Strahlenther Onkol (2021) 197:505–519 https://doi.org/10.1007/s00066-021-01781-6 ORIGINAL ARTICLE Early morbidity and dose–volume effects in definitive radiochemotherapy for locally advanced cervical cancer: a prospective cohort study covering modern treatment techniques Yvette Seppenwoolde1,2,4 · Katarina Majercakova1,2,3 · Martin Buschmann1,2 · Elke Dörr1,2 · Alina E. Sturdza1 · Maximilian P. Schmid1 · Richard Pötter1,2 · Dietmar Georg1,2 Received: 13 October 2020 / Accepted: 30 March 2021 / Published online: 30 April 2021 © The Author(s) 2021 Abstract Purpose Predicting morbidity for patients with locally advanced cervix cancer after external beam radiotherapy (EBRT) based on dose–volume parameters remains an unresolved issue in definitive radiochemotherapy. The aim of this prospective study was to correlate patient characteristics and dose–volume parameters to various early morbidity endpoints for different EBRT techniques, including volumetric modulated arc therapy (VMAT) and adaptive radiotherapy (ART). Methods and materials The study population consisted of 48 patients diagnosed with locally advanced cervix cancer, treated with definitive radiochemotherapy including image-guided adaptive brachytherapy (IGABT). Multiple question- naires (CTCAE 4.03, QLQ-C30 and EORTC QLQ-CX24) were assessed prospectively for patients treated with different EBRT techniques, including online adaptive VMAT. Contouring and treatment planning was based on the EMBRACE pro- tocols. Acute toxicity, classified as general, gastrointestinal (GI) or genitourinary (GU) and their corresponding dose–volume histograms (DVHs) were first correlated by applying least absolute shrinkage and selection operator (LASSO) and subse- quently evaluated by multiple logistic binomial regression. Results The treated EBRT volumes varied for the different techniques with ~2500 cm3 for 3D conformal radiotherapy (3D-CRT), ~2000 cm3 for EMBRACE-I VMAT, and ~1800 cm3 for EMBRACE-II VMAT and ART. In general, a wors- ening of symptoms during the first 5 treatment weeks and recovery afterwards was observed. Dose–volume parameters significantly correlating with stool urgency, rectal and urinary incontinence were as follows: bowel V40Gy < 250 cm3, rectum V40Gy < 80% and bladder V40Gy < 80–90%. Conclusion This prospective study demonstrated the impact of EBRT treatment techniques in combination with chemother- apy on early morbidity. Dose–volume effects for dysuria, urinary incontinence, stool urgency, diarrhea, rectal bleeding, rectal incontinence and weight loss were found. Keywords Image-guided adaptive radiotherapy · Radiotherapy · Brachytherapy · Organs-at-risk toxicity · Quality of life Introduction  Dietmar Georg dietmar.georg@meduniwien.ac.at Radiation oncology plays a major role in the treatment 1 of locally advanced cervical cancer. Current standard of Department of Radiation Oncology, Comprehensive Cancer Center, Medical University of Vienna/AKH Wien, Währinger care is concurrent radiochemotherapy combining exter- Gürtel 18–20, 1090 Vienna, Austria nal beam radiotherapy (EBRT) with cisplatin followed by 2 a brachytherapy (BT) boost, preferably by image-guided Christian Doppler Laboratory for Medical Radiation Research for Radiation Oncology, Medical University of Vienna, adaptive brachytherapy (IGABT; [1]). Vienna, Austria Advanced EBRT techniques, especially those based 3 Department of Radiation Oncology, Hospital de la Santa Creu on intensity modulation, enabled improved organ-at-risk i Sant Pau, Barcelona, Spain (OAR) sparing and consequently decreased the incidence 4 Department of Radiation Oncology, Erasmus MC Cancer of severe toxicity (G3, G4) [2, 3]. Intensity modulated ra- Institute, Rotterdam, The Netherlands diotherapy (IMRT) and volumetric modulated arc therapy K 506 Strahlenther Onkol (2021) 197:505–519 (VMAT) with small margins are challenging in cervical other diagnosis of tumor and patients capable to treatment cancer, because of large intra- and interfraction motion and study compliance. [4–7]. Therefore, adaptive radiotherapy (ART) that adjusts All 48 included patients were treated with definitive treatment plans according to organ movements to further radiochemotherapy including consecutive MRI-guided reduce irradiated volumes has become a main research IGABT. A total of 44 patients (92%) received chemother- interest [8–15]. apy. Patient-specific parameters (PsP) like smoking, alco- Despite these advancements, 60–70% of patients still ex- hol, chemotherapy (both regimen and the number of com- perience early lower grade (G1–2) side effects, of which pleted cycles), the use of para-aortic nodal fields (PAN) and 30% develop into late morbidity [16, 17]. Several publica- age were collected as well. The patients’ characteristics of tions concluded that especially gastrointestinal (GI), geni- this ethics-committee-approved study are summarized in tourinary (GU) and vaginal/sexual problems have an im- Fig. 1. pact on quality of life (QoL) [8, 18, 19]. Furthermore, Target and organ-at-risk (OAR) delineation as well as observational studies demonstrated that radiochemotherapy treatment planning were based either on EMBRACE-I and leads to more physical, psychological and sexual sequelae later EMBRACE-II guidelines [29], though not all patients [18, 20–22], especially in premenopausal patients. How- were participating in those studies. The bowel bag was re- ever, none of those studies correlated side effects with dose contoured as in EMBRACE-II guidelines. EBRT was ad- distributions. ministered with a total dose of 45 Gy/1.8 Gy per fraction. Unraveling the EBRT effects from brachytherapy and After the introduction of VMAT, a 55–60 Gy simultaneous chemotherapy effects is challenging. At the time of