Nab-pacl*taxel, cisplatin, and capecitabine versus cisplatin and gemcitabine as first line chemotherapy in patients with recurrent or metastatic nasopharyngeal carcinoma: randomised phase 3 clinical trial (2024)

Study design

This was a multicentre, open label, randomised, controlled, phase 3 study conducted in eligible patients at four hospitals located in China (supplementary table A). Figure 1 shows a flowchart illustrating the study design. All patients signed written informed consent before enrolment.

Participants

Patients were eligible if they had histological or cytological confirmation of nasopharyngeal carcinoma. Other key inclusion criteria included primary metastatic disease or metastatic disease after curative radiotherapy, no previous receipt of systemic therapy for metastatic disease, Eastern Cooperative Oncology Group performance status of 0 or 1, at least one measurable lesion according to Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1), adequate organ function, an estimated life expectancy of at least 12 weeks, and age ≥18 years. We excluded patients with a history of other malignancies, central nervous system metastases, severe coexisting illness, or pregnancy. The complete eligibility criteria are given in the trial protocol, available as an online supplement.

Randomisation and masking

A computer generated random number code was provided for randomisation of enrolled patients at the Sun Yat-sen University Cancer Center. Random distribution sequences were generated using a permuted block of flexible size (2, 4, 6, or 8), with no stratification factors. Randomisation details were provided in sequentially numbered, opaque, sealed envelopes prepared by a data management team unaware of the patient assignment process. The therapy assignment was not masked to investigators, patients, and other treating oncologists, but the central evaluation of the imaging data by an independent review committee was done in a masked manner.

Procedures

Before randomisation, patients received a comprehensive screening assessment within four week intervals between enrolment and start of treatment. Eligible patients were randomly assigned (1:1) to the nab-TPC or gemcitabine plus cisplatin cohort. Patients assigned to the nab-TPC cohort received nab-pacl*taxel (200 mg/m²) on day 1, cisplatin (60 mg/m²) on day 1, and capecitabine (1000 mg/m² twice daily) on days 1-14 of each three week cycle for up to six cycles, followed by capecitabine maintenance for a maximum of two years. Patients assigned to the gemcitabine plus cisplatin cohort received gemcitabine (1g/m²) on days 1 and 8 and cisplatin (80 mg/m²) on day 1 of every three week cycle for up to six cycles, followed by best supportive care. Trial assigned chemotherapy was continued until progression of disease, intolerable toxicity, investigator’s decision, or withdrawal of consent, whichever occurred first. Crossover between the two chemotherapy regimens was permitted at disease progression. Details of the pretreatment screening evaluations and chemotherapy dose modifications during the trial are available in the trial protocol.

We assessed radiological responses of tumours every six weeks during the course of chemotherapy and every 12 weeks thereafter, according to RECIST v1.1, with 18F-fluorodeoxyglucose positron emission tomography-computed tomography or enhanced magnetic resonance imaging or computed tomography scan. We used the National Cancer Institute Common Terminology Criteria, version 4.0, to evaluated adverse events during treatment and for a period of 90 days after the last dose of study medication.

Endpoints

Progression-free survival served as the primary endpoint, which we assessed as the time from randomisation to disease progression according to RECIST v1.1 evaluated by the independent review committee, or death, whichever occurred first. Secondary endpoints consisted of overall survival, objective response rate, and safety. We defined overall survival as the time from randomisation to death from any cause and the objective response rate as the proportion of patients showing a best overall response, including complete and partial response.

Statistical analysis

We derived the sample size from the previously reported median progression-free survival (7.0 months) in a phase 3 trial in the setting of patients with recurrent or metastatic nasopharyngeal carcinoma treated with a gemcitabine plus cisplatin regimen as the first line option.7 We calculated that we would need to randomise approximately 134 patients (67 patients per group), providing 80% power to detect the underlying hazard ratio of 0.42 for the primary analysis of progression-free at a two sided α level of 5% with one interim analysis. An interim analysis of progression-free survival, as prespecified, was conducted approximately three years after the first patient was randomised.

We did all analyses within the intention-to-treat population. Safety analyses included all patients who received at least one dose of the study treatment. We presented continuous variables as medians with interquartile ranges and compared them by using the Mann-Whitney test. We assessed categorical variables by using either the χ² test or Fisher’s exact test for comparison. We estimated survival curves and rates for time-to-event endpoints (overall survival, progression-free survival, and duration of response) by using the Kaplan-Meier method. We used stratified log-rank tests to compare survival between the two treatment cohorts. We used a Cox proportional hazards model to calculate hazard ratios and their corresponding 95% confidence intervals, with stratification according to the study design. We verified the assumption of proportional hazards by examining the Schoenfeld residuals. In addition, we did post hoc subgroup analyses stratified by various factors, including age (<50 v ≥50 years), gender (male v female), smoking history (yes v no), cancer stage (primary metastases v recurrent), number of metastatic organs (1 v ≥2), presence of liver metastasis (present v absent), presence of lung metastasis (present v absent), presence of bone metastasis (present v absent), and levels of Epstein-Barr virus DNA (<2000 v ≥2000 copies/mL). We tested treatment-by-covariate interaction by using the Cox proportional hazards model as the basis for the interaction study.

