7. Radiotherapy of endometrial carcinoma – Endometrial Cancer

Background

The role of external pelvic irradiation in stage I endometrial cancer (endometrioid histology) has been repeatedly assessed in meta-analyses of randomized trials by Kong et al. [579]. The most recent version of the meta-analysis is from 2012 [580]. The evidence assessment by the Guideline group was based on the ASTRO Guideline published in 2014 [581], which was based on the systematic literature search from 1980 to 2011, and a supplementary update search. As this most recent version of the meta-analysis by Kong et al. 2012 [580] could therefore not yet be included in the ASTRO Guideline used for adaptation, although both have a very similar basis of primary studies, it is now given special consideration. 

The meta-analysis considered 8 studies involving 4,273 patients with stage I endometrial cancer. When considering subgroups, it should be noted that in the primary studies as well as in meta-analyses, older TNM and FIGO stage classifications were used in some cases; in the recommendations and background texts of this Guideline, the corresponding current stages according to the TNM 7 classification of 2010/11 [582] are always mentioned. In the TNM 8 classification [583] valid since 2017, no changes have occurred for endometrial cancer. In the comparison of external beam radiotherapy vs. no external beam radiotherapy (with balanced distribution of the use of vaginal brachytherapy), there was no effect on overall survival (primary endpoint of the meta-analysis), disease-specific survival, or distant metastasis for the overall stage I group, but there was a significant decrease in locoregional recurrence rate (HR = 0.36, 95%-CI = 0.25–0.52; p < 0.001).

In the stage I “low risk” subgroup (defined for meta-analysis as stage IA, i.e., maximal myometrial infiltration inner half, and G 1–2), the primary endpoint overall survival was not analyzed.

In the stage I “intermediate risk” subgroup (defined for meta-analysis as stage IB, i.e., infiltration of the outer half of the myometrium or G3, only one of the two factors present or as defined by the respective study), no significant effects of external beam radiotherapy on overall survival or disease-specific survival were found.

For the stage I “high risk” subgroup (defined for meta-analysis as stage IB, i.e., infiltration of the outer half of the myometrium, and G3 or as defined by individual studies), there was no significant effect of external beam radiotherapy on overall survival (HR = 0.91, 95%-CI = 0.60–1.39; I2 = 0%; p = 0.67) or disease-specific survival (HR = 0.84, 95%-CI = 0.51–1.40; I2 = 0%; p = 0.51). In the meta-analysis, the delivery of external beam radiotherapy was associated with increased rates of acute toxicities (RR = 4.68, 95%-CI = 1.35–16.16; I2 = 0%; p = 0.01 for grades 3 and 4) and late toxicities (RR = 2.58, 95%-CI = 1.61–4.11; I2 = 0%; p < 0.001). Regarding quality of life, the meta-analysis refers to the evaluation on the randomized PORTEC-1 trial [584], see below.

Regarding the results of external beam radiotherapy in stage I (endometrioid histology), the Dutch study PORTEC-1 is presented as an example for which 15-year data [585] have not yet been included in the ASTRO Guideline, but are included in the described meta-analysis [580]. In PORTEC-1, 714 patients in the treatment period 1990–1997 after hysterectomy without lymphadenectomy in stages IB, i.e. more than 50% myometrial infiltration with G1–2 or IA with myometrial infiltration (myometrial infiltration < 50%) with G2–3, were randomized to: external pelvic irradiation with 46 Gy (2-dimensional irradiation techniques) vs. observation alone. 

At median follow-up of 13.3 years [585], a 15-year locoregional recurrence rate of 5.8% (with irradiation) vs. 15.5% (without) was found in the overall collective (HR = 3.46; 95%-CI 1.93–6.18; log-rank test p < 0.0001), and the vaginal recurrence rate was 2.5% (with pelvic irradiation) vs. 11% (without) (no p-value given). The 15-year overall survival was 52% (with radiation) vs. 60% (without) (HR = 0.84; 95%-CI 0.67–1.06; log-rank test, p = 0.14), and the 15-year recurrence-free survival was 50% (with radiation) vs. 54% (without) (p = 0.94).

In the “high-intermediate risk” subgroup of the PORTEC-1 trial (defined as the presence of at least two of the following: G3, age > 60 years, stage IB, i.e., myometrial infiltration > 50%), 15-year overall survival was 41% with radiation vs. 48% without (p = 0.35), and disease-specific mortality risk was 14% vs. 13% (p value not reported). In this risk group, external beam irradiation particularly reduced the 15-year locoregional recurrence rate, from 21% to 7% (percentages not given in text, read from Kaplan-Meier curve).

A cross-sectional quality of life analysis was performed on the PORTEC-1 study [584]. No baseline data were obtained at the start of therapy, but 351 patients were surveyed over the long-term (median 13.3 years after therapy) with quality-of-life questionnaires (SF-36 and individual questions on bladder and bowel symptoms and sexual function from EORTC organ modules); the response rate was 70%. There was significantly increased symptomatology regarding urinary incontinence, diarrhea and fecal incontinence in the radiation arm vs. observation, as well as worsened quality of life in the areas of physical function and role function. 

In the interpretation, it should be taken into account that the form of 2D radiotherapy used in the study (1990–1997) has been replaced in Germany since the mid-1990s by the gentler 3D conformal radiotherapy (CT-guided radiation planning with 3D contouring of target volumes and adjacent organs).