BT, boost (SIB, 2.2–2.4 Gy per fraction) was given to positive the full EBRT dose might not have yet been delivered and pelvic or para-aortal lymph nodes. EBRT was followed by is mostly in the lower range of the tolerance dose (TD50) IGABT in 4 fractions with 2 applications, with the planning for rectum and bladder [23]. On the contrary, and based aim to achieve a D90% of 85 Gy (EQD2) for the CTVHR. on the authors’ experience, the recommended dose–volume The EBRT technique was not explicitly predefined for parameter for small bowel (V45Gy < 195 cm3) is exceeded for this study to allow for a progression of treatment technol- most cervix patients. For small bowel tolerance, no differ- ogy over time. EBRT treatment evolved over time from 3D- ence in the incidence of small bowel toxicity was found for CRT, via single plan VMAT to an ART protocol based on doses from 5 to 40 Gy [24], while another study concluded a library approach (plan of the day = PotD), encompassing that V16Gy should be <290 cm3 for patients without prior ab- VMAT plans for empty and full bladder plus a motion ro- dominal surgery to prevent >G2 acute diarrhea [25]. For bust VMAT plan. Daily CBCT information was utilized to bladder, grade 3 late toxicity was found to occur for doses select the most suitable library plan [15]. A drinking proto- >50 Gy [26] and the TD50 of patient reported symptoms is col with the aim of comfortably full-bladder was routinely often >85 Gy [27]. Late rectal injury was found to be rare used in all groups. An empty rectum was advised, but no in current dose ranges [28]. specific rectum protocol was provided. The aim of this mono-institutional observational study was to assess early morbidity as a function of EBRT dose Acute toxicity scoring distributions for different techniques, varying from 3D-CRT to VMAT-ART, the latter being based on a bladder filling/ Acute toxicity was evaluated by Common Terminology uterus motion model. Early side effects were assessed using Criteria for Adverse Events (CTCAE) 4.03 (physician- multiple questionnaires (patient and physician reported). reported; PhysRO) [30] complemented by local report- Dose–volume parameters as well as patient demographics ing items (daily micturition, stool consistency and stool were correlated with early morbidity at different time points urgency) and patient-reported (PRO) Quality of Life Ques- with the aim to determine dose–effect relations and predic- tionnaire (QLQ)-C30 and EORTC QLQ-CX24 question- tive factors. naires. The evaluation was performed before treatment, weekly during treatment, and 1, 6 and 12 weeks after finishing EBRT. The patients evaluated the grade of the Materials and methods symptoms as none, a little, quite a bit and very much. For each PhysRO question, corresponding PRO questions were Patient cohorts and treatment determined and analyzed. The evaluation of stool consistency was based on Bristol The inclusion criteria were: age ≥18 years, histologi- stool form scale, assessing the stool type from 1–7, where cally proven locally advanced cervix cancer FIGO Ib–IVb 1–3 means constipation (scored separately), 4–5 normal, (para-aortic lymph node metastasis), no previous radio- or 6 mushy and 7 liquid stool [31]. Urine frequency 5–7 times chemotherapy, patients suitable for definitive treatment, no a day was considered normal. For daily micturition the fol- K Strahlenther Onkol (2021) 197:505–519 507 Fig. 1 Distribution of patients over the FIGO stages, number of completed chemotherapy cycles, simultaneous integrated boost (SIB), extended fields (PAN) and irradiation technique. Embrace- I//II patients were delineated and planned according to the respec- tive study protocols, but not all were included in those studies. PotD plan of the day, 3DCRT 3D conformal radiotherapy lowing groups were introduced: <5, 6–7, 8–14, >14 times the frequencies of their use in the clinic, resulting in one a day. Synchronous to the introduction of VMAT, extra pa- average weighted overall value. tient-related questions were added (QLQ 55–59) to provide more detailed information. The questions were grouped into Time series general, GI and GU toxicity and were analyzed accordingly. At all analyzed time points, reported side effects were cor- Dose–volume parameters rected for their corresponding baseline values in order to avoid bias from pre-existing complaints or of symptoms that The following dose–volume (DV) parameters were derived may be due to the disease itself. While the maximum EBRT from all EBRT plans: Sigmoid V40Gy, Rectum V40Gy and dose before the start of BT can be linked with the 5 week V30Gy, BladderV40Gy and V30Gy (%), Bowel V40Gy and V30Gy time point, the chemotherapy (CHT) influence (e.g., nausea (cc), Body V43Gy (cc). For all ART patients, the DV param- and diarrhea) was difficult to separate. Although almost all eters for each of the three VMAT plans were weighted with patients received similar chemotherapy regimens, the fol- lowing PsP classification for CHT regimen was evaluated: K 508 Strahlenther Onkol (2021) 197:505–519 0 = no chemotherapy, 0.5 = capecitabine, 1 = cisplatin. An regression at two timepoints: just before the start of BT and additional factor with the number of completed chemother- at the end of treatment. apy cycles was included as well. Statistics Results All statistical calculations were performed utilizing R From October 2014 until February 2018, 48 patients with (www.R-project.org) [32, 33]. In case multiple parameters median age of 54 (range 31–74) and FIGO stage Ib–IVb, correlated with an effect, e.g., bladder V30Gy and V40Gy, only were included in the analysis of early side effects (Fig. 1). the most significant univariate parameter was selected be- Almost all patients