Adjuvant intensity modulated whole-abdominal radiation therapy for high-risk patients with ovarian cancer FIGO stage III: final results of a prospective phase 2 study

The OVAR-IMRT-02 study is a multi-center single-arm phase-II-trial (investigator-initiated trial). Between 2010 and 2015, 20 patients with FIGO stage III ovarian cancer were treated with WART using IMRT with a dose of 30 Gy (daily fractions of 1.5 Gy 5 times per week) within a consolidation concept. Patient characteristics have already been described previously [39] and are shown in Table 1. All patients had maximal cytoreductive surgery (including at least total abdominal hysterectomy, bilateral adnexectomy, omentectomy, debulking of tumor masses) with postoperative residual tumor of < 1 cm followed by chemotherapy. Consolidative WART should not start later than 10 weeks after the last cycle of chemotherapy. Performance of IMRT using helical tomotherapy has been described previously [37, 38]. In short, WART was applied as helical IMRT using a tomotherapy device (TomoTherapy, Madison, WI) with 6-MeV photons. Control of positioning accuracy was performed daily with integrated megavoltage computed tomography for tomotherapy (3.5 MV). The clinical target volume included the whole peritoneal cavity extending from diaphragm to Douglas cavity, the liver surface, and the pelvic and para-aortic node regions. The planning target volume (PTV) encompassed an axial margin of 1.5 cm around the clinical target volume (2.5 cm in the cranial direction). Organs at risk (OARs) were kidneys, liver (except the 1-cm outer border), lungs, bones (vertebral bodies and pelvic bones), heart, and spinal cord. A dose of 30.0 Gy was prescribed to the median of the PTV. Inverse radiation dose planning was performed according to general recommendations (International Commission on Radiation Units and Measurements Report 50, 1999). There were no strict dose constraints for OARs used for optimization. The goal of optimization was to deliver a dose distribution in the PTV as homogeneous as possible, in addition to maximal sparing of OARs with a high priority on liver, kidney, and bone marrow. Tolerated maximum doses to OARs were not to exceed the tolerance dose 5/5 for each organ [25, 26]. Dosimetric information about the dose distribution in the target volume (PTV) and the OARs have already been reported previously [39].

Inclusion criteria were: histologically confirmed ovarian cancer stage FIGO III, maximal cytoreductive surgery (including at least total abdominal hysterectomy, bilateral adnexectomy, omentectomy, debulking of tumor masses) with postoperative residual tumor of < 1 cm (R0, R1, R2 with maximal diameter of largest tumor residual of 1 cm), adjuvant chemotherapy with complete remission after chemotherapy, Karnofsky performance score > 60, age > 18 years, and written informed consent.

Patients were followed for 36 months after WART. Follow-up visits were scheduled at 6 weeks and 3, 6, 9, 12, 15, 18, 24, 30, and 36 months after treatment. Each visit included update of case history, documentation of adverse events according to Common Terminology Criteria for Adverse Events (CTCAE) version 3.0, assessment of quality of life using the European Organization for Research and Treatment of Cancer QLQ-C30 questionnaire, gynecologic examination, transvaginal ultrasound, and a blood sample including tumor marker checks (CA- 125). In addition, pelvic-abdominal CT or MRI scans were performed 6, 12, and 24 months after treatment. The primary endpoint was treatment tolerance, defined as the lack of any CTCAE °4 toxicity. Secondary endpoints were rate of therapy disruption, rate of therapy abortion, acute toxicity (< 6 weeks after end of WART), chronic toxicity (> 6 weeks after end of WART), quality of life, PFS and OS. The study was approved by the local Ethics Committee of Heidelberg University. Written informed consent was obtained from each participant before entering the trial.

The study was planned using an adaptive two-stage design allowing for a sample size recalculation after the interim analysis. Details on the applied design and sample size recalculation can be found in the article presenting the results on the short-term endpoints [39]. The first part included the confirmatory analysis, where the tested null hypothesis stated that the true rate of patients for whom no CTCAE °4 toxicity occurs during radiation therapy and until 6 weeks after its termination amounts to at most 70%. In this article, results on the secondary endpoints regarding follow-up data are presented. Different types of toxicities occurring 6 weeks after termination of WART are analyzed using absolute and relative frequencies. Quality of life is evaluated for each follow-up visit using mean and standard deviation. Kaplan-Meier method is used to estimate survival rates of PFS and OS. For PFS and OS, time was calculated from start of WART until disease recurrence or death. Statistical analyses were performed using the Statistical Analysis System, version 9.4 (SAS Institute, Cary NC), and figures were prepared in R, version 3.5.0 [40].

The study is registered with ClinicalTrials.gov (NCT01180504).

