A Systematic Post-QUANTEC Review of Tolerance Doses for ...

International Journal of
Radiation Oncology
biology physics

www.redjournal.org

CME

Physics Contribution

A Systematic Post-QUANTEC Review of ToleranceDoses for Late Toxicity After Prostate CancerRadiation TherapyCaroline E. Olsson, PhD,*,y Andrew Jackson, PhD,z

Joseph O. Deasy, PhD,z and Maria Thor, PhDz

*Department of Radiation Physics, Institute of Clinical Sciences, Sahlgrenska Academy at Universityof Gothenburg, Gothenburg, Sweden; yRegional Cancer Center West, Western Sweden HealthcareRegion, Gothenburg, Sweden; and zDepartment of Medical Physics, Memorial Sloan Kettering CancerCenter, New York, New York

Received Mar 27, 2018, and in revised form Jul 27, 2018. Accepted for publication Aug 4, 2018.

Summary

From a total of 33 studies(N Z 10,149) this post-QUANTEC systematic re-view has identified noveldose/volume tolerances for17 distinct gastrointestinal(GI), genitourinary, and sex-ual dysfunction symptomsfollowing radiation therapy(RT) for prostate cancer.Synthesized dose/volumetolerances were assessedacross studies for Defecationurgency, Diarrhea, Fecal in-continence, Proctitis, andRectal bleeding post

NotedAn online CME test for this article

academy.astro.org.

Reprint requests to: Maria Thor, PhD, Depar

Memorial Sloan Kettering Cancer Center, 485 L

NY 10017. Tel: (646) 888-8013; E-mail: thorm

This research was funded in part throug

HealtheNational Cancer Institute Cancer Ce

CA008748 (M.T., A.J., and J.O.D.). Grants from

Clinic Cancer Foundation in Gothenburg, the Sw

Research, the Kamprad Family Foundation for E

Int J Radiation Oncol Biol Phys, Vol. 102, No.

0360-3016/� 2018 Elsevier Inc. All rights rese

https://doi.org/10.1016/j.ijrobp.2018.08.015

Downloaded for Anonymous User (n/a) at CONSORFor personal use onl

Purpose: The aims of this study were to systematically review tolerance doses for latedistinct gastrointestinal (GI), genitourinary (GU), and sexual dysfunction (SD) symp-toms after external beam radiation therapy (EBRT) alone and treatments involvingbrachytherapy (BT) for prostate cancer after Quantitative Analysis of Normal TissueEffects in the Clinic (QUANTEC) and ultimately to perform quantitative syntheses ofidentified dose/volume tolerances represented by dose-volume histogram (DVH)thresholds, that is, statistically significant (P � .05) cutoff points between symptomaticand asymptomatic patients in a certain study.Methods and Materials: PubMed was scrutinized for full-text articles in Englishafter QUANTEC (January 1, 2010). The inclusion criteria were randomizedcontrolled trials, case-control studies, or cohort studies with tolerance doses for latedistinct symptoms �3 months after primary radiation therapy for prostate cancer(N> 30). All DVH thresholds were converted into equivalent doses in 2-Gy fractions(EQD2a/b) and were fitted with a linear or linear-quadratic function (goodness of fit,R2). The review was registered on PROSPERO (CRD42016042464).Results: From 33 identified studies, which included 36 to 746 patients persymptom domain, the majority of dose/volume tolerances were derived for GItoxicity after EBRT alone (GI, 97 thresholds; GU, 8 thresholds; SD, 1 threshold).

can be taken at https://

tment of Medical Physics,

exington Ave, New York,

@mskcc.org

h National Institutes of

nter Support Grant P30

the King Gustaf V Jubilee

edish Society for Medical

ntrepreneurship Research

& Charity, and the Research and Development Council of West Region

County (Sweden) supported C.E.O.

Conflict of interest: none.

Supplementary material for this article can be found at https://doi.org/

10.1016/j.ijrobp.2018.08.015.

AcknowledgmentsdThe authors acknowledge Konstantina Matsoukas

at Memorial Sloan Kettering Cancer Center for conducting all initial

literature searches. They thank all authors of individually published studies

who clarified essential information on request.

5, pp. 1514e1532, 2018

rved.

TIUM MEDICAL LIBRARIES - ISRAEL -Medical Center Rambam from ClinicalKey.com by Elsevier on June 12, 2019.y. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

Delta:1_given name

Delta:1_surname

Delta:1_given name

Delta:1_surname

https://academy.astro.org

https://academy.astro.org

mailto:[email protected]



http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijrobp.2018.08.015&domain=pdf


http://www.redjournal.org

Volume 102 � Number 5 � 2018 Post-QUANTEC prostate cancer review 1515

external-beam RT (EBRT)

D

and also across these symp-toms leading to a curve foroverall GI toxicity(R2 Z 0.98). Restrictingtreatments for EBRT usingthe lower boundaries of thiscurve are likely to limitoverall GI toxicity. Furtherinvestigative initiatives areencouraged to untanglesimilar effects for symptomsof the remaining domains.

ownloaded for Anonymous User (n/a) at CONSORFor personal use on

For 5 symptoms (defecation urgency, diarrhea, fecal incontinence, proctitis, andrectal bleeding), relationships between dose/volume tolerances acrossstudies (R2 Z 0.93 [0.82-1.00]), and across symptoms, leading to a curve foroverall GI toxicity (R2 Z 0.98), could be determined. For these symptoms,mainly rectal thresholds were found throughout low and high doses(10 Gy � equivalent dose in 2-Gy fractions using a/b Z 3Gy (EQD23) � 50 Gyand 55 Gy � EQD23 � 78 Gy, respectively). For BT with or without EBRT,dose/volume tolerances were also mainly identified for GI toxicity (GI, 14 thresh-olds; GU, 4 thresholds; SD, 2 thresholds) with the largest number of DVH thresh-olds concerning rectal bleeding (5 thresholds).Conclusions: Updated dose/volume tolerances after QUANTEC were found for 17GI, GU, or SD symptoms. A DVH curve described the relationship between dose/volume tolerances across 5 GI symptoms after EBRT alone. Restricting treatmentsfor EBRT alone using the lower boundaries of this curve is likely to limit overall GItoxicity, but this should be explored prospectively. Dose/volume tolerances for GUand SD toxicity after EBRT alone and after BT with or without EBRT were scarceand support further research including data-sharing initiatives to untangle the dose/volume relationships for these symptoms. � 2018 Elsevier Inc. All rights reserved.

Introduction

Over the past 3 decades, 2 major efforts have beenconducted to better understand radiation therapy (RT)dose/volume tolerances for organs at risk (OARs) withthe overall aim to limit the occurrence of severe latetoxicity. The first consensus values were proposed in1991.1 In 2010, the Quantitative Analysis of NormalTissue Effects in the Clinic (QUANTEC) group presentedevidence-based results made available from the accu-mulation of 3-dimensional (3D) treatment planning data.2

For prostate cancer, these efforts summarize dose/volumetolerances for OARs after external beam RT (EBRT) forcommonly reported gastrointestinal (GI) and genitouri-nary (GU) toxicities and sexual dysfunction (SD).1,3-5

However, dose/volume tolerances for distinct symptomswere typically not distinguished because these recom-mendations were based on toxicity scoring systems thatcombine symptoms (eg, by the Radiation TherapyOncology Group/European Organisation for Researchand Treatment of Cancer6 and National Cancer InstituteCommon Terminology Criteria for Adverse Events7

systems). Furthermore, these QUANTEC reports on GIand GU toxicity and SD were derived for only the wholerectum, bladder, and penile bulb, respectively, and didnot explicitly focus on dose/volume tolerances aftertreatments involving brachytherapy (BT).3-5

Since the publication of the 3 QUANTEC reports,dose/volume tolerances for late and distinct symptomsand other OARs have become available. The objective ofthis systematic review was to summarize dose/volumetolerances between distinct late GI and GU toxicities, SDsymptoms, and relevant OARs after prostate cancer RTseparately for EBRT alone and for BT with or withoutEBRT.

TIUM MEDICAL LIBRARIES - ISRly. No other uses without permission.

Methods and Materials

Search strategy

This systematic review adheres to the Preferred ReportingItems for Systematic Reviews and Meta-Analyses state-ment,8 and protocol details were registered on PROSPERO(http://www.crd.york.ac.uk/PROSPERO/display_record.asp?IDZCRD42016042464). The PubMed database wasscrutinized for full-text articles in English published up toDecember 31, 2017, focusing on studies reporting dose/volume tolerances for late toxicities only after prostatecancer RT. For GI toxicity and SD after EBRT alone, thesearch interval started on January 1, 2010, immediatelyafter the publication of the QUANTEC reports.3,5 Becausethe QUANTEC reports focused neither on GU toxicity afterprostate cancer RT nor on the effects after BT,4 the searchinterval for GU toxicity after EBRT alone and for GI andGU toxicity, as well as SD, after BT with or without EBRTstarted after the publication of the report by Emami et al1

(i.e., January 1, 1992). The search strategy is detailed inAppendix E1 (available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).