received concomitant weekly cisplatin forehand. We only selected specific parameters that seemed 40 mg/m2, 3 patients received capecitabine (Xeloda, Genen- to be logical into the analyses. Patients with missing values tech Inc, South San Francisco, CA, USA) and four patients for one of the side effects were excluded from the analysis did not receive chemotherapy because of renal insufficiency. only if that specific effect was under evaluation. The number of completed cycles is shown in Fig. 1. In all, For relevant feature selection of DV and PsP, LASSO 22 patients smoked, 7 had quit and 17 patients were non- was used [34]. LASSO (least absolute shrinkage and selec- smokers. tion operator) is a regression analysis method that performs Three patients received 3D-CRT, 31 patients were treated both variable selection and regularization in order to en- with VMAT with fixed margins based on a single CT scan hance the prediction accuracy and interpretability of the and 14 patients with large uterus motion received ART. A statistical model it produces. As the number of incidences total of 18 patients received SIB-VMAT plan and 22 re- was sparse, especially in the more severe categories, higher ceived extended field irradiation (with VMAT). scores were added to the lower category because glmnets’ At the 3-month follow-up (FU), 43 patients had com- LASSO works best for binomial parameters (glmnet is an plete remission, 1 patient incomplete remission, 2 patients R package that fits a generalized linear model via penalized had distant systemic progressive disease and 2 patients were maximum likelihood). lost to long-term follow-up. During further FU beyond this For the final parameters selected by LASSO, multiple study (median 16 months), 2 patients developed local fail- logistic binomial regression was performed with only the ure, 1 pelvic nodal failure, 2 para-aortic nodal failures and selected parameters to maximize the information and to 10 patients developed distant metastases beyond para-aortic minimize the effect of missing values. In case a clear thresh- lymph nodes. old dose was present, a Fisher’s exact test was performed. The significance of DV and PsP for the continuous vari- able weight loss was determined by multivariate stepwise Fig. 2 Boxplots of volumes irra- diated to at least 43 Gy (V43Gy) in cc for the different irradiation techniques (EBRT treatment evolved over time from 3D conformal radiotherapy (3D- CRT), via single plan VMAT (based on EMBRACE I and later EMBRACE II guidelines) to an ART protocol based on a library approach (plan of the day = PotD)). The Robust equiv- alent is provided as a reference for the larger volume the PotD patients would have been irradi- ated to, in case they would not have been adapted every fraction K Strahlenther Onkol (2021) 197:505–519 509 Fig. 3 Distribution of dose–volume parameters. PotD plan of the day. 3DCRT 3D conformal radiotherapy Irradiated EBRT volumes volume variations (nodal SIB, pelvic and para-aortal lymph nodes (PLN)). In Fig. 3, the range and distribution of the As shown in Fig. 2, the volumes exposed to 43 Gy (95% of various dose–volume parameters is depicted. prescribed dose) varied for the different EBRT techniques. Table 1 summarizes the analysis of the DV parameters V43Gy was largest for the few patients receiving 3D-CRT for primary OAR. On average, ART reduced the irradiated (~2500 cm3), somewhat lower for patients enrolled in the normal tissue volumes to 43 Gy by 80 cm3. More specif- EMBRACE I study (~2000 cm3) treated with VMAT, and ically, bowel tissue irradiated to ≥40 Gy was reduced by lowest for patients planned according to EMBRACE II, 17 cm3 and bladder volume irradiated receiving ≥40 Gy by treated with either VMAT or ART (~1800 cm3). In general 5%, respectively. the V43Gy for extended (PAN) fields was larger: ~2250 cm3 vs ~1600 cm3 over the entire patient group. However, for the Patient and physician reported side effects recent techniques, the volume of the extended fields were often smaller than the small fields of the older techniques. The PRO evaluation was found to be more sensitive in terms In only 8% of ART treatment sessions, the (larger) robust of both incidence and grading of toxicity for all side ef- plan was chosen, while 35% of the delivered ART fractions fects, but most prominent for dysuria (see Appendix). Ta- were based on the full bladder plan and 57% for the empty ble 2 and Fig. 4 provide an overview of the side effects bladder plan. In other words, in 92% of all ART fractions that were found to correlate with one or more DV param- V43Gy was smaller than it would have been otherwise. eters. In general, a worsening of symptoms during the first Differences between EBRT treatment techniques and 5 weeks of treatment and recovery afterwards was observed. the underlying target definition protocol (EMBRACE I No G4–G5 toxicity was reported. vs II) were larger and more significant for the higher dose–volume parameters. The volume differences between ART and EMBRACE II patients were often due to target K 510 Strahlenther Onkol (2021) 197:505–519 Table 1 Differences and significance of dose–volume parameters for Appendix). Only for the patients who showed persistent di- the plan of the day patients compared to their robust plan arrhea in the follow-up just after treatment was a dose effect Average Difference p-value (paired t-test) of the Bowel parameter V30Gy obtained, regardless of BT. Sigmoid V40Gy –3% 0.08 CHT also played a role in stool urgency. Almost full recov- Bowel V30Gy –13 cc 0.2 ery was present in the longer follow-up, as for the other pa- Bowel V40Gy –17 cc 0.007 rameters. The significant dose–volume parameter for stool Rectum V30Gy 0% 0.7 urgency was the Bowel V40Gy, with