Also after data closure of the meta-analysis [580] as well as the ASTRO Guideline [581] the long-term evaluation of a Norwegian study based on previous publications [586] considered in both was published [587]: in the period 1968–1974, 568 patients in stage I in both arms received vaginal radium brachytherapy, with randomization to external pelvic irradiation (40 Gy, counter-field technique with partial block-out from 20 Gy, partly Cobalt-60) vs. no external irradiation. The overall collective showed identical overall survival of median 20.5 years, which also corresponds to the median follow-up time of this collective, in both arms in younger patients (< 60 years at diagnosis), but significantly better survival without external pelvic irradiation, which was attributed to an increased risk of second malignancy in the arm with pelvic irradiation (HR = 2.02; 95%-CI 1.30–3.15 in the younger subgroup). 

Recent analyses of secondary tumor risk including the PORTEC-1 and PORTEC-2 endometrial cancer trials show no increased risk of secondary malignancies after external pelvic irradiation at median follow-up of 13.0 years [588].

A special position within stage I is occupied by the combination of stage IB with simultaneous presence of G3. This constellation of a “high-risk”-stage I was, among others, not included in the PORTEC-1 and PORTEC-2 studies. Parallel to the PORTEC-1 study, 104 consecutive patients with this constellation were included in a registry study between 1990 and 1997 and received postoperative pelvic irradiation with a mean total dose of 46 Gy, of which 99 patients were evaluable [589]. 

For the registry collective, the following 5-year results were reported (in parentheses, the comparative value for the subgroup of n = 137 with IC [old] but grading G2 from the arm with external pelvic irradiation of PORTEC-1): Overall survival 58% (85%), death from endometrial cancer 30% (6%), vaginal recurrence 5% (2%), pelvic recurrence 7% (0%), distant metastasis 23% (7%). These registry data document good pelvic tumor control in stage IB G3 patients receiving pelvic irradiation despite an unfavorable survival prognosis.

In the current meta-analysis [580], no survival benefit of pelvic irradiation is observed for the stage I “high risk” subgroup (defined as stage IB with G3 or as defined by the respective study) (two studies, total 334 patients, HR = 0.91; 95%-CI = 0.60–1.39; I2 = 0%; non-significant reduction in risk of death p = 0.67). 

The two studies included in the current meta-analysis on this topic [590], [591] will be considered in detail. The GOG 99 study randomized 392 patients (endometrioid type only) with myometrial infiltration (i.e., stage IA with myometrial infiltration or stage IB) and stage II, each of any grade, postoperatively to postoperative pelvic irradiation alone at 50.4 Gy (2D irradiation technique, counterfield or 4-field technique, in some cases Cobalt-60) vs. observation [590]. 

During the course of the study, a subgroup “high-intermediate risk” (HIR) was identified and considered separately. This was defined as follows: age >70 years with at least one of the factors G2–3, lymphovascular invasion, infiltration of the outer third of myometrium, or age >50 with at least two of these factors, or any age with all three factors. For this HIR group, overall survival (HR = 0.73, 90% CI = 0.43–1.26) was not significantly improved in the arm with external pelvic irradiation (4-year overall survival 88% vs. 74%, p = 0.35). The rate of initial pelvic recurrence at four years was decreased by pelvic irradiation from 13% to 5% (HR = 0.37; 90% CI 0.12–1.11).

The ASTEC/EN.5 study represents the pooled analysis of two initially separately planned randomized trials from Europe and North America [591]. Between 1996 and 2005, 905 patients at 112 centers of stage I or IIA with “intermediate or high risk” (definition: IA with G3 or IB with any grading or papillary serous/clear cell) were randomized to external pelvic irradiation with 40–46 Gy vs. observation. Vaginal brachytherapy was performed according to center strategy (actual use in both arms equally frequent, 53% vs. 54%). For the overall group, 5-year overall survival was 83.5% with external pelvic irradiation and 83.9% with observation (HR = 1.05; 95%-CI 0.75–1.48; p = 0.77). 

In a defined high-risk group (IB with G3 or II (endocervical glandular invasion only) with G3 or papillary serous/clear cell or II (cervical stromal invasion) (the latter outside the inclusion criteria but also evaluated), disease-specific 5-year recurrence-free survival (i.e., death from other causes not counted as an event) was 73.7% vs. 88.8% with intermediate risk. However, within the high-risk group, with the use of vaginal brachytherapy in approximately half of the cases in both arms, no effect of external pelvic irradiation was seen on overall survival (HR 1.07; 5-year rate = 3%, 95%-CI = –6% to 10%) or on disease-specific survival (HR 1.01; 5-year rate = 3%, 95%-CI = –5% to 9%).

In summary, the current meta-analysis shows that external pelvic irradiation for stage I endometrial cancer (endometrioid type) reduces the locoregional recurrence rate (including vaginal and other pelvic recurrences) to about one-third of the control group over the long-term, but thus does not positively affect endpoints of overall or disease-specific survival, even in the meta-analyzable subgroups. In the subgroup “high risk” it must be taken into account that the two studies included in the meta-analysis (see above) used significantly different risk definitions; in the larger study brachytherapy was used in equal proportions in both arms, and both studies tended to show contrasting results. Only in the GOG-99 study was a central reference pathological assessment performed, which is relevant in light of the massive downgrading in subsequent reference pathological assessment in the PORTEC-1 and PORTEC-2 studies [592]. 

