Propensity score matching analysis of a phase II study on simultaneous modulated accelerated radiation therapy using helical tomotherapy for nasopharyngeal carcinomas

Currently, simultaneous modulated accelerated radiation therapy (SMART) is the most widely used intensity modulated radiation therapy (IMRT) pattern in the treatment of nasopharyngeal carcinomas (NPC) [1]. SMART can simultaneously delivery different doses to different targets and improve local control rate (LCR) through increasing fraction dose in the primary tumour and metastatic nodes and shortening the overall treatment time (OTT) to reduce post-process accelerated repopulation of tumour cells. Some studies have confirmed that SMART using Helical TomoTherapy (HT) system has significant dosimetric advantages in the treatment of NPC [2, 3]. More than 600 NPC patients have been treated with HT system at our centre. Based on previous 70Gy/33F pattern, we conducted in September 2011 a prospective phase II study, P67.5 study, with a prescription dose of 67.5Gy in 30 fractions to the primary tumour and positive lymph nodes [4]. Due to increased fraction dose and shortened OTT, the corrected biological effective dose (BED) to the primary tumour and positive lymph nodes increased from 62Gy to 62.9Gy, while that to late reaction tissues (LRTs) decreased from 99.7Gy to 97.9Gy (α/β = 5Gy), which could theoretically improve local control rate while reducing radiation injury. The study was approved by the research ethics board of the Chinese PLA General Hospital with an official number of S2014-048-01, and with a registration code of ChiCTRONC-14,004,895. To confirm the safety and feasibility of the P67.5 study, we retrospectively analyzed the data of our previous P70 study with a prescription dose of 70Gy in 33 fractions to the primary tumour and positive lymph nodes and used propensity score matching method (PSM) [5] to screen the cases and exclude the impact of confounding factors.

HT is a kind of advanced technology of radiation therapy and the treatment model of “rotation - step in - shoot” is on behalf of a type of highly efficient and high accurate IMRT [10]. Since our centre installed the first HT unit in china in September 2007, over 3000 cases had been treated by Match 2016. The P67.5 study was a non-randomized single-centre prospective study which aimed to evaluate the safety and feasibility of a new fractionation pattern, and the control group (P70 study) was a retrospective study with classical fractionation. In order to minimize the impact of confounding factors, we used PSM method and effectively corrected the hybrid bias in N stage and clinical stage. The final general characteristics of patients in both groups tended to be balanced.

The RTOG 0225 study [11] laid the fractionation of 70Gy/33F with SMART technology to become the standard IMRT pattern of NPC and the LCR reached 92.6% at 2-year. Our centre conducted P70 study with the same fractionation mode and achieved a 3-year LRRFS of 92.7%. Although this result was consistent with many other studies, we tried to optimize the fractionation pattern. In theory, the best radiation therapy plan should be under the premise of tolerance of OARs to achieve maximum destruction of tumour tissue. Because the regeneration of LRTs is slow and generally not affected by the total time of radiation therapy, the biological effects of radiation to early responding tissues (ERTs) are similar to that of tumour tissues, all ways to improve local control is bound to increase ERT damage. During radiation therapy, acute side-effects occur in oral cavity mucosa, pharyngeo-esophageal mucosa and other ERTs often become the main factors affecting the treatment compliance. The incidence of grade 2–4 oral mucositis was 29.4%, 36.8% and 4.4%, respectively in the RTOG 0225 study. However, with dosimetric advantages and image guided radiation therapy (IGRT) realized with megavoltage computed tomography (MVCT) equipped on the gantry, radiation-induced acute injuries in ERT is decreased with HT technique. The incidence of grade 2–3 mucositis and esophageo-tracheitis in P70 group was only 56.8%, 3.2% and 52.1%, 0.5%, respectively, without grade 4 side-effects. If the BED remains the same, increased fractional dose and shortened OTT end to a decreased prescription dose, which would result in the following advantages: 1) Improve LCR; Many studies have shown tumour cells appeared accelerated repopulation during the late period of radiation therapy and the total dose should compensate 0.6Gy for every extra day of the OTT (equal to γ/α value) [12‐14], so appropriate shorten the OTT could improve LCR. 2) Reduce dose to OARs; In P67.5 group, maximum doses of brainstem, spinal cord, eyeball, lens, optic nerve and the mean dose of temporomandibular joint, oral cavity, pharyngeo-esophageo-trachea were significantly lower than in P70 group. 3) Reduce costs; The treatment cost reduced by about 3.9%, and the costs of accommodation, food and transportation were correspondingly reduced too. 4) Improve equipment utilization; Physical depreciation of machinery reduces about 9.1% and the saved medical resources can be used to treat additional 8 patients a year. In P67.5 group, the incidence of acute toxicities such as oral mucositis and esophageo-tracheitis was 8.8% and 2.7%, respectively, without significant difference compared to that in P70 study, even with more patients receiving CCRT. All of the above results confirm that the fractionation pattern of 67.5Gy/30 was safe and feasible.

Improving the survival rate was still one of the intentions of the P67.5 study. Compared with P70 study, the absolute value of the 3-year LRRFS, DMFS, DFS and OS in P67.5 study was improved by 1.3%, 2.1%, 0.7% and 3.5%, respectively. Although statistical significance was not achieved, we observed a trend of improvement in the 3-year OS, which was confirmed by multivariate analysis. Univariate analysis of all cases showed that T stage was the only factor affecting the LRRFS and increasing the fractional dose did not improve the LCR, but it was known that the good overall outcome of NPC and the use of SMART technology could both result in a good LCR [11, 15‐18], so a 3-year LRRFS of 94% in P67.5 study was acceptable. T stage not only affected LRRFS, but together with UICC stage also affected DFS, which showed that the progression of the disease was closely related to the severity of the primary tumour and the clinical stage.

