Metronomic oral vinorelbine doublet chemotherapy with carboplatin in treatment of advanced lung cancer: a feasibility and safety study

Background

Non-small cell lung cancer (NSCLC) is the second most common cancer worldwide.¹ In Denmark 4,658 new cases were diagnosed yearly from 2012 to 2016.² The treatment goal for patients with metastatic NSCLC is to relieve symptoms and prolong survival while maintaining the best possible quality of life. Palliation, therefore, needs to be balanced against survival and toxicity. In most countries platinum based combination chemotherapy has been the standard first line treatment in NSCLC for decades, and in Denmark platinum has been used in combination with vinorelbine in NSCLC (www.dolg.dk).

When chemotherapy is given according to conventional principles, the treatment is typically given in high doses every three weeks, as the goal is to maximize the therapeutic outcome. However, because of the toxicity, there is a need for treatment-free intervals between the cycles, which could provide space and time for the development of resistant clones and regrowth of the tumour cells.³ Furthermore, the immunosuppressive effect of the chemotherapy can lead to infections and the need of antibiotic therapy and hospitalization and can ultimately be fatal.

Compared to classical cytotoxic chemotherapy, metronomic treatment regimens may have a complementary mechanism of action, targeting the tumour vasculature, thus counteracting tumour regrowth that may occur between chemotherapy cycles; in addition, metronomic chemotherapy suppresses regulatory T cells and induces the maturation of dendritic cells, thereby leading to an anti-tumour immune response.^4,5 Metronomic regimens, in which small, frequent doses of chemotherapy are administered, have been suggested to lower treatment-associated toxicities.⁶ The high frequency of administration aims to expose tumour cells continuously to the drug, preventing recovery between cycles and possibly improving tumour control.⁷ Preclinical studies, as well as clinical observations, suggest that metronomic regimen increases the effect of a drug, not only by a direct effect on cancer cells, but also by activating endothelial cells in the tumour vasculature and thereby the tumour anti-angiogenesis. By giving smaller, but more frequent, doses of the drug, higher dose intensity, without corresponding side effects, is obtained.⁸

In a randomized phase II study, conducted in NSCLC patients, toxicity and efficacy between intravenous and oral administered vinorelbine, seemed comparable, and safety profiles of both formulations appeared to be qualitatively similar.⁹ Easy oral administration of vinorelbine, allows for flexible treatment schedules including a more frequent and metronomic dose application, and makes it an obvious candidate for metronomic studies.

Many studies have focused on comparing the efficacy of different drugs,¹⁰ but little interest has been given to the posology.^11–16 Phase II studies with weekly intravenous vinorelbine 25-30 mg/m² as first line therapy to advanced breast cancer patients showed response rates of 41-50%, and acceptable toxicity profile.^11–14 A previous phase I dose finding study in pre-treated NSCLC patients showed that metronomic daily oral vinorelbine as monotherapy, given in cycles of four weeks with one week of drug holiday, was safe and feasible in recommended dose level of 30-40 mg daily.¹⁷ Platinum-based doublet chemotherapy in a phase I dose finding study determined that cisplatin 85 mg/m² every third week in combination with oral vinorelbine 60 mg three times weekly was feasible and active in advanced NSCLC.¹⁶ Upon registration of a drug there is often limited knowledge of the best dosing and there could be much to gain in this field. Metronomic chemotherapy is not yet well validated in the clinic, and the optimal dosage and regimens are yet to be identified.

There are no published studies with daily oral vinorelbine in combination with carboplatin for treatment of patients with advanced non-small cell lung cancer. The current study aims to test a daily dosage of vinorelbine 20/30 milligram (mg) in combination with carboplatin focusing on the safety aspect and tolerability. If this outcome is achieved this dosing schedule would be usable in future randomised combination studies with carboplatin including maintenance arm with vinorelbine in lung cancer treatment.