Kong et al. [579] point out in their discussion of the meta-analysis that the number of high-risk patients in the reported studies of external pelvic irradiation is relatively small (n = 334 in two studies) and it cannot be ruled out that the meta-analysis has insufficient power to detect a survival benefit. 

Recent analyses document a particular risk profile for the group with “substantial LVSI” (highest degree of lymphovascular vascular invasion in three-tiered system): in the PORTEC-1 and PORTEC-2 study collectives (no LK dissection there), 4.8% of patients had “extensive LVSI” [593]. After 5 years, the following pelvic recurrence rates were observed: no adjuvant therapy 30.7%, vaginal brachytherapy 27.1%, external pelvic irradiation 4.3%. Thus, in stage pT1pNx with “extensive LVSI” – there is a strong reduction in the high pelvic recurrence risk with external pelvic irradiation, regardless of grading. However, a prospective evaluation of the significance of LVSI is not available to date.

Background

The recommendations of the guidelines suitable for adaptation are based on the results of randomized trials that investigated vaginal brachytherapy with external pelvic irradiation or with observation alone or different dose concepts of brachytherapy. The update search found only two other relevant studies: a population-based analysis of forgoing any radiotherapy [607] and a retrospective analysis of the feasibility of vaginal brachytherapy in parallel with chemotherapy [608]. In a European multicenter study [609], 645 low-risk patients (defined as stage IA by current classification with G1–2, endometrioid histology only) were randomized to HDR brachytherapy alone (Iridium-192 or Cobalt-60) of the upper two-thirds of the vagina with total doses of 18 to 24 Gy (single dose 3 to 8 Gy), with dose prescription to 5 mm tissue depth (one center of six: LDR technique cesium-137, 40 Gy) vs. observation. At a median follow-up of 5.7 years, recurrence rates were as follows: vaginal 3.1% vs. 1.2% with brachytherapy (p = 0.114), pelvic 0.9% without brachytherapy vs. 0.3% with brachytherapy (p = 0.326), and distant 0.6% without brachytherapy vs. 2.2% with brachytherapy (p = 0.087) (confidence intervals not shown).

Late side effects were assessed according to RTOG/EORTC classification (https://www.rtog.org/). Significantly more vaginal adverse events (mild atrophy, dryness, sporadic mucosal bleeding) occurred in the vaginal brachytherapy arm (p = 0.00004), but these were rare at 8.8% (vs. 1.5% in the observation arm) and predominantly grade 1. There was a trend regarding more urogenital side effects in the brachytherapy arm (total 2.8% vs. 0.6%, p = 0.063), but no difference in intestinal sequelae (0.9% vs. 0.6%); side effects of grade 3 or higher did not occur at all.

The Dutch PORTEC-2 study [600] tested whether in stage I “high intermediate risk” vaginal brachytherapy alone can secure comparable good locoregional, especially vaginal, tumor control compared to external pelvic irradiation, with more favorable toxicity and quality of life profiles. Included were: age >60 years with stage IB and G1–2 or with stage IA and G3. Patients with combination IB and G3 were not included. Stage II (endocervical gland involvement only) was included unless the combination >50% myometrial infiltration and G3 was present. 427 patients were randomized to external pelvic irradiation alone (46 Gy in fractions of 2 Gy each, CT-guided 3D radiation planning) vs. vaginal brachytherapy alone to the upper half of the vagina (HDR 3 x 7 Gy, dose prescription to 5mm tissue depth, 1 x per week; or 30 Gy LDR or 28 Gy MDR in each session) between 2002 and 2006 after hysterectomy without routine lymphadenectomy. The following 5-year recurrence rates (external pelvic irradiation vs. vaginal brachytherapy, respectively) were reported: vaginal 1.6% vs. 1.8% (HR = 0.78, 95%-CI 0.17–3.49; p = 0.74), locoregional 2.1% vs. 5.1% (HR = 2.08, 95%-CI 0.71–6.09; p = 0.17), distant 5.7% vs. 8.3% (HR = 1.32, 95%-CI 0.63–2.74; p = 0.46). 

The 5-year overall survival was 79.6% after external beam radiation and 84.8% after brachytherapy (HR = 1.17, 95%-CI 0.69–1.98; p = 0.57); disease-free survival was 78.1% vs. 82.7% (HR = 1.09, 95%-CI 0.66–1.78; p = 0.74). Gastrointestinal toxicities of grades 1 to 2 (RTOG/EORTC classification) were present at the end of therapy in 53.8% (external beam) vs. 12.6% (brachytherapy); significant differences existed in this regard until 24 months after therapy. From 6 months on, more vaginal atrophy (especially grade 2) was continuously detectable in the brachytherapy arm. 

Recently, the 10-year data of the PORTEC-2 study were published [342]: while overall the approximate equivalence of vaginal brachytherapy to external pelvic irradiation was confirmed in the studied stage groups, histopathologically or molecularly defined subgroups benefiting from external irradiation could be elaborated. After 10 years, overall survival was 69.5% (vaginal brachytherapy) vs. 67.6% (external irradiation) (HR 0.94, 95%-CI 0.67-1.32; p=0.72), pelvic recurrence rate was 2.5% vs. 0.5% (HR 5.07, 95%-CI 0.59-43.41; p=0.1) and vaginal recurrence rate was 3.0% vs. 1.5% (HR 1.68, 95%-CI 0.40-7.03; p=0.47).

