Patients aged ≥75 years and above diagnosed with previously untreated cervix cancer who presented to PMCC between January 1998 and December 2010 for radiotherapy as part of their treatment (either definitive or adjuvant postoperative radiotherapy) were eligible for the study. Patients with known metastatic disease upon initial presentation, or referred for disease recurrence were excluded. All patients were clinically staged using the International Federation of Gynecology and Obstetrics (FIGO) staging criteria. Patients’ baseline Eastern Cooperative Oncology Group (ECOG) performance status and comorbidities were also recorded and utilised to generate a modified Charlson comorbidity index (CCI) [10]. Patients being considered for curative treatment underwent magnetic resonance imaging (MRI) to assess the local extent of disease and position emission tomography-computed tomography (PET-CT) or PET for nodal staging and to exclude distant metastasis. Of 126 patients that met eligibility criteria, MRI was performed in 97 patients and PET was done in 85 patients. Generally, MRI and PET information was not used to change clinical FIGO staging, except where MRI and PET findings would put patients in stage 4b category. Within the 4b category, patients were eligible for this analysis provided all known disease was able to be encompassed within the pelvic field, including metastasis or direct invasion into pelvic bone only or disease involving inguinal nodes. Baseline characteristics of patients are described in Table 1.

All patients received external beam radiotherapy (EBRT) as adjuvant (n=10), definitive (n=81), or palliative hypofractionated regimen (n=35). Definitive and adjuvant radiotherapy technique have been previously described [41112]. Briefly, EBRT consisted of a four-field technique with 18 MV photons. Adjuvant radiotherapy consisted of 45 to 54 Gy of EBRT. Definitive radiotherapy consisted of 40 to 45 Gy of EBRT followed by BT. BT was given with either tandem and ovoids (T&O) or tandem and cylinder. EBRT was given with concurrent weekly cisplatin (40 mg/m²) or carboplatin (area under curve (AUC) 2) in the majority of patients, unless contraindicated, as shown in Tables 2, 3. Lymph node (LN) negative patients and patients with metastatic nodes confined to the pelvis were treated with pelvic fields in prone position on a belly board. The upper border of pelvic field was at L5–S1. The patients with common iliac or para-aortic (PA) nodes were treated supine with 4-field extended field radiotherapy (EFRT). The upper border of the PA field was kept 4 cm above the upper most involved node. Involved nodes received a rectangular nodal boosts dose of 6 to 10 Gy in 2 Gy fractions.

Of the 35 patients treated with palliative intent, radiotherapy consisted of a series of 1–3 the so-called “quad-shots” (QSs) [13]. Each QS consisted of 14 Gy in four fractions separated by 4 weeks interval without the use of concurrent chemotherapy. Where suitable, patients also received BT following EBRT. Radiation effectiveness of single QS (given in four fractions) in 2 Gy fraction equivalent radiation dose (EQD2) was 15.8 Gy for acute effects (tumour tissue) and 18.2 Gy for late effects [14]. Those receiving 3 QS (EQD2=47.3 Gy acute and 54.6 Gy late effect) to the pelvis and with the addition of two BT received EQD2 of approximately 78 Gy to the primary tumour volume [14].

Thirteen patients received three QS followed by BT, whilst one patient following pelvic surgery for a rectal cancer was treated with BT alone without the EBRT. Patients with disease limited to pelvis were treated with three QS and BT, provided their ECOG performance status was adequate. The remaining patients who were deemed unsuitable either by treating physician due to performance status or patient refusal of more radical therapy, received 1–2 QS EBRT followed by 1–2 BT or just QS EBRT (1–3) only.

All patients were reviewed 4 to 6 weeks postcompletion of radiotherapy and thereafter every 3 to 4 months for the first 3 years, every 6 months for the fourth and fifth years and annually until 10 years. A posttreatment PET or PET-CT was performed 6 months after completion of curative RT to assess response to treatment. Follow-up information about disease status and late complications were prospectively collected. Late treatment-related toxicities were scored according to modified World Health Organization/Radiation Therapy Oncology Group criteria. Once the disease recurrence was confirmed, patients were treated with radiotherapy, systemic chemotherapy or best supportive care as appropriate. Date and cause of death, if known, were documented.

Patient follow-up data was censored at 1st Aug 2012, as it was the earliest of the last follow-up date among those who were alive and not lost to follow-up.

