Between May 2011 and September 2023, 46 patients were scheduled to receive palliative pelvic RT. Six patients were excluded from the analysis due to a switch to a curative concept (n = 3), refusal of RT (n = 2), and death before start of treatment (n = 1). The outcomes for a total of 40 patients with a median age of 69 years and a median follow-up time of 9.7 months (range 0.1–73.1) were analyzed.

The most common primary tumor was endometrial cancer (n = 16; 40.0%), followed by cervical cancer (n = 10; 25.0%), vulvar cancer (n = 8; 22.2%), and ovarian cancer (n = 6; 15.0%).

Previous therapy included oncological surgery in 25 patients with a median time of 20 months (range 2–134 months) prior to RT, including a hysterectomy in 16 women. No patient received simultaneous systemic therapy during RT.

Prior RT was documented in 8 patients with a median time interval of 17 months (range 4–223 months) and a median RT dose of 54 Gy (range 37–104 Gy; EQD2 α/β = 3; n = 6 EBRT, n = 2 vaginal brachytherapy, and n = 1 EBRT and endocervical brachytherapy).

The median Karnofsky performance score before the start of RT was 70% (range 50–90%). The median time from the first consultation at our department to the start of RT was 8 days (range 0–177). To a median pelvic PTV of 804 cm3 (range 82–3814 cm3), a median total dose of 39 Gy (range 20–45 Gy) in a single dose of 3 Gy (range 3–4 Gy) in 13 fractions (range 5–15) was applied. For 5 patients with vulvar cancer the PTV additionally included the inguinal lymph nodes. Figure 1 shows an example of a palliative hemostatic pelvic IMRT treatment.

An 83-year-old woman with palliative hemostatic RT for endometrial cancer receiving 6‑MV photon intensity-modulated radiotherapy with 39 Gy in 13 fractions for a planning target volume of 1634 cm3 resulting in cessation of bleeding after 12 days (equivalent dose in 2‑Gy fractions for α/β of 10 = 24 Gy) (a: axial, b: coronar and c: sagittal computed tomography slices with the resulting radiotherapy plan)

Radiotherapy treatment included IMRT for 28 (70.0%) and 3D-CRT techniques for 12 patients (30.0%). Premature termination of RT was present in 4 (10.0%) patients in favor of best supportive care. Detailed patient and treatment characteristics with dose fractionation regimes of all patients are presented in Tables 1 and 2.

Inpatient treatment during RT was necessary in 77.5% of the women. Fifteen (37.5%) patients received a transfusion of red blood cells (range 2–12) during RT, and 10 (25.0%) patients required additional application of vaginal tamponade with local hemostatic agents. Complete cessation of bleeding was achieved in 80.0% (n = 32) of all patients after a median time of 16 days (range 1–78 days) and a median EQD2 (α/β = 10) dose of 39 Gy (range 3–49 Gy). Patients treated with a higher total RT dose had cessation of bleeding significantly more often (p < 0.0001), with a cut-off value of at least EQD2 (α/β = 10) = 36 Gy.

Before the start of RT, 9 patients (22.5%) suffered from a symptomatic vein thrombosis or acute pulmonary embolism requiring therapeutic anticoagulation, and 5 patients received prophylactic anticoagulation. Successful cessation of bleeding was significantly less frequent in patients receiving anticoagulation concurrently with radiation (p < 0.0001) and in patients with infield re-irradiation concepts (p = 0.018). The applied RT technique (p = 0.168), PTV volume (p = 0.973), tumor entity (p = 0.252), and additional application of vaginal tamponade with local hemostatic agents (p = 0.361) had no significant influence on the occurrence of bleeding cessation.

Twenty-three (57.5%) patients suffered from pain (grade I: n = 12, grade II: n = 5, grade III: n = 6) before the start of RT. At the first follow-up after palliative RT, relief from pain was documented in 14 (60.9%), one woman presented with a deterioration of pain symptoms (n = 1, 4.3%), and the remaining patients remained stable.

Acute RT-induced toxicity included only low-grade gastrointestinal symptoms in 42.5% of patients (grade I n = 12, grade II n = 5) and genitourinary problems in 30.0% (grade I n = 11, grade II n = 6). No higher-grade RT-induced toxicity occurred. No significant correlation between the applied dose and toxicity could be found. Of note, 3 patients reported gastrointestinal symptoms (grade III n = 3, grade IV n = 1) prior to RT, with a deterioration of pre-existing symptoms with abdominal infection exacerbation in five women (grade III n = 3, grade IV n = 2) during RT.

At the end of the observation period, 17 (42.5%) patients were still alive. The 6‑month, 1‑year, and 2‑year OS rates were 66.9%, 60.8%, and 30.0%, respectively. Local failures were detected in 12 women (30.0%) after a median time to relapse of 3.9 months (range 1.4–14.5), resulting in 6‑month, 1‑year, and 2‑year LC rates of 66.9%, 60.8%, and 57.7%, respectively. The LC was significantly superior in patients who received a total applied EQD2 dose of at least 36 Gy (p = 0.011).

Chest and abdominal imaging were performed for distant staging purposes for a subgroup of 30 patients, resulting in 6‑month, 1‑year, and 2‑year DC rates of 47.6%, 37.0%, and 26.4%, respectively. The most common progressive metastases were in the extrapelvic lymph nodes (n = 8), liver (n = 7), peritoneum (n = 6), lung (n = 2), bone (n = 2), soft tissue (n = 2), and spleen (n = 1).