Methods

Study design

This study was investigator-initiated and designed as a prospective non-randomized, open-label, single site, safety, and feasibility study. It was planned to recruit 20 patients. Patients with incurable NSCLC, who were chemo-naïve and candidates for 1^st-line chemotherapy (PD-L1 < 50%), were included. Patients included in the study received standard dose of carboplatin AUC 5 every 3^rd week and metronomic oral Navelbine® (20/30 mg) daily for 12 weeks. After four cycles of chemotherapy, if partial response or stable disease, patients with non-squamous carcinoma received pemetrexed as maintenance chemotherapy according to national standard treatment recommendation (www.dolg.dk). Patients were followed by computer tomography (CT) scans for evaluation every second month and evaluated according to Response Evaluation Criteria In Solid Tumours (RECIST).¹⁸ The study design is shown in Figure 1. Each patient received first cycle as carboplatin AUC 5 day 1 and Navelbine® 20 mg daily in first 3 weeks, and if there was acceptable toxicity, the dose of Navelbine® was escalated to 30 mg daily. A maximum of four cycles were delivered. For safety reason the first six patients enrolled had World Health Organisation (WHO) performance status (PS) 0-1. The next fourteen patients included were allowed to have PS 0-2.

Figure 1. Study design.

There was incorporated an interim analysis in the study. We proceeded with the analysis when the 10^th patient had received 2 cycles (6 weeks) of treatment or if 4 patients experienced an unacceptable toxicity (UT; see definition in appendix A in extended data in figshare datafile¹⁹), despite the dose reduction of Navelbine and carboplatin or in case of any grade 5 toxicity.

Dose adjustment plan and definition: A dose adjustment of chemotherapy was permitted according to treatment modification guidelines at the Department of Oncology, Aarhus University Hospital (see guidelines for treatment modification in appendix A in extended data in figshare datafile.¹⁹). Dose adjustment was allowed in the case of significant haematological and/or non-haematological toxicity. Generally, dose adjustment and/or treatment delay was decided at the respective visit based on symptoms, physical examination and haematological values (see Dose Adjustment section, appendix A in extended data in figshare datafile¹⁹).

Patients were instructed to report any toxicity that occurred between cycles or in between visits. Moreover, during the entire treatment period patients kept a diary. They registered dose and time of taking Navelbine and brought their diary on every visit. The patients were asked to register any adverse event in the diary. There were scheduled five visits during the 12 weeks of treatment. Regarding antiemetics there were no restrictions. Metoclopramide was given in addition to the standard antiemetics (oral prednisolone 75 mg day 1, 50 mg day 3-4 and 25 mg day 5; aprepitant 125 mg day 1, 80 mg day 2-3 and ondansetron 16 mg day1-2), when treating with a standard carboplatin and vinorelbine regimen. Significant effort has gone into using the same antiemetics across patients in the study.

Safety assessment: Symptoms and haematological values were measured at baseline and then every third week during the four cycles. Additional haematological values were measured at day 12 ± 2 during the first cycle to register any haematological side effects. If haematological toxicity occurred the patient was contacted by telephone.

All toxicities were graded according to the Common Terminology Criteria for Adverse Events (CTCAE version 4.03), during the first 84 days (12 weeks) of the treatment period. Toxicity was registered as patient-reported outcome (via online records/AmbuFlex®), every 3^rd week during the treatment period. The focus was on the following: febrile neutropenia, neutropenia (reported only by study-nurse or physician according to blood tests), any fever, hospitalization, vomiting, diarrhoea, stomatitis, anorexia, dyspnoea, weight loss, thromboembolic events and any other adverse events. All unacceptable toxicities (UTs) such as febrile neutropenia, thrombocytopenia or any non-haematological grade 3 and 4 toxicity were defined (see unacceptable toxicities (UTs), appendix A in extended data in figshare datafile¹⁹).

Statistical methods

Results

20 patients with inoperable or disseminated, pathologically proven NSCLC, from May 2018 to February 2019, were included. The median age was 70.5 years (range 49-83). 19 patients (95%) had adenocarcinoma. Patient characteristics are shown in Table 1. Two patients withdrew their consent within one week; one decided not to receive any palliative treatment; the other decided to receive standard treatment, as the patient could not cope with protocol requiring an extra blood sample during the first cycle. These two patients are only included in the intent-to-treat population in the overall survival analyses.