Failure was defined as either persistent disease following radiotherapy or recurrence of disease. The date of failure was taken as the date of sign of any failure. Sites of failure were recorded as: (1) local: residual or recurrent disease at cervix or uterus; (2) pelvic: all failures below L5–S1 including nodal, parametrial, and vaginal; (3) abdominal: PA nodes above L5–S1; and/or (4) distant: including supraclavicular, and inguinal and systemic spread. For first failure at multiple sites, the dominant site of failure was determined according to the hierarchy, in descending order: distant, abdominal, pelvic, and local.

Patient characteristics were described using descriptive statistics. OS was defined as the duration of time between the date of diagnosis and the date of death from any cause; RFS was defined as the duration of time between date of diagnosis and date of first failure. Kaplan-Meier survival curves were calculated for OS and RFS. Times to events were censored by the closeout date. Prognostic factors for OS and RFS were evaluated using Cox proportional hazards regression models. Moreover, subgroup analyses were performed stratified by treatment intent. Treatment toxicities were expressed as proportions.

There were 194 eligible patients based on FIGO staging criteria. Fifty-two patients had disease beyond the PA nodes and 16 patients had small cell and clear cell histology. These were excluded. The analysis was restricted to the remaining 126 patients, of these, 113 were squamous cell carcinoma (SCC) and 13 had adenocarcinoma histology.

The median follow-up time was 37 months (interquartile range, 10.5 to 71.1 months). All observations were censored on the 1st August 2012. Three patients were lost to follow-up at 6 and 7 years and one patient at 125 days from the date of their histological diagnosis.

The mean age was 81.5 years (SD, 4.6 years) and ranged from 75 to 98. The FIGO stage 1, 2, and 3&4 (as a single group) was estimated in 26%, 41%, and 28% respectively. Ninety-one patients were treated with curative intent and 35 patients received palliative radiotherapy. There were statistically significant (p<0.05) differences in most of the characteristics between the patients treated with curative intent and those treated with palliative intent.

At the closeout date, 20 of the 21 patients were alive and had no evidence of disease and 102 patients had died. Of those who died, 43% died from other causes without any relapse of cervix cancer, 44% died of disease and three patients were lost to follow-up. Only one of 35 patients treated with palliative intent was alive at 5 years without any evidence of disease.

OS of the entire cohort of patients at 3 years was 52.7% (95% CI, 43 to 61) and at 5 years was 41.2% (95% CI, 32 to 49). RFS at 3 years was 62.88% (95% CI, 53 to 70) and at 5 years was 59.51% (95% CI, 49 to 67).

Ninety-one patients were treated with curative intent. OS in this group at 3 years was 66.63% (95% CI, 55 to 75) and at 5 years was 54.48% (95% CI, 43 to 64). RFS at 3 years was 75.86% (95% CI, 65 to 83) and at 5 years was 71.57% (95% CI, 60 to 80) (Fig. 1). RFS at 3 years according to various prognostic factors are described in Table 4.

Thirteen of 35 patients who received hypofractionated (palliative) radiotherapy remained disease free in the pelvis until last follow-up or death. Five patients survived for 7.5, 7, 6, 5, and 4 years respectively without any evidence of disease. All five patients were node-negative and were staged as either FIGO stage 1b or 2a. The remaining eight died of unrelated causes whilst one died of distant metastasis at 3 years without any signs of locoregional disease (Table 5, cause of death).

Nearly all patients treated with CCRT with curative intent completed treatment. Thirty-five percent of patients treated with CCRT experienced a drop in their hemoglobin below 120 g/L. Two patients required a treatment break due to acute toxicity. One patient had intractable diarrhoea but subsequently completed treatment after a short break. The other patient developed neutropenic sepsis and subsequently died of uncontrolled sepsis. Late toxicity details are shown in Table 6. Half the number of patients remained completely asymptomatic. None of the patients had grade 4 toxicity.

Twenty-four (26%) of 91 patients experienced relapse of their disease. Failure rates for primary, pelvis, and PA and distant sites were 7%, 7%, 16%, and 16% respectively. Only one patient with FIGO stage 3b failed exclusively at the primary site. All other failures were multisite. The failure rate in patients with positive nodes was 44%. The failure rate in node negative patient was 24%.