The PORTEC-2 trial demonstrated a significant decrease in pelvic recurrence rates compared to brachytherapy for the p53-positive or L1CAM-positive or LVSI-positive subgroups (defined as “extensive LVSI” in the three-stage system) after external pelvic irradiation. Pelvic recurrence rates after 5 years were approximately 30% (brachytherapy) vs. approximately 0% (external irradiation) in the p53-positive group (p< 0.001), and approximately 25% (brachytherapy) vs. approx. 0% (external irradiation) (p< 0.0.1) and in the group LVSI-positive approx. 30% (brachytherapy) and approx. 12% (external irradiation) (p< 0.001) (exact values not given, reading from graph). Thus, for the above groups, there is an increased risk of pelvic recurrence after postoperative vaginal brachytherapy alone, which is why postoperative external pelvic irradiation is now recommended as an alternative.

While the main benefit of postoperative vaginal brachytherapy has previously been seen in lowering the risk of vaginal recurrence, several recent registry studies consistently suggest a resulting benefit in overall survival as well. Al-Hili et al. looked at overall survival in 132,393 stage patients from 2004 to 2013 (National Cancer Database: https://www.facs.org/quality-programs/cancer/ncdb) [598]. For group IB G1-2, vaginal brachytherapy showed the best 5-year survival at 89% with significant advantage over observation (83%; HR 0.64, 95%-CI0.56-0.73); p< 0.0001) and also over external pelvic irradiation (87%, p=0.0004). For group IA G3, vaginal brachytherapy achieved a 5-year survival of 87% (vs. observation 83%; HR 0.81, 95%-CI 0.64 -1.02;p=0.07). Similarly, in an analysis of 44,309 stage I patients with lymphadenectomy (National Cancer Database 2003-2011), Rydzewski et al. showed a significant benefit of vaginal brachytherapy in overall survival compared with observation (HR 0.62, 95%-CI 0.51-0.74, p< 0.001), which was not present for pelvic irradiation (HR 0.93, 95%-CI 0.77-1.11, p=0.409) [610]. Thus, it can be assumed that vaginal brachytherapy also has a demonstrable beneficial effect on overall survival in the stage groups in which it has been previously recommended for prophylaxis of vaginal recurrence due to a favorable benefit-risk profile.

The quality-of-life analysis of the PORTEC-2 study was conducted as a longitudinal study over the first 24 months after therapy using the EORTC QLQ-C30 questionnaire and individual questions from organ modules with an overall response rate of 81%, and 53% at the 2-year time point [599]. There was a statistically significant and clinically relevant advantage for the vaginal brachytherapy arm at different time points in the areas of diarrhea, impairment of daily life due to bowel symptoms, and social function. It is concluded that in the considered collective, vaginal brachytherapy can achieve excellent vaginal tumor control, locoregional control comparable to external pelvic irradiation and comparable recurrence-free and overall survival, with less gastrointestinal toxicity and better quality of life.

The results of generally foregoing any radiotherapy in stage I low-risk (defined here as <50% myometrial infiltration with G1–2) or intermediate-risk (>50% myometrial infiltration with G1–2 or <50% myometrial infiltration with G3) patients were prospectively studied nationwide in Denmark during the 1998 to 1999 treatment period (survival data in Bertelsen et al. [607]; recurrence data in Ortoft et al. [611]). At a median follow-up of 13.8 years, 6.3% recurrences (2.9% vaginal, 1.2% pelvic, 1.2% abdominal, and 1.0% distant) were observed at low risk. At intermediate risk, recurrences occurred without radiotherapy in 21.6% (9.5% vaginal, 4.8% pelvic, 2.2% abdominal, 5.2% distant). When this intermediate group was subdivided into “high-intermediate” (>50 years with G2 with >2/3-myometrial infiltration or >70 years with G2–3 with >2/3-myometrial infiltration) and “low-intermediate” (the remainder), the recurrence rates for “high-intermediate” vs. “low-intermediate”: total 25.8% vs. 16.2%, vaginal 11.4% vs. 7.0%, pelvic 6.1% vs. 3.0%, abdominal 3.0% vs. 1.0%, and distant 5.3% vs. 5.1%. 

The investigators of this study [607] consider the recurrence rates in the intermediate-risk stage I group to be acceptable because the 5-year overall survival in this population of 78% is comparable to the 79% observed for this high-risk population in an earlier period (1986–1988) with even broader use of radiotherapy. However, this evaluation insufficiently accounts for the burden of salvage therapies and for trends in overall life expectancy relevant to the overall survival endpoint.

The randomized GOG 249 trial compared pelvic irradiation (45-50 Gy) with the combination of vaginal brachytherapy plus 3 cycles of paclitaxel 175 / carboplatin AUC 6 [606] in stage I endometrioid with risk factors (IB or G2-3 or LVSI, age-dependent 1-3 factors had to be present), stage II endometrioid, and stages I to II serous or clear cell (without positive peritoneal cytology ). Of 601 patients randomized, 75% were stage I, and the rate of lymphadenectomies performed was 89%. After 5-year recurrence-free survival was 76% in both arms (HR 0.92, 95%-CI 0.65 – 1.30; p=0.31), 5-year overall survival was 87% for pelvic irradiation and 85% for brachytherapy plus chemotherapy (HR 1.04, 95%-CI 0.66 – 1.63; p=0.57). While rates of vaginal recurrence (2.5%) and distant recurrence (18%) were the same in both arms, pelvic irradiation halved the rate of pelvic or para-aortic recurrence (4% vs. 9%, HR 0.47, 95%-CI 0.24 – 0.94). Since for the majority of included patients none of the study arms met the previous S3 Guideline recommendation, current recommendations for action can hardly be derived. From a radiotherapeutic perspective, the study demonstrates the value of external pelvic irradiation in reducing the locoregional risk of recurrence, even after lymphadenectomy and in comparison to brachytherapy plus chemotherapy.