a V40Gy ~ 250 cm3. For Rectum V40Gy –2% 0.08 rectal incontinence, a shallow dose–effect relation was ob- Bladder V30Gy –6% 0.2 served, with very low incidence even at higher irradiated Bladder V40Gy –5% 0.1 volumes; a Rectum V40Gy above 80% increased the (sparse) Body V43Gy –80 cc 0.0001 incidence significantly in combination with CHT. Rectal bleeding seemed to be an EBRT effect as this side ef- fect increased in week 4–5 (4 patients, 8%) and recov- Acute gastrointestinal side effects ered quickly after treatment (6-week and 3-month follow- up) without showing worsening after brachytherapy. Rectal From all gastrointestinal (GI) collected symptoms (nausea, bleeding was absent for Rectum V30Gy < 96%, and increased vomiting, stool consistency, diarrhea, constipation, bloat- to 20% for V30Gy above 96%. Due to the low total number ing, flatulence, stool urgency, rectal tenesmus, rectal mucus, of incidences, this was borderline significant (Fisher’s exact rectal bleeding, rectal incontinence and proctitis) only stool test: p-value = 0.054). consistency, diarrhea, stool urgency, rectal bleeding and rec- tal incontinence revealed dose effects in the multivariate lo- Acute genitourinary side effects gistic regression analysis. Stool consistency was found to be influenced by age, CHT and smoking (Table 2). The major- From all collected genitourinary (GU) side effects (see ity of patients developed diarrhea already in the first weeks Appendix A), only dysuria with its highest incidence in of treatment, which indicates an early RT effect combined week 5 (18 patients, 38%) and bladder incontinence at the with CHT and smoking with the accumulated received dose time points week 6 and 1 week after treatment completion being still small. During the treatment a slight recovery was (14 patients, 29%) remained significant in the multivari- observed (for the dynamics of side effects over time, see ate logistic regression. Both symptoms were influenced by Table 2 Overview of dose effects that were selected by LASSO are marked with “+” and their (lowest) significance (in numbers) obtained from subsequent multivariate logistic regression analysis Side effect Report Grade Bladder Rectum Rectum Bowel Body Alcohol CHT SmokingAge Vomiting Time V40Gy V30Gy V40Gy V40Gy V43Gy point Stool CTCAE ≥1 – – + 0.015 – – – – + – W5 urgency PROa ≥2 – – – 0.02 – – – + + – Stool con- – Fluid – 0.009 – 0.015 – – – – – – W5 sistency stool Diarrhea PRO ≥1 – – – 0.008 – – + + – – Fu1w Rectal in- CTCAE ≥1 – – 0.006 + – – + – – – W6 continence PRO – – 0.0006 – – – 0.002 – – – Rectal CTCAE ≥1 – 0.05 – – – – – – – – W5 bleedingb PRO Dysuria PRO ≥1 0.03 – – – – – + + – – W6 Bladder in- CTCAE ≥1 + – – – – + + 0.0005 0.0003 – Fu6w continence PROa 0.001 – – – – + + + 0.003 – Weight loss Mea- Con- – – – – 0.0007 0.01 – – – – Before sured tinu- brachy ously – – – – 0.006 0.02 – – – + After brachy No significant correlation between number of chemotherapy cycles and any of the general parameters W5 week 5, W6 week 6, Fu1w Follow-up 1 week after finishing of the treatment, Fu6w Follow-up 6 weeks after finishing of the treatment, CHT chemotherapy, PRO patient-reported outcome, CTCAE Common Terminology Criteria for Adverse Events a Data available only for patients after October 2015 b Sparse data but clear cutoff at RectumV30Gy of 96%: below: no incidence K Strahlenther Onkol (2021) 197:505–519 511 Fig. 4 Dose–volume effect rela- tions for CTCAE gastrointesti- nal (upper graphs) and patient reported genitourinary (lower graphs) parameters. The dose intervals are divided (binned) in 3 equal parts, the points are plotted at the average dose value of each bin and represent the ratio of patients with and with- out symptoms. The error bars are the 68% confidence intervals (Chi-square) and depend on the number of patients in the bin. The black dots are the individual patients without symptoms; red dots are patients with symptoms. The logistic regression uses all data at once, so the binning does not influence the significance and was done only for display purposes CHT, smoking and age and in case of bladder incontinence, Discussion also by alcohol use (Table 2). Despite the low incidence of bladder incontinence, a clear Bladder V40Gy dependency The EBRT treatment techniques in cervix cancer radiother- (consistent between questionnaires) was observed. For blad- apy have evolved during the decades. Several studies com- der incontinence, the TD50 of V40Gy was obtained for an pared 2D, 3D and IMRT technique, providing the same irradiated volume of 80–90%. local control and overall survival with significantly reduced gastrointestinal and genitourinary toxicity in IMRT treat- Acute general side effects ment group [35–37]. The current pilot study was designed as a prospective Just before the start of BT, there was a significant cor- observational and hypothesis generating one, with the main relation between weight loss and the EBRT body V43Gy objective to explore correlations between acute toxicity and (Table 2). After the first BT fraction, weight loss increased DVH parameters in this context. These correlations can later rapidly with 4 kg on average, due to limited oral intake dur- be used to reduce early effects as much as possible and ing BT (which implies over 48 h parenteral nutrition) and to avoid consequential late side effects, since some acute the correlation of weight loss and EBRT dose became less symptoms can persist over a longer period of time and be- prominent. Lumbar pain recovered during treatment com- come chronic. Despite the limited number of patients and pared to baseline, with the lowest incidence in week 4; low frequency of reported acute side effects, it was possi- afterwards it increased after epidural anesthesia