Late toxicities are shown in Fig. 1. Late side effects of intensity modulated WART mostly consisted of lower limb edema. Seven patients (38.9%) developed °1 edema, one patient (5.6%) °2 and °3 edema, respectively. No late gastrointestinal toxicities greater than °2 were observed. Five patients (27.8%) reported °1 and one patient (5.6%) °2 diarrhea, respectively. Two patients (11.1%) showed enteritic symptoms °1. One patient developed cystitis °1 (5.6%). Other side effects mostly consisted of fatigue °1-°2, constipation °1, nausea °1-°2, abdominal pain °1-°2, edema of the abdomen °1 and pruritus °1. For 3 patients, an ileus has been reported during follow-up, which was mainly associated with disease recurrence.

Hematological toxicities are shown in Fig. 2. There were no severe late side effects regarding renal or hepatic function. One patient (5.6%) showed elevated creatinine °1. In total, 12 patients showed slightly elevated liver enzymes with 8 patients (44.4%) showing elevated SGOT (serum glutamic oxaloacetic transaminase) °1, 9 patients (50%) and one patient (5.6%) showing elevated SGPT (serum glutamic pyruvic transaminase) °1 and °2, respectively. Three patients (16.7%) showed elevated gamma-Glutamyltransferase (γGT) °3, which is a parameter for cholestasis. Furthermore, 50%/11.1% of patients showed elevated γGT °1/°2 and 55.6%/5.6% of patients showed elevated alkaline phosphatase (AP) °1/°2, respectively. One patient (5.6%) was observed with isolated elevation of bilirubin °2.

Baseline function scale scores were available for all 20 patients; baseline global health status scores were available for 19 patients. Data for week 4 during and 6 weeks after WART and month 3, 6, 9, 12, 15, 18, 24, 30 and 36 were available for 19, 19, 18, 18, 17, 15, 12, 11, 9, 8 and 8 patients, respectively. Box plots of function scale scores and global health status scores at baseline and all follow-up visits are shown in Fig. 3. Patients presented with a mean global health status score of 57.9 +/− 15.6 at baseline. We could already show that the mean global health status score decreased by 18.1 points (95% CI 7.1–29.0) 4 weeks after starting WART [39]. However, 6 months after WART the score had normalized completely with a mean score of 57.9 +/− 26.3. The score even increased above baseline level during further follow-up with a maximum mean score of 69.8 +/− 14.7 after 30 months. Physical functioning score at baseline was 74.1 +/− 19.2. The lowest score could be observed 4 weeks after starting WART, with continuous increase back to baseline after 9 months with a mean score of 76.1 +/− 19.2, which even increased further with a maximum mean score after 36 months of 83.3 +/− 11.7. Role functioning score at baseline was 58.3 +/− 33.1. The lowest score could be observed 4 weeks after starting WART, with continuous increase back to normal already 6 weeks after WART with a mean score of 60 +/− 33.5. The maximum mean score of 75.6 +/− 25.1 was observed after 12 months. Emotional functioning score at baseline was 63.9 +/− 26.4, which decreased to a minimum mean score of 55.7 +/− 29.4 4 weeks after starting WART and showed an increase back to baseline level 6 weeks after WART with a mean score of 65.8 +/− 22.9. The maximum score could be observed after 9 months with a mean score of 71.6 +/− 22.3. Cognitive functioning score at baseline was 84.2 +/− 21.3. The lowest score was observed after 6 months with a mean score of 71.3 +/− 33.7, which increased nearly back to baseline after 36 months with a mean score of 83.3 +/− 25.2. Social functioning score at baseline was 61.7 +/− 26.5, which decreased to a minimum score of 43.9 +/− 36.1 4 weeks after starting WART, with continuous increase back to baseline after 3 months. The maximum mean score of 80.6 +/− 19.9 was observed after 15 months.

Eleven patients (55%) experienced disease recurrence, all of them intraperitoneally. 8 (72.7%) of all first recurrences were localized inside the peritoneal cavity only, 3 (27.3%) were localized inside as well as outside. Different patterns of intraperitoneal recurrence were observed: 2 (18.2%) patients presented with malignant ascites only, 5 (45.5%) with a new macroscopic tumor lesion and 3 (27.3%) with a combination of both. One patient (9.1%) developed a new macroscopic tumor lesion together with malignant pleural effusion. All in all, 7 patients showed up with distant metastases during follow-up. Details of recurrence patterns and salvage treatment are listed in Table 2. Partial remission after salvage therapy of first recurrence could be achieved in 5 cases, 3 patients showed progressive disease and 4 patients received further second line therapies because of progression in the course of the follow-up.

Median time to recurrence was 25.3 months. Estimated 1-, 2- and 3-year-PFS was 74, 51 and 40%, respectively (Fig. 4). Three patients died during follow-up after 5, 8.3 and 16.9 months, respectively. Two of them died because of disease progression, 1 patient died because of operative complications. Ten patients (50%) were known to be alive at the end of the study, 7 patients were censored in the course of the trial. Estimated 1-, 2- and 3-year-OS was 89, 83 and 83%, respectively (Fig. 5).