Especially under the impression of equivalence of the two therapy arms with regard to oncological outcome in overall survival, acute and late toxicity are of particular relevance: acute toxicity was higher in the chemotherapy + brachytherapy arm and patient reported outcomes showed a maximum at 11 weeks in the FACIT fatigue subscale in the chemotherapy arm (p < 0.001) and it took 8 months to reach baseline again. In the percutaneous radiotherapy arm, fatigue had already subsided after 11 weeks. A similar picture was seen for neurotoxicity, which measured by FACT/GOG-Ntx subscale was still relevantly higher after 4, 11 weeks and even after 8 months and only returned to baseline after 14 months (p< 0.001). Especially probably under the aspect of toxicity, the authors concluded that percutaneous radiotherapy was the appropriate adjuvant therapy in this situation. Toxicity data from the PORTEC-3 study [601] also point to the risk of long-lasting neurotoxicity, which is highly relevant for quality of life. Measured with EORTC QLQ-C30 OV 28, late neurotoxicity was strongly increased in the chemotherapy arm with 27.8 vs. 13.2% (p< 0.0001).

Specifically for stage II, the value of vaginal brachytherapy and external pelvic irradiation (± vaginal brachytherapy) was considered in a recent meta-analysis of 15 cohort studies involving 1070 patients [605]. For the group of studies with performance of lymph node dissection or sampling in at least 90% of cases, the use of external pelvic irradiation (± vaginal brachytherapy) compared with brachytherapy alone showed a strong reduction in locoregional recurrence risk (HR 0.17, 95%-CI 0.05-0.49; p=0.0009) and also a significant improvement in overall survival (HR 0.41, 95%-CI 0.17-0.99; p< 0.05). The authors concluded that external pelvic irradiation should be considered in stage II if risk factors (G3, myometrial involvement > 50% or LVSI) are present.

Also for stage II, a Danish registry study (2005-2012) was able to document the effect of external beam irradiation when lymphadenectomy was performed area-wide, as external beam irradiation was not performed from 2010 and brachytherapy was not performed at all [604]. The 5-year overall survival with vs. without pelvic irradiation was 80.3% vs. 71.7% (HR 0.66, 95%-CI 0.39-1.11), the 5-year rate of vaginal recurrence 7.3% vs. 14.2% (p< 0.05), and the 5-year rate of other pelvic recurrence 2.8% vs. 12.7% (p< 0.05).

While these two analyses on stage II did not consider molecular subgroups, they demonstrate a potential of postoperative external pelvic irradiation to safeguard locoregional tumor control in a proportion of stage II patients – even if lymphadenectomy was performed.

Performance of vaginal brachytherapy

Vaginal brachytherapy is performed with a special cylindrical applicator. The diameter and length of the applicator are based on the anatomical conditions of the patient. The applicator is inserted in lithotomy position. The irradiation itself should be performed with the patient’s legs extended. The length of the residual vagina should be measured and documented before insertion of the cylinder. The target volume is individualized and should generally include the proximal third of the vaginal stump. The dose specification for standardized plans is given in 5mm tissue depth. Ideally, imaging (vaginal ultrasound, CT or MRI) will be used to determine the thickness of the vaginal epithelium and the distance to the rectal wall and adjust the dose specification if necessary.

For vaginal brachytherapy, there are different dosing and fractionation regimens that are used. There are no phase 3 trials meeting current standards that have comparatively evaluated the dosing of vaginal brachytherapy. A randomized trial by Sorbe compared 6 x 2.5 Gy versus 6 x 5.0 Gy. This showed no significant differences in terms of local recurrence rate. The rate of vaginal shortening and bleeding was significantly increased in the 6 x 5 Gy arm at 5 years.  In the study, the upper 2/3 of the vagina was irradiated in a short period of 8 days in all patients. Irradiation of 2/3 of the vagina should not be performed today. Therefore, the 6 x 2.5 Gy regimen can only be conditionally recommended. 

For vaginal brachytherapy alone, doses between 15-25 Gy should be applied in 3-4 fractions using HDR – brachytherapy. In the Portec II study, a dosage of 3 x 7.0 Gy at 5mm tissue depth once weekly was used.  This dosage is also recommended in the ESGO/ESTRO/ESP Guideline 2020/21. Equivalent dose schedules are 4 x 6.0 Gy or 5 x 5.0 Gy 1-2 weekly. For vaginal brachytherapy as boost after percutaneous radiotherapy, 8 – 11 Gy are applied in 2-3 fractions. In general, the regimen of 2 x 5.0 Gy is used. Vaginal brachytherapy as a boost should be performed at the end of percutaneous radiotherapy.

Background

Relevant studies on the value of radiotherapy for endometrioid endometrial carcinoma (type I) in stages III to IVA relate to its effectiveness compared with chemotherapy alone or as an element of combined (sequential or simultaneous) radiochemotherapy.