Study selection criteria

Identified abstracts and full-text articles were indepen-dently screened and cross-checked by 2 investigators (MTand CEO). Studies qualifying for inclusion were case-control and cohort studies, as well as randomizedcontrolled trials, all reporting on distinct late symptomsassessed �3 months after completion of photon-based RTadhering to typically applied regimens for localizedprostate cancer (“Publication Type” according to thePopulation, Intervention, Comparison, and Outcome

AEL -Medical Center Rambam from ClinicalKey.com by Elsevier on June 12, 2019.Copyright ©2019. Elsevier Inc. All rights reserved.

http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42016042464

http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42016042464



Olsson et al. International Journal of Radiation Oncology � Biology � Physics1516

Worksheet9). The exclusion criteria were (1) articles not inEnglish or unavailable in full text; (2) no data on therelationship between OAR doses and symptoms (with anexception for GU toxicity, in which prostate dose wasaccepted as a surrogate for prostatic urethral dose); (3)reanalysis not adding to previously published data; (4)fewer than 30 patients included; (5) only a combination ofsymptom “scores” reported; (6) acute toxicity or combinedacute and late toxicity; and (7) data not interpretable evenafter the authors of individual studies were contacted.

To report study quality, we used an 8-criterion assess-ment, including descriptions of study individuals, follow-up, toxicity scoring, and statistics (yielding a score of 0-8,with >6 regarded as good quality; 4-6, fair quality; and <4,poor quality).10 Studies that assessed symptoms usingretrospective reviews instead of prospective collections (eg,case-report forms or questionnaires) were penalized bysubtracting 1 from the total score.

Dose-volume harmonization

Correction for fractionation effectsFor each study, dose/volume tolerances were assessed fromdose-volume histogram (DVH) thresholds. A threshold isdefined as a statistically significant dose/volume cut pointdistinguishing symptomatic from asymptomatic patients.Doses were converted into equivalent doses in 2-Gy frac-tions (EQD2a/b)

11 using a/bZ 3 Gy for GI toxicity and SDand a/b Z 6 Gy for GU toxicity, consistent with theQUANTEC studies.3-5 For studies exploiting multiplefractionation schemes, the most conservative scheme(lowest EQD2a/b) was used. For low-dose-rate BT, theEQD2a/b formalism accounting for continuous RT wasapplied.12

If no DVH thresholds were reported in a study, theywere derived at a 10% predicted risk from the study-specific dose-response curves (judged as clinically relevant)or from the study-specific averaged DVHs stratified be-tween symptomatic and nonsymptomatic patients. If onlyabsolute volume DVH thresholds were reported, these wereconverted, on the basis of population- and/or cohort-averaged OAR volumes, to relative volume thresholds(with the exception that the maximum dose to a volume�2 cm3 was set equal to a relative volume of 2% with theassumption of this being a sufficiently sized volume forreliably estimating maximum doses within a treatmentplanning system).

If none of these data could be determined from thestudies, the authors were contacted and asked to supply thisinformation. Finally, if exactly the same DVH thresholdwas identified for multiple symptom grades or severitieswithin a study, that of the lowest severity is reported (e.g.,the same threshold for grade 1 or higher and grade 2 orhigher would lead to only grade 1 or higher being stated).All DVH thresholds are given as EQD2a/b/relative volume

Downloaded for Anonymous User (n/a) at CONSORTIUM MEDICAL LIBRARIES - ISRAFor personal use only. No other uses without permission. C

(e.g., 50 Gy/25% indicates �25% of the concerned struc-ture volume receiving �50 Gy).

Quantitative synthesisIf a study included multiple dose-symptom associationmethods (i.e., combinations of a hypothesis test, univariateanalysis, and/or multivariate analysis), the DVH thresholdderived by the most comprehensive method was used (withhypothesis test being least comprehensive and multivariateanalysis being most comprehensive).

Initially within the quantitative syntheses, a linear orlinear-quadratic function with or without robust linear leastsquares (bisquare weights or least absolute residuals[LARs]) was fitted to all identified DVH thresholds for asymptom. All fits were constrained to adhere to an EQD2a/bof 0 Gy at a 100% relative volume; in addition, an EQD2a/bin the near vicinity of 80 Gy was constrained to a 0%relative volume. For each symptom, the function con-forming to these constraints and yielding the lowest sum ofsquares due to error (SSE) was considered final, and theassociated adjusted coefficient of determination (R2) wasreported and used as a goodness-of-fit measure. If thresholdoutliers were apparent for a certain symptom, as judgedgraphically, the same function was refitted excluding theseoutliers. Consequently, if such a refit resulted in a lowerSSE, that fit was considered final.

Ultimately, for each symptom with �3 DVH thresholdsidentified from �2 different studies excluding mean dosethresholds, unless otherwise stated, 2 to 3 across-studysymptom-specific thresholds were identified within thelow- to intermediate-dose region and within the high-doseregion. These center-of-mass DVH thresholds (COMTDVH)were given equal importance and averaged as the meanvalue over each dose-volume interval for included studies(COMTDVH; EQD2a/b/relative volume). If possible, allCOMTDVHs for symptoms of the same domain were com-bined into a meta-DVH (i.e., a DVH curve for overalldomain-specific toxicity using a fitting procedure analo-gous to that described earlier). All analyses were conductedin MATLAB (version R2016a; The MathWorks, Natick,MA).

Results

A total of 168 full-text articles were identified, of which 33fulfilled the inclusion criteria. Of these 33 studies, 22concerned EBRT alone (n Z 6313 treated between 1993and 2014) and 11 concerned BT with or without EBRT(n Z 3836 treated between 1995 and 2010; Fig. 1;Tables 1-3; Tables E1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). All studies pre-sented with at least “fair” study quality (meanscore � standard deviation, 6 � 1). The studies concerningEBRT alone included a median of 275, 557, and 56 patients(minimum of 90, 262, and 42 patients per study) for GItoxicity, GU toxicity, and SD, respectively. The

EL -Medical Center Rambam from ClinicalKey.com by Elsevier on June 12, 2019.opyright ©2019. Elsevier Inc. All rights reserved.



Search strategy step Identified studies

Identified studies

Excluded studies

I.Initial search

II.Abstract screening

III.Full-text screening

1954

1275

168

33

679

1107

135

Unique studies

Full-text studiesassessed for eligibility

Finally includedstudies

Duplicates

not fulfilling inclusion criteria

Not fullfilling inclusion criteria:Full-text N/A (N = 4)No OAR DVH metrics (N = 32)Re-analysis of previous studies (N = 18)Less than 20 patients (N = 5)Aggregated symptom scores (N = 62)Acute toxicity only (N = 5)Data not interpretable (N = 5)Uncommon RT approach (N = 4)

••••••••

web4C=FPO

Fig. 1. Schema of study selection procedure. Altogether, the authors of 23 different studies were contacted for additionaland/or clarifying information in search strategy step II. Of these, the authors of 17 studies (74%) responded. Abbreviations:DVH Z dose-volume histogram; N/A Z not available; OAR Z organ at risk; RT Z radiation therapy.


corresponding patient numbers for BT with or withoutEBRT were 118, 197, and 44 patients (minimum of 37, 36,and 41 patients per study).

EBRT alone

GI toxicity: Symptom-specific tolerance dose patternsDefecation urgencyFour studies fulfilled the inclusion criteria for defecationurgency in which 5 scoring systems had been used(n Z 1037; Table 1).13-16 All 14 identified rectal DVHthresholds had been assessed using a defecation urgencydefinition relating to its frequency (Table 1; Tables E1 andE2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). The thresholds were located between 13and 71 Gy/8% and 70%, and the final fit for this symptom,which included all thresholds, was a linear-quadratic(þLAR) function (R2 Z 0.82; Fig. 2; Table E3, availableonline at https://doi.org/10.1016/j.ijrobp.2018.08.015).

DiarrheaFor diarrhea, only 2 studies fulfilled the inclusion criteria(n Z 1026; Table 1).13,17 These studies had both defineddiarrhea as diarrheal frequency (grade 1 or 2 or higher), and

Downloaded for Anonymous User (n/a) at CONSORTIUM MEDICAL LIBRARIES - ISRFor personal use only. No other uses without permission.

the associated 9 anal canal, anorectal, or rectal DVHthresholds were located within 23 to 62 Gy/20% to 80%(Table 1; Tables E1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). A linear-quadratic(þLAR) function described the association between thesethresholds (R2Z 0.86, SSEZ 1008), and excluding the analcanal threshold at 30 Gy/20% but retaining the anorectal andrectal thresholds within 23 to 62 Gy/30% to 80% yielded thefinal fit for diarrhea (R2 Z 1.00; Fig. 2; Table E3, availableonline at https://doi.org/10.1016/j.ijrobp.2018.08.015).

Fecal incontinenceAmong the 8 studies fulfilling the inclusion criteria forfecal incontinence, 5 symptom definitions had been used,but for the 13 identified thresholds, the definition of fecalincontinence overall related to the use of pads either oc-casionally or daily (n Z 2237; Table 1).13-16,18-21 Thethresholds were primarily derived from the rectum andwere reported within 10 to 65 Gy/14% to 100%. Of thesethresholds, 1 each pertained to the anal wall, anal sphincterregion, external anal sphincter, and iliococcygeal muscle(Table 1; Tables E1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). Retaining all thresholdsresulted in a final linear (þLAR) fecal incontinence fit(R2 Z 0.93; Fig. 2; Table E3, available online at https://doi.