18 patients were evaluable for toxicity analysis as shown in the flowchart in Figure 2. These eighteen patients received a total of 62 cycles of carboplatin in addition to daily oral vinorelbine. The median dosage of oral vinorelbine received per patient was 1780 mg (range 260-2370 mg). 13 patients received all four cycles of chemotherapy (see appendix B, fig 5 in extended data in figshare datafile²¹). One patient received only three cycles, as the physician clinically suspected progression, although the following evaluating CT-scan showed stable disease. Four patients received only 1-2 cycles of chemotherapy; one patient had clinical progression (without verifying CT scan); one patient had decline in PS and was terminated from study (see flowchart, Figure 2); two patients had clinical disease progression during the treatment time and performed an evaluation CT scan in advance.

Figure 2. Study flowchart.

The majority of patients had grade 1 or 2 toxicity. Frequent toxicities (any grade) included fatigue 13 (72%), diarrhoea 13 (72%), constipation 13 (72%), nausea 9 (50%) and dyspnoea 12 (67%). Some patients had more than one toxicity. Grade 3 toxicities were dyspnoea 2 (11%), nausea 3 (17%) and fatigue 3 (17%) as shown in toxicity Table 2. These mentioned toxicities led to a brief break in daily vinorelbine intake and dose reduction according to dose adjustment guidelines (in appendix A in extended data in figshare datafile).¹⁹ Two patients (11%) had febrile neutropenia as a grade 4 toxicity: one patient recovered after three weeks and continued the treatment in reduced dose; the other patient recovered after six days, but did not receive any treatment hereafter, according to investigator’s decision, as the PS disqualified the patient from any further treatment (exacerbation of patient’s chronic obstructive lung disease). Both febrile neutropenia events occurred at day 12 and 13 after the first cycle, respectively. None of the patients included in the study had cumulative haematological toxicity. No patients in the study had grade 5 toxicity.

Six (46%) of the thirteen patients who received all four cycles did not have reduction in vinorelbine doses (20/30 mg), even though the total dosage varies per patient from 2200-2370 mg, as treatment cycle period varied from 19-23 days. Five patients received a reduced dosage of vinorelbine from the second cycle (stayed at 20 mg without dose escalation to 30 mg): one patient had a single rise in alanine aminotransferase (ALT) grade 2 at first cycle; one had diarrhoea grade 2; one had fatigue and weight loss grade 3; one had dose interruption in first cycle because of febrile neutropenia grade 4 and started the second cycle at 20 mg (a dose reduction) and one patient continued at 20 mg in all treatment periods, because of the patient’s age, 80 years (physician’s decision). Two patients received full doses of chemotherapy up to the third cycle but had an interruption of 3-5 days in vinorelbine intake at the end of the fourth cycle, because of fatigue and nausea grade 2. Four patients did not receive more than 1 to 2 cycles due to degrading in PS and/or disease progression as mentioned above.

The total dose of vinorelbine administered per patient per cycle and response characteristics is shown in appendix B, fig 5, in extended data in figshare datafile.²¹

Median follow-up time was 22.1 months (95% CI: 18.9 – 25.2 months). At the end of the study, four patients were alive. Sixteen patients had eligible progression free survival (PFS) data for analysis. Median PFS was 4.8 months (95% CI: 4.3 – 5.2 months). Median overall survival was 14.1 months (95% CI: 5.9 – 18.6 months) for the entire population, see Figure 3. Five patients (28%) out of 18 patients had partial response (PR), and six patients (33%) had stable disease (SD). Five patients (28%) had progressive disease (PD) in evaluation scans. One of these patients had regression in target lesion but was considered PD because of newly developed malignant pleural exudate (as marked with asterisk in the waterfall plot, Figure 4). Two patients could not go through evaluation with CT scans, but developed clinical symptoms of progression and decline in PS. Response rate (RR) in this feasibility and safety study was 61% (11 out of 18 patients).

Figure 3. Kaplan-Meier curve for overall survival in intent to treat population (n = 20).

Figure 4. Waterfall plot shows responses in 16 patients, all with evaluation CT scans.

*Patient had regression in evaluable target lesion, but CT scan showed a new lesion:

malignant pleural exudate.

