Introduction

Globally, lung cancer is the leading cause of cancer death in men and the second leading cause in women (Parkin et al. 2005). At diagnosis, most patients with non-small-cell lung cancer (NSCLC) present with locally advanced/metastatic disease (Ettinger et al. 2006) which is associated with a poor prognosis, with 2-year survival rates of 14 and 11% for stage IIIB and stage IV or recurrent disease, respectively (Schiller et al. 2002).

Platinum-based chemotherapy doublets are standard first-line therapy for patients with stage IV disease or stage IIIB disease with malignant pleural effusion (MPE), and can be combined with targeted therapies (D’Addario et al. 2010; National Comprehensive Cancer Network 2010). However, improvements in survival with standard therapy are modest and the median survival for these patients still does not exceed 1 year (Fisher and D’Orazio 2000). Platinum-based chemotherapy together with concurrent thoracic radiotherapy is recommended in patients with un-resectable stage IIIB NSCLC (Crino et al. 2010; National Comprehensive Cancer Network 2010), but progress with this approach appears to have reached a plateau with few patients surviving beyond 5 years. Over the last decade, there has been increasing interest in the development of therapeutic cancer vaccines to help prolong survival and improve quality of life when used alongside current standard therapies (Gridelli et al. 2009; Pilla et al. 2009).

The BLP25 liposome vaccine (L-BLP25, Stimuvax®; Merck KGaA, Darmstadt, Germany) is a therapeutic cancer vaccine targeting the mucin 1 (MUC1) glycoprotein antigen that is overexpressed and aberrantly glycosylated in many cancers, including lung, breast, prostate and colorectal tumors (Ho et al. 1993; Zotter et al. 1988). Early evidence suggests that L-BLP25 has promise for the management of NSCLC (Butts et al. 2005; Palmer et al. 2001). A phase IIB, open-label, randomized trial compared treatment with L-BLP25 plus best supportive care (BSC) with BSC alone in 171 patients with stage IIIB or IV NSCLC who had no evidence of disease progression following first-line chemotherapy or chemoradiotherapy. Initial analyses reported that the median survival time was 17.4 months with L-BLP25 plus BSC compared with 13.0 months with BSC alone, although the difference between treatment arms was not statistically significant (adjusted hazard ratio [HR] 0.739, 95% confidence interval [CI] 0.509–1.073, P = 0.112). The difference in survival time was greatest in patients with stage IIIB loco-regional (LR) disease (adjusted HR 0.524, 95% CI 0.261–1.052), with the median survival time not being reached in the L-BLP25 arm (Butts et al. 2005).

The results of the phase IIB study and earlier studies indicated that L-BLP25 was well tolerated in patients with stage IIIB or IV NSCLC (North and Butts 2005; Palmer et al. 2001). In the phase IIB study, no unexpected safety issues were observed, and most adverse events were considered to be disease-related and unrelated to the study drug. The most common side-effects attributable to the vaccine were mild flu-like symptoms and mostly mild injection-site reactions. One serious adverse event (pneumonia) was classified as possibly related to treatment in the L-BLP25 group (Butts et al. 2005).

Here, we report updated survival data and analyses for the intent-to-treat population (n = 171) and stage IIIB MPE/IV (n = 106) and stage IIIB LR (n = 65) subgroups in the phase IIB study.

Patients and methods

Full details of patient eligibility criteria, study design and outcome measures have been described previously (Butts et al. 2005).

Patients

Consenting patients aged ≥18 years with stage IIIB or IV NSCLC and who had stable disease or an objective clinical response after first-line treatment with chemotherapy alone or chemotherapy and radiotherapy (with a washout period of ≥3 weeks prior to study initiation) were eligible for inclusion in the study. Participants were also required to have an Eastern Cooperative Oncology Group (ECOG) performance status ≤2.

Patients who had undergone surgery within 4 weeks of study entry, or who had received immunotherapy within 4 weeks of study entry or immunosuppressive drugs within 3 weeks of study entry, were excluded. Other exclusion criteria included known brain metastases, autoimmune or immunodeficiency disorders, significant hepatic, renal or cardiac disease, splenectomy or active infection.

Study design

This was a randomized, open-label, parallel-group, controlled, multicenter study (EMR 63325-005; ClinicalTrials.gov study identifier NCT00157209). Eligible patients were randomized 1:1 to receive L-BLP25 plus BSC or BSC alone. Randomization was also stratified by disease status (either stage IIIB LR or stage IV and stage IIIB with MPE).

