PDF
Citation
Print
Introduction
Since the introduction of methotrexate, vinblastine, adriamycin, and cisplatin (MVAC), which demonstrated superior therapeutic efficacy compared to cisplatin monotherapy [1], platinum-based combination chemotherapy has long been the standard first-line treatment for advanced urothelial carcinoma (aUC). Although MVAC and gemcitabine plus cisplatin (GP) are commonly used standard treatments [2], the optimal number of cycles for these regimens has not been established. These treatments are associated with hematologic and nonhematologic toxicities, often leading to dose delays and dose reductions. Furthermore, some adverse events of cisplatin are cumulative and long-lasting [3]. Patients with urothelial cancer are often elderly and susceptible to complications such as infections (especially urosepsis) and nephrotoxicity. Therefore, if disease progression does not occur after 4-6 cycles of first-line chemotherapy, clinicians typically cease treatment and monitor patients until the initiation of second-line therapy becomes necessary due to disease progression.
At the time this study was designed, no standard salvage treatment had demonstrated a statistically significant survival benefit. The relative rarity of metastatic urothelial cancer and the fragility of the patient population have hindered the conduct of large-scale phase III clinical trials for salvage chemotherapy, resulting in a limited number of agents being evaluated in prospective clinical trials. For second-line treatment, taxanes and vinflunine were recommended by clinical guidelines [4]. However, in real-world practice, patients with urothelial cancer often experience rapid clinical deterioration after or during first-line treatment, which often prevents them from receiving second-line therapy [5].
Given the potential for rapid deterioration during treatment breaks, maintenance therapy using a relatively effective and well-tolerated salvage regimen for patients without disease progression after standard first-line treatment is an attractive approach. In non-small cell lung cancer, several clinical trials have demonstrated maintenance therapy after first-line platinum-based chemotherapy can lead to improved progression-free survival (PFS) and overall survival (OS) [6,7].
Although taxanes and vinflunine have been most frequently recommended in the salvage setting for aUC, their efficacy as single agents has been modest [8-11]. Additionally, both taxanes and vinflunine are associated with cumulative neurotoxicity, limited their suitability for maintenance therapy. Pemetrexed, however, has shown moderate efficacy with a tolerable toxicity profile and cumulative toxicity when tested in urothelial cancer [12,13]. Pemetrexed has also demonstrated promising activity as a first-line chemotherapy agent combination with cisplatin for aUC [14]. Given its efficacy and safety profile in urothelial carcinoma, pemetrexed appears to be a promising candidate for maintenance therapy. Based on this background, we designed this trial to evaluate the role of pemetrexed as maintenance therapy in patients with aUC who have completed first-line platinum-based chemotherapy without disease progression.
Materials and Methods
1. Study design
PREMIER was a multicenter, open-label, prospective, randomized phase III trial designed to evaluate the efficacy of pemetrexed as switch maintenance therapy compared with surveillance in patients with unresectable locally advanced, recurrent, or metastatic urothelial cancer who had not experienced disease progression after 4 to 6 cycles of first-line chemotherapy. The primary endpoint of the trial was PFS as assessed by Response Evaluation Criteria in Solid Tumor (RECIST) v1.1 criteria, while secondary endpoints included response rate, safety and tolerability of maintenance pemetrexed, and OS of each group.
2. Patients
Eligible patients were those aged 20 years or older with histologically or cytologically confirmed urothelial carcinoma (UC) of bladder, ureter, renal pelvis, or urethra. They had to have locally advanced, recurrent, or metastatic disease considered unsuitable for curative-intent local treatment. Patients were required to have completed 4-6 cycles of cisplatin-based or carboplatin-based first-line chemotherapy (GP, gemcitabine plus carboplatin [GC], classical MVAC) or dose-dense MVAC with no evidence of disease progression on a computed tomography (CT) scan performed within 3 weeks following the last chemotherapy cycle. Additional inclusion criteria were measurable disease per RECIST v1.1 (when first-line chemotherapy was administered), an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0-2, and adequate bone marrow (absolute neutrophil count ≥ 1,500/mm3, hemoglobin ≥ 9 g/dL, and platelet ≥ 100,000/mm3), hepatic (bilirubin < 1.5 mg/dL, transaminases < 1.5×upper limit of normal, alkaline phosphatase < 2.5×upper limit of normal) and renal function (serum creatinine < 1.5 mg/dL or creatinine clearance ≥ 45 mL/min).
