Since 1999, the standard treatment of locally advanced cervical cancer (LACC) has been pelvic radiation with concurrent cisplatin, with an absolute improvement of 12% in overall survival compared with radiotherapy alone [14]. It is important to investigate better treatment strategies considering that approximately 40% of patients experience recurrence within 5 years.
NACT before definitive radiotherapy has been generally perceived as not beneficial or even detrimental because of the greater toxicity of the chemotherapy regimen [15], higher recurrence rate, inferior PFS and lower survival rate [7, 16, 17]. Previous trials investigating the theory of accelerated repopulation suggested that in rapidly proliferating cancer, especially the protracted schedules, there is a large gap between completing chemotherapy and radiotherapy, suboptimal regimens are used, and tumor regrowth may be accelerated after chemotherapy, thereby limiting the effectiveness of NACT [18, 19]. However, data seemed to indicate the importance of dose-intensity and cycle lengths [5]. Cervical tumors are rapidly proliferating with a median doubling time of only 4-4.5 days and a high growth fraction [20, 21]. After a few cell divisions, the tumor volume may be restored, but the tumor cells may be less sensitive to chemotherapy and potentially to conventional radiotherapy due to the changed growth kinetics. With these assumptions, the schedule of chemotherapy may play an important role: a short and high dose-intense chemotherapy cycle seems optimal to minimize tumor repopulation with resistant cells.
Trials using short cycle chemotherapy with a chemotherapy cycle length of < 14 days and cisplatin dose intensities ≥ 25 mg/m2 per week appear to be associated with an improvement in survival compared with those using a more prolonged cycle interval [5]. The initial results from two phase II studies [8, 9] have been reported on patients who received NACT using weekly paclitaxel (60–80 mg/m2) and carboplatin (AUC = 2) for 6 weeks followed by CCRT. Following NACT, a response rate of 67.8–72.7% was achieved, mostly partial responses. Post-CCRT, the response rate was approximately 90%. Grade 3–4 hematologic toxicity was observed in approximately 20% of patients. A 3-year overall survival rate of 67% was observed in stage IB2-IVA patients. The observations are encouraging. The approach is now being evaluated in a randomized trial (weekly paclitaxel 80 mg/m2 and carboplatin AUC 2 for 6 weeks) (NCT01566240: INTERLACE trial) [22].
For chemotherapy, cisplatin is widely accepted in cervical cancer management both in neoadjuvant and salvage treatment, and it remains the most active single agent [5]. The triweekly administration of paclitaxel with cisplatin is considered the most effective regimen for metastatic cervical cancer [23]. In the GOG study published by Moore et al., the combination of cisplatin and paclitaxel was superior to cisplatin alone with respect to objective response rate, progression-free survival and sustained quality of life [24]. Among patients who had not received prior cisplatin, cisplatin-based chemotherapy might be significantly better than carboplatin-based therapy [25, 26]. Thus, we chose cisplatin with paclitaxel as our NACT regimen. Increased doses of cisplatin and paclitaxel are associated with survival improvement [27]. In our dose-finding phase, patients initially received NACT using weekly paclitaxel (75 mg/m2) and cisplatin (60 mg/m2) for 4 weeks followed by CCRT. Considering the severe hematological toxicity, we adjusted the paclitaxel dose to 60 mg/m2 and the cisplatin dose to 40 mg/m2. Based on these data, we formed our trial to evaluate the addition of NACT to standard CCRT.
The results from our trial confirmed that a short course of dose-dense weekly NACT with paclitaxel and cisplatin followed by radical CCRT is feasible with acceptable toxicity, as it did not compromise chemoradiotherapy, with 92% (46/50) of patients initiating and completing the radiation phase within the required time and 86% (43/50) of patients receiving at least 4 cycles of concomitant cisplatin. The observed response rate to this short course of chemotherapy, as assessed radiologically, was associated with an encouraging outcome. The response rates after NACT (79.2%) and after CCRT (90.0%) in our trial were similar to those reported in other studies[8, 9]. The three-year PFS and OS might be higher than those reported in other studies (studies using dose-dense NACT followed by the CCRT strategy are listed in Table 4).
