Patient demographics and other treatment parameters

Toxicity

In general, chemotherapy with CCRT was well tolerated. The acute systemic toxicities in both arms are detailed in [Table 2], scored according to the RTOG Acute Radiation Morbidity Scoring Criteria. None of the patients died during treatment or within 1 month thereof. The most common Grade 2 or higher toxicity was gastrointestinal (30% in Arm A and 7% in Arm B, P = 0.052). The most common hematological toxicity of Grade 2 or higher was anemia seen in 6% in Arm A and 5% in Arm B (P = 0.321). There was no incidence of nephrotoxicity or neuropathy of Grade 2 or higher.

Comparison of patients in the two arms according to the grade of adverse reaction during treatment

Owing to the short duration of follow-up available in this study, late toxicities could not be assessed.

Treatment response at 6 months

Arm A had two (10%) patients with a local residual disease with no distant metastases. Two patients with residual disease expired 4 months after the last cycle of chemotherapy, and 17 patients were lost to follow-up. In Arm B, 21 (57%) patients had a PR. One patient developed bone metastasis and one suffered liver metastasis along with the residual disease. Two patients expired 3 months after ICRT, and 12 patients were lost to follow-up. There was a statistical difference between the two arms regarding failure (P = 0.004) [Table 4]. There were no recurrences seen in the limited follow-up duration in both arms.

Pattern of failure

Discussion

We performed a prospective study to compare neoadjuvant and adjuvant chemotherapies with concurrent chemoradiation versus concurrent chemoradiation alone in the treatment of locally advanced carcinoma cervix regarding the treatment response and toxicities. After taking into consideration the inclusion and exclusion criteria, 116 patients were taken into the study. The patients were randomized into two arms, study (n = 54, Arm A) and control (n = 62, Arm B). Of 116 patients, 83 received full treatment as planned.

Age of the patients in our study ranged from 30 to 68 years with a mean age of presentation being 50 years in Arm A and 49 years in Arm B. Majority of the patients in both arms was between 40 and 49 years. In this study, there was no impact of the age on outcome between the groups, as in the studies by Wong et al. and Vrdoljak et al. [2],[3] Statistically, no significant difference was observed in the distribution of patients between the two arms according to KPS scores (P = 0.05). Thus, there was no impact of KPS on outcome between the two groups in our study. Choi et al. reported that the performance status of a majority of the patients is relatively poor in developing countries including India.[4]

We included only histologically proven squamous cell carcinoma cases in our study, as the most common histology variant in cervical cancer is squamous cell carcinoma (about 90%). Regarding tumor differentiation, the most common type observed was well differentiated in Arm A (56%) and moderately differentiated in Arm B (60%). In this study, there was no impact of tumor differentiation on outcome in both arms. Choi et al., Wong et al., and Vrdoljak et al. have shown no significant correlation between survival and tumor behavior with the degree of differentiation of squamous cell carcinoma or adenocarcinoma of the cervix. Although Reagan and Fu revealed the prognostic value of histologic differentiation in patients treated with irradiation, Crissman et al. failed to observe a correlation between histologic parameters and patient survival.[5],[6]

A higher number of patients (21 of 48) had PR in Arm B as compared to Arm A of this study, perhaps explained by the greater number having Stage IIIB (53%); however, there was no significant difference between the distribution of patients among various stages between the two arms (P = 0.064). This is in agreement with Choi et al. Most of the studies show a decrease in pelvic control rate and survival rate with increase in the stage.[7],[8] The patterns of care study stated a 5-year survival rate of 74% for Stage I, 56% for Stage II, and 33% for Stage III patients of cervical cancer after radiotherapy.[9] Another report of 1988 found 89% of patient survival in Stage IB, 85% in IIA, 76% in IIB, 62% in IIIA, 50% in IIIB, and 20% in Stage IV.[10]