The progress of trial was overseen by an independent data monitoring committee, empowered to determine whether early termination of the trial was warranted. To uphold an overall type I error rate of 0.05 throughout the trial, early termination followed the O’Brien-Fleming type boundary (with an α level of 0.005). An interim analysis done on 31 October 2022 led the independent data monitoring committee to terminate the trial on the basis of the findings. This report presented the data of survival and adverse events.

We used R software (version 4.0.5) and SPSS (version 24.0) for statistical analyses. The codes used for analysis are provided as an online supplement. We deemed a significance threshold of 0.05 or lower for two sided P values to indicate statistical significance.

Patient and public involvement

Although not initially involved in designing the trial, all participants were informed of the trial’s objectives and contents on recruitment. Patients were not involved in the study’s execution or subsequent report preparation. When we began the research, patient and public participation was not a routine practice in our specialty in this region.

Patients

Between October 2019 and August 2022 a total of 81 eligible patients with recurrent or metastatic nasopharyngeal carcinoma across four hospitals were randomly assigned to treatment with either nab-TPC (n=41) or gemcitabine plus cisplatin (n=40) (fig 1). Baseline demographics and disease characteristics were balanced between the treatment groups (table 1). The median age was 48 (interquartile range 36-57) years, and 79% of patients were male. Non-keratinising carcinoma was predominant among the patients (98%). Approximately two thirds of patients had previously undergone induction and concurrent chemoradiotherapy, and all patients had distant metastases. Thirty three (80%) patients in the nab-TPC cohort and 33 (83%) patients in the gemcitabine plus cisplatin cohort received six cycles of chemotherapy. The initial cost for each group is shown in supplementary table B.

After progression was documented, more than half of the patients received second line or third line chemotherapy, with 23 (56%) patients in the nab-TPC cohort and 26 (65%) patients in the gemcitabine plus cisplatin cohort receiving such treatment (supplementary table D). The predominant regimen used was anti-programmed cell death protein-1 inhibitor, administered to 21 (51%) patients in the nab-TPC cohort and 20 (50%) patients in the gemcitabine plus cisplatin cohort. Additionally, a portion of patients in both cohorts received crossover treatment—12 (29%) in the nab-TPC cohort and 12 (30%) in the gemcitabine plus cisplatin cohort.

Efficacy

As of the interim analysis cut-off date (31 October 2022), the median follow-up period was 15.8 (interquartile range 11.6-24.9) months. For the primary endpoint of progression-free survival, as evaluated independently, 27 (66%) of 41 patients in the nab-TPC cohort and 32 (80%) of 40 patients in the gemcitabine plus cisplatin cohort had disease progression or died. The median progression-free survival was 11.3 (95% confidence interval 9.7 to 12.9) months for the nab-TPC cohort and 7.7 (6.5 to 9.0) months for the gemcitabine plus cisplatin cohort. The difference in median progression-free survival between the two cohorts was statistically significant (hazard ratio 0.43, 95% confidence interval 0.25 to 0.73; P=0.002) (fig 2). The proportional hazards assumption for progression-free survival was unmet (supplementary figure A). In multivariate analyses, treatment group was an independent prognostic factor for progression-free survival (supplementary table F). For patients whose pre-treatment Epstein-Barr virus DNA was ≥2000 copies/mL, the nab-TPC regimen significantly improved their progression-free survival (hazard ratio 0.42, 0.21 to 0.84; P=0.01, supplementary figure B). However, this survival benefit was not observed for patients whose pre-treatment Epstein-Barr virus DNA was <2000 copies/mL (hazard ratio 0.45, 0.17 to 1.19; P=0.11). Improvement in progression-free survival was consistent in patients with or without liver metastases (supplementary figure C). Figure 3 shows the outcomes of the primary endpoint, independently assessed progression-free survival, in the primary analyses for post hoc subgroups.

