1

Radiobiological Rationale of Hypofractionation, Clinical Relevance,

Risk of Late Toxicities, and

Prediction Possibilities 

Barry S. Rosenstein, Ph.D.

Departments of Radiation Oncology,Mount Sinai School of Medicine and

NYU School of Medicine

Deuxieme Rencontre du Cercle Des Oncologues Radiotherapeutes du Sud Meridien

Juan les Pins26 Juin 2009

2Claudius Regaud

(1870-1940)

http://www.pasteur.fr/infosci/archives/

e_reg0.html

3Hall, Radiobiology for the

Radiologist, 2000

4

Henri Coutard

(1876-1950)radonc.ucsd.edu/.../

historyImages/Coutard.jpg

5

DISCUSSION.-DR. MAURICE LENZ (New York):

It had been realized for a long time that large doses were essential for clinical arrest of cancer by roentgenotherapy. This could frequently not be carried out because of concomitant roentgen ray injury to adjacent normal tissues, especially indeeply situated and not markedly radiosensitive malignant tumors. Coutard reduced this handicap by applying to practice the principle of fractionatingand protracting the total dosage over a longer period. This he did- at the suggestion, and on the basis, of experimental work carried out on ram's testesby Regaud.

6Strandquist Acta Radiol 55 (suppl):1, 1944

7

Total Dose = (NSD) T 0.11 N 0.24

 N = number of fractionsT = overall time

NOMINAL STANDARD DOSE (NSD) SYSTEM

Ellis, Br J Radiol 44:101, 1971

8

What’s Wrong with the NSD System

and the Resulting Time, Dose and Fractionation

(TDF) Tables?

9

1.T0.11 predicts a large increase of isoeffect dose at first, then increasing more slowly. The biological fact is just the opposite: it shows no increase at first and then a rapid rise of isoeffect dose as proliferation accelerates.

2.The time factor is underestimated for tumors and early‑responding tissues.

3.The time factor is overestimated for late‑responding tissues.

THE TIME FACTOR

10

1.N0.24 does not predict the severe late damage that occurs for larger fraction sizes.

2.The impact of fractionation is underestimated for late-responding tissues and possibly some forms of cancer.

3.Fraction size, not number, is the important parameter.

THE FRACTION NUMBER

11

TDF tables were too easy to use without thinking rigorously about the impact of fraction size, proliferation rates and the potential for incomplete repair between fractions.

GENERAL CRITICISM

12

How can we more accurately estimate the

impact of fraction size for tumor control as well as early and late effects?

13

BIOLOGICALLY EFFECTIVE DOSE (BED)

BED = nd[1+(d//)]

n = number of fractionsd = dose per fraction/ = parameter, in units of Gy, characteristic of the impact of fraction size on the particular tissue or tumor

BED = (total dose)(relative effectiveness)

BED is the quantity by which different fractionation regimens can be compared.

14

Where do / values come from?

15Withers, Cancer 55:2086,

1985

16

Calculation of /

Values

(n1d1)[1+(d1//)] = (n2d2)[1+(d2//)]

/ = (D2d2-D1d1)/(D1-D2)

http://www.dkfz-heidelberg.de/en/medphys/appl_med_rad_physics/images/Biology_1_re.jpg

18Hall, Radiobiology

for the Radiologist, 2000

19

Normalized Total Dose (NTD)

or2 Gy Equivalent Dose

The total dose delivered in 2 Gy fractions that corresponds to a particular BED.

NTD = BED/[1+(2 Gy//)]

20

What is the impact of treatment time?