Patients were excluded if they had received prior systemic chemotherapy for palliative purposes before or after first-line cisplatin-based or carboplatin-based chemotherapy. However, prior intravesical chemotherapy or immunotherapy was permitted. For patients with recurrent disease who had received prior adjuvant or neoadjuvant chemotherapy, the last administration had to be at least 6 months before the start of first-line chemotherapy. Other exclusion criteria included disease progression during or after first-line platinum-based chemotherapy, prior exposure to pemetrexed as systemic anticancer therapy, known central nervous system metastasis, a diagnosis of any serous secondary malignancy within the past 2 years (except for adequately treated basal cell or squamous cell carcinoma of the skin), early gastric carcinoma, early-stage thyroid carcinoma, insignificant prostate carcinoma, or in situ carcinoma of the cervix uteri), serious hypersensitivity reaction to pemetrexed its excipients (mannitol, hydrochloric acid, or sodium hydroxide), other serious medical or psychiatric conditions, and being pregnant or breastfeeding.
3. Treatment plan
Patients were enrolled and treated by the investigators at participating study center on an outpatient basis. After completing 4-6 cycles of first-line chemotherapy, patients who did not exhibit disease progression on CT scans conducted within 3 weeks of the last chemotherapy administration were randomized within 6 weeks to the maintenance pemetrexed group or the observation group. Treatment commenced within 2 weeks after randomization. Patients were randomly allocated in a 1:1 ratio to either the pemetrexed group or the observation group, and randomization was stratified based on total score (0-60 point vs. 61-120 point vs. 121 or more) derived from the post-treatment prognostic nomogram for patients with metastatic UC completing first-line cisplatin-based chemotherapy, as suggested by Galsky et al. [15], comprising baseline PS, baseline number of visceral metastatic sites, baseline white blood cell counts, and response to treatment. Randomization was performed centrally through an interactive web-based randomization system managed by the data center, and random table was generated by an independent statistical group. Group allocations were disclosed to the data center staff member only after all the participant details had been entered into the system and then the patient randomization number and assignment to treatment groups were provided by this system. Patients, investigators, and their teams, and those analyzing data were not masked to study treatment.
Patients in the pemetrexed maintenance group received pemetrexed 500 mg/m2 mixed in 100 mL normal saline as a 10-minute intravenous infusion on day 1 of each 21-day cycle, along with vitamin supplementation (folic acid 1,000 μg daily orally beginning 7 days before treatment initiation and vitamin B12 1,000 μg administered intramuscularly 7 days before treatment initiation, then every 3 cycles). Subsequently, vitamin B12 could be administered on the same day of pemetrexed infusion. Dexamethasone 4 mg was given orally twice daily for 3 days starting the day before treatment to minimize cutaneous reactions. Treatment continued until disease progression or intolerable toxicity, up to maximum of 16 cycles.
In the pemetrexed group, subsequent cycles were not initiated unless treatment-emergent toxicity had recovered to grade 1 or less (except anemia and alopecia). Dose adjustments or delays were permitted according to the protocol guidelines. Depending on the severity and recurrence of toxicity, the pemetrexed dose could be reduced from 500 to 400 mg/m2 or from 400 to 300 mg/m2. Patients requiring a third dose reduction or a treatment delay of more than 3 weeks were withdrawn from the study.
Patients allocated to the surveillance group received best supportive care according to institutional practice guidelines and underwent regular clinic visits with CT assessments as per the assessment protocol.