Table 1 Study-specific information relating to finally selected studies for GI toxicity after EBRT alone

Symptom Structure Reference N D (d), Gy RTSymptomcutoff

Symptomdefinition at

lowestinvestigated

gradeSymptomrate, %

MedianFU, y Modeling

EQD23/relativevolume,Gy/%

Defecationurgency

R

1 13 288 64-74 (2.0) 3D CRT �1 �1/wk 47 5.0 UVA 56/4044/5533/65

�2 �1/d 23 69/152 14 170 50-74

(2.5-3.1)3D CRT/IMRT

�2 At leastintermittenturgency orregular

49 3.0 H test 13/56

3 15 90 76-78(1.8-2.0)

3D CRT/IMRT

�1 Rarely or never 5 2.0 H test 71/8

65/1459/2553/3042/3531/4618/70

4 16 212(N1)

78 (2.0) 3D CRT �1 Immediate toiletvisit <1/mo

17 3.6 MVA 32/55

277(N2)

70 (2.0) Hold stools >1 h 39 6.4 63/40

Diarrhea AC5 17 708 66-74 (2.0) 3D CRT �1 Increase of <4/d 40 6.0 MVA 30/20AR6 17 708 66-74 (2.0) 3D CRT �1 Increase of <4/d 40 6.0 MVA 32/66

26/74R 25/761 13 318 64-74 (2.0) 3D CRT �1 Usually 23 5.0 UVA 56/40

44/5533/6523/80

�2 Always 13 62/30Fecalincontinence

AW

7 18 379 64-74 (2.0) 3D CRT �1 Occasional (paduse)

15 5.0 H test 48/53

ASR8 16 277

(N2)70-78 (2.0) 3D CRT �1 Any pad use 7 6.4 MVA 23/70

EAS9 20 256 78 (2.0) IMRT �2 Pad use for

unwilling lossof stools >2/wk

8 �3 MVA 10/75

IM10 20 256 78 (2.0) IMRT �2 Pad use for

unwilling lossof stools >2/wk

8 �3 MVA 49/100

(continued on next page)


Downloaded for Anonymous User (n/a) at CONSORTIUM MEDICAL LIBRARIES - ISRAEL -Medical Center Rambam from ClinicalKey.com by Elsevier on June 12, 2019.For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

Table 1 (continued )



lowestinvestigated

gradeSymptomrate, %



R11 19 515 70-78

(1.8-2.0)3D CRT �1 Unintentional

stool dischargesometimes

11 7.0 H test 31/80

1 13 318 64-74 (2.0) 3D CRT �1 Occasional (paduse)

15 5.0 UVA 56/40

33/652 14 170 50-74

(2.5-3.1)3D CRT/IMRT

�2 Intermittent (paduse)

5 �2.0 H test 18/69

3 15 90 76-78(1.8-2.0)

3D CRT/IMRT

�1 Rarely or never 4 2.0 H test 65/14

59/2553/30

12 21 232 70 (2.0) 3D CRT �2 Intermittent (paduse)

4 5.0 UVA 10/96

Gastrointestinalpain

ASR

8 16 212(N1)

70-78 (2.0) 3D CRT �1 Need to strain<1/mo

8 3.6 MVA 10/15

AR6 17 655 66-74 (2.0) 3D CRT �1 Occasional and

mild/nonnarcotic

65 6.0 MVA 30/54

29/7125/7812/9311/93

R 10/8012 21 232 70 (2.0) 3D CRT �2 Intermittent and

tolerable/regularnonnarcotic

6 5.0 UVA 68/4

62/16Incomplete

evacuationAC

5 17 671 66-74 (2.0) 3D CRT �1 Occasionalmultipleevacuations ornot “all done”approximately1/wk

55 6.0 MVA 34/27

31/2130/2129/3328/3816/6315/6314/6313/6312/63







lowestinvestigated

gradeSymptomrate, %



Mucus ASR8 16 277

(N2)70 (2.0) 3D CRT �1 In stools 1/mo 8 6.4 MVA 70/5

Proctitis R1 13 388 64-74 (2.0) 3D CRT �1 Minor symptoms 47 5.0 UVA 62/30

23/80�2 Simple outpatient

management22 69/15

65/3056/4044/5533/65

2 14 170 50-74(2.5-3.1)

3D CRT/IMRT

�2 Superficial�1 cm3 oroccasionalsteroids

14 3.0 H test 64/4

28/69Rectalbleeding

AC

5 17 702 66-74 (2.0) 3D CRT �1 Occult, stoolsoftener, oriron therapy

53 6.0 MVA 40/26

38/2934/30

AR 32/326 17 710 66-74 (2.0) 3D CRT �1 Occult, stool

softener, oriron therapy

53 6.0 MVA 62/25

61/2960/3059/3358/35

13 23 512 68-78 (2.0) IMRT �2 Loss of blood >1/d requiringsingle lasercoagulation ortransfusion

6 3.0 MVA 61/42

14 20 256 78 (2.0) IMRT �2 Loss of blood >1/d requiringsingle lasercoagulation ortransfusion

5 �3.0 MVA 69/18

RW15 25 205 70-78 (2.0) IMRT �2 Moderate,

persistent,symptomatic,medical orminor surgicalinterventionrequired

6 7.4 UVA 55/24

43/3332/4523/70







lowestinvestigated

gradeSymptomrate, %



R16 22 141 65-70 (2.5) 3D CRT �2 Simple outpatient

management14 4.0 UVA 65/20

17 24 173 78-80 (2.0) IMRT �1 Mild orinterventionother than ironsupplementsnot indicated

6 4.8 H test 63/57

18 17 702 66-74 (2.0) 3D CRT �1 Occult, stoolsoftener, oriron therapy

53 6.0 MVA 63/30

62/3160/3459/35

11 19 515 70-78(1.8-2.0)

3D CRT �1 <2/wk 7 7.0 UVA 71/6

1 13 361 64-74 (2.0) 3D CRT �1 Occasional 44 5.0 UVA 56/4044/5533/6569/1562/30

2 14 170 50-74(2.5-3.1)

3D CRT/IMRT

�2 Occasional >2/wk oroccasionaltransfusion

6 3.0 H test 18/69

19 26 159 74 (2.0) 3D CRT �2 Slight bleeding 7.5 2.7 - 64/220 27 261 70-80 (2.0) 3D CRT/

IMRT�1 Occult, stool

softener, oriron therapy

10 5.2 MVA 75/2

21 28 129 70-76 (2.0) 3D CRT/IMRT

�2 Moderate,persistent,symptomatic,medical orminor surgicalinterventionrequired

21 1.6 UVA 61/17

3 15 90 76-78(1.8-2.0)

3D CRT/IMRT

�1 Rarely 4 2.0 UVA 78/5

71/8Stool

frequencyAR

13 23 512 68-78 (2.0) IMRT �2 Increase of >3/d 6 3.0 MVA 37/686 17 575 66-74 (2.0) 3D CRT �1 2-4/d or �2

antidiarrhealagents/wk

53 2.5 MVA 24/24

22/3822/2820/4019/43

IM10 20 256 78 (2.0) IMRT �2 Increase of >3/d 11 �3.0 MVA 37/100LAM23 20 256 78 (2.0) IMRT �2 Increase of >3/d 11 �3.0 MVA 32/100







lowestinvestigated

gradeSymptomrate, %



R1 13 344 64-74 (2.0) 3D CRT �1 2-4/d or �2

antidiarrhealagents/wk

39 5.0 UVA 56/40

44/55

Abbreviations: AC Z anal canal; AR Z anorectum; ASR Z anal-sphincter region; AW Z anal wall; CRT Z conformal radiation therapy;

D Z prescribed dose; d Z dose/fraction; EAS Z external anal sphincter; EBRT Z external beam radiation therapy; EQD23 Z equivalent dose in 2-Gy

fractions using a/b Z 3Gy; FU Z follow-up time; GI Z gastrointestinal; H Z hypothesis; IM Z iliococcygeal muscle; IMRT Z intensity modulated

radiation therapy; LAMZ levator ani muscle; MVAZ multivariate analysis; N1Znumber of patients in cohort 1; N2Z number of patients in cohort 2;

R Z rectum; RT Z radiation therapy; RW Z rectal wall; 3D Z 3-dimensional; UVA Z univariate analysis.


org/10.1016/j.ijrobp.2018.08.015). In addition to these 13DVH thresholds, associations between fecal incontinence(occasional pad use) and the anal wall and the analsphincter mean dose were identified at 20 Gy and 23 Gy,respectively (Table E4, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).18

GI painThree studies were included for GI pain (n Z 1099;Table 1).16,17,21 Of the 9 identified DVH thresholds, 6 hadbeen assessed from an occasional GI pain definition (grade1 or higher). Thresholds were represented by 9 analsphincter regional, anorectal, or rectal thresholds within 10to 68 Gy/4% to 93% (Table 1; Tables E1 and E2, availableonline at https://doi.org/10.1016/j.ijrobp.2018.08.015). A fitto the 9 thresholds was not considered given that there wereonly data from 2 studies in the DVH space for this symp-tom (Fig. E1, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).

ProctitisThe majority of the 9 proctitis DVH thresholds that wereidentified from 2 studies had been assessed for proctitis definedas requiring intervention (outpatient management and/oroccasional steroids), that is, grade 2 or higher (5 of 9 thresholds,n Z 558; Table 1).13,14 The final fit to all thresholds locatedwithin 23 to 69 Gy/4% to 80% was a linear-quadratic (þLAR)function (R2 Z 0.93; Fig. 2; Tables E1-E3, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015).