In the course of treatment in our study, 19 patients defaulted in Arm A. During EBRT, three patients defaulted due to adverse reactions, and two patients defaulted in the initial 10 days due to reasons not known. Two patients left the treatment in between EBRT and ICRT, and three patients absconded after two fractions of ICRT. Nine patients defaulted after the first cycle of adjuvant chemotherapy due to adverse gastrointestinal reactions. In Arm B, 14 patients defaulted. During EBRT, five patients defaulted due to adverse reactions and three patients defaulted in the initial 12 days due to personal reasons. Three patients left treatment in between EBRT and ICRT, and three patients defaulted after taking one fraction ICRT for reasons not known. The number of defaulters in Arm A was more as compared to Arm B, possibly owing to the increased duration of assigned treatment.

Comparison of disease status after completion of EBRT in this study revealed no residual disease in 82% of patients in Arm A as compared to 48% in Arm B. It was statistically significant (P = 0.01). The actual difference in treatment response was evident after completion of treatment. After completion of radiotherapy, response assessment was done 1 month later; and then, the patients were assessed monthly for 3 months followed by 3-monthly assessments. In the chemotherapy arm, adjuvant chemotherapy was started 3 weeks after ICRT completion, and response checked a month after the last chemotherapy cycle. The patients with no response or progressive disease on regular follow-up were subjected to salvage chemotherapy as found appropriate by the treating radiation oncologists. The response assessment could be done for 35 patients of Arm A and 48 patients of Arm B. The majority of the patients in Arm A (n = 33, 94%) achieved CR, PR in two patients (6%), with no patient having no response or progressive disease at 6 months after completion of treatment. In Arm B, 27 patients (56%) achieved CR, PR in 21 patients (44%), with no patient having no response or progressive disease. The CR was higher in study arm (Arm A) as compared to control arm (Arm B), and the difference was statistically significant (P = 0.01).

Wong et al. compared the treatment effects of epirubicin-based CCRT plus adjuvant epirubicin for five cycles (110 patients) versus radiotherapy alone (110 patients). Their results showed that CCRT and adjuvant chemotherapy produced better local control and higher survival. In Vrdoljak et al., the CR rate for adjuvant chemotherapy group was 100%. The duration of treatment did not influence local control. The main reason why this study differs from others may be a small number of patients and very short follow-up period.

We analyzed treatment toxicities regarding hematological, gastrointestinal, genitourinary, and skin reactions. The most common adverse reaction was gastrointestinal and genitourinary in both arms. Frequencies of Grade 2 and Grade 3 gastrointestinal and genitourinary reactions were high in the study arm (Arm A). The frequency of hematological toxicity was equal in both groups. However, these were statistically not significant (P = 0.052). Most of the gastrointestinal reactions were vomiting and diarrhea. In Wong et al., median follow-up for the whole group at the time of the analysis was 96 months. Leukopenia and thrombocytopenia were significantly more common in patients who received CCRT plus adjuvant chemotherapy, whereas the frequency of other toxicities had no difference in the two groups. There was no occurrence of Grade 3 or higher vomiting, diarrhea, proctitis, and/or hematuria in our study. In the study by Vrdoljak et al. on CCRT plus consolidation chemotherapy, two cases of rectovaginal fistula, one of the vesicovaginal fistulas, and two of ureteral obstruction were observed. The major shortcoming of the present study is lack of adequate follow-up to see the late toxicity and response.

The mean duration of follow-up for Arm A was 9 months whereas it was 7 months for Arm B. The longer follow-up duration of Arm A could not be attributed to any specific reason. The difference was statistically significant. In the study arm, no patient developed metastasis, but two patients expired. One died of lower respiratory tract infection; however, the second mortality's cause could not be ascertained. In the control arm, two patients developed metastasis and two expired. It was statistically significant. We need further longer follow-up period to assess the disease-free survival and overall survival. In Wong et al., as the median follow-up duration was 96 months, the response was assessed regarding local control, 3-year actuarial disease-free survival, and pattern of failure. The results of our study show that the local control and complication rates are comparable with those reported by Wong et al., Vrdoljak et al., and Choi et al.