21Withers et al., Acta

Oncologica 27:131, 1988

22

ALLOWANCE FOR CELLULAR PROLIFERATION

BED=nd[1+(d//)] – [(loge2) (T-Tk)/Tpot ]

= (Total Dose) (Relative Effectiveness) – (loge2/) (Number Cell Doublings During Treatment)

T - total treatment timeTk ("kick-off" time) - time at which compensatory proliferation or accelerated repopulation begins. - parameter associated with cellular radiosensitivityTpot – time required for cells comprising the tumor or normal tissue to double in number

23

LATE EFFECTSPROTOCOL DESCRIPTION BED

(Gy4

)

2 Gy Eq Dose

25 fx X 2 Gy (32 days) Standard whole breast 75 5023 X 2 Gy (30 days) Standard whole breast 69 46

12 X 3.94 Gy (37 days)

Scandinavian 1970s post-mastectomy

94 6313 X 3.2 Gy (32 days) The Start Trialists’ Group

Lancet Oncol 9:331, 200875 50

15 x 2.67 Gy (18 days)

The Start Trialists’ Group Lancet 371:1098, 2008

67 4516 x 2.6 Gy (22 days) Whelan et al JNCI 94:1143,

200271 47

10 x 3.85 Gy (5 days) RTOG 0413; rtog.org/members/protocols/0413/041

3.pdf76 51

15 x 2.7 Gy (19 days) Formenti et al. JCO 16:2236, 2007 (NYU Prone AIMRT)

68 455 x 6 Gy (10 days) Formenti et al. IJROBP 60:493,

2004 (NYU Prone Partial Breast)

75 50

24

TUMOR CONTROL

PROTOCOL DESCRIPTION BED (Gy4

)

2 Gy Eq Dose

25 fx x 2 Gy (32 days) Standard whole breast 71 4730 fx x 2 Gy (39 days) Standard with 7 fx X 2 Gy

boost84 56

12 x 3.94 Gy (37 days)

Scandinavian 1970s post-mastectomy

88 5913 x 3.2 Gy (32 days) The Start Trialists’ Group

Lancet Oncol 9:331, 200871 47

15 x 2.67 Gy (18 days)

The Start Trialists’ Group Lancet 371:1098, 2008

67 4516 x 2.6 Gy (22 days) Whelan et al JNCI 94:1143,

200271 47

10 x 3.85 Gy (5 days) RTOG 0413; rtog.org/members/protocols/

0413/0413.pdf

76 51

15 x 3.2 Gy (19 days)

5 x 6 Gy (10 days)

Formenti et al. JCO 16:2236, 2007 (NYU Prone AIMRT)

Formenti et al. IJROBP 60:493, 2004 (NYU Prone Partial

Breast)

8675

5750

25

Are BED Calculations Appropriate for Protocols Using Fraction Sizes >8Gy

PROBABLY NOT!

/ ratios determined for protocols using fraction sizes <8Gy•Does not adequately take into account microvasculature; doses >8Gy produce endothelial cell damage due to activation of acid sphingomyelinase triggering apoptosis•Cancer stem cells resistant to doses <8Gy•Radiosurgery doses that produce adequate tumor control would have been predicted as insufficient

BED CALCULATIONS, ALTHOUGH A USEFUL GUIDE FOR RESEARCH PURPOSES OR TO SERVE AS A

YARDSTICK BY WHICH TO JUDGE NEW FRACTIONATION SCHEMES, ARE NOT TO BE CONSIDERED A

SUBSTITUTE FOR CLINICAL JUDGEMENT AND TRAINING.

27

THE Gene-PARE PROJECT

Genetic Predictors of Adverse Radiotherapy Effects 

28

HYPOTHESISPeople who are carriers of certain single nucleotide polymorphisms (SNPs) may be more likely to develop adverse radiation effects compared with individuals who do not possess these DNA markers.

29

OVERALL GOALS

1. To develop an assay capable of predicting, with a high level of sensitivity and specificity, which cancer patients are most likely to develop radiation injuries resulting from treatment with a standard radiotherapy protocol.

2. To obtain information to assist with the elucidation of the molecular pathways responsible for radiation-induced normal tissue toxicities through identification of genes possessing SNPs associated with the development of adverse effects.