Rectal bleedingFrom 13 studies, the majority of DVH thresholds werederived from occasional rectal bleeding followed by rectalbleeding requiring intervention (laser coagulation ortransfusion at least once), which corresponded to grade 1 orhigher and grade 2 or higher, respectively (n Z 3682;Table 1).13-15,17,19,20,22-28 Rectal bleeding presented withthe largest number of identified DVH thresholds across allsymptoms: 33 in total within 18 to 78 Gy/2% to 70%.Eighteen of these thresholds were assessed for rectal doses,


whereas the anal canal and the rectal wall were each rep-resented by 4 thresholds and the anorectum was representedby 7 thresholds (Table 1; Tables E1-E3, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015). A final rectalbleeding fit was a linear-quadratic (þLAR) function to allthresholds except the 4 anal canal thresholds (R2 Z 0.95;Fig. 2; Table E3, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). Furthermore, refitting the finalfunction to all data, excluding the anal canal thresholds butincluding the 25 QUANTEC-identified DVH thresholds,3

resulted in an inferior fit (R2 Z 0.68; Fig. 3A). In addi-tion to the 33 identified DVH thresholds, a rectal wall meandose finding with moderate rectal bleeding was reported at43 Gy (Table E4, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).25

Other GI toxicitiesAmong the 4 stool frequency studies, all 10 identified DVHthresholds had been assessed for the number of stoolsreleased per day (n Z 1687; Table 1).13,17,20,23 These DVHthresholds were identified between 19 and 56 Gy/24% and100% and had been assessed from 5 OARs (Table 1; TablesE1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). No final fit was assessed for thissymptom because there was no apparent relationship be-tween thresholds (Fig. E1, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). In addition, an associa-tion with the anorectal wall mean dose at 51 Gy wasidentified (Table E4, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).23 Thresholds for other GI tox-icities included 1 high-dose anal sphincter association at70 Gy/5% for mucus (in stools monthly)16 and 10 low tointermediate anal canal DVH thresholds in the range 12 to34 Gy/21% to 63% for incomplete evacuation (multiple andoccasional; Table 1).17

GI toxicity: Meta-DVH across symptomsThe generated COMTDVHs based on dose/volume tolerancesfor defecation urgency, diarrhea, fecal incontinence, proctitis,and rectal bleeding were calculated within the low- to




















Table 2 Study-specific information relating to finally selected studies for GU toxicity and SD after EBRT alone

Symptomdomain Symptom Structure Reference N

D (d),Gy RT

Symptomcutoff


lowestinvestigated

gradeSymptomrate, %



GU Dysuria BW24 30 746 66-74

(2.0)3DCRT

�1 Intermittent andtolerable

25 5.0 UVA 64/19

�2 Persistent andintense

8 5.0 UVA 64/19

Frequency BW24 30 746 66-74

(2.0)3DCRT

�2 Increase >2 timesnormal but<1/h

29 5.0 UVA 39/41

Hematuria B25 29 262 74-80

(2.0)IMRT �2 Gross 14 4.0 MVA 74/5

BW24 30 746 66-74

(2.0)3DCRT

�1 Microscopic ornormalhemoglobin

14 5.0 UVA 63/41

Obstruction B26 31 276 70

(2.0)3DCRT

�1 Bear down orpush to urinate<50% of times

9 6.4 MVA 66/20

BW27 32 557 68-78

(2.0)3DCRT

�2 Urethralnarrowing withhydronephrosisrequiringcatheterizationand/ortransurethralresection ordilation

10 5.9 MVA 81/1

Urgency B26 31 276 78

(2.0)3DCRT

�1 Rarely having torestrainurinatingbefore reachingtoilet

10 3.6 MVA 83/2*

SD Erectiledysfunction

PB

30 33 70 70(2.0)

3DCRT

�2 Quality ofnocturnalerection firmenough forintercourseapproximately50% of times

37 4.0 UVA 70/2

Abbreviations: B Z bladder; BW Z bladder wall; CRT Z conformal radiation therapy; D Z prescribed dose; d Z dose/fraction; EBRT Z external

beam radiation therapy; EQD23 Z equivalent dose in 2-Gy fractions using a/b Z 3Gy; FU Z follow-up time; GU Z genitourinary; IMRT Z intensity

modulated radiation therapy; MVA Z multivariate analysis; NA Z not available; PB Z penile bulb; RT Z radiation therapy; SD Z sexual dysfunction;

3D Z 3-dimensional; U Z urethra; UVA Z univariate analysis.

* Assumption of a maximum dose at 105% equal to 82 Gy.


intermediate-dose region (21-44 Gy/46%-76%) and within thehigh-dose region (54-70 Gy/10%-42%). A linear-quadratic fitto these points suggested good agreement between symptom-specific COMTDVHs (R

2 Z 0.98; Fig. 4; Table E3, available


online at https://doi.org/10.1016/j.ijrobp.2018.08.015) andfurther suggested that these aligned along a meta-DVH curvefor overall GI toxicity. Most of the COMTDVHs were derivedfrom rectal doses (71%) extending from the low- to high-dose



Table 3 Study-specific information relating to finally selected studies for GI and GU toxicity as well as SD after BT with or withoutEBRT

Symptomdomain Symptom Structure Reference N D (EBRT þ BT), Gy d (EBRT þ BT), Gy

GI Anorectal pain R31 35 118 46 þ 19.5 2 þ 6.5

Defecation urgency R31 35 118 46 þ 19.5 2 þ 6.5

Diarrhea R31 35 118 46 þ 19.5 2 þ 6.5

Fistula R32 40 2368 (1604, 764) 44, 46 þ 144, 124

(125I, 103Pd)-

Rectal bleeding R33 38 37 40/45 þ 110 2/1.8 þ 110

31 35 118 46 þ 19.5 2 þ 6.534 36 216 45-51 þ 18-25 3 þ 5-10.5

RW35 39 212 160 160R mucosa36 37 64 45-50.4 þ 18 1.8 þ 9

Stool frequency R31 35 118 46 þ 19.5 2 þ 6.5

GU Incontinence U36 37 64 45-50.4 þ 18 1.8 þ 9

Retention U39 41 36 38 9.5

Stricture U38 42 330 46 þ 18 2 þ 6

SD Erectile dysfunction PB40 44 46 (37/9) 145, 115 (125I, 103Pd) -

41 45 41 (11/30) 145, 125 (125I, 103Pd) -

Abbreviations: BTZ brachytherapy; DZ prescribed dose; dZ dose/fraction; EBRTZ external beam radiation therapy; EQD23 Z equivalent dose in

2-Gy fractions using a/b Z 3Gy; FU Z follow-up time; GI Z gastrointestinal; GU Z genitourinary; H Z hypothesis; HDR Z high dose rate; LDR Zlow dose rate; MVA Z multivariate analysis; NA Z not available; PB Z penile bulb; R Z rectum; RT Z radiation therapy; RW Z rectal wall;

SD Z sexual dysfunction; U Z urethra; UVA Z univariate analysis.

* Assumption of a maximum dose at 105% equal to 82 Gy.


region, followed by primarily high anorectal doses (13%)and intermediate anal canal and rectal wall doses (5%for both). As judged from this DVH, the more sensitivesymptom, that is, the symptom with the most conservativeassociated COMTDVH, was defecation urgency. The regionbelow and above this meta-DVH can be regarded as low riskand high risk of development of overall GI toxicity,respectively.


GU toxicityTwo studies each identified a high-dose bladder orbladder wall DVH threshold at 74 Gy/5%29 and at 63 Gy/41%30 for gross or microscopic hematuria (n Z 262 andn Z 746; Table 2; Tables E1 and E2, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015). For thebladder wall, a high-dose threshold was also identified at64 Gy/19% for intermittent or persistent dysuria, and an



Table 3 Study-specific information relating to finally selected studies for GI and GU toxicity as well as SD after BT with or withoutEBRT (continued)

RTSymptomcutoff

Symptom definitionat lowest investigated grade

Symptomrate, %


EQD23/relativevolume, Gy/%

EBRT þ HDR (192Ir) �1 Occasional and mild or nonnarcotic 30 6.0 UVA 48/28

EBRT þ HDR (192Ir) �1 Occasional urgency 70 6.0 UVA 58/451/1950/1542/3025/62

EBRT þ HDR (192Ir) �1 Increase of <4/d 35 6.0 UVA 50/13

EBRT þ LDR NA Presence of fistula (yes or no) <1 6.0 H test 93/2, 175/2*

EBRT þ LDR (125I) �2 Moderate, persistent, symptomatic,medical or minor surgical intervention required

14 3.5 H test 150/2

EBRT þ HDR (192Ir) �2 Occasionally >2/wk or occasional transfusion 20 6.0 UVA 52/12EBRT þ HDR (192Ir) �2 Moderate, persistent, symptomatic,

medical or minor surgical intervention required9 2.9 MVA 68/5

EBRT þ LDR (125I) 2 Simple outpatient management 10 2.5 H test 180/2

EBRT þ HDR (192Ir) �1 Rarely 25 1.5 H test 66/2

EBRT þ HDR (192Ir) �2 4-8/d or >2 antidiarrheal agents/wk 12 6.0 UVA 52/12

EBRT þ LDR �1 Leaked urine 1/wk or no pads 22 1.5 H test 185/2

HDR (192Ir) �3 >1 catheterization/d or urologicintervention indicated

11 3.0 H test 164/12

148/22

EBRT þ HDR (192Ir) 2 Symptomatic and/or dilation orurethrotomy required

8 2.0 UVA 109/10

LDR (125I, 103Pd) 1 Erection sufficiently firm forvaginal penetration (yes or no)

50 2.8 H test 36/50, 55/50*

LDR (125I, 103Pd) NA Maintaining erection sufficient forintercourse with or without sildenafil (yes or no)

27 1.7 MVA 8/90, 10/90*


intermediate threshold was identified at 39 Gy/41% for 2times normal urinary frequency.30 In 1 of the 2 cohortsinvestigated by Thor et al,31 a high-dose threshold at83 Gy/2% was reported for rare urinary urgency and thebladder; in the other cohort, a high-dose bladderthreshold for obstruction (bearing down during the ma-jority of times when urinating) was observed at 66 Gy/20% (n Z 207 and n Z 276). Another threshold forobstruction (requiring intervention), but for the bladderwall, was observed at 81 Gy/1% (n Z 557)32; in that


study, the same obstruction definition was also associ-ated with the trigone point dose at 42 Gy (Table E4,available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).