Despite the LCR has been guaranteed by the wide application of IMRT in NPC, distant metastasis was still the first reason of treatment failure. In recent years, a large number of clinical evidence suggested that CCRT could improve the survival rate of patients with locally advanced NPC and the 5-year DMFS attained up to 74.7% – 85.8% [19‐21], at the same time anti-EGFR Mab treatment also made clinical benefit in NPC patients [7, 22]. In both groups, CCRT was the standard treatment for locally advanced NPC patients, while anti-EGFR Mab treatment was also performed and the 3-years DMFS was 92.5%, the same with the literatures, but 21 cases developed distant metastases, almost double number of the cases with loco-regional failure. Whether ICT could improve the survival of patients with locally advanced NPC was still controversial, some studies have shown its benefits. The phase II study conducted by Ferrari et al. [23] confirmed that patients with locally advanced NPC received induction regimen of cisplatin and fluorouracil (PF) followed by cisplatin-based CCRT, had improved LCR and OS. Hui et al. [24] added ICT with DP regimen (docetaxel 75 mg/m² + cisplatin 75 mg/m²) and showed a significant improvement of 3-year OS and a trend of improvement of 3-year PFS and DMFS compared with CCRT alone regimen (cisplatin 40 mg/m2 per week). The phase III study conducted by Sun et al. [25] conformed that addition of TPF induction chemotherapy (docetaxel 60 mg/m², cisplatin 60 mg/m² intravenously every 3 weeks and fluorouracil 600 mg/m² per day as a continuous 120 h infusion) to CCRT significantly improved the 3-year failure-free survival compared with CCRT alone (80% vs. 72%, p = 0.034) in locoregionally advanced nasopharyngeal carcinoma with acceptable toxicities. Based on the results of the above studies, we were more inclined to use ICT + CCRT regimen hoping to improve the survival and the use rate of ICT + CCRT regimen in P67.5 group was as high as 90.5%, while that in P70 group was only 13.0%. However, there was no statistical significance in 3-year LRRFS, DMFS, DFS and OS between patients with ICT + CCRT regimen and CCRT alone, and the same result was obtained by other recent prospective randomized studies [26‐28]. In Xu’s study [29], it was found that ICT only improved the DMFS and OS in patients with N3 disease, so what kind of patients with local advanced NPC could benefit from ICT might need more studies. In addition, in our study, age was another factor affecting survival rate and the 3-year DMFS, DFS and OS in patients aged ≥50 years were significantly lower than that in patients aged <50 years, which was also shown in Qiu’s study [30].

In failure patients of NPC, active salvage therapy might achieve prolonged survival, or even radical cure. Zhou et al. [31] reirradiated 53 locally recurrent patients with IMRT (67.9Gy) combined with cisplatin-based chemotherapy and the 2-year OS and progression-free survival (PFS) were 58.7% and 52.3%, respectively. Goto et al. [32] reirradiated 50 locally relapsed patients using HT plus concurrent chemotherapy and got similar results. It has been recognized that platinum-based chemotherapy as the first-line treatment achieved an objective response (OR) up to 50-90% in metastatic NPC [33], and could obtain an OR of 22-75% even as a second-line treatment [34]. Zheng et al. [35] retrospectively analyzed three kinds of treatment in patients with metastatic NPC and found that salvage chemotherapy plus palliative radiation therapy or other localized treatment resulted in better survival than chemotherapy alone or supportive treatment, and the 2-year DMFS reached to 57.7%, while that in the other two groups was only 32.7% and 1.6%, respectively. Currently there was no standard treatment for relapsed NPC. Zheng et al. [35] suggested that active salvage therapy should be necessary, and systemic treatment should be combined with local treatment, and local treatment should not be limited to the nasopharynx but extended to the appropriate metastatic lesions. In this study, six regional relapse patients, all received salvage therapy, had the best prognosis with a survival rate as high as 67%. The prognosis of local recurrence was worse, 5 (45%) of 11 these patients received salvage therapy and 7 cases (64%) died. The worst prognosis was happened in distant metastatic patients, 11 cases (48%) receiving salvage therapy, 18 (86%) died. The incidence and mortality of the above three failure patterns were comparable in both groups. It was noted that there were 5 patients without loco-regional recurrence or distant metastasis died of hemorrhage in this study, which was rarely reported in the literatures. In the study of Lin et al. [36], among the 370 patients of NPC, only one died of local hemorrhage. Nasopharyngeal hemorrhage is one of the common complications after radiation therapy, which is relatively easy to control, and uncontrollable hemorrhage is often associated with local recurrence. At the beginning of the P67.5 study, we realized the importance of nasal care and regular review after radiation therapy and only one patient died of hemorrhagic till now. The difference in this failure pattern between the two groups, led to a significant difference (p = 0.037) in the 3-year OS analyzed in multivariate analysis.

Through increasing the fractional dose and shorten the treatment time, the P67.5 study achieved excellent short-term outcomes and potential clinical benefits, with acceptable acute and late toxicities. The long-term outcomes are under investigation.