30

If we can identify SNPs associated with radiosensitivity and develop a predictive assay, what can be done with this information?

•Receive a strictly surgical treatment, if feasible

•Receive more of a conformal treatment (i.e. IMRT, protons, etc.)

•Could be ideal radiotherapy candidates as their cancers may also be radiosensitive; standard treatment overdoses

31

SINGLE NUCLEOTIDE

POLYMORPHISM

SNP

32

Humans are 99.9% Identical Genetically

33

...but, 100% of Humans Differ Genetically

35

PERSONALIZED MEDICINE

The use of detailed information about a person’s genotype in order to select a medication, therapy or preventative measure that is particularly suited specifically to that individual.

36

RADIOGENOMICS

Predicting radiotherapy response

of cancer patients based upon genetic

profiles

37Mark et al., Nature Reviews Genetics 9, 356-369, 2008

38

CANDIDATE GENE STUDIES

1998-2008

39

RADIATION RESPONSE RADIATION RESPONSE GENES SCREENEDGENES SCREENED

ATMATM – – regulation of cell cycle checkpoints following irradiationregulation of cell cycle checkpoints following irradiation

SOD2SOD2– Response to reactive oxygen speciesResponse to reactive oxygen species

RAD21RAD21– Repair of DNA double strand breaksRepair of DNA double strand breaks

TGFB1TGFB1– Fibrosis, proliferation, differentiation, angiogenesis and Fibrosis, proliferation, differentiation, angiogenesis and

wound healingwound healing

XRCC1XRCC1– Base excision repairBase excision repair

XRCC3XRCC3– Recombinational repair of DNA double strand breaksRecombinational repair of DNA double strand breaks

40

DANISH POST-MASTECTOMY BREAST CANCER PATIENTS

WHO RECEIVED A HYPOFRACTIONATED

PROTOCOL

41

42

Low lymphocyte apoptosis = increased late side effects

Cumulative incidence of grade 2 or more late side effects according to radiation-induced CD8 T-lymphocyte apoptosis in 399 patients

Low apoptotic response (≤16%)

Intermediate apoptotic response (16-24%)

High apoptotic response (>24%)

Ozsahin et al, Clin Cancer Res 2005

43

Distribution of SNPs According to Radiation-Induced Late Effects

44

THE PROBLEMS WITH

CANDIDATE GENE STUDIES

45

1. Although a number of studies have detected correlations between possession of a minor SNP allele with an increased incidence of radiation toxicity, the results of early studies have not always been validated in subsequent work.

46

2. There is relative ignorance of the full spectrum of genes and proteins that are associated with the development of radiation injury.

47

3. Even if all of the important genes that encode the essential protein products associated with radiation toxicity were included in candidate gene studies, it is not certain whether all of these genes would possess SNPs that would both alter protein function and be present at a high enough frequency in the population to be of importance.

48

4. Critical SNPs associated with radiosensitivity may not even be located within genes, but in regulatory portions of the DNA.

49

Genome-Wide SNP Association Studies

50

THE AGNOSTIC APPROACH

51

52

53IlluminaAffymetrix

SNP (Genotyping) Arrays

54

55

Cost of SNP Genotyping

1998 - $4 per SNP2009 - $0.0004 per SNP

~10,000-fold decrease in the cost of SNP genotyping in the past decade!!!

56

We need to develop a large consortium to create

BIGBiorepositories of tissue samples and Databanks

derived from well-characterized irradiated subjects

57

Creation of an International Consortium to Establish a Radiotherapy Patient Biorepository/DatabankNIH- Sponsored Conference, March 18, 2008

Presenter Institution RT Populations that may be Contributed to the Biorepository/Databank

David Azria CRLC Val d’Aurelle, Montpellier, France

Breast and prostate cancer patients treated in CO-HO-RT, PHRC and BONBIS European cooperative trials

