range, 78e100%).
3.3. Efficacy
A total of 34 patients were evaluable for efficacy as per RECIST: 22 BRCA1-mutated and 12 BRCA2-mutated patients. Five patients (4 BRCA1-mutated and one BRCA2-mutated patients) were non evaluable because they had no target lesions according to RECIST (n 3) or because they received no trabectedin (n 2).
Of 22 evaluable BRCA1-mutated patients, 2 had confirmed PR as per RECIST by independent expert review (ORR 9.1%; 95%CI, 1.1e29.2%), with DR of 1.5e1.7 months; both were triple negative. Two more patients had unconfirmed PR (9.1%), and 9 had disease stabilization that lasted longer than 4 months in 2 (9.1%) (Table 2 CI, confidence interval; DR, duration of response; ORR, overall response rate; PD, progressive disease; PR, partial response; PRnc, unconfirmed partial response; SD, stable disease. a Patients evaluable for efficacy as per RECIST. Five patients were not evaluable because they had no target lesions evaluable per RECIST (n ¼ 3) or because they received no trabectedin infusions before discontinuation (n ¼ 2).
Fig. 2. Maximum variation of tumor lesions in BRCA1-mutated (n ¼ 21) and BRCA2- mutated (n ¼ 11) breast cancer patients treated with trabectedin. No variation could be calculated in one BRCA1-mutated patient because she died due to the disease and in one BRCA2-mutated patient because she had disease progression in non-target lesions. PD, progressive disease; PR, partial response; RECIST, Response Evaluation Criteria In Solid Tumors.
Four of 12 evaluable BRCA2-mutated patients had confirmed PR as per RECIST (ORR 33.3%; 95%CI, 9.9e65.1%), and DR was 1.4e6.8 months. All 4 patients had tumors positive for hormone receptors (all for progesterone receptor, and 3 also for estrogen receptor). HER2 overexpression was absent in all of them. In addition, disease stabilization was found in 4 patients and lasted longer than 4 months in 3 of them (25.0%) (Table 2 and Fig. 2). Of note, one BRCA2-mutated patient with inflammatory infiltrating ductal adenocarcinoma had disease stabilization for 4.7 months.
Median PFS was 2.5 months (95%CI, 1.4e4.1 months) in BRCA1- mutated patients and 4.7 months (95%CI, 0.7 months-not reached at the time of last follow-up) in BRCA2-mutated patients (Fig. 3).
3.4. Safety
Safety profile was similar regardless of the germline mutation type. Most trabectedin-related adverse events (AEs) were mild/ moderate. As expected, the most common trabectedin-related AEs were fatigue (n ¼ 11, 44% in BRCA1-mutated patients, and n ¼ 9, 75% in BRCA2-mutated patients) and nausea (n ¼ 10, 40% in BRCA1- mutated patients, and n ¼ 7, 58% in BRCA2-mutated patients) (Table 3). Grade 3/4 AEs comprised febrile neutropenia (n ¼ 3),tological abnormality in both subtypes regardless of relationship: 13 (52%; grade 4: 12%) BRCA1-mutated patients and 10 (83%; grade 4: 58%) BRCA2-mutated patients (Table 3). Grade 3/4 thrombocytopenia occurred in 4 (16%; grade 4: 8%) BRCA1-mutated patients and 5 (42%; grade 4: 17%) BRCA2-mutated patients. Grade 3/4 anemia was found in only one or two patients per subtype.
Transaminase increases were the most frequent severe biochemical abnormalities in both patient subtypes (Table 3). Grade 3/4 alanine aminotransferase and aspartate aminotrans- ferase increases were found in 9 (36%; grade 4: 4%) and 3 (12%) BRCA1-mutated patients, and in 6 (50%, grade 4: 8%) and 4 (33%) BRCA2-mutated patients, respectively.