and bed ble to isolate dose-effects in GU and GI toxicity and some rest during the days of BT. The recovery of lumbar pain patient specific factors. Within this cohort there were differ- in the follow-up should be evaluated with caution, as some ent types of treatment planning, i.e., 3D-CRT, single plan confounding factors need be taken into account such as VMAT and ART. This heterogeneity in treatment fields and lymphadenectomy, epidural anesthesia, bed rest during the nodal dose delivered, provided a larger range of dose pa- days of BT. rameters than a study of each single technique would have done. The evolution of EBRT treatment techniques in our center over time is clearly reflected in clinical results of K 512 Strahlenther Onkol (2021) 197:505–519 this study. As the variables “treatment technique” and “ir- Most of significant parameters were found at timepoints radiated volume” are correlated and the subgroups were before BT. For the parameters assessed at a later time- small, the clinical results obtained did not reveal a dif- point (i.e., Bladder incontinence at the 6-week follow-up) ference in toxicity between VMAT-ART and VMAT. On the EBRT dose effect was highly significant even without the other hand, there was a selection bias since VMAT- considering BT dose. ART was intentionally limited to patients classified as The influence of the chemotherapy regime is difficult “movers” with a more complicated and changing anatomy. to assess as almost all patients got cisplatin in this study. Another contributing factor was target definition, mainly Both the PsP parameter CHT or the number of completed in terms of concepts (EMBRACE-II protocol). A similar chemotherapy cycles seem to influence the severity of trend in treated volumes was observed recently in the evo- side effects a little, but did not remain significant in the lution of EMBRACE-I/-II study when evaluating treatment multivariate analysis (except for rectal incontinence). Prior technique, treatment protocol and para-aortic irradiation studies reported on acute side effects with and without [38]. chemotherapy. Morbidity was described as elevated [44] or Despite all technological advancements, accurate assess- similar [45] in the combination treatment. ment of the actually delivered doses remains challenging. For patients undergoing cervix cancer treatments involv- Because of daily differences in OARs volumes, DVHs of ing radiation therapy, numerous other studies related to treatment plans are not representative for the whole treat- QLQ evaluation were published [8, 18, 20–22]. The most ment. Therefore, the analysis of the V43Gy is probably a bet- reported GU symptoms were urinary frequency, cystitis and ter means to assess treatment plans for EBRT of cervix incontinence [9]; the most reported GI symptoms were di- cancer, as OARs move to different dose regions. In the cur- arrhea, stool urgency and rectal incontinence [10]. The ma- rent study V43Gy correlated linearly with weight loss during jority of patients treated in those studies received 3D-CRT, the treatment. i.e. 74% in [8], 63% in [10] and [9]. In an IMRT study EBRT of cervix cancer with its rather challenging target comprising 50 patients treated with 45–50 Gy diarrhea was volume, surrounded by critical OARs that have a direct im- the most common symptom and the cut-off point for small pact on QoL, is one of the main clinical sites where ART bowel ≥G2 toxicity was V45Gy > 150 cm3 (65% vs. 33% in is explored. However, there is only a limited number of V45Gy < 150 cm3) [46]. Although toxicity was evaluated by studies regarding prospective assessment of physician and RTOG scale, the results concur with QUANTEC constraints patient reported outcome of acute morbidity in this context V45Gy < 195 cm3 [47]. [14, 39]. Although toxicity was evaluated with standard- In our cohort, where 94% received VMAT and 31% ized questionnaires, a considerable discordance was found VMAT-ART, low grade diarrhea, stool urgency, rectal for patient reported vs physician reported outcome, mainly incontinence, bleeding from GI and dysuria and bladder in dysuria evaluation. This should be taken into account incontinence from GU symptoms were the most common because side effects evaluated by physicians as low grade, toxicities. Only stool urgency, rectal and bladder inconti- could have an important impact on QoL [40]. nence and weight loss showed a significantly increasing When assessing patient reported outcomes, it is of utmost distribution of incidences over the entire dose range (Fig. 3) importance to perform a baseline assessment as many pa- Furthermore, a correlation between dysuria with Bladder tients report GI, GU or general symptoms before treatment V40Gy, stool urgency and incontinence with Bowel and (Appendix, Table B.1; [8]). The incidence of GI symptoms Rectum V40Gy was observed. Based on these data, the fol- was already two times higher in the second week of treat- lowing treatment planning objectives are recommended to ment, which can be explained as a direct effect of small minimize stool urgency, rectal and urinary incontinence: intestines exposure, changes in irrigation, epithelial atro- bowel V40Gy ≤ 250 cm3, rectum V40Gy ≤ 80% and bladder phy and chemotherapy [41–43], and not just by the irradi- V40Gy ≤ 80–90%, respectively. These results concur reason- ated volume. The incidence peak was in week 5–6 for the ably with planning aims in EMBRACE-II [29], which were majority of the symptoms; however almost complete recov- derived from a cohort of representative dose distributions. ery was present at the 3 month follow-up. These findings With ART, less body volume was irradiated to mid and concur with data published by Heijkoop et al. [14]. low dose volume