SDOnly 1 study reported a DVH threshold for SD. More spe-cifically, a relationship between erectile dysfunction(nocturnal erection sufficient for intercourse on the majorityof occasions) and the penile bulb at 70 Gy/2% (nZ 70) was




[14]: N = 170[25]: N = 205[17]: N = 702[13]: N = 361[17]: N = 710[17]: N = 702[23]: N = 512[28]: N = 90[24]: N = 173[26]: N = 159[22]: N = 141[20]: N = 256[15]: N = 90[19]: N = 515[27]: N = 261

R2 = 0.95

[13]: N = 388

[13]: N = 388

R2 = 0.93

[14]: N =170

[20]: N = 256[21]: N = 232[14]: N = 170[16]: N

2 = 277

[19]: N = 515[13]: N = 318[18]: N = 379[20]: N = 256[15]: N = 90

R2 = 0.93

[13]: N = 318

[13]: N = 318

[17]: N = 708[17]: N = 708

R2 = 1.00

[14]: N = 170[15]: N = 90[16]: N

1 = 212

[13]: N = 288[16]: N

2 = 277

[13]: N = 288

R2 = 0.82

Defecation urgency

5%47%

49%17%

39%

23%

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

100

100

100

100

100

80 80

80

80

80

80

80

80

80

80

EQD23 [Gy]

EQD23 [Gy]

EQD23 [Gy]

EQD23 [Gy]

EQD23 [Gy]

60 60

60

60

60

60

60

60

60

10%7%4%

8%

5%14%

21%

53%

6%

53%

6%

6%44%

6%

53%

15%

15%

11%

4%

5%

8%

7%8%

Rectal bleeding

22%14%

47%Fecal

incontinence

Diarrhea

13%

23%40%

40%

Proctitis

40 40

40

40

40

40

40

40

40

40 60

20 20

20

20

20

20

20

20

20

20

0 00

00

0

00

00

web4C=FPO

Fig. 2. Identified dose-volume histogram thresholds for the 5 gastrointestinal symptoms after external beam radiationtherapy alone for which the quantitative synthesis was considered. Study-specific symptom rates are given next to the first(lowest dose) dose-volume histogram threshold in each study. Each symbol is study specific (with the exception of rectalbleeding); open and closed symbols refer to grade 1 or higher and grade 2 or higher, respectively. Black dotted lines representfinal fits (the associated R2 is presented in the key). Regarding color coding, red indicates the anal canal; blue, anorectum;magenta, anal sphincter; purple, external anal sphincter; brown, iliococcygeal muscle; black, rectum; and green, rectal wall.Abbreviations: EQD23 Z equivalent dose in 2-Gy fractions using a/bZ 3Gy; Rel volZ relative volume. (A color version ofthis figure is available at https://doi.org/10.1016/j.ijrobp.2018.08.015.)


identified (Table 2; Tables E1 and E2, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015).33 In addition,an association between erectile dysfunction (firm enough forintercourse >50% of times) and penile bulb mean dose at21 Gy was reported (n Z 41; Table E4, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015).34


BT with or without EBRT

GI toxicity: Symptom-specific tolerance dose patternsFive studies had each reported a rectal DVH thresholdfor rectal bleeding defined primarily as being moderateand requiring surgery or transfusion (n Z 647;





[35]: N = 52

[3K]: N = 553

[3Z]: N = 107

[34]: N

2 = 315

[34]: N

1 = 261

[317

]: N = 245

[316

]: N = 154

[311

]: N = 163

[14]: N = 170

[17]: N = 710[17]: N = 702[23]: N = 512[28]: N = 129[24]: N = 173[26]: N = 159[22]: N = 141

[20]: N = 256[19]: N = 515[15]: N = 90[27]: N = 261

R2 = 0.95

R2 = 0.68

[13]: N = 361

[25]: N = 205

4-6%7-10%

11-19%21-25%

76-79Gy33-53%

Open: >Grade 1Closed: >Grade 2

A

B

C

D

E

3DCRT3DCRT/IMRT

IMRT

<66.1Gy69-70.3Gy

74-75Gy72.4-73.4Gy

100

100

100

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

Rel.

vol [

%]

EQD23 [Gy]

EQD23 [Gy] EQD23 [Gy]

EQD23 [Gy]EQD23 [Gy]

100

100

80

80

80 80

80

80

80

80

80

80

60

60

60

60

60

60

60

60

60

60

40

40

40

40

40

40

40

40

40

40

20

20

20

20

20

20

20

20

20

20

00

00

00

00

00

web4C=FPO

Fig. 3. The 29 identified dose-volume histogram thresholds for rectal bleeding after external beam radiation therapy alone(excluding the 4 anal canal thresholds; compare with Fig. 2) and the 25 dose-volume histogram thresholds for grade 2 orhigher rectal bleeding that were identified by Quantitative Analysis of Normal Tissue Effects in the Clinic (QUANTEC)3 areshown. QUANTEC studies are marked with a bright green border, with different symbols for different sub-QUANTECstudies; subscript numbers refer to the reference list of Michalski et al3; subscript K refers to Koper et al in Michalski etal3; and subscript Z refers to Zapatero et al. All 54 thresholds were stratified with respect to organ at risk (blue, anorectum;black, rectum; green, rectal wall; black dotted line, final fit for rectal bleeding; gray dashed line, fit including 25 QUANTECthresholds) (A); rectal bleeding symptom cutoff (B); rectal bleeding rate separated into 5 intervals (C); radiation therapytechnique (D); and prescription dose level (EQD210) separated into 5 intervals (after QUANTECdthe median value of allexploited prescription levels) (E). Abbreviations: CRT Z conformal radiation therapy; EQD23 Z equivalent dose in 2-Gyfractions using a/b Z 3Gy; IMRT Z intensity modulated radiation therapy; Rel vol Z relative volume; 3D Z 3-dimensional. (A color version of this figure is available at https://doi.org/10.1016/j.ijrobp.2018.08.015.)


Table 3; Tables E1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).35-39 The 5thresholds were located within 52 to 68 Gy/2% to12%35-37 and within 150 to 180 Gy/2%.38,39 No func-tion was explored to fit these 5 thresholds given lack of


data between the 2 clusters of DVH thresholds(Fig. E1, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). Thresholds for other GI toxic-ities included intermediate rectal thresholds at 25 to58 Gy/4% to 62% for occasional anorectal pain,







Fit; R2 = 0.98Fit: 95%CIFit: 95%CI

Meta DVH: GI toxicity

High risk of GI toxicity

Low risk of GI toxicity

Fecal IncontinenceDiarrheaDefecation urgencyProctitisRectal bleeding

EQD23[Gy]

Rel.

vol [

%]

100

80

60

40

20

00 20 40 60 80

web4C=FPO

Fig. 4. Estimated center-of-mass DVH thresholds(COMTDVHs) for defecation urgency, diarrhea, fecal in-continence, proctitis, and rectal bleeding after externalbeam radiation therapy alone and the fitted metaedose-volume histogram (DVH) for overall gastrointestinal (GI)toxicity based on COMTDVHs for each of these symptoms(black dotted line). Dashed black lines show 95% confi-dence intervals (CIs). The majority of data from which allCOMTDVHs were derived stemmed from rectal doses(71%), followed by high anorectal doses (13%). The regionbelow and above this meta-DVH suggests low risk and highrisk, respectively, of development of overall gastrointestinaltoxicity; the circles denote the 2 clusters of COMTDVHs.The color coding indicates symptom-specific sensitivity;for example, the tolerance dose for the most sensitivesymptom of defecation urgency is red. Abbreviations:EQD23 Z equivalent dose in 2-Gy fractions using a/b Z3Gy; Rel vol Z relative volume. (A color version of thisfigure is available at https://doi.org/10.1016/j.ijrobp.2018.08.015.)


occasional defecation urgency, diarrhea (increase indaily stools), and stool frequency (4-8 times per day)35

(n Z 118; Table 3). For fistula, 1 threshold at 93 Gy/2% (n Z 2368) was identified.40

GU toxicity and SDUrinary retention (requiring catheterization or urologicintervention) was represented by 2 high urethral DVHthresholds at 148 to 164 Gy/12% to 22%,41 and urinaryincontinence (not requiring pads) was represented by 1urethral threshold at 185 Gy/2%37 (n Z 36 and n Z 64;Table 3; Tables E1 and E2, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015). In a considerablylarger cohort (nZ 330) and for urinary stricture (dilation orurethrotomy required), 1 threshold at 109 Gy/10% was


found.42 For GU pain (during urination <50% of times), anassociation with the urethral mean dose was reported at91 Gy (n Z 368; Table E4, available online at https://doi.org/10.1016/j.ijrobp.2018.08.015).43 Two smaller studieshad each identified 1 DVH threshold for erectile dysfunc-tion (inability to maintain erection for intercourse) and thepenile bulb at 36 Gy/50%44 and at 8 Gy/90%45 (nZ 46 andn Z 41).