4. Discussion
This exploratory, hypothesis-generating analysis assessed the efficacy of trabectedin in patients with MBC who have germline BRCA1 or BRCA2 mutations. The characteristics of the treated pop- ulation reflect the general MBC patient population with BRCA mu- tations. Thus, BRCA1 mutation was more frequent than BRCA2 mutation, and BRCA1-mutated patients were younger, with more frequent triple negative tumors and more extended disease at diagnosis [7].
Higher antitumor activity was reported in this exploratory analysis for trabectedin in MBC patients with BRCA2 mutations (ORR 33.3%) than in those with BRCA1 mutations (ORR 9.1%). Duration of response ranged for 1.4e6.8 months in BRCA2-mutated patients and for 1.5e1.7 months in BRCA1-mutated patients. More BRCA2-mutated patients had disease stabilization lasting longer than 4 months (25.0% vs. 9.1%), and their median progression-free survival was also longer: 4.7 months in BRCA2-mutated patients vs. 2.5 months in BRCA1-mutated patients. Trabectedin was well tolerated in both patient subtypes.
The efficacy of trabectedin in MBC patients carrying the germ- line deleterious BRCA1 and BRCA2 mutations can be explained by its mechanism of action. Trabectedin is a selective inhibitor of RNA polymerase II, thus inhibiting active transcription at elongation step and producing DSBs as a downstream event [8e10]. DNA breaks generated by trabectedin have been shown to be repaired mainly by the HR machinery, since the drug is substantially more active against HR-deficient cell lines [11].
The higher antitumor activity observed in MBC patients with BRCA2 mutation with respect to BRCA1-mutated patients in this analysis could be the result of different functionality between BRCA1 and BRCA2 proteins. Although both tumor suppressor pro- teins work in HR, they have different roles in the cell that are not interchangeable. Besides the well-known role in HR, BRCA2 is necessary to prevent the formation of RNA-DNA hybrids (R-loops) that takes place during the elongation step of the transcription by RNA polymerase II [12]. Therefore, BRCA2-deficient cells lacking an efficient transcription elongation would add an additional burden to the transcription process already troubled by trabectedin [12], and thus represent the most sensitive target of this drug.
The small sample size of this exploratory analysis prevented drawing firm conclusions about the role of the germline mutation type in the differential activity of trabectedin in BRCA1-mutated and BRCA2-mutated MBC patients, and the effect of the different prior platinum exposure cannot be ruled out. While exploratory, these findings are consistent with the activity of a new synthetic analog of trabectedin, lurbinectedin (PM01183), which is very similar structurally and shares the mechanism of action with tra- bectedin. Likewise, lurbinectedin is more effective in HR-deficient cells [13,14]. A phase II clinical trial of lurbinectedin in pretreated MBC patients with germline BRCA mutations (n 31 BRCA1- mutated patients and n 23 BRCA2-mutated patients) also showed greater activity of lurbinectedin in BRCA2-mutated patients, ORR 61% vs. ORR 26% in BRCA1-mutated patients (median PFS of 5.9 and 2.7 months, respectively) [15].
Fig. 3. Swim-lane plot showing progression-free survival of BRCA1-mutated (n ¼ 22) and BRCA2-mutated (n ¼ 12) breast cancer patients treated with trabectedin. CI, confidence interval; PD, progressive disease; PFS, progression-free survival; PR, partial response; PRnc, unconfirmed partial response; SD, stable disease.
The results of the present exploratory analysis compare favor- ably with those reported in heavily pretreated advanced BRCA- mutated breast cancer with the poly (ADP-ribose) polymerase (PARP) inhibitor olaparib (13e41%) [16,17] and with single-agent veliparib (25%) [18]. These results suggest a potential role for tra- bectedin in the treatment of tumors with defective DNA damage repair, especially those carrying BRCA2 mutations.
5. Conclusions
Trabectedin showed higher antitumor activity in relapsed MBC patients with BRCA2 mutations than in those with BRCA1 muta- tions. A new compound with a similar mechanism of action, the synthetic analog lurbinectedin, has also shown high activity against MBC in BRCA2-mutated patients. Further studies in larger pop- ulations are planned to confirm these results.