In the updated analysis done on 3 June 2023, we found that 31 (76%) patients in the nab-TPC cohort and 35 (88%) patients in the gemcitabine plus cisplatin cohort had experienced disease progression or died according to an independent review committee. The median progression-free survival, as assessed by the independent review committee, was 11.9 (10.0 to 13.8) months in the nab-TPC cohort and 7.6 (6.6 to 8.7) months in the gemcitabine plus cisplatin cohort. The hazard ratio was 0.39 (0.24 to 0.65; P<0.001) (supplementary figure D). Sixteen deaths were reported, with eight (20.0%) deaths in the gemcitabine plus cisplatin cohort and eight (20%) deaths in the nab-TPC cohort. The overall survival data were still immature for both cohorts at the time of analysis.

Tumour responses, as assessed by the independent review committee during the interim analysis, are outlined in supplementary table E. We observed complete responses in five (12%) patients in the nab-TPC cohort and four (10%) patients in the gemcitabine plus cisplatin cohort, as well as partial responses in 29 (71%) patients in the nab-TPC cohort and 21 (53%) patients in the gemcitabine plus cisplatin cohort. The nab-TPC cohort had a significantly higher objective response rate than the gemcitabine plus cisplatin cohort (34/41 (83%) v 25/40 (63%); P=0.05). Among responders, the median duration of response was notably longer in the nab-TPC cohort (10.8 (8.8 to 12.8) months) than in the gemcitabine plus cisplatin cohort (6.9 (5.5 to 8.2) months), with a hazard ratio of 0.42 (0.22 to 0.81; P=0.009) (fig 2). However, the disease control rate was similar in the two treatment cohorts (1/41 (98%) v 3/40 (92.5%); P=0.36). Tumour burden and response in patients with baseline positive or negative plasma EBV DNA levels are shown in supplementary table C.

Safety

Table 2 shows the treatment related adverse events that occurred in ≥5% of patients. The most common adverse events were haematological and gastrointestinal. However, the overall incidence of grade 3 or 4 adverse events was low in both of cohorts (table 2). Treatment related grade 3 or 4 adverse events that differed significantly between the nab-TPC and gemcitabine plus cisplatin cohorts were leucopenia (10% v 33%); P=0.02), neutropenia (15% v 40%; P=0.01), and anaemia (2% v 20%; P=0.01). The rate of discontinuation due to drug related adverse events was 2% (1/41) in the nab-TPC cohort and 3% (1/40) in the gemcitabine plus cisplatin cohort. Dose reductions occurred in five (12%) patients in the nab-TPC cohort and 10 (25%) patients in the gemcitabine plus cisplatin cohort. No treatment related deaths were reported during the trial.

Comparison with other studies

In a landmark phase 3 trial conducted in 2016 in patients with recurrent or metastatic nasopharyngeal carcinoma, gemcitabine plus cisplatin showed superior progression-free survival compared with a fluorouracil plus cisplatin regimen, thereby establishing gemcitabine plus cisplatin as a preferred first line regimen option.67 However, the study found a median progression-free survival of seven months with the gemcitabine plus cisplatin regimen. In our recently reported prospective phase 3 trial, using capecitabine maintenance following pacl*taxel, cisplatin, and capecitabine (TPC) chemotherapy for disease control in metastatic nasopharyngeal carcinoma, the median progression-free survival extended to 35.5 months. Interestingly, we noticed that the response rate with TPC was higher than for the gemcitabine plus cisplatin regimen before capecitabine maintenance was initiated (74% v 64%), strongly indicating superior tumour control efficacy with TPC chemotherapy.10 Nevertheless, nab-pacl*taxel has shown a higher complete remission rate than conventional taxanes, with anaphylaxis being rare.14 Owing to the promising antitumour effects of the nab-TPC regimen, we conducted a controlled, randomised, phase 3 trial to assess its antitumour efficacy compared with a standard gemcitabine plus cisplatin regimen for recurrent or metastatic nasopharyngeal carcinoma.

The median progression-free survival was 7.7 months for the gemcitabine plus cisplatin group in this study, which was similar to that reported in the previous studies in recurrent or metastatic nasopharyngeal carcinoma.915 Notably, the nab-TPC treatment cohort showed a significant improvement, with a median increase of 3.6 months in progression-free survival among the intention-to-treat population. This improvement represents a 57% reduction in immediate risk of disease progression or mortality compared with the gemcitabine plus cisplatin treatment group. We note that median progression-free survival of nab-TPC regimen in this trial was shorter than the previously reported result. Several reasons contributed to this difference. First of all, the results from a post hoc subgroup analysis showed that patients without liver metastasis would more likely to achieve improved survival through capecitabine maintenance therapy.10 More than half of patients (63.4%) enrolled in this trial had liver metastasis, compared with 38.5% of patients in the previous study.1016 Secondly, and importantly, the method of calculating progression-free survival is different. In this trial, we assessed it from randomisation (before chemotherapy started) to disease progression, whereas in the previous study it was counted from the time of randomisation (after achievement of disease control with the TPC regimen) to disease progression. In addition, patients who progressed during TPC treatment were excluded from capecitabine maintenance therapy in the previous study.