levels than with the corresponding robust From the general symptoms, lumbar pain seems to have plan. It can be speculated that this could contribute to mor- a dose effect in multivariate analysis, but some confounding bidity reduction. However, para-aortic extended fields may factors need to be taken into account. For example, if pa- have a larger impact on normal tissue exposure than the tients report lumbar pain at the baseline, it could be a result additional organ sparing through ART. Therefore, a future of lymphadenectomy or the tumor itself. comparison of adaptive and nonadaptive techniques should The BT as confounding factor in DVH evaluation of all include a correction for extended field irradiation. GI, GU and general symptoms was not taken into account. K Strahlenther Onkol (2021) 197:505–519 513 The ongoing EMBRACE-II study aims to further reduce morbidity by reducing the irradiated volume through re- duced PTV margins and IMRT/VMAT following dose vol- ume constraints for OARs. The EMBRACE-II study with a considerable patient cohort can certainly contribute to answer this clinical and scientific question and provide cor- relations with late morbidity, including a novel approach to distinguish between late transient and long-lasting side effects [48]. Conclusion This prospective study demonstrated the impact of treat- ment technique on quality of life (QoL) of cervix cancer patients undergoing radiochemotherapy. In general, a peak of impaired QoL was observed in treatment weeks 5–6, al- though this was transient with a recovery at the 3-month fol- low-up. The following external beam radiotherapy (EBRT) planning objectives may be used to reduce early morbidity: bowel V40Gy < 250 cm3 for stool urgency, rectum V40Gy < 80% for rectal incontinence and bladder V40Gy < 80–90% for uri- nary incontinence. Open Access This article is licensed under a Creative Commons At- tribution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, pro- vide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/. Appendix A Collected and evaluated side effects GI: stool consistency, diarrhea, constipation, bloating, flat- ulence, stool urgency, rectal tenesmus, rectal mucus, rectal hemorrhage, rectal Incontinence, proctitis. GU: micturition during the day, nocturia, urinary ur- gency, bladder tenesmus, bladder bleeding, dysuria, bladder incontinence and cystitis. General side effects: weight loss, fatigue, insomnia, limb edema, tinnitus, paresthesia, back pain, pelvic pain. Dynamics of those side effects that were analyzed and were associated with a dose-effect K 514 Strahlenther Onkol (2021) 197:505–519 Fig. A.1 Dynamics of side ef- fects K Strahlenther Onkol (2021) 197:505–519 515 Fig. A.1 (Continued) K 516 Strahlenther Onkol (2021) 197:505–519 Fig. A.1 (Continued) K Strahlenther Onkol (2021) 197:505–519 517 Appendix B Acknowledgements The financial support by the (Austrian) Federal Ministry for Digital and Economic Affairs and the National Founda- tion for Research, Technology and Development is gratefully acknowl- Table B.1 Side effects (>grade 1) present before start of treatment edged. None or few 10–50% of the pa- >50% of the pa- Funding The financial support by the (Austrian) Federal Ministry for tients tients Digital and Economic Affairs and the National Foundation for Re- Edema limbs Pelvic pain Fatigue search, Technology and Development is gratefully acknowledged. Tinnitus Constipation Insomnia Funding Open access funding provided by Medical University of Vi- Paresthesia Micturition daytime Back pain enna. Nausea Micturition night- Flatulence time Vomiting Dysuria Vaginal discharge Declarations Stool consistency Bladder inconti- Conflict of interest Y. Seppenwoolde, K. Majercakova, M. Buschmann, nence E. Dörr, A.E. Sturdza, M.P. Schmid, R. Pötter and D. Georg declare Diarrhea Vaginal pain that they have no competing interests. Stool urgency Vaginal hemorrhage Ethical standards The trial was registered at ethics committee of the Rectal tenesmus Vaginal inflamma- Medical University of Vienna under the EK-Nr: 2009/2014. All indi- tion viduals gave consent to participate. Rectal mucus Bloating Rectal hemorrhage Rectal incontinence References Proctitis Urinary urgency 1. Cibula D, Pötter R, Planchamp F, Avall-Lundqvist E, Fischerova D, Bladder spasm/ Haie Meder C et al (2018) The European society of gynaecolog- tenesmus ical oncology/European society for radiotherapy and oncology/ European society of pathology guidelines for the management Bladder hemorrhage of patients with cervical cancer. Radiother Oncol 127:404–416. Cystitis https://doi.org/10.1016/j.radonc.2018.03.003 Vaginal mucositis 2. Du XL, Tao J, Sheng XG, Lu CH, Yu H, Wang C et al (2012) In- tensity-modulated radiation therapy for advanced cervical cancer: a comparison of dosimetric and clinical outcomes with conventional radiotherapy. Gynecol Oncol 125:151–157. https://doi.org/10.1016/ j.ygyno.2011.12.432 3. Gandhi AK, Sharma DN, Rath GK, Julka PK, Subramani V, Sharma S et al (2013) Early clinical outcomes and toxicity of intensity modulated versus conventional pelvic radiation therapy for locally advanced cervix carcinoma: a prospective randomized study. Int J Radiat Oncol Biol Phys 87:542–548. https://doi.org/10. 