4. Assessment of response and toxicity
Baseline evaluations included a complete medical history, physical examination, ECOG PS assessment, laboratory measurements including a complete blood count (CBC), serum chemistry and electrolytes, chest X-ray, and electrocardiography. For women of childbearing potential, a urine beta human chorionic gonadotropin test was required. Tumor assessments using chest and abdominopelvic CT were performed during screening and subsequently every 9 weeks by the investigator according to RECIST ver. 1.1. If patients did not experience disease progression after the third disease evaluation, subsequent tumor assessments were performed every 12 weeks. Safety assessments included regular monitoring of vital signs, physical examinations, ECOG PS, CBC, serum chemistry and electrolytes, adverse events, and documentation of all concomitant medications and therapies, performed on D1 of each cycle. Toxicity was evaluated according to National Cancer Institute Common Terminology Criteria for Adverse Events ver. 4.03.
5. Statistical analysis
This study was designed to demonstrate the superiority of pemetrexed maintenance therapy over observation in terms of PFS. The sample size was determined to detect an improvement in PFS from 1.5 months with observation [8] to 2.6 months with pemetrexed [13], with a two-sided α=0.05, β=0.2 (power=80%), and assumption of 10% loss to follow-up, requiring a total of 156 patients (78 per group).
All randomized patients were included in the efficacy and safety analyses (intention-to-treat analysis). Baseline characteristics and safety profiles were summarized descriptively. Kaplan-Meier estimates and the log-rank test were utilized for analyzing the time-to-event endpoints, while hazard ratios (HRs) were estimated using the Cox proportional hazards model. PFS was defined as the time from randomization to the first occurrence of either disease progression or death from any cause. OS was defined as the time from randomization to death from any cause. The duration of follow-up was assessed using the reverse Kaplan-Meier method. All analyses were conducted with SPSS Statistics for Windows ver. 20.0 (IBM Corp.) and Kaplan-Meier plots were made by using GraphPad Prism 10. The R package Publish (R Foundation for Statistical Computing) was used to run the post-hoc PFS and OS subgroup analysis.
Results
1. Patient characteristics
This trial was terminated early due to poor patient accrual after the approval of avelumab as maintenance therapy. From October 2016 to December 2022, a total of 97 patients were randomized to either pemetrexed maintenance group (n=49) or the observation group (n=48) (Fig. 1).
Table 1 demonstrates baseline characteristics. The median age was 69 years (range, 43 to 90 years) in the pemetrexed group and 66 years (range, 33 to 82 years) in the observation group, with a predominance of males (63.3% and 72.9%, respectively). The first-line treatment regimen included GP in 83.7% and 83.3% of patients, GC in 12.2% and 6.2% of patients, and MVAC in 4.1% and 10.4% of patients in the pemetrexed and observation groups, respectively. Most patients received 6 cycles of first-line treatment. The response rates to first-line therapy were complete response (CR) in 8.2% of patients in the pemetrexed group and 18.8% in the observation group, partial response (PR) in 69.4% and 66.7%, and stable disease (SD) in 22.4% and 14.6%, respectively. The most common sites of metastasis were abdominopelvic lymph nodes (61.2% in the pemetrexed group and 64.6% in the observation group) and the lung (36.7% and 29.2%, respectively). The Galsky score, which was used as a stratification factor, was well balanced between the two arms. In the overall study population, 31 patients (63.3%) in pemetrexed group and by 29 patients (60.4%) in the observation group received subsequent anticancer therapy, including immunotherapy, platinum-based chemotherapy, and taxane (Table 2).
2. Efficacy
In the pemetrexed group, the median number of administered cycles was 6 (range, 1 to 16). The most common reason for discontinuation was disease progression (53.1%), followed by completion of the planned cycles (22.4%). After a median follow-up of 36.4 months, the median PFS was 6.0 months (95% confidence interval [CI], 3.4 to 8.5) in the pemetrexed group compared with 2.3 months (95% CI, 1.8 to 2.7) in the observation group (log-rank p=0.044), with a HR of 0.64 (95% CI, 0.41 to 0.99) (Fig. 2A). The median OS was 18.1 months (95% CI, 6.8 to 29.4) in the pemetrexed group versus 17.9 months (95% CI, 16.1 to 19.7) in the observation group (p=0.913; HR, 1.03; 95% CI, 0.61 to 1.73) (Fig. 2B). In the post-hoc exploratory subgroup analysis, there was a tendency toward better PFS with pemetrexed across all subgroups, while no difference in OS was observed in any subgroup (Fig. 3). The best response to maintenance pemetrexed included CR in five patients (10.2%), PR in six (12.2%), SD in 23 (46.9%), progressive disease in 14 patients (28.6%), and data not available for one patient (2.0%).