In a randomized Italian multicenter trial [612] from 1990 to 1997, 491 patients with stages IB with G3 or II with >50% myometrial infiltration with G3 or stage III, 65% of cases were in the latter) were randomized postoperatively to pelvic irradiation with 45–50 Gy vs. chemotherapy with five cycles of cisplatin 50/doxorubicin 45/cyclophosphamide 600. 5-year overall survival (69% vs. 66%, p = 0.85) and 5-year progression-free survival (63% vs. 63%, p = 0.64) (HR = 1.04 95%-CI 0.72–1.50) were nearly identical for radiotherapy vs. chemotherapy. The following recurrence rates were reported (radiotherapy vs. chemotherapy, respectively): distant (extra-abdominal or liver) 21%/16%, pelvic 7%/11%, distant and pelvic 5%/5%. The data indicate improved locoregional control in the radiotherapy arm and improved distant tumor control in the chemotherapy arm, suggesting the benefit of combining both elements.

In the GOG 122 [613] trial, 396 patients with stage III or IV disease after hysterectomy with surgical staging and leaving no more than 2 cm of residual tumor were randomized to whole-abdomen radiation with 30 Gy (single dose 1.5 Gy) followed by pelvic dose boost to 45 Gy vs. chemotherapy alone with seven cycles of doxorubicin 60/cisplatin 50 followed by one cycle of cisplatin. 

Despite randomization, the stage distribution was unequal between arms, e.g., nodal positive 45.1% in the radiotherapy arm and 58.2% in the chemotherapy arm. The 5-year progression-free survival was 42% with chemotherapy and 38% with radiotherapy, and the 5-year overall survival was 53% vs. 42% (p-values not reported). Due to the imbalance in stage distribution, the authors performed a stage-adjusted analysis, which showed an advantage for the chemotherapy arm in 5-year overall survival (55% vs. 42%, HR = 0.68, 95%-CI 0.52–0.89; p = 0.004) and 5-year progression-free survival (50% vs. 38%, HR = 0.71, 95%-CI 0.55–0.91; p = 0.007). 

The following recurrence rates were reported (radiotherapy vs. chemotherapy, respectively): total 54% vs. 50%, pelvic 13% vs. 18%, abdominal 16% vs. 14%, extra-abdominal or liver 22% vs. 18%. There were 4% therapy-associated deaths in the chemotherapy arm and 2% in the radiotherapy arm. The study has methodologic flaws regarding stage-adjusted evaluation after randomization and the use of whole-abdomen radiation, which is now obsolete.

The potential of combined sequential radiochemotherapy is documented by the pooled analysis of two randomized trials [614]. Here, 534 patients with stage I (“with risk profile requiring adjuvant therapy”, unspecified) and, based on amendments, also stage II, IIIA (positive peritoneal cytology only), and IIIC (positive lymph nodes only, with no macroscopic residual) were randomized to postoperative external pelvic irradiation with at least 44 Gy and optional vaginal brachytherapy (this used in 38%, largely balanced in both arms) vs. same radiotherapy followed by chemotherapy with 4 cycles of doxorubicin 50/cisplatin 50 (amendment: also paclitaxel 175/epirubicin 60, doxorubicin 40/carboplatin AUC5, paclitaxel 175/carboplatin AUC 5–6). The treated collective consisted of stage IA (28%), IB (36%), II (14%), and III (20%), respectively, according to current FIGO classification, with non-endometrioid histologies present in 29% of cases. In the primary endpoint of progression-free survival, a benefit was seen for the sequential combination at 78% vs. 69% at five years (HR = 0.63, 95%-CI 0.44–0.89; p = 0.009), and in overall survival there was a trend in favor of the combination at 82% vs. 75% (HR = 0.69, 95%-CI 0.46–1.03; p = 0.07). This benefit was predominantly due to effects in the endometrioid carcinoma subgroup, with 5-year overall survival 84% vs. 74% (HR = 0.60, 95%, CI 0.36–1.00; p = 0.05). No significant benefits of the combination were seen in the serous and clear cell carcinoma group.

In the current Guideline update, the results of the PORTEC-3 and GOG 258 randomized trials and several large registry studies can be included to evaluate the place value of radiotherapy in stage III to IVA.

In the PORTEC-3 trial, 660 patients with stage IA (with myometrial infiltration) G3 to IIIC were randomized to pelvic irradiation with 48.6 Gy (plus brachytherapy for cervical involvement) vs. the same radiotherapy with simultaneous chemotherapy with cisplatin 50 mg/m2 at weeks 1 and 4 and sequentially four cycles of paclitaxel 175 / carboplatin AUC5 [602]. Formally, then, this study tested the benefit of additional chemotherapy in relation to the standard of external pelvic irradiation. Approximately 30% in stage I, 25% in stage II, and 45% in stage III were included. The addition of chemotherapy improved 5-year overall survival from 76.1% to 81.4% (HR 0.70, 95%-CI 0.51-0.97;p adjusted=0.034), and 5-year recurrence-free survival from 69.1% to 76.5% (HR 0.70 (95%-CI 0.52-0.94; p adjusted=0.016). The effect was generated very predominantly in stage III, where there was improvement in 5-year overall survival from 68.5% to 78.5% (HR 0.63, 95%-CI 0.41-0.99; p adjusted=0.043) and 5-year recurrence-free survival from 58.4% to 70.9% (HR 0.61, 95%-CI 0.42 – 0.89; p adjusted=0.011). The addition of chemotherapy tended to reduce the rate of distant recurrences (22.1% vs. 29.4%, HR 0.75, 95%-CI 0.56 – 1.01; p=0.057), but there was virtually no improvement in pelvic (5.5% vs. 8.5%) or vaginal (2.1% in both arms) recurrence. Relative to defined molecular subgroups, the addition of chemotherapy to pelvic irradiation achieved an improvement in 5-year overall survival for p53 mutation (64.9% vs. 41.8%, HR 0.55, 95%-CI 0.30 – 1.00, p-adjusted = .0.49), but not for pole mutation (100% vs. 96%, HR 0.02, 95%-CI < 0.01 – 105, p adjusted 0.637), in MMR deficiency (78.6% radiochemotherapy vs. 84.0% radiotherapy, HR 1.33, 95%-CI 0.64 -2.75, p adjusted 0.446) or without specific mutation (89.3% vs. 87.6%, 95%-CI 0.26 – 1.77, p adjusted = 0.434).