Discussion

This systematic review was based on dose/volume toler-ances for late and distinct RT-induced toxicities afterprostate cancer RT from 10,149 patients and 33 studies. Weidentified 97 dose/volume tolerances, 8 dose/volume tol-erances, and 1 dose/volume tolerance as quantified by DVHthresholds for GI, GU, and SD toxicity, respectively, in 13OARs after EBRT alone. The corresponding numbers afterBT with or without EBRTwere 14, 4, and 2 in 4 OARs. Anoverall GI toxicity DVH curve based on primarily rectaldoses was generated by fitting the tolerance doses acrossdefecation urgency, diarrhea, fecal incontinence, proctitis,and rectal bleeding after EBRT alone in the low- tointermediate-dose region and the high-dose region(R2 Z 0.98).

The report by Emami et al1 and the QUANTEC reports2

presented normal tissue dose/volume tolerance data afterEBRT alone (with 4400 combined citations in peer-reviewed articles as of May 25, 2018). The QUANTECreports embraced published studies up to 2009 and, forprostate cancer, mainly covered relationships betweencombined symptom scores and rectal and penile bulb dosesfor GI toxicity and SD, respectively.3,5 The QUANTECeffort still provides the major guidelines in treating cancerswithout severe RT-induced toxicity. The GU-specificQUANTEC report did not include dose/volume tolerancesafter prostate cancer EBRT because of heterogeneity inpublished results.4 Our systematic review spans from afterthe publication of these QUANTEC reports (2010) to 2017for GI and SD toxicity after EBRT alone and spans fromafter the report by Emami et al (1992) to 2017 for GUtoxicity after EBRT alone and for the 3 investigatedsymptom domains after BT with or without EBRT. Weidentified dose/volume tolerances for 15 and 11 distinctsymptoms after EBRT alone and after BT with or withoutEBRT, respectively. For both regimens, the majority ofDVH thresholds were reported for GI toxicity and, inparticular, for rectal bleeding.

Among all GI toxicities, rectal bleeding after EBRTalone was the only distinct symptom for which dose/vol-ume tolerances were suggested by the associated QUAN-TEC report. Their 5 high-dose constraints (EQD23/relativevolume of 50 Gy/50%, 60 Gy/35%, 65 Gy/25%, 70 Gy/20%, and 75 Gy/15%) were presented to limit the occur-rence of grade 2 or higher rectal bleeding to <15%.3 Eventhough the majority of the dose/volume tolerances









identified in our study13-15,17,19,20,22-28 and those reportedby QUANTEC3 were derived primarily from rectal doses,the post-QUANTEC data were generally located betweenthose reported by QUANTEC for EQD23 <56 Gy, followedby a wider spread in the post-QUANTEC data within56 Gy < EQD23 < 80 Gy. This finding could be explainedto some extent by the lower number of post-QUANTECdose/volume tolerances identified for the low- tointermediate-dose region compared with QUANTEC data(7 vs 16 thresholds) and the higher number for the high-dose region (20 vs 9 thresholds; Fig. 3A), but it is also aconsequence of increasingly conformal treatment tech-niques over time enabling higher therapeutic doses withlower OAR doses (median EQD210, 73.4 Gy vs 70.7 Gy;Fig. 3D and E).

Associated differences were also reflected when refittingour final rectal bleeding function to all data also includingthe raw data behind the QUANTEC constraints (i.e., the 25QUANTEC DVH thresholds), which reduced the fittingstatistics from a satisfying fit to a modest fit (R2 Z 0.95 vsR2 Z 0.68). The differences between the post-QUANTECand QUANTEC dose/volume tolerances can partly beexplained by differences in rectal bleeding severity orgrades because we have summarized the results for bothgrade 1 or higher and grade 2 or higher rectal bleeding incontrast to QUANTEC, which summarized tolerance dosesfor grade 2 or higher rectal bleeding exclusively (Fig. 3B).This is also reflected for lower doses in the refitted curve(including QUANTEC data), which is more conservativethan the final rectal bleeding fit (Fig. 3A). It should, how-ever, be kept in mind that the more ideal stratificationwould involve the actual symptom definition rather than aspecific grade given the use of various rectal bleeding as-sessments performed either by physicians or by patients. Inthe majority of our identified studies, rectal bleedinghad been defined as occurring occasionally, which corre-sponded to grade 1 or higher. Even though patient-reportedoutcomes are included in the post-QUANTEC data only,which can be expected to lead to a higher rate compared tothat of the same physician-assessed symptom,46 it isinteresting to note that the median rectal bleeding rate for agrade 2 or higher definition was 6% in the post-QUANTECstudies versus 15% in the QUANTEC studies (Fig. 3C).

Across all 5 GI symptoms after EBRT alone,the overall best fit was that applied to diarrhea,13,17

followed by rectal bleeding,13-15,17,19,20,22-28 fecal inconti-nence,13-16,18-21 proctitis,13,14 and last, defecationurgency13-16 (R2 Z 1.00, 0.95, 0.93, 0.93, and 0.82,respectively). The final linear-quadratic fits for diarrhea,rectal bleeding, and proctitis were similar in shape butresulted in a somewhat lower tolerance for rectal bleedingin the low- to intermediate-dose levels (relative volumesalong the curves at EQD23 Z 30 Gy and EQD23 Z 50 Gywere 67% vs 70%-72% and 44% vs 46%-47%, respec-tively; Fig. 2). Within the same dose range, the final linearfit for fecal incontinence and, even more so, the linear-quadratic fit for defecation urgency resulted in even lower


tolerances (relative volumes at EQD23 Z 30 Gy andEQD23 Z 50 Gy were �63% and �38%, respectively).These sensitivity patterns are also captured in the meta-DVH (i.e., the function fitted to the synthesizedCOMTDVHs), suggesting that treatments designed to thelower boundaries could limit overall GI toxicity (Fig. 4).Physicians designing treatments adhering to these should,however, bear in mind a patient’s individual risk of devel-opment of a certain symptom given variables related todisease, patient, and/or treatment not accounted for hereand their plausible interaction with dose.

Without any definitive consensus, the GU- and SD-specific QUANTEC reports implied that high bladderdoses4 and intermediate penile bulb doses5 are importantfor the development of GU toxicity and erectile dysfunc-tion, respectively, after EBRT alone. Unfortunately, dose/volume tolerance data for these toxicities are still scarce,and quantitative synthesis adhering to the criteria of thissystematic review was not possible. Yet, on the basis of theidentified dose/volume tolerances, recent data have sug-gested that a high or maximum bladder dose or bladderwall dose is critical for dysuria, hematuria, obstruction, andurinary urgency29-32 whereas a somewhat lower dose to thebladder wall is associated with urinary frequency.30 Only 1high-dose penile bulb threshold was reported for erectiledysfunction.33 Dose/volume tolerances for other GU andSD symptoms were not identified.

Dose/volume tolerances after BT with or without EBRThave not been investigated to the same extent as those afterEBRT alone. From the small number of identified DVHthresholds within the scope of this review, it is also clear thatdata on BTwith or without EBRTare still much less reported.A quantitative synthesis for BT with or without EBRT was,therefore, not possible. The largest number of dose/volumetolerances, a total of only 5, was identified for rectalbleeding,35-39 for which associations between both the lowerand higher rectal doses were observed, similarly to afterEBRTalone. None of the studies in which BT plus EBRTwasapplied had used intensity modulated RT for the EBRTcomponent. Moreover, the OARs investigated were limited tothe rectum (GI), urethra (GU), and penile bulb (SD).Assessing DVH thresholds after BT with or without EBRT ismore challenging than after EBRT alonedsimilar to a2-phase EBRT alone treatmentdbecause the componentsshould be summed in a radiobiologically sound manner.47,48

Furthermore, the applied fractionation correction formulamust be isotope specific and BT approach specific and able toaccount for effects at a higher dose/fraction (approximately>6 Gy).49 Among the identified studies involving BT plusEBRT, only 1 study reported summarized dose/volumetolerances after first having corrected for the fractionationeffect of each component,38 whereas others typically reporteddose/volume tolerances separately for each modality.