3. Safety
Adverse events in patients allocated to the pemetrexed group are presented in Table 3. The most frequent adverse events were anemia (30.6%), fatigue (18.4%), and neutropenia (12.2%). Notably, there were no reported cases of febrile neutropenia. With prophylactic dexamethasone use, skin rash occurred infrequently (4 patients, 8.2%). Overall, the adverse events were primarily grade 1 or 2 and were manageable with routine supportive care. One patient experienced grade 4 acute kidney injury, which was caused by urinary tract obstruction due to disease progression and resolved after a percutaneous nephrostomy.
Discussion
This phase III trial aimed to investigate the efficacy of maintenance pemetrexed in patients with aUC whose disease had not progressed after first-line platinum-based chemotherapy. The trial was prematurely terminated due to slow patient accrual after the approval of avelumab as a maintenance therapy. In this study, patients in the maintenance pemetrexed group achieved a median PFS of 6.0 months (95% CI, 3.4 to 8.5) and a median OS of 18.1 months (95% CI, 6.8 to 29.4). These results compare favorably with the observation group, which showed a median PFS of 2.3 months (95% CI, 1.8 to 2.7 months) and a median OS of 17.9 months (95% CI, 16.1 to 19.7) (PFS: p=0.044; HR, 0.64; OS: p=0.913; HR, 1.03). This study met its primary endpoint, demonstrating that maintenance pemetrexed significantly delayed disease progression. Additionally, pemetrexed maintenance was associated with a favorable toxicity profile, with manageable adverse events.
The trial was designed in 2015, with patient enrollment beginning in September 2016. At that time, there was no universally accepted second-line treatment option for aUC, and immunotherapy was just beginning to emerge as a potential treatment modality. Therefore, this study focused on evaluating cytotoxic chemotherapy as a maintenance therapy. The ideal agent for maintenance therapy should possess palliative efficacy and a manageable safety profile without cumulative toxicity. Although gemcitabine is widely used as a first-line chemotherapy agent for UC, gemcitabine monotherapy as a second-line treatment has demonstrated only modest activity and a substantial incidence of grade 3 or higher toxicity, particularly hematologic toxicity [16,17]. A retrospective study on gemcitabine maintenance therapy in UC reported considerable grade 3 or higher toxicity (17% grade 3 neutropenia and 3% grade 3 febrile neutropenia) [18]. Given these safety concerns and the weekly administration schedule of gemcitabine, we opted to design a switch maintenance strategy rather than using gemcitabine.
Prior to this study, no phase III clinical trial has demonstrated a survival advantage for any salvage chemotherapy regimen compared to another or to best supportive care as second-line treatment following first-line platinum-based chemotherapy. Various drugs have been tested in phase II trials, but their efficacy was generally modest, with an overall response rate (ORR) ranging from 5% to 28% [9-13,19,20]. Vinflunine was the only agent tested in a randomized phase III trial [8], but failed to demonstrate a statistically significant benefit in an intention-to-treat analysis. Pemetrexed showed an ORR of 28% in one trial [13], along with a favorable toxicity profile without cumulative neurotoxicity, making it a suitable candidate for maintenance therapy.