In our study, the nab-TPC regimen achieved a significant 20% improvement in objective response rate compared with the gemcitabine plus cisplatin group during the initial phase of induction chemotherapy in patients with recurrent or metastatic nasopharyngeal carcinoma (82.9% v 62.5%; P=0.05). Thus, this nab-TPC regimen represents an important option for patients in need of clinically meaningful tumour shrinkage. The objective response is commonly regarded as an independent predictor of survival in patients with solid tumours and may serve as a potential indicator for overall survival and progression-free survival. Meanwhile, objective response rate is strongly correlated with progression-free survival in metastatic nasopharyngeal carcinoma.1718 Our data are consistent with previous findings regarding the association between objective response rate and progression-free survival. Furthermore, we observed that the progression-free survival curve in the two groups started to separate at the end of the induction chemotherapy phase (fig 2), indicating the superior efficacy of nab-TPC versus the gemcitabine plus cisplatin regimen even in the induction phase. In addition, the significantly improved duration of response was also observed in the nab-TPC group (median duration of response 10.8 v 6.9 months; P=0.009). Likewise, as part of the nab-TPC regimen, maintenance capecitabine therapy may also play a role in improving progression-free survival. These results suggest that the nab-TPC regimen possesses more potent and enduring anticancer effects than the gemcitabine plus cisplatin regimen and could be considered as a front line treatment option for patients with recurrent or metastatic nasopharyngeal carcinoma.

Recently, three phase 3 trials reported a significant improvement in progression-free survival for gemcitabine and cisplatin plus programmed cell death protein 1 inhibitors compared with gemcitabine and cisplatin alone.91519 These findings suggest that the combination of programmed cell death protein 1 inhibitors with gemcitabine plus cisplatin has emerged as the new standard treatment for patients with recurrent or metastatic nasopharyngeal carcinoma in the front line setting. Furthermore, investigating whether the use of nab-TPC as chemotherapy backbone, combined with programmed cell death protein 1 inhibitors, improves efficacy compared with gemcitabine plus cisplatin with programmed cell death protein 1 inhibitors would be worthwhile

The adverse events profiles of nab-TPC and gemcitabine plus cisplatin in this trial were as expected. No new toxicity signals were noted during the trial. All the adverse events were generally manageable. Treatment related haematological adverse events were more frequent with the gemcitabine plus cisplatin regimen, whereas hand-foot syndrome was associated with nab-TPC regimen. However, the observed rate of hand-foot syndrome is comparable to that reported in previous clinical trials.10 The proportion of patients discontinuing the study drugs was similar between the two cohorts.

Limitations of study

This trial has some inevitable limitations. Firstly, as all participants were sourced from endemic areas in China where non-keratinising nasopharyngeal carcinoma comprises more than 95% of cases, whether the findings can be extrapolated to non-endemic regions without further validation remains uncertain. Secondly, although the trial has achieved its primary endpoint of progression-free survival, the overall survival data have not yet fully matured. A meta-analysis of individual patient data suggested that progression-free survival was strongly correlated with overall survival in locoregional nasopharyngeal carcinoma.20 However, measurement of overall survival is potentially diluted by administration of crossover treatment in our study (29% of patients in the nab-TPC group and 30% of patients in the gemcitabine plus cisplatin group). Furthermore, with an increasing number of treatment options for recurrent or metastatic nasopharyngeal carcinoma and considering that regimens containing programmed cell death protein 1 inhibitor are effective in subsequent treatment, whether the same correlation exists is unclear.21 Further follow-up is needed to ascertain any potential benefit in overall survival. Thirdly, a thorough analysis of potential predictive biomarkers is essential to identify patients who may derive benefit from the nab-TPC regimen as a first line treatment.

Conclusion

In summary, our findings show, for the first time, that the nab-TPC regimen offers a statistically significant enhancement in progression-free survival compared with the current standard-of-care gemcitabine plus cisplatin regimen in recurrent or metastatic nasopharyngeal carcinoma as a first line treatment. The nab-TPC regimen has a manageable safety profile. These results advocate for the inclusion of the nab-TPC regimen as a standard treatment option for patients with recurrent or metastatic nasopharyngeal carcinoma and propose its suitability as a control arm in future randomised clinical trials.

Ethical approval

This study was approved the Chinese Ethics Committee of Registering Clinical Trials (ChiECRCT20190198) and was conducted in accordance with Good Clinical Practice and the Declaration of Helsinki. All participants provided written informed consent.