To our knowledge, there has not been any previousattempt to perform a meta-analysis with a quantitativesynthesis of dose/volume tolerances represented by a rangeof DVH thresholds. Typical meta-analysis and quantitative



synthesis approaches involve assessing heterogeneity forthe same measure.50 Such an approach was not suitablehere because of dose levels identified for multiple measures(i.e., volumes). We therefore devised an alternative strategyin which the relationship among dose/volume tolerancesfrom different studies was assessed by the goodness of fit ofthe applied linear or linear-quadratic functions. In addition,these dose/volume tolerances were combined into 2 to 3unbiased summary metrics (COMTDVHs) and finally com-bined into a symptom-specific COMTDVH, forming anoverall GI toxicity meta-DVH curve. Equal importance wasconsidered across individual study DVH thresholds, butalternative approaches could involve, for example,weighting with respect to study size in combination withstudy quality as given by the investigated 9-grade score.However, any additional weighting procedure was consid-ered outside the scope of this work. Ideally, stratified ana-lyses with respect to additional aspectsdfor example,symptom definition (the number of different definitionstypically did not exceed 2 for any of the investigatedsymptoms) or RT techniquedwould have shed further lighton the robustness of the fitted curves, including the meta-DVH, but this could not be performed given limited data.A crude stratification with respect to symptom definition,that is, refits of the final functions to either grade 1 orhigher (defecation urgency, diarrhea, fecal incontinence,rectal bleeding) or grade 2 or higher (proctitis) data, how-ever, indicated no substantial improvement in the goodnessof fit (final vs refitted R2 of 0.82 vs 0.85 for defecationurgency, 1.00 vs 1.00 for diarrhea, 0.93 vs 0.92 for fecalincontinence, 0.95 vs 0.96 for rectal bleeding, and 0.93 vs0.93 for proctitis). In addition, because the majority oftreatments concerned 3D conformal RT (CRT) or 3D CRT/intensity modulated RT, identified dose/volume tolerancescould not be studied with respect to the applied RTtechnique. The robustness of dose/volume tolerances withrespect to variables other than dose was further not assessedgiven that only 2 studies included such data (anticoagu-lants20 as well as cardiovascular disease and surgery23; all 3predisposing for rectal bleeding]. Future efforts are, how-ever, encouraged to untangle dose-response relationships inthe presence of all potentially influencing factors.

When dose/volume tolerance data were synthesized forthis review, 135 studies did not fulfill the inclusion criteria.The most common reason was absence of distinct symp-toms (46%) (i.e., concerned toxicities were typicallyassessed as multiple symptoms), making it impossible todistinguish the degree to which a reported symptom per-tained to this toxicity exclusively, followed by unavailableDVH data (24%) and reanalysis not adding to previousresults (13%). Five studies were excluded because ofinsufficient reporting of results, which prevented interpre-tation of DVH thresholds, and for almost 50% of theincluded studies (16 of 33), assumptions regarding DVHthresholds were made (Normal Tissue Complication Prob-ability calculations were performed in 3 studies,20,23,26

averaged DVHs were stratified between patients with and


without toxicity in 5 studies,8,14,21,25,31 and theabsolute volume threshold/maximum dose to a volumeof �2 cm3 was converted to relative values for 6studies27,32,33,37,39,41). In future and associated efforts, in-vestigators are, therefore, strongly encouraged to investi-gate and report DVH thresholds for OARs moreconsistently: (1) State the fractionation scheme (if multiple,denote number of patients [or percentage] per scheme); (2)report the DVH threshold, complication rate, and number ofpatients above and below the identified threshold; (3)include a DVH summary measure (e.g., the mean dose orthe generalized equivalent uniform dose in addition tothresholds), enabling established meta-analysis approachesand true validation of identified dose-response relation-ships; and (4) make the data publicly available as alreadyencouraged by QUANTEC almost 10 years ago51

(formatting and standardization guidelines for such effortswere recently published by the Medical Physics editorialteam52).

In addition to investigating the potential effects ofcovariates other than dose for the development of a certainsymptom, nomenclature and contouring standardization ofOARs is important.53,54 Among the included studies, thereare examples in which the OARs were defined either byanatomic landmarks or with respect to distances above and/or below tumor volumes (Table E1, available online athttps://doi.org/10.1016/j.ijrobp.2018.08.015). This reviewhas indicated that dose/volume tolerances most commonlyconcern physician-assessed rectal bleeding determinedfrom no versus any symptom severity. Although this maybe sound from a statistical modeling point of view (thelower the symptom cutoff, the higher the rate), future dose-response modeling efforts should rather focus on symptomsand severities that matter to patients. Concerning RTtechnique dependence, especially for rectal bleeding afterEBRT alone, this review could only provide incrementalimprovement compared with QUANTEC given that mostthresholds were still assessed after 3D CRT. However, morerecent 3D CRT techniques are to a larger extent moreconformal, leading to the considerable decrease in rectalbleeding rates documented in this update. Given this, it isplausible that the response to higher doses remains un-changed for GI toxicities and that data from future studieswill primarily provide more information about the low- tointermediate-dose range and associated OAR volumes.

Conclusions

This systematic review provides evidence in support ofrectal dose/volume tolerances after prostate cancer EBRTalone for defecation urgency, diarrhea, fecal incontinence,proctitis, and rectal bleeding within both the low- tointermediate-dose range (10 Gy � EQD23 � 50 Gy) andhigh-dose range (55 Gy � EQD23 � 78 Gy). The moresensitive symptom of defecation urgency implies that if itstolerance can be respected, the risk of development of other




GI toxicities is low. Even though RT-induced injuries ofvarious GI symptoms differ,55-58 our data suggest that theirdose/volume tolerances are primarily given by rectal doses.Designing treatments to meet the constraints presentedalong our meta-DVH curve (Fig. 4) to reduce GI toxicityshould be further exploited in a prospective setting. Datafor other GI toxicities, as well as GU toxicity and SD afterEBRT alone, are still scarce and even less reported after BTwith or without EBRT. To accelerate the understanding ofdose/volume tolerances for specific symptoms after state-of-the-art RT regimens, researchers are encouraged tocollect and report such data, including standardization andsharing initiatives.

References

1. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to

therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109-

122.

2. Bentzen SM, Constine LS, Deasy JO, et al. Quantitative Analyses of

Normal Tissue Effects in the Clinic (QUANTEC): An introduction to

the scientific issues. Int J Radiat Oncol Biol Phys 2010;76(suppl 3):

3-9.

3. Michalski JM, Gay H, Jackson A, et al. Radiation dose-volume ef-

fects in radiation-induced rectal injury. Int J Radiat Oncol Biol Phys

2010;76(suppl 3):S123-S129.

4. Viswanathan AN, Yorke ED, Marks LB, et al. Radiation dose-volume

effects of the urinary bladder. Int J Radiat Oncol Biol Phys 2010;

76(suppl 3):116-122.

5. Roasch M III, Nam J, Gagliardi G, et al. Radiation dose-volume

effects and the penile bulb. Int J Radiat Oncol Biol Phys 2010;

76(suppl 3):130-134.

6. Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy

Oncology Group (RTOG) and the European Organization for

Research and Treatment of Cancer (EORTC). Int J Radiat Oncol Biol

Phys 1995;31:1341-1346.

7. Common Terminology Criteria for Adverse Events (CTCAE) version

4.0. Available at: https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_

2010-06-14_QuickReference_5x7.pdf.

8. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for

systematic reviews and meta-analyses: The PRISMA statement.

PLoS Med 2009;6:1-6.

9. Forrest JL, Miller SA. Evidence-based decision making in dental

hygiene, education, practice, and research. J Dent Hyg 2001;75:50-

63.

10. Viswanathan M, Ansari MT, Berkman ND, et al. Assessing the risk of

bias of individual studies in systematic reviews of health care in-

terventions. Available at: https://www.ncbi.nlm.nih.gov/pubmedheal

th/PMH0077771/pdf/PubMedHealth_PMH0077771.pdf.

11. Bentzen SM, Dorr W, Gahbauer R, et al. Bioeffect modeling and

equieffective dose concepts in radiation oncologydterminology,

quantities and units. Int J Radiat Oncol Biol Phys 2012;105:266-268.

12. Dale RG. The application of the linear-quadratic equation to frac-

tionated and protracted radiotherapy. Br J Radiol 1985;58:515-528.

13. Gulliford SL, Foo K, Morgan RC, et al. Dose-volume constraints to

reduce rectal side effects from prostate radiotherapy: Evidence from

MRCRT01 trial ISRCTN 47772397. Int J Radiat Oncol Biol Phys

2010;76:747-754.

14. Hamlett L, Mcpartlin AJ, Maile EJ, et al. Parametrized rectal dose

and associations with late toxicity in prostate cancer radiotherapy. Br

J Radiol 2015;88:1-8.

15. Stenmark MH, Conlon ASC, Johnson S, et al. Dose to the inferior

rectum is strongly associated with patient reported bowel quality of


life after radiation therapy for prostate cancer. Radiother Oncol 2014;

110:291-297.

16. Thor M, Olsson CE, Oh JH, et al. Relationships between dose to the

gastrointestinal tract and patient-reported symptom domains after

radiotherapy for localized prostate cancer. Acta Oncol 2015;54:1326-

1334.

17. Ebert MA, Kerwyn F, Haworth A, et al. Gastrointestinal dose-

histogram effects in the context of dose-volume constrained pros-

tate radiation therapy: Analysis of data from the RADAR prostate

radiation therapy trial. Int J Radiat Oncol Biol Phys 2015;91:595-

603.

18. Buettner F, Gulliford SL, Webb S, et al. The doseeresponse of the

anal sphincter regiondAn analysis of data from the MRC RT01 trial.