Thus, compared with the standard arm of pelvic radiotherapy (± brachytherapy), the PORTEC-3 trial defined simultaneous radiochemotherapy followed by sequential chemotherapy as a new standard with survival benefit, especially in stage III and in the group with p53 mutation.

The GOG 258 trial compared chemotherapy alone with 6 cycles of paclitaxel 175 / carboplatin AUC 6  against radiochemotherapy with 2 doses of cisplatin 50 simultaneously followed by 4 cycles of paclitaxel 175 / carboplatin AUC 5  [606] in stages III to IVA (with <2 cm residual tumor) of any type and stages I to II serous-clear cell. Data on overall survival are not yet available from this trial. However, no significant advantage for radiochemotherapy has yet been demonstrated in 5-year progression-free survival (after radiochemotherapy 59%, after chemotherapy 58%; HR 0.9 90% CI 0.74-1.10, p=0.20). Recurrence patterns were significantly different: lower 5-year rates of vaginal recurrence (2% vs. 7%, HR 0.36, 95%-CI 0.16 -0.82) and pelvic or para-aortic lymph node recurrence (11% vs. 20%, HR 0.43, 95%-CI 0.28 – 0.66), but higher rates of distant recurrence (27% vs. 21%, HR 1.36, 95%-CI 1.00-1.86) were seen in the radiochemotherapy arm.

As in the GOG 249 trial, with equivalence of the two treatment arms in terms of oncologic outcome based on the data to date, acute and late toxicity is of particular importance: acute grade 3, 4, and 5 toxicity was more favorable at 58% vs. 63% in the radiochemotherapy arm; in particular, grade 4 toxicity was much more pronounced at 14% vs. 30% in the chemotherapy arm, and grade 5 toxicity was present only in the chemotherapy arm. The rate of lymphedema was also more favorable in the radiochemotherapy arm at 7 vs. 15% (p< 0.05), which may be related to the lower rate of pelvic and para-aortic recurrence. In the evaluation, the sometimes massive symptomatology of pelvic as well as para-aortic tumor recurrences should also be considered. In this case, a therapeutic approach that includes radiotherapy is more advantageous, even independent of the other oncological outcomes. It remains to be seen whether the lower rate of distant metastases after chemotherapy alone leads to improved survival [606].

In the absence of recent data from randomized trials on the effect of radiotherapy on overall survival in stages III to IVA, a recent analysis from the National Cancer Database from 2004 to 2016 of 13,270 patients who received polychemotherapy can be considered [615]. In Cox-adjusted analysis of overall survival, for endometrioid carcinomas, the addition of external beam radiotherapy showed an improvement in stage III (HR 0.87, 95%-CI 0.79-0.96, p=0.004), which was most pronounced in stage IIIC when subgroups were considered (HR 0.84, 95%-CI 0.75 – 0.95, p=0.003). In stage IVA endometrioid the effect was not significant with small group size (HR 0.38, 95%-CI 0.10 – 1.41, p=0.15). In non-endometrioid carcinomas, the addition of external beam radiotherapy improved overall survival in stage III overall (HR 0.80, 95%-CI 0.72-0.88, p< 0.0001), particularly in subgroups IIIB (HR 0.52 95%-CI 0.32-0.86) and IIIC (HR 0.79, 95%-CI 0.70 – 0.88, p< 0.0001). Thus, registry data support a survival benefit of external beam radiotherapy in subgroups of the collective considered in GOG 258 (for which survival data are not yet available), particularly for stage IIIC endometrioid.

Provided that sequential (not simultaneous) chemotherapy and radiotherapy are performed postoperatively in stage III to IVA patients, the results of current registry studies favor the sequence chemotherapy followed by radiotherapy. An analysis of the US National Cancer Database (2004-2014, n=5795 stage III-IVA patients) looked at overall survival after chemotherapy followed by radiotherapy vs. the reverse sequence. This was significantly better at 5 years for initiation with chemotherapy (80.1%) compared to initiation with radiotherapy (73.3%, p< 0.001) [616].

Background

When external pelvic irradiation is indicated, vaginal brachytherapy should not automatically be performed in addition. In the PORTEC-1 study, a vaginal recurrence rate of only 2.5% after 15 years was observed in the arm with external pelvic irradiation alone (without additional vaginal brachytherapy) [585]. 

In PORTEC-2, vaginal recurrence occurred in only 1.6% after 5 years following pelvic irradiation alone [600]. In the absence of randomized trials of pelvic irradiation ± vaginal brachytherapy, a recent review [617] considered retrospective studies on the topic. A recommendation of combination pelvic irradiation and brachytherapy was made for stage II and IIIB patients, each with close or positive incision margins.

