1. Understanding Radiotherapy-Induced Second Cancers
and Potentially Reducing
Understanding Radiotherapy-Induced Second Cancers 
David Brenner and Igor Shuryak
Center for Radiological Research
Columbia University New York
Understanding radiogenic cancers resulting from contemporary radiotherapy protocols

2. There is increasing concern about radiotherapy-related second cancers
15-year relative survival rate for patients treated for breast or prostate cancer is 75% (c.f., 58% for breast in 2001)
Estimated risk of developing a radiation-induced second cancer for 10+ year prostate RT survivors treated with RT around the 1980s was ~1.5%**
As younger patients are treated, and with longer life expectancy, RT-induced second malignancies will likely assume increasing importance
** SEER analysis Brenner et al (2000)

3. There is also an increasing realization that lifetime cancer risks due to radiation exposure in middle age may be larger than we thought
4000
From BEIR-VII (2006)
2000
Shuryak et al 2010
BEIR
BEIR-VII
Age at Exposure

4. 2,000 prostate cancer patients treated with RT (1984 to 2005) vs
2,000 prostate cancer patients treated with RT (1984 to 2005) vs. matched prostate cancer patients who underwent surgery
14%
28%
Data from William Beaumont Hospital
Huang et al 2011

5. Estimating second cancer risks after contemporary radiotherapy
Retrospective epidemiology necessarily relates to RT protocols several decades ago
Different prescription doses
Different fractionation schemes / dose rates
Different normal-tissue dose distributions
relates

6. Example: Second Cancers: IMRT vs. 3-D conformal RT
Compared to the older 3-D conformal radiotherapy, modern IMRT techniques minimize the amount of normal tissue getting high doses
But IMRT does result in larger volumes of normal tissue getting lower doses (more fields and more leakage)
Which is preferable in terms of second cancers?
Small volumes of normal tissue getting high doses (3D-CRT)
Larger volumes of normal tissue getting low doses (IMRT)

7. Example: Second Cancers: IMRT vs. 3-D conformal RT
Key is the shape of the dose-response relationship for radiation-induced carcinogenesis...
High doses don’t matter
High doses do matter OR
Cancer Risk
total dose
fewer
total dose
total dose
IMRT minimizing high doses doesn’t help
IMRT’s extra lower doses are bad
IMRT minimizing high doses helps
IMRT’s extra lower doses less important

8. The standard model of carcinogenesis at high doses: Competition between oncogenic transformation & cell killing
SURVIVAL
TRANSFORMATION
DOSE
Gray 1965
SURVIVAL
ONCOGENIC TRANSFORMATION

9. RT-induced breast cancer
However, recent epidemiology suggests that the risks are not small at large doses
RT-induced breast cancer
Hodgkins data:
Travis 03,
Van Leeuwen 03

10. RT-induced lung cancer
However, recent epidemiology suggests that the risks are not small at large doses
RT-induced lung cancer
Repopulation
Hodgkins data:
Gilbert 2003

11. Cell numbers during RT and subsequent normal-tissue repopulation
End RT
is this with treatment on weekends?
Radiation-induced pre-malignant cells
Sachs & Brenner 2005

12. Cancer risks at high doses: A 3rd significant mechanism
Proliferation of pre-malignant cells during organ repopulation
proliferation of normal cells *during* treatment is also important, because it increases the number of cells at risk during later fractions
We know enough about repopulation mechanisms to be able to add them to the standard (Gray) model of radiation-induced cancer at high doses
Sachs & Brenner 2005

13. How to calculate cancer risks at high doses, which are organ-specific, age-specific, and gender-specific....
Estimate the low dose (~2 Gy) age- gender- and organ-specific relative risks from A-bomb survivors
Use standard models to “convert” these low dose relative risks to apply to Western population / individual of given age and gender
Extrapolate these low-dose risks to fractionated high doses using mechanistic models (initiation / killing / repopulation)
Sachs & Brenner 2005

14. Radiation-induced breast cancer: Excess relative risk at high doses
Mean exposure age: 23
use inititiation/killing/repoulation model – you have room
Brenner et al 2006 JNCI 98: (2006)
PNAS 102: (2005)

15. Radiation-induced lung cancer: Excess relative risk at high doses
Mean exposure age:45
use “initiation, killing, repopulation” – you have room
Brenner et al: JNCI 98: (2006)
PNAS 102: (2005)

16. Example: Second Cancers: IMRT vs. 3-D conformal RT
Key is the shape of the dose-response relationship for radiation-induced carcinogenesis...
High doses do matter
High doses don’t matter
Cancer Risk
total dose
fewer
total dose
total dose
IMRT minimizing high doses doesn’t help
IMRT’s extra lower doses are bad
IMRT minimizing high doses helps
IMRT’s extra lower doses less important