Table 4
Characteristics of the studies (dose-dense NACT followed by CCRT)
Author
|
Year
|
Location
|
Sample size
|
Average age
|
FIGO stage
|
NACT regimen
|
Response rate to treatment
|
Follow-up
(months)
|
Survival
|
Post-NACT (%)
|
Post-CCRT (%)
|
3-year PFS
|
3-year OS
|
Our trial
|
2022
|
China
|
50
|
53 (35–68)
|
IIB-IVA
|
Weekly cisplatin (40 mg/m2) and paclitaxel (60 mg/m2) for 4 weeks
|
CR (10.4%)
CR + PR (79.2%)
|
CR (72.0%)
CR + PR (90.0%)
|
23.9
|
73.6%
|
83.9%
|
Singh RB [8]
|
2013
|
India
|
28
|
51 (35–67)
|
IIB-IVA
|
Weekly paclitaxel (60 mg/m2) and carboplatin (AUC = 2) for 6 weeks
|
CR (7.1%)
CR + PR (67.8%)
|
CR (85.7%)
CR + PR (92.8%)
|
12
|
-
|
-
|
McCormack M [9]
|
2013
|
UK
|
46
|
43 (23–71)
|
IB-IVA
|
Weekly paclitaxel (80 mg/m2) and carboplatin (AUC = 2) for 6 weeks
|
CR (4.5%)
CR + PR (72.7%)
|
CR (67.4%)
CR + PR (90.7%)
|
39.1
|
68%
|
67%
|
Lymph node involvement tended to be correlated with inferior outcome. In the present study, patients with stage IIB-IIIB had a higher probability of achieving CR (87.5%) than those with stage IIIC1 (64.3%) and IIIC2 (66.7%). Earlier (stage IIB-IIIB) FIGO stage was related to superior PFS (P = 0.0098) and OS (P = 0.0128). The mPFS and mOS of FIGO IIIC2 were merely 13 months and 23.5 months, respectively. This result suggested that other effective strategies should be investigated for a subset of patients with stage IIIC disease, especially those with IIIC2 disease.
The use of NACT before radiotherapy could potentially eradicate subclinical distant metastasis, reduce the tumor size and correct pelvic anatomy distortion, and ultimately allow better delivery of radiation. Although the long-term survival benefits compared with CCRT alone remain uncertain, our therapeutic strategy incorporating dose-dense TP chemotherapy and CCRT yielded favorable outcomes. After a median follow-up of 28 months, the 3-year OS rate was 83.9%, and the 3-year PFS rate was 73.6%.
The reasons for a possible detrimental effect of neoadjuvant treatment are unclear. In general, NACT was well tolerated, with only 4% (2/50) of patients experiencing any grade 3/4 late adverse events and no treatment-related deaths. A possible explanation for the detrimental results of a particular population might be the toxicity associated with NACT that compromised the ability to deliver concurrent CCRT.
The delay to initiate definitive chemoradiation because of neoadjuvant treatment might be detrimental. Most of the patients in our trial were treated subsequently by CCRT within 2 weeks after NACT. Two patients delayed the initiation of chemoradiation therapy for 3 and 4 weeks. Grade 3/4 hematological toxicity during NACT was 40%, which was similar to the report by Singh et al. (49.7%) but was considerably higher than that reported by McCormack et al. (11%). Most patients recovered soon after medication or blood transfusion, except one patient (stage IIIC1r) who had grade 4 neutropenia with high fever after NACT. She was assessed for SD after NACT and PD three months after CCRT (the local lesion was in complete remission but developed metastasis in the lung, intestinal membrane and inguinal lymph node). The majority of nonhematological toxicities were gastrointestinal toxicity, fatigue and alopecia (using cold caps may significantly reduce the occurrence of alopecia [28]), which did not delay subsequent CCRT. However, one had grade 4 hepatic insufficiency after 4 cycles of NACT and was evaluated as PR after NACT and PD three months after CCRT (local lesion was reduced but developed metastasis in the vagina and inguinal lymph node). The gap between NACT and CCRT was significantly influenced by the development of grade 3/4 neutropenia during NACT. The use of prophylactic G-CSF may help in maintaining dose intensity in future studies using this treatment protocol.
According to the recommendation [29], patients who took more than 8 weeks to complete radiotherapy might also have detrimental outcomes. There were more adverse events during CCRT (52% of patients had grade 3/4 hematological toxicity, and 32% of patients had grade 3/4 nonhematological toxicity). Thus, the incorporation of additional neoadjuvant chemotherapy into the standard treatment regimens is likely to result in increased toxicity. In our trial, one patient who discontinued brachytherapy for grade 4 neutropenia could not have improved and developed metastasis in the para-aortic and pelvic lymph nodes nine months after treatment. Another patient postponed brachytherapy for 2 weeks due to a large amount of vaginal bleeding with severe anemia and thrombocytopenia and developed metastasis in the lung and intestinal membrane three months after CCRT and died thereafter.
Moreover, in nonresponding patients, there might be inherent resistance to both chemotherapy and radiotherapy. Additionally, delayed access to CCRT might be detrimental. The combined analysis showed that a better clinical response and pathologic response to NACT were associated with favorable PFS and OS [30]. Stable disease post-NACT has also been identified by others as a poor prognostic sign [31]. In our trial, 10 patients (20.8%) had stable disease at the end of NACT, 4 of these patients had progressive disease 3 months post-CCRT, and 2 had progressive disease during follow-up, 4 of which subsequently died from their disease. NACT response was related to superior PFS (HR 0.14; P = 0.0003) and OS (HR 0.07; P = 0.0011) compared with NACT nonresponse.
Finally, tumor cells may have acquired resistance throughout NACT. Some studies have suggested that previous exposure to cisplatin could result in cross-resistant cellular clones and may lead to increased DNA repair and platinum-induced radioresistance [32–34]. Molecular studies of resistance pathways before and after neoadjuvant treatment may help to inform us about this finding.