Background

Evidence regarding the effects of postoperative radiotherapy in type II carcinomas is scarce, as these have mostly been brought in as a smaller subgroup together with high-risk type I patients due to their rarity. However, recent adjuvant therapy trials have relevantly included patients with type II histology, especially with serous carcinomas (PORTEC-3: 16% serous, GOG-249: 15% serous, GOG-258: 18% serous; see above for detail of individual studies). In the PORTEC-3 trial, the addition of concurrent chemotherapy to the standard of percutaneous radiotherapy showed a significant advantage in overall survival for the serous carcinoma group (after 5 years 59.7% vs. 47.9%, HR 0.42, 95%-CI 0.22-0.80) [602]. For the GOG-249 trial comparing postoperative percutaneous radiotherapy and vaginal brachytherapy with three cycles of chemotherapy, subgroup analysis for patients with serous or clear cell carcinoma did not identify a significant advantage for either strategy [618]. In the GOG-258 trial, which tested the addition of percutaneous radiotherapy to chemotherapy alone and for which evaluations of overall survival are not yet available, subgroup analysis of recurrence-free survival for serous carcinoma has not yet shown a clear signal in favor of either arm.

The effect of the therapeutic element radiotherapy in serous carcinomas was considered in a recent meta-analysis of retrospectively collected data [619], numerically dominated by large registry studies from SEER and NCDB. In 9,354 patients evaluated, a highly significant survival benefit (HR = 0.72, 95%-CI 0.63–0.84; p < 0.0001) was reported for combination radiotherapy plus chemotherapy versus chemotherapy alone, which was similar for both limited and advanced stages. Of the patients for whom the type of radiotherapy was known, most had been treated with percutaneous radiotherapy +/- vaginal brachytherapy [619]. This suggests a preference for percutaneous radiotherapy in the high-risk group of type II carcinomas.

For patients with endometrial carcinoma who are inoperable for medical reasons, radiotherapy alone represents a treatment approach with curative intent. 

In the absence of randomized trials, the Gynecological Cancer Group of the European Organization for Research and Treatment of Cancer currently produced a systematic review describing the use of radiotherapy for this indication and the results s [620]. 

A total of 2,694 patients from 25 case series were considered. These were treated with brachytherapy alone (51%) or the combination of brachytherapy plus percutaneous radiation (47%). At five years, disease-specific survival was 78.5%, local control 79.9%, and overall survival, reflecting pre-existing comorbidities, 53.2%. The risk of late sequelae ≥ grade 3 was 2.8% (brachytherapy only) and 3.7% (combination). Based on these data, brachytherapy alone is recommended for stage I grade 1 only, and the combination of percutaneous irradiation and brachytherapy is recommended for the remainder of stage I and for stages II to IV in cases of internal cancer inoperability.

Background

Data are available on the benefit of adjuvant pelvic irradiation compared with no adjuvant therapy for stage I or II carcinosarcoma from the EORTC trial 55874 [622]. Within a mixed cohort of 224 patients with uterine sarcomas, 92 women with carcinosarcoma were randomized to the pelvic irradiation arm with 50.4 Gy (single dose 1.8 Gy, starting within 8 weeks postoperatively) or to the observation arm. At a median follow-up of 6.8 years, improved local control with reduction of local recurrence rate from 47% to 24% was observed for the subgroup of patients with carcinosarcoma, but without significant survival benefit.

Whole-abdomen irradiation with 30 Gy, followed by pelvic dose boost up to 50 Gy (in each case partially with 2 x 1 Gy per day), did not achieve significant differences in overall survival or recurrence rate in the randomized GOG 150 trial (206 patients, of whom 31% were stage I, 45% stage III) compared with chemotherapy with cisplatin, ifosfamide, and mesna, so such an extended volume of irradiation is not recommended [623].

Retrospective evaluations of US registry data (National Oncology Database or Surveillance, Epidemiology and End Results Database) repeatedly show positive effects for adjuvant pelvic irradiation (± brachytherapy) in large collectives with carcinosarcoma on locoregional recurrence-free survival [624] and also on overall survival [522], [625]. Another US epidemiologic study examining the benefit of lymphadenectomy [626], showed a favorable effect on overall survival (HR = 0.64, 95%-CI 0.56–0.73), but pelvic radiotherapy showed only nonsignificant improvements in the group with lymphadenectomy (HR = 0.92, 95%-CI 0.76–1.11) and without lymphadenectomy (HR = 0.87, 95%-CI 0.72–1.05), respectively. The effect of radiotherapy in the overall collective was not reported. 

Given the repeatedly confirmed beneficial effects of radiotherapy on local control, postoperative radiotherapy should usually be indicated for carcinosarcoma.

As part of the guideline revision, an update search on postoperative radiotherapy for carcinosarcoma was conducted in 2020, the results of which were published in a formal non-systematic review  [627]. No new prospective data were found in the update search from 2010, but nine analyses of US or European registry data were found that provided results on overall survival in varying detail depending on stage and radiotherapy modality used (vaginal brachytherapy, external beam or combination).

From the stage-specific registry data, the following options were offered ( +: recommended, 0: unclear, -: not recommended):

When carrying out radiotherapeutic measures, the recommendations of the S3 Guideline “Supportive therapy in oncological patients” [628] should be taken into account. See also Recommendation 9.9 in Chapter 9.

Supportive therapy is an integral part of the treatment concept. Side effects can occur as acute changes during or directly after therapy or as late effects.