The treatment landscape of aUC has evolved rapidly. Avelumab became the standard maintenance therapy after the JAVELIN Bladder 100 trial, which demonstrated significant PFS and OS benefits with avelumab maintenance plus best supportive care compared with best supportive care alone (PFS, 5.5 months vs. 2.1 months; HR, 0.54 [95% CI, 0.46 to 0.64]; p < 0.001; median OS, 23.8 months vs. 15.0 months; HR, 0.76 [95% CI, 0.63 to 0.91]; p=0.0036) [21,22]. In our trial, pemetrexed showed a promising effect in delaying disease progression (median PFS, 6.0 months in the pemetrexed group vs. 2.3 months in the observation group) comparable to avelumab (median PFS, 5.5 months in the avelumab group vs. 2.1 months in the best supportive care group). However, there was no significant difference in OS. The discrepancy between PFS and OS might be attributable to subsequent treatments after progression and baseline patient characteristics. The pemetrexed group had somewhat unfavorable baseline characteristics, including a higher median age (69 vs. 66 years), fewer patients with locally advanced cancer (2% vs. 12.5%), and fewer patients achieving CR after first-line treatment (8% vs. 19%).
Other trials have also explored switch maintenance therapy in aUC. The SOGUS 2011/02 MAJA trial investigated maintenance vinflunine compared with best supportive care. Vinflunine maintenance resulted in better PFS (median PFS, 6.5 months vs. 4.2 months; p=0.031), but there was no significant difference in OS (median OS, 16.7 months vs. 13.2 months; p=0.182) [23,24]. The most common grade 3 or 4 adverse events in the vinflunine group were neutropenia (18%), fatigue (16%), and constipation (14%). The adverse event profile was better than second-line vinflunine, likely due to the inclusion of patients with ECOG PS 0-1 and a starting dose adjusted based on depending on patients characteristics. Although both the MAJA trial and our study are underpowered, the results suggest that maintenance therapy may have a role in delaying progression. Considering the adverse events, pemetrexed appears to be a more favorable choice for maintenance therapy.
This study has several limitations. First, the premature termination of the trial resulted in a smaller sample size than initially planned, rendering the study statistically underpowered. Second, we did not collect quality-of-life data, which would have provided additional insights into the patient experience during treatment. Third, the standard treatment paradigm for aUC changed during the enrollment period, limiting the applicability of our findings to current clinical practice. Although the PFS benefit of pemetrexed maintenance is noteworthy, these results may not be directly translatable to the real-world setting given the shift in first-line treatment strategies.
Looking forward, the role of maintenance therapy in aUC may be redefined within novel treatment frameworks. In 2023, the standard first-line therapy for aUC changed for the first time in 30 years. The EV302 trial demonstrated that pembrolizumab combined with enfortumab vedotin (EV) achieved impressive response rates, PFS, and OS (68%, 12.5 months, and 31.5 months, respectively) [25]. Consequently, previous strategies, such as platinum induction followed by avelumab maintenance, are no longer standard. Future research should focus on optimizing first-line treatment regimens to enhance efficacy and reduce toxicity. Given the PFS benefit observed with pemetrexed, combining maintenance pemetrexed with pembrolizumab could be explored after induction therapy with EV and pembrolizumab. This approach may mitigate some of the cumulative toxicities associated with EV, such as skin toxicity, neuropathy, and fatigue [26], which led to treatment discontinuation in 22% of patients in the EV302 trial (any grade: 97%; grade ≥ 3: 56%) [25]. There are also intriguing preclinical data suggesting that pemetrexed may enhance the efficacy of immunotherapy [27,28]. On the other hand, pemetrexed maintenance can also be considered in a later line. Although there is a lack of consensus on the subsequent treatment after EV plus pembrolizumab, European Society of Medical Oncology guidelines recommend platinum-based chemotherapy as a second-line therapy [29]. If platinum-based chemotherapy is used, pemetrexed maintenance can be followed in patients without disease progression. Currently, we do not know which patients may benefit from pemetrexed maintenance, so its use should be approached with caution and validation in this setting is needed.
The PREMIER trial demonstrated that switch maintenance therapy with pemetrexed significantly prolonged PFS in patients with aUC who had not progressed after first-line platinum-based chemotherapy, with a favorable safety profile. Further investigation into combination maintenance therapies involving pemetrexed is warranted.
XML Download