1016/j.ijrobp.2013.06.2059 4. van de Bunt L, Jürgenliemk-Schulz IM, de Kort GA, Roesink JM, Tersteeg RJ, van der Heide UA (2008) Motion and deformation of the target volumes during IMRT for cervical cancer: what margins do we need? Radiother Oncol 88:233–240. https://doi.org/10.1016/ j.radonc.2007.12.017 5. Taylor A, Powell ME (2008) An assessment of interfractional uterine and cervical motion: implications for radiotherapy tar- get volume definition in gynaecological cancer. Radiother Oncol 88:250–257. https://doi.org/10.1016/j.radonc.2008.04.016 6. Beadle BM, Jhingran A, Salehpour M, Sam M, Iyer RB, Eifel PJ (2009) Cervix regression and motion during the course of external beam chemoradiation for cervical cancer. Int J Radiat Oncol Biol Phys 73:235–241. https://doi.org/10.1016/j.ijrobp.2008.03.064 7. Chan P, Dinniwell R, Haider MA, Cho YB, Jaffray D, Lockwood G et al (2008) Inter- and intrafractional tumor and organ movement in patients with cervical cancer undergoing radiotherapy: a cine- matic-MRI point-of-interest study. Int J Radiat Oncol Biol Phys 70:1507–1515. https://doi.org/10.1016/j.ijrobp.2007.08.055 8. Kirchheiner K, Pötter R, Tanderup K, Lindegaard JC, Haie- Meder C, Petrič P et al (2016) Health-related quality of life in locally advanced cervical cancer patients after definitive chemora- diation therapy including image guided adaptive brachytherapy: an K 518 Strahlenther Onkol (2021) 197:505–519 analysis from the EMBRACE study. Int J Radiat Oncol Biol Phys Oncol Biol Phys 73:1501–1509. https://doi.org/10.1016/j.ijrobp. 94:1088–1098. https://doi.org/10.1016/j.ijrobp.2015.12.363 2008.06.1905 9. Jensen NBK, Pötter R, Kirchheiner K, Fokdal L, Lindegaard JC, 23. Bentzen SM, Constine LS, Deasy JO, Eisbruch A, Jackson A, Kirisits C et al (2018) Bowel morbidity following radiochemother- Marks LB et al (2010) Quantitative analyses of normal tissue ef- apy and image-guided adaptive brachytherapy for cervical cancer: fects in the clinic (QUANTEC): an introduction to the scientific physician- and patient reported outcome from the EMBRACE issues. Int J Radiat Oncol Biol Phys 76:S3–9. https://doi.org/10. study. Radiother Oncol 127:431–439. https://doi.org/10.1016/j. 1016/j.ijrobp.2009.09.040 radonc.2018.05.016 24. Xu B, Guo Y, Chen Y, Lu H, Tang T, Yue Z et al (2015) Is the 10. Fokdal L, Pötter R, Kirchheiner K, Lindegaard JC, Jensen NBK, irradiated small bowel volume still a predictor for acute lower Kirisits C et al (2018) Physician assessed and patient reported uri- gastrointestinal toxicity during preoperative concurrent chemo- nary morbidity after radio-chemotherapy and image guided adap- radiotherapy for rectal cancer when using intensity-modulated tive brachytherapy for locally advanced cervical cancer. Radiother radiation therapy? Radiat Oncol 10:257. https://doi.org/10.1186/ Oncol 127:423–430. https://doi.org/10.1016/j.radonc.2018.05.002 s13014-015-0566-6 11. Ahmad R, Hoogeman MS, Bondar M, Dhawtal V, Quint S, De 25. Lee TF, Huang EY (2014) The different dose-volume effects of Pree I et al (2011) Increasing treatment accuracy for cervical can- normal tissue complication probability using lasso for acute small- cer patients using correlations between bladder-filling change and bowel toxicity during radiotherapy in gynecological patients with cervix-uterus displacements: proof of principle. Radiother Oncol or without prior abdominal surgery. Biomed Res Int 2014:143020. 98:340–346. https://doi.org/10.1016/j.radonc.2010.11.010 https://doi.org/10.1155/2014/143020 12. Bondar L, Hoogeman M, Mens JW, Dhawtal G, de Pree I, Ah- 26. Viswanathan AN, Yorke ED, Marks LB, Eifel PJ, Shipley WU mad R et al (2011) Toward an individualized target motion manage- (2010) Radiation dose-volume effects of the urinary bladder. Int ment for IMRT of cervical cancer based on model-predicted cervix- J Radiat Oncol Biol Phys 76:S116–22. https://doi.org/10.1016/j. uterus shape and position. Radiother Oncol 99:240–245. https://doi. ijrobp.2009.02.090 org/10.1016/j.radonc.2011.03.013 27. Thor M, Olsson C, Oh JH, Petersen SE, Alsadius D, Bentzen L 13. Heijkoop ST, Langerak TR, Quint S, Bondar L, Mens JW, Heij- (2016) Urinary bladder dose-response relationships for patient- men BJ et al (2014) Clinical implementation of an online adap- reported genitourinary morbidity domains following prostate can- tive plan-of-the-day protocol for nonrigid motion management in cer radiotherapy. Radiother Oncol 119:117–122. https://doi.org/10. locally advanced cervical cancer IMRT. Int J Radiat Oncol Biol 1016/j.radonc.2016.01.013 Phys 90:673–679. https://doi.org/10.1016/j.ijrobp.2014.06.046 28. Serrano NA, Kalman NS, Anscher MS (2017) Reducing rectal in- 14. Heijkoop ST, Nout RA, Quint S, Mens JWM, Heijmen BJM, jury in men receiving prostate cancer radiation therapy: current per- Hoogeman MS (2017) Dynamics of patient reported quality of life spectives. Cancer Manag Res 9:339–350. https://doi.org/10.2147/ and symptoms in the acute phase of online adaptive external beam CMAR.S118781 radiation therapy for locally advanced cervical cancer. Gynecol 29. EMBRACE (2020) EMBRACE studies and EMBRACE research. Oncol 147:439–449. https://doi.org/10.1016/j.ygyno.2017.08.009 https://www.embracestudy.dk/. Accessed 17 May 2020 15. Buschmann M, Majercakova K, Sturdza A, Smet S, Najjari D, 30. National Cancer Institute (2010) Common terminology criteria for Daniel M et al (2018) Image guided adaptive external beam ra- adverse events (CTCAE ) (Version 4.0) diation therapy for cervix cancer: evaluation of a clinically im- 31. Lewis SJ, Heaton KW (1997) Stool form scale as a useful guide to plemented plan-of-the-day technique. Z Med Phys 28:184–195. intestinal transit time. Scand J Gastroenterol 32:920–924. https:// https://doi.org/10.1016/j.zemedi.2017.09.004 doi.org/10.3109/00365529709011203 16. Dörr W, Hendry JH (2001) Consequential late effects in normal tis- 32. R core team (2017) R: A language and environment for statistical sues. Radiother Oncol 61:223–231. https://doi.org/10.1016/S0167- computing. R Foundation of Statistical Computing, Vienna, Austria 8140(01)00429-7 33. Friedman AJ, Hastie T, Simon N, Tibshirani R, Hastie MT (2020) 17. Wang CJ, Leung SW, Chen HC, Sun LM, Fang FM, Huang E et al Lasso and elastic-net regularized generalized linear models. https:// (1998) The correlation of acute toxicity and late rectal injury in cran.r-project.org/web/packages/glmnet/glmnet.pdf. Accessed 17 radiotherapy for cervical carcinoma: evidence suggestive of conse- May 2020 quential late effect (CQLE). Int J Radiat Oncol Biol Phys 40:85–91. 