Discussion

This study showed that metronomic oral vinorelbine 20/30 mg daily in combination with carboplatin AUC 5 was safe and feasible. The most frequent toxicities such as fatigue, diarrhoea, and constipations have been manageable.

Oral vinorelbine as metronomic monotherapy has been evaluated, using different dose regimens, such as fractionated regimen (day 1, 3 and 5) and daily intake.^22–26 In patients with NSCLC the daily intake of oral vinorelbine up to 40 mg, as monotherapy, was well-tolerated.^22,23 In a study, Tufman et al. showed no relevant accumulation in the pharmacokinetic analysis, using vinorelbine as monotherapy in a daily dosing of 30-40 mg.¹⁵ Oral vinorelbine is rapidly absorbed at 80 mg/m² (Tmax 1.4 ± 0.7 hours) and shows a bioavailability of 43% ± 14%.⁸ The absorption is not affected to a significant extent by food.²⁷

Other studies have focused on metronomic oral vinorelbine as combination chemotherapy. A dose finding phase I doublet chemotherapy study of oral vinorelbine 50 mg 3 times weekly in combination with cisplatin 80 mg/m² (day 1 every 3 weeks) showed no haematological toxicities. With escalation of vinorelbine up to 70 mg 3 times weekly plus 85 mg/m² cisplatin per cycle, grade 3 or 4 toxicities were dose limiting.¹⁶ The Hellenic Oncology Research Group (HORG), in a phase II study, evaluated the safety and efficacy of metronomic vinorelbine in combination with cisplatin as first line treatment in patients with advanced NSCLC.²⁸ Patients were treated with cisplatin (80 mg/m²) in combination with oral metronomic vinorelbine (60 mg, three times weekly) in cycles of 21 days. They reported myelosuppression as the main adverse event with grade 3 and 4 neutropenia occurring in five (14.3%) and six (17.1%) patients, respectively.²⁸

Several breast cancer studies, with metronomic vinorelbine concomitant with capecitabine, have also found that vinorelbine three times weekly is safe and has unchanged efficacy.^29–31 Two of these studies indicate that it might be possible to improve both efficacy and safety with daily metronomic therapy.

There are no published studies of daily oral vinorelbine, without any planed break, in combination with carboplatin for treatment of NSCLC.

In the current study, we examined whether the doublet chemotherapy with carboplatin every third week concomitant with daily oral vinorelbine in 12 weeks was safe and feasible. Fatigue, constipation, and diarrhoea grade 2 was the most frequent reason for short interruptions in daily intake or dose reduction of vinorelbine to 20 mg daily. We found that the above-mentioned doublet regimen in treatment of NSCLC patients is feasible with both 20 mg daily with escalation to 30 mg daily or only 20 mg daily in the entire treatment period. The observed partial response or stable disease (in 11 out of 18 patients) and disease progression were not correlated with the daily dose of vinorelbine as shown in appendix B, fig 5, in extended data in figshare datafile.²¹ The study design, carboplatin in combination with daily oral vinorelbine 20 mg (or 20/30 mg) without planned breaks, could be useful in future studies, that includes metronomic doublet regimens.

During the first cycle haematological quantities were also measured at day 12 ± 2 and showed no significant reduction in neutrophil counts. Therefore, routine measuring of haematological quantities at day 12 ± 2 in the first cycle may not be necessary. The tolerability profile, with grade 3; fatigue 17%, nausea 17%, grade 4; febrile neutropenia 11% and no grade 5 toxicity, was acceptable and comparable with historical doublet regimen with cisplatin every third week and metronomic vinorelbine 60 mg, 3 times weekly, in first line chemotherapy in NSCLC.²⁸ Progression free survival of 4.8 months in this study was comparable with other studies using platinum based standard doublet treatments with vinorelbine (PFS: 4.2 months) or pemetrexed (PFS: 4.3 months).³²

This study was designed to determine the safety and clinical feasibility and was not powered to conclude on efficacy. Therefore, the trend of higher RR and longer OS is associated with a high level of uncertainty.

In conclusion, carboplatin in combination with metronomic daily oral vinorelbine 20/30 mg without any pre-planned breaks is a safe and feasible option in the treatment of patients with NSCLC. The toxicity rates, PFS and OS were comparable with standard regimens.

Data availability