17. Such models can do a reasonable job of modeling radiotherapy-induced second-cancer risks for many sites
BLADDER
BREAST
CNS
COLON
LUNG
PANCREAS
RECTUM
STOMACH
THYROID
Brenner et al 2009

18. Lifetime absolute risks, as a function of age at exposure
Blue = BEIR VII (2006)
Red = 2010 analysis
Excess lifetime risks per 0.1 Gy per 105 persons
Shuryak et al JNCI 2010

19. Based on these approaches, we can make predictions of second-cancer risks for modern radiotherapeutic protocols
Koh et al 2007
Bilateral breast DVH
30 year old female, 35 Gy mantle RT, 20 fractions
Breast cancer ERR after 20 years
30 year old female, 20 fractions +
Contributions of different doses to the overall risk
ERR = 2.1 [1.1, 6.1]

20. A potential application: Reducing Second Breast Cancers

21. Large genetically-based second-cancer risk in breast-cancer survivors
A potential application: Reducing Second Breast Cancers
1. Second breast cancer in the contralateral breast
Data from Freedman et al 2005
Large genetically-based second-cancer risk in breast-cancer survivors
Mean age at 1st cancer: 57
Brenner et al. JCO 2007

22. A potential application: Reducing Second Breast Cancers
2. Second breast cancer in the ipsilateral breast
In the ipsilateral breast, the risk of a genetically-based second-cancer has been essentially eliminated
Data from Freedman et al 2005
Brenner et al. JCO 2007

23. Why is there no genetically-based second-cancer risk in the ipsilateral breast?
Likely explanation is related to the ~46 Gy fractionated dose to the ipsilateral breast
Only about 1 in 106 cells will survive this fractionated dose
So assuming there at most a few thousands of background pre-malignant stem cells in the breast, they will all be sterilized

24. Prophylactic mammary irradiation (PMI) to the contralateral breast
If whole breast irradiation has eliminated all the background pre-malignant stem cells in the ipsilateral breast ....
prophylactic mammary irradiation (PMI) to the contralateral breast would have the potential to eliminate the large background risk in that breast
PMI would need much lower dose than the ~46 Gy ipsilateral breast dose, as we are only trying to kill relatively small numbers of pre-malignant cells, not millions of tumor cells

25. Irradiating healthy normal tissue?????
The contralateral breast of a breast cancer survivor is not a healthy normal tissue

26. What PMI dose to the contralateral breast would be needed?
So a realistic PMI fractionated dose would be around 20 Gy
Much lower than the standard post-lumpectomy RT dose
Need to consider the risk of radiation-induced cancer
Predicted PMI-induced breast cancer risk is ~4% at 20 yrs
So if PMI eliminates a ~15% contralateral breast cancer risk, it would have a favorable benefit / risk ratio
Brenner et al. JCO 2007

27. Experimental investigations of PMI
MMTV-PyVT mice
Relative risk of breast cancer after PMI
Relative Breast Cancer Risk
PMI Dose

28. PMI for BRCA1/2 carriers
Second contralateral breast cancer in BRCA1/2 carriers is very frequent.... ~40% at 15 years
The benefit / risk balance for contralateral PMI is probably even more favorable for BRCA1/2 carriers, but there are uncertainties
Major pluses for BRCA1/2 carriers are that PMI is
estrogen independent
a breast conserving option, compared with prophylactic contralateral breast mastectomy

29. Implications for current partial breast irradiation approaches?
Should we be adding a whole-breast PMI dose to current partial breast irradiation techniques?

30. Prophylactic Mammary Irradiation Conclusions
Low-dose PMI of the contralateral breast, given at the same time as conventional post-lumpectomy RT, may significantly reduce the large risk of second cancer in the contralateral breast of breast cancer survivors
Independent of estrogen status
Cost effective
Need to balance the risk of radiation-induced cancer but overall PMI is likely to have a favorable benefit / risk balance
Benefit / risk ratio is likely to be still better for BRCA1/2 patients, who are subject to very large second-cancer risks
PMI is a breast-conserving option, c.f. prophylactic contralateral breast mastectomy

31. Overall Conclusions
As long-term cancer survival rates increase, there are increasing concerns about radiation-induced second cancers
Better models are giving us a better understanding about whether we need to be more concerned about large doses to small volumes of normal tissue, or about smaller doses to larger volumes…
We can potentially use our understanding of radiation-induced cancers to combat a major problem, contralateral second breast cancer, through prophylactic mammary irradiation