Radiother Oncol 2012;103:347-352.

19. Fellin G, Rancati T, Fiorino C, et al. Long term rectal function after

high-dose prostate cancer radiotherapy: Results from a prospective

cohort study. Radiother Oncol 2014;110:272-277.

20. Schaake W, van der Schaaf A, van Dijk LV, et al. Normal tissue

complication probability (NTCP) models for late rectal bleeding,

stool frequency and fecal incontinence after radiotherapy in prostate

cancer patients. Radiother Oncol 2016;119:381-387.

21. Thor M, Apte A, Deasy JO, et al. Dose/volume-response relations for

rectal morbidity using planned and simulated motion-inclusive dose

distributions. Radiother Oncol 2013;109:388-393.

22. Benadjaoud MA, Blanchard P, Schwartz B, et al. Functional data

analysis in NTCP modeling: A new method to explore the radiation

dose-volume effects. Int J Radiat Oncol Biol Phys 2014;90:654-663.

23. Defraene G, van den Bergh L, Al-Mamgami A, et al. The benefits of

including clinical factors in rectal normal tissue complication prob-

ability modeling after radiotherapy for prostate cancer. Int J Radiat

Oncol Biol Phys 2012;82:1233-1242.

24. Drean G, Acosta O, Ospina JD, et al. Identification of a rectal sub-

region highly predictive of rectal bleeding in prostate cancer IMRT.


25. Inokuchi H, Mizowaki T, Norihisa Y, et al. Clinical effect of multi-

leaf collimator width on the incidence of late rectal bleeding after

high-dose intensity-modulated radiotherapy for localized prostate

carcinoma. Int J Clin Oncol 2016;21:156-161.

26. Liu M, Moiseenko V, Agranovich A, et al. Normal Tissue Compli-

cation Probability (NTCP) modeling of late rectal bleeding following

external beam radiotherapy for prostate cancer: A test of the

QUANTEC-recommended NTCP model. Acta Oncol 2010;49:1040-

1044.

27. Opsina JD, Zhu J, Chira C, et al. Random forests to predict rectal

toxicity following prostate cancer radiation therapy. Int J Radiat

Oncol Biol Phys 2014;89:1024-1031.

28. Someya M, Hori M, Tateoka K, et al. Results and DVH analysis of

late rectal bleeding in patients treated with 3D-CRT or IMRT for

localized prostate cancer. J Radiat Res 2015;56:122-127.

29. De Langhe S, De Meerleer G, De Ruyck K, et al. Integrated models

for the prediction of later genitourinary complaints after high-dose

intensity modulated radiotherapy for prostate cancer: Making

informed decisions. Radiother Oncol 2014;112:95-99.

30. Yahya N, Ebert MA, Bulsara M, et al. Urinary symptoms following

external beam radiotherapy of the prostate: Dose-symptom correlates

with multiple-event and event-count models. Radiother Oncol 2015;

177:277-282.

31. Thor M, Olsson C, Oh JH, et al. Urinary bladder dose-response re-

lationships for patient-reported genitourinary morbidity domains

following prostate cancer radiotherapy. Radiother Oncol 2016;119:

117-122.

32. Heemsbergen W, Al-Mamgani A, Witte MG, et al. Urinary

obstruction in prostate cancer patients from the Dutch trial (68 Gy vs

78 Gy): Relationships with local dose, acute effects, and baseline

characteristics. Int J Radiat Oncol Biol Phys 2010;78:19-25.

33. Thor M, Olsson CE, Oh JH, et al. Radiation dose to the penile

structures and patient-reported sexual dysfunction in long-term

prostate cancer survivors. J Sex Med 2015;12:2388-2397.


http://refhub.elsevier.com/S0360-3016(18)33546-6/sref1




















https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_5x7.pdf

https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_5x7.pdf







https://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0077771/pdf/PubMedHealth_PMH0077771.pdf

https://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0077771/pdf/PubMedHealth_PMH0077771.pdf























































































34. McDonald AM, Baker CB, Shekar K, et al. Reduced radiation

tolerance of penile structures associated with dose-escalated hypo-

fractionated prostate radiotherapy. Urology 2014;84:1383-1388.

35. Moulton CR, House MJ, Lye V, et al. Prostate external beam

radiotherapy combined with high-dose-rate brachytherapy: Dose-

volume parameters from deformably-registered plans correlate with

late gastrointestinal complications. Radiat Oncol 2016;11:144.

36. Okamoto M, Ishikawa H, Ebara T, et al. Rectal bleeding after high-

dose-rate brachytherapy combined with hypofractionated external-

beam radiotherapy for localized prostate cancer: The relationship

between dose-volume histogram parameters and the occurrence rate.

Int J Radiat Oncol Biol Phys 2012;82:211-217.

37. Pinkawa M, Fischedick K, Treusacher P, et al. Dose-volume impact

in high-dose-rate iridium-192 brachytherapy as a boost to external

beam radiotherapy for localized prostate cancerdA phase II study.


38. Kikuchi K, Nakamura R, Tanji S, et al. Three-dimensional sum-

mation of rectal doses in brachytherapy combined with external

beam radiotherapy for prostate cancer. Radiother Oncol 2013;107:

159-164.

39. Snyder KM, Stock RG, Hong SM, et al. Defining the risk of devel-

oping grade 2 proctitis following 125I prostate brachytherapy using a

rectal dose-volume histogram analysis. Int J Radiat Oncol Biol Phys

2001;50:335-341.

40. Wallner K, Sutlief S, Bergsagel C, et al. Severe rectal complications

after prostate brachytherapy. Radiother Oncol 2015;114:272-275.

41. Ghadjar P, Keller T, Rentsch CA, et al. Toxicity and early treatment

outcomes in low- and intermediate-risk prostate cancer managed by

high-dose-rate brachytherapy as a monotherapy. Brachytherapy

2009;8:45-51.

42. Hindson BR, Millar JL, Matheson B. Urethral strictures following

high-dose-rate brachytherapy for prostate cancer: Analysis of risk

factors. Brachytherapy 2013;12:50-55.

43. Pettersson N, Olsson C, Tucker SL, et al. Urethral pain among

prostate cancer survivors 1 to 14 years after radiation therapy. Int J

Radiat Oncol Biol Phys 2013;85:e29-e37.

44. Merrick GS, Wallner K, Bulter WM, et al. A comparison of radiation

dose to the bulb of the penis in men with and without prostate

brachytherapy-induced erectile dysfunction. Int J Radiat Oncol Biol

Phys 2001;50:597-604.

45. Wright JL, Newhouse JH, Laguna JL, et al. Localization of neuro-

vascular bundles on pelvic CT and evaluation of radiation dose to


structures putatively involved in erectile dysfunction after prostate

brachytherapy. Int J Radiat Oncol Biol Phys 2004;59:426-435.

46. Tom A, Bennett AV, Rothenstein D, et al. Prevalence of patient-

reported gastrointestinal symptoms and agreement with clinician

toxicity assessments in radiation for anal cancer. Qual Life Res 2018;

27:97-103.

47. Boman E, Kapanen M, Pickup L, et al. Importance of deformable

image registration and biological dose summation in planning of

radiotherapy retreatments. Med Dosim 2017;42:296-303.

48. Olsson C, Johansson KA. Do we need fractionation-corrected

doses in sequential two-phase treatments? A quantification of

dose differences between non-corrected and corrected combined

non-uniform dose distributions in normal tissue. Acta Oncol 2010;49:

1253-1260.

49. Astrahan M. Some implications of linear-quadratic-linear radiation

dose-response with regard to hypofractionation. Med Phys 2008;35:

4161-4172.

50. Higgins JPT, Thompson SG, Deeks JJ, et al. Measuring inconsistency

in meta-analyses. BMJ 2003;327:557-560.

51. Deasy JO, Bentzen SM, Jackson A, et al. Improving normal tissue

complication probability models: The need to adopt a “data-pooling”

Int J Radiat Oncol Biol Phys 2010;76(suppl 3):151-154.

52. Williamson JF, Das SK, Goodsitt MS, et al. Introducing the medical

physics dataset article. Med Phys 2017;44:349-350.

53. Nyholm T, Olsson C, Agrup M, et al. A national approach for

automated collection of standardized and population-based radiation

therapy data in Sweden. Radiother Oncol 2016;119:344-350.

54. Gillepsie EF, Panjawi N, Golden DW, et al. Multi-institutional ran-

domized trial testing the utility of an interactive three-dimensional

contouring atlas among radiation oncology residents. Int J Radiat

Oncol Biol Phys 2017;98:547-554.

55. Andreyev J. Gastrointestinal symptoms after pelvic radiotherapy: A

new understanding to improve management of symptomatic patients.

Lancet Oncol 2007;8:1007-1017.

56. Grodsky MB, Sidani SM. Radiation proctopathy. Clin Colon Rectal

Surg 2015;28:103-111.

57. Krol R, Smeenk RJ, van Lin EN, et al. Systematic review: Anal and

rectal changes after radiotherapy for prostate cancer. Int J Colorectal

Dis 2014;29:273-283.

58. Maeda Y, Høyer M, Lundby L, et al. Faecal incontinence following

radiotherapy for prostate cancer: A systematic review. Radiother

Oncol 2011;98:145-153.
























































































A Systematic Post-QUANTEC Review of Tolerance Doses for ...

Documents