Patients randomized to the L-BLP25 arm received a single intravenous dose of cyclophosphamide 300 mg/m2 (up to a maximum dose of 600 mg) 3 days before the first L-BLP25 treatment, in order to augment the immune response to L-BLP25 vaccine (MacLean et al. 1996). Patients randomized to the L-BLP25 arm received subcutaneous vaccinations of L-BLP25 (930 μg; administered as four 0.5 ml subcutaneous injections)—a lyophilized liposomal product consisting of the 25-amino-acid BLP25 lipopeptide, an immunoadjuvant monophosphoryl lipid A and three lipids (cholesterol, dimyristoyl phosphatidylglycerol and dipalmitoyl phosphatidylcholine), prepared by addition of sodium chloride, 0.9% to form a suspension—weekly at weeks 0, 1, 2, 3, 4, 5, 6 and 7. Maintenance vaccinations were administered at 6-week intervals starting at week 13, at the investigators’ discretion. The nominal dose of L-BLP25 is different to that previously stated (Butts et al. 2005; Butts et al. 2010). In clinical studies prior to 2008, the density determination of L-BLP25 prior to freeze-drying of the final powdered product led to the content of L-BLP25 being declared as 1,000 μg. However, the more recent density determination means that the majority of phase IIb study participants actually received 930 μg antigen as opposed to the 1,000 μg that was originally declared.

BSC was provided at the investigators’ discretion to patients in both arms of the study, and could include psychosocial support, analgesics and nutritional support. When indicated for the treatment of progressive disease, patients could receive second-line chemotherapy or palliative radiation.

The study was conducted in accordance with the Declaration of Helsinki. The protocol was approved by institutional review boards.

Statistical analysis

As detailed in the previous report (Butts et al. 2005), an intent-to-treat approach was used, such that all randomized patients were included in the survival analysis. Standardized assessment of disease progression was not mandated as part of the study design.

Survival time was defined as the time from the date of randomization to death. Analysis of survival duration was performed using the Cox proportional hazard regression model adjusted for treatment assignment, response to first-line treatment and disease stage at study entry. The study was powered to detect a difference in survival of 5 months (assumptions: 12 months survival in the L-BLP25 arm and 7 months in the BSC arm; HR 0.583; two-sided type I error rate of 0.05; power of 80%). The primary survival analysis based on the intent-to-treat population and its timing were event-based, occurring after a minimum of 108 events. Multiple testing was not predefined in the study protocol and so, for subgroup analyses by disease stage at study entry, HRs with 95% CIs are reported rather than P-values.

The cut-off date for the primary analysis summarized above was March 2004. In this paper, we describe an updated survival analysis in the same patient population with 2 years of additional follow-up, with a cut-off date of March 2006. We also present the 3-year survival rates in the intent-to-treat population and stage IIIB MPE/IV and stage IIIB LR subgroups. The P-values for differences in 3-year survival rates are descriptive.

Results

Patient characteristics

A total of 171 patients were enrolled: 88 in the L-BLP25 arm and 83 in the BSC arm. As reported previously, the two treatment arms were well matched at baseline (Butts et al. 2005). In the intent-to-treat population, median age at randomization was 59 years; 56% of patients were male; and ECOG performance status was 0 in 31%, 1 in 64%, and 2 in 5% of patients. With regards to the stratification according to disease stage subsets, the majority of patients had NSCLC stage IIIB MPE or stage IV (62%) disease; the remainder had stage IIIB LR disease (38%).

Survival analysis

At the time of this survival analysis (March 2006), 21 of 88 (24%) patients randomized to L-BLP25 plus BSC remained alive compared with 10 of 83 (12%) patients randomized to BSC alone. Median follow-up times in both treatment arms exceeded 3 years, both for the intent-to-treat population and stratified subsets.

In the intent-to-treat population, median overall survival for patients in the L-BLP25 arm was 17.2 months compared with 13.0 months for patients in the BSC arm (HR 0.745, 95% CI 0.533–1.042) (Fig. 1a). The 3-year survival rate was 31% for the L-BLP25 arm vs. 17% for the BSC arm (P = 0.035) (Fig. 2).

