This tree-structured survival model implied that patients should be stratified first by the number of metastatic lymph nodes and second by UGT1A1 genotype to help to determine the risk of recurrence. We believe it may be valid to administer CPT-11/NDP chemotherapy for patients with one or no lymph node metastases and, secondarily, UGT1A1 polymorphism. A conditional inference tree is an effective way to determine and rank prognostic factors [18, 19].
We found that cervical cancer patients with one or no metastatic lymph nodes are less likely to experience recurrence after CPT-11/NDP therapy. It has been reported that the number of metastatic pelvic lymph nodes (≤3 vs. >3) is a significant prognostic factor in patients treated with radical surgery followed by postoperative CCRT. Further, no significant survival difference is observed between patients without metastasis and those with 1–3 metastatic lymph nodes [20]. Park and Bae reported that the 5-year OS rates for patients with stage IB-IIA cervical cancer and 0, 1, and ≥2 positive metastatic lymph nodes were 91, 80, and 47%, respectively (P = 0.006) [21]. Inoue and Morita reported that the 5-year OS rates for patients with stage IB-IIB cervical cancer and 0, 1, 2-3, and ≥4 positive metastatic lymph nodes were 89, 81, 41, and 23%, respectively [22]. Sakuragi et al. reported that the cumulative 5-year OS rates for patients with 1 and ≥2 positive metastatic lymph nodes were 84.9 and 26.5%, respectively, with no significant difference between cumulative OS rates of patients with 0 and those with 1 positive node [23]. Therefore, ≥2 positive metastatic lymph nodes might be an important prognostic factor, rather than just lymph node positivity.
Chemotherapy and surgery may be useful for patients with one or no lymph node metastasis. Nevertheless, we need to consider CCRT as adjuvant therapy, rather than chemotherapy alone, for patients with two or more lymph node metastases. We consider that stratification of treatment based on the number of the lymph node metastases is preferable.
In patients who have radiation history, chemotherapy is the only course of treatment recommended when local recurrence is found in the vicinity of the pelvic cavity. We believe that secondary surgery or radiation therapy should be administered for local recurrence if patients have no history of radiation [11]. There are some arguments in support of chemotherapy or CCRT as initial adjuvant treatment after radical hysterectomy; however, the research is inconclusive. We also believe that consolidation chemotherapy might lead to a better prognosis for patients with locally advanced cervical cancer if they were initially treated with CCRT [24].
This study implied that the UGT1A1 polymorphism might also stratify patients and act as a predictive prognostic factor for the efficacy of CPT-11/NDP for cervical cancer patients. The UGT1A1 genotype has previously been implicated as a prognostic marker for CPT-11 therapy in colorectal cancer [25]. Some controversial studies have suggested a limited survival benefit in patients who were UGT1A1-poor metabolizers due to UGT1A1 polymorphisms [26, 27], although this association has been inconsistently reported [28].
In our study, 43% of patients treated by chemotherapy experienced grade 3 or higher neutropenia, and 13.7% of patients experienced diarrhea and vomiting. Neutropenia and diarrhea are common adverse effects of CPT-11. The UGT1A1 genotype is known to be a useful predictor for adverse effects [29]. In our study, we categorized patients into a wild-type group and a polymorphism group (*1/*6, *1/*28, *6/*6, *28/*28, and *6/*28), including a few patients with homozygotic or compound heterozygotic polymorphism (5.9% and 2.0%, respectively). Patients with UGT1A1 polymorphisms tended to experience grade 3 or 4 neutropenia more frequently than those with wild-type UGT1A1 (p = 0.09, no significance). This finding is relatively consistent with reports that patients with UGT1A1 homozygotic (*6/*6 or *28/*28) and compound heterozygotic (*6/*28) polymorphisms tend to experience adverse effects of CPT-11 [16, 30]. UGT1A1*28 and UGT1A1*6 have been well studied in UGT1A1 polymorphisms as regards CPT-11 pharmacokinetics and pharmacodynamics. Particularly for Caucasians, UGT1A1*28 seems to be a good predictor of neutropenia (all CPT-11 doses) and diarrhea (dose [125 mg/m2]), and UGT1A1*28 is also significantly associated with an increased risk of diarrhea in Asian patients at a CPT-11 dose of 125 mg/m2. However, in Asian populations, the UGT1A1*6 variant is more common and appears to be a more accurate predictor of neutropenia (all irinotecan doses) and diarrhea [31].
Our retrospective analysis revealed there was a significant difference in PFS between the UGT1A1 wild-type and polymorphism groups when we analyzed only patients with one or no lymph node metastases. Although we recommend CPT-11/NDP to patients with one or no lymph node metastases and UGT1A1 polymorphism, our data do not support recommending this regimen to other patients. Nevertheless, we did not compare the efficacy and adverse effects of CPT-11/NDP to those of CCRT or other regimens, including paclitaxel/carboplatin or paclitaxel/cisplatin. Therefore, we should conduct a prospective study to test the more favorable prognostic effect of the CPT-11/NDP regimen for the UGT1A1 polymorphism group compared to the wild-type group in cervical cancer patients with one or no lymph node metastases after radical hysterectomy.