Yuhchyau Chen University of Rochester Medical Center, Rochester, NY

Patients treated by investigators who are members of CURED (Cancer Survivorship Research and Education)

Karen Drumea Rambam Medical Center, Haifa, Israel

Breast, prostate, head&neck and cervical cancer patients treated at the Rambam

Silvia Formenti NYU Medical Center, New York, NY

Breast cancer patients treated under NYU protocols

Debra Friedman Fred Hutchinson Cancer Research Center,

Seattle, WA

Patients treated with total body irradiation at the Fred Hutchinson Cancer Center

Bruce Haffty_________________

Germaine Heeren

UMDNJ-New Brunswick, NJ___________________________

ESTRO, Brussels, Belgium

Breast cancer patients treated at UMDNJ, Yale and Korea_______________________________________________

Patients enrolled in the GENEPI Biorepository

Alice HoMemorial Sloan-Kettering Cancer

Center, New YorkBreast cancer patients treated at MSKCC

Mayumi Iwakawa National Institute of Radiological Sciences, Chiba, Japan

Breast, prostate, head&neck and cervical cancer patients enrolled in the RadGenomics Biorepository

Shannon MacDonald

Massachusetts General Hospital Boston, MA

Breast and pediatric cancer patients treated at MGH

Thomas Merchant St. Jude Children’s Research Hospital, Memphis, TN

Pediatric patients treated at St. Jude

Mahmut Ozsahin Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland

Cancer patients treated under EORTC protocols

Matthew Parliament Cross Cancer Institute/University of Alberta, Edmonton, Canada

Prostate, breast and head&neck cancer patients treated at the Cross Cancer Institute

Barry Rosenstein Mount Sinai School of Medicine, New York, NY

Patients enrolled in the Gene-PARE biorepository, the U.K. Start trial and RTOG

58

The

RaPiDInternational Consortium

Radiotherapy Patient Biorepository and Databank

59

INTERNATIONAL RADIOGENOMICS CONSORTIUM

17-18 November, 2009

Manchester, UK

GOAL

To provide a collaborative structure for the international radiation oncology research community to pool data to discover the genetic basis for individual differences in susceptibility for the development of radiation injuries resulting from radiotherapy.

60

STUDY INVESTIGATORS

American Collaborators

Mount Sinai School of Medicine NYU School of MedicineDavid Atencio, Ph.D. Silvia Formenti, M.D.Ryan Burri, M.D. Harry Ostrer, M.D.Jamie Cesaretti, M.D.Grace Fan, M.D. Yale/UMDNJSheryl Green, M.D. Bruce Haffty, M.D.Alice Ho, M.D.Lynda Kusnetz, B.A.Karen Loeb, M.D.Christopher Peters, M.D.Sheila Peters, B.A. Richard Stock, M.D. Nelson Stone, M.D.Sylvan Wallenstein, Ph.D.

61

INTERNATIONAL COLLABORATORS

Aarhus University Hospital, Denmark CRLC Val d'Aurelle, FranceJan Alsner, M.D. David Azria, M.D., Ph.D.Nicolaj Andreassen, M.D. Nigel Crompton, Ph.D.Jens Overgaard, M.D. Jean-Bernard Dubois, Ph.D.Marie Overgaard, M.D Andrew Kramar, Ph.D.

Françoise Mornex, M.D.Institute for Cancer, England André Pèlegrin, M.D.Roger A’Hern, M.Sc.Soren Bentzen, Ph.D. (Wisconsin) CHUV, Lausanne, SwitzerlandLone Gothard, H.N.D. René-Olivier Mirimanoff, M.D.Jo Haviland, M.Sc. Mahmut Ozsahin, M.D., Ph.DRoger Owen, M.D.Georges Sumo, M.S. Rambam Medical Center, IsraelMark Sydenham, B.Sc. Abraham Kuten, M.D.John Yarnold, M.B.B.S. Karen Drumea, M.D.

62http://jamescbender.com/bendersblog/Default.aspx