Fig. 1
figure 1

Overall survival in L-BLP25 and best supportive care (BSC) treatment arms after extended follow-up in a total patient group, b stage IIIB MPE/IV subgroup, and c stage IIIB LR subgroup (Survival based on Kaplan–Meier estimates) BSC best supportive care, CI confidence interval, HR hazard ratio, LR loco-regional, MPE malignant pleural effusion

Full size image
Fig. 2
figure 2

Three-year survival rates (% patients surviving) in the total patient population, patients with stage IIIB MPE/IV disease, and patients with stage IIIB LR disease BSC best supportive care, LR loco-regional, MPE malignant pleural effusion. P-values are descriptive statistics for chi-square testing comparing 3-year survival rates with L-BLP25 and BSC

Full size image

Median overall survival in the stage IIIB MPE/IV subgroup was 15.1 months in the L-BLP25 arm vs. 12.9 months in the BSC arm (HR 0.878, 95% CI 0.587–1.313) (Fig. 1b). The 3-year survival rate was 19% for the L-BLP25 arm vs. 11% for the BSC arm (P = 0.278) (Fig. 2).

In the stage IIIB LR subgroup, median overall survival was 30.6 months in the L-BLP25 arm and 13.3 months in the BSC arm (HR 0.548, 95% CI 0.301–0.999) (Fig. 1c). The 3-year survival rate was 49% for the L-BLP25 arm vs. 27% for the BSC arm (P = 0.070) (Fig. 2).

Discussion

Most patients with NSCLC are diagnosed at an advanced stage (stage IIIB/IV) when surgical cure is not possible and the prognosis is poor (Ettinger et al. 2006; Shepherd et al. 2000). The results of the updated analysis presented here confirm the findings from the original analysis in suggesting that L-BLP25 is an active agent in NSCLC, and are of great interest with regards to patients with stage III NSCLC who are candidates for aggressive treatment with chemoradiation.

For the intent-to-treat population including all patients (stage IIIB LR and IIIB MPE/IV NSCLC), the initial and updated analyses were consistent in indicating median survival that was ~4 months longer with L-BLP25 compared with BSC alone (4.4 and 4.2 months for the initial and updated analyses, respectively). In the subgroup of patients with stage IIIB LR disease, at the time of the initial analysis, the median survival time was 13.3 months in the BSC arm but had not been reached in the L-BLP25 arm, with 2-year survival rates of 60% with L-BLP25 compared with 37% for BSC alone (Butts et al. 2005). The updated survival analysis provides further support for a potential, clinically meaningful survival advantage in this group of patients, with a median survival time that was more than doubled with L-BLP25 compared with BSC alone (median survival of 30.6 months with L-BLP25 and 13.3 months with BSC alone). The 3-year survival rate for patients with stage IIIB LR disease was almost doubled in the L-BLP25 arm as compared with the BSC alone arm (49% vs. 27%).

The stage IIIB LR population in this study represents a small, and probably select, sample. The difference in survival between the two treatment groups in this subgroup may be related to chance caused by the small sample size or some unrecognized selection factors. However, given that these patients were randomized ≥3 weeks after completion of chemoradiotherapy, the median survival in the control arm is consistent with other published data. Furthermore, the median survival in the L-BLP25 arm is among the best survival data series in the published literature (Albain et al. 2002; Gandara et al. 2003; Hanna et al. 2008).

There are a number of possible explanations for why the survival advantage with L-BLP25 appears to be substantially less in patients with stage IIIB MPE/IV as compared with stage IIIB LR disease. First, rapid disease progression and short survival times in patients with stage IIIB MPE/IV may not allow sufficient time for development of a robust immunologic response following vaccination. Secondly, while the majority (85%) of patients with LR disease had received radiotherapy prior to study entry (Butts et al. 2005), it is not indicated for metastatic disease. Radiotherapy of tumors can induce a cascade of pro-immunogenic effects involving both the innate and adaptive immune responses (Formenti and Demaria 2009; Roses et al. 2008). Therefore, prior radiotherapy in the patients with LR disease may have contributed to the effects of immunotherapy with L-BLP25. Finally, the extent of immunosuppression in patients with stage IV disease may negatively impact on the development of an adequate immune response after vaccination.

The updated survival analysis provides further support for L-BLP25 as a promising therapy for patients with stage IIIB LR NSCLC. A randomized, double-blind, placebo-controlled phase III trial (START—stimulating targeted antigenic responses to NSCLC) is ongoing in patients with un-resectable stage III NSCLC (excluding patients with metastatic disease or MPE) who have completed treatment with chemoradiotherapy (ClinicalTrials gov 2010). The primary objective of the trial is to compare the survival duration of all randomized patients by treatment arm. Secondary objectives are to compare the two treatment arms for: time to symptom progression; time to disease progression; survival at 1, 2 and 3 years; and safety. START will enroll nearly 1,500 patients at 275 centers worldwide. The results of this trial are eagerly awaited, as they will help to define the role of L-BLP25 in the clinical management of patients with late-stage NSCLC.