34. Friedman J, Hastie T, Tibshirani R (2010) Regularization paths https://doi.org/10.1016/S0360-3016(97)00560-9 for generalized linear models via coordinate descent. J Stat Softw 18. Park SY, Bae DS, Nam JH, Park CT, Cho CH, Lee JM et al (2007) 33:1–22. https://doi.org/10.18637/jss.v033.i01 Quality of life and sexual problems in disease-free survivors of 35. Gandhi AK, Sharma DN, Rath GK, Julka PK, Subramani V, cervical cancer compared with the general population. Cancer Sharma S, Manigandan D, Laviraj MA, Kumar S, Thulkar S (2013) 110:2716–2725. https://doi.org/10.1002/cncr.23094 Early clinical outcomes and toxicity of intensity modulated versus 19. JKirwan JM, Symonds P, Green JA, Tierney J, Collingwood M, conventional pelvic radiation therapy for locally advanced cervix Williams CJ (2003) A systematic review of acute and late toxicity carcinoma: a prospective randomized study. Int J Radiat Oncol Biol of concomitant chemoradiation for cervical cancer. Radiother On- Phys 87:542–548. https://doi.org/10.1016/j.ijrobp.2013.06.2059 col 68:217–226. https://doi.org/10.1016/S0167-8140(03)00197-X 36. Naik A, Gurjar OP, Gupta KL, Singh K, Nag P, Bhandari V (2016) 20. Bjelic-Radisic V, Jensen PT, Vlasic KK, Waldenstrom AC, Singer S, Comparison of dosimetric parameters and acute toxicity of inten- Chie W (2012) Quality of life characteristics inpatients with cervi- sity-modulated and threedimensional radiotherapy in patients with cal cancer. Eur J Cancer 48:3009–3018. https://doi.org/10.1016/j. cervix carcinoma: a randomized prospective study. Cancer Radio- ejca.2012.05.011 ther 20:370–376. https://doi.org/10.1016/j.canrad.2016.05.011 21. Ferrandina G, Mantegna G, Petrillo M, Fuoco G, Venditti L, 37. Lin Y, Chen K, Lu Z, Zhao L, Tao Y, Ouyang Y, Cao X (2018) In- Terzano S (2012) Quality of life and emotional distress in early tensity-modulated radiation therapy for definitive treatment of cer- stage and locally advanced cervical cancer patients: a prospective, vical cancer: a meta-analysis. Radiat Oncol 13:177. https://doi.org/ longitudinal study. Gynecol Oncol 124:389–394. https://doi.org/ 10.1186/s13014-018-1126-7 10.1016/j.ygyno.2011.09.041 38. Berger T, Seppenwoolde Y, Pötter R, Assenholt MS, Lindegaard JC, 22. Korfage IJ, Essink-Bot ML, Mols F, van de Poll-Franse L, Kruit- Nout RA et al (2019) Importance of technique, target selection, wagen R, van Ballegooijen M (2009) Health-related quality of life contouring, dose prescription, and dose-planning in external beam in cervical cancer survivors: a population-based survey. Int J Radiat radiation therapy for cervical cancer: evolution of practice from K Strahlenther Onkol (2021) 197:505–519 519 EMBRACE-I to II. Int J Radiat Oncol Biol Phys 104:885–894. pared with pelvic and para-aortic radiation for high-risk cervical https://doi.org/10.1016/j.ijrobp.2019.03.020 cancer. N Engl J Med 340:1137–1143. https://doi.org/10.1056/ 39. Reijtenbagh DMW, Godart J, Mens JWM, Heijkoop ST, Heemsber- NEJM199904153401501 gen WD, Hoogeman MS (2020) Patient-reported acute GI symp- 45. Saad A, Lo SS, Han I, Keole S, Lee CP, Tekyi-Mensah S et al toms in locally advanced cervical cancer patients correlate with (2004) Radiation therapy with or without chemotherapy for cervi- rectal dose. Radiother Oncol 148:38–43. https://doi.org/10.1016/j. cal cancer with periaortic lymph node metastasis. Am J Clin Oncol radonc.2020.03.035 27:256–263. https://doi.org/10.1097/01.coc.0000092564.16409.cd 40. Kirchheiner K, Nout R, Lindegaard J, Petrič P, Limbergen EV, 46. Simpson DR, Song WY, Moiseenko V, Rose BS, Yashar CM, Jürgenliemk-Schulz IM et al (2012) Do clinicians and patients Mundt AJ et al (2012) Normal tissue complication probability agree regarding symptoms? A comparison after definitive ra- analysis of acute gastrointestinal toxicity in cervical cancer patients diochemotherapy in 223 uterine cervical cancer patients. Strahlen- undergoing intensity modulated radiation therapy and concurrent ther Onkol 188:933–939. https://doi.org/10.1007/s00066-012- cisplatin. Int J Radiat Oncol Biol Phys 83:e81–6. https://doi.org/10. 0156-6 1016/j.ijrobp.2011.12.012 41. Tanderup K, Georg D, Pötter R, Kirisits C, Grau C, Linde- 47. Kavanagh BD, Pan CC, Dawson LA, Das SK, Li XA, Ten gaard JC (2010) Adaptive management of cervical cancer ra- Haken RK et al (2010) Radiation dose-volume effects in the stom- diotherapy. Semin Radiat Oncol 20:121–129. https://doi.org/10. ach and small bowel. Int J Radiat Oncol Biol Phys 76:S101–7. 1016/j.semradonc.2009.11.006 https://doi.org/10.1016/j.ijrobp.2009.05.071 42. Shadad AK, Sullivan FJ, Martin JD, Egan LJ (2013) Gastroin- 48. Kirchheiner K, Pötter R, Nout RA, Schwartz-Vittrup A, Holzner B, testinal radiation injury: symptoms, risk factors and mechanisms. Bentzen SM et al (2020) Late, persistent, substantial, treat- World J Gastroenterol 19:185–198. https://doi.org/10.3748/wjg. ment-related symptoms after radiation therapy (LAPERS): a new v19.i2.185 method for longitudinal analysis of late morbidity—applied in the 43. Andreyev J (2005) Gastrointestinal complications of pelvic radio- EMBRACE study. Int J Radiat Oncol Biol Phys 106:300–309. therapy: are they of any importance? Gut 54:1051–1054. https:// https://doi.org/10.1016/j.ijrobp.2019.10.027 doi.org/10.1136/gut.2004.062596 44. Morris M, Eifel PJ, Lu J, Grigsby PW, Levenback C, Stevens RE et al (1999) Pelvic radiation with concurrent chemotherapy com- K