Fractionation Survival

Question 1

What is the definition of accelerated fractionation?

Answer 1

Any schedule that gives over 10Gy/week

Question 2

What is the definition of accelerated hyperfractionation?

Answer 2

Any schedule that gives >1 Gy per day

Question 3

What are the three kinds of boost schedules?

Answer 3

1) Concomitant boost: 1 fx to large area, 1 fx to boost area, daily (2fx daily)
2) SIB: 1 dose to large area, larger dose to boost area, daily (1fx daily)
3) Sequential: conedown boost

Question 4

What are the 4 Rs of radiotherapy that are modulated by fractionation?

Answer 4

Repair
Reoxygenation (cycling hypoxia)
Redistribution (in cell cycle to sensitive phase)
Repopulation

Question 5

What is the difference between “sublethal” and “potentially lethal” damage?

Answer 5

SLD: damage that is not lethal on its own, but when added with more damage can be lethal (e.g. ssDNA break)

PLD: damage that is lethal during cell division but has the potential to be repaired if cell division is stalled

Question 6

How are SLD and PLD experimentally derived?

Answer 6

SLD: SpLit Dose (lethality as a function of time between doses)
PLD: PLating Delay (lethality as a function of time to cell division)

Question 7

What is the time sequence of repair?

Answer 7

Half-life approx 1h in cell culture

Essentially complete by 6h

Question 8

Poisson statistics describe a ______ number of random events happening to a ______ number of subjects, averaging to a _______ number of events per subject.

Answer 8

1) large
2) large
3) small

Question 9

For Poisson statistics, for an average of X events per cell, what is the percentage that have

Answer 9

X events = (1 - e^-X)

Thus for 1 event per cell
e^-1 = 0.37 have <1 event
0.63 have at least 1 event

For 2 events per cell
e^-2 =.14 have <2 events
0.86 have at least 2 events

For 2.3 events per cell
e^-2.3 = 0.10
etc.

Question 10

What is D0 defined as?

Answer 10

D0 is the radiation dose that results in 37% survival, assuming one lethal event per cell leads to death.

Question 11

How is tumor control probability calculated?

“What is the TCP for 0.01 cells left per patient?”

Answer 11

Using Poisson statistics, setting X as 0.01 cells per patient:
TCP = e^-.01 (chance for <0.01 cells alive per patient)
TCP = 0.99 = 99% chance of tumor control

Question 12

Describe the single-hit, multi-target model. What is the equation for this?

Answer 12

Every cell has multiple independent targets (‘n’), all targets must be hit to kill the cell (hitting a single target is sublethal).

SF (Dose) = 1 - [1-(e^-D/D0)]^n

Question 13

Where is D0 calculated on a survival curve?

Answer 13

In the linear portion. Thus, should actually calculated D0 as the additional dose required to drop survival from 0.1 to 0.037

Question 14

How is the extrapolation number n calculated?

Answer 14

The linear part of the survival curve (where D0 is calculated) is extrapolated back through the Y-axis. this gives a relative approximation of the shoulder of the curve – a wider shoulder begets a larger extrapolation number

Question 15

How is the quasi-threshold dose calculated?

Answer 15

The linear part where D0 is identified, the extrapolation line drawn back to the Y-intercept. The extrapolation number is always >1. Where the extrapolation line crosses a value of 1 is the Dq. The Dq is significant because this is where sublethal damage occurs – below this dose there is likely negligible clonogenic inactivation

Dq = D0 * ln(n)

Question 16

What is the dose that will reduce a surviving fraction to 37% of what it was before? To 10%?

Answer 16

D0 = where 37% survive
D10 = 2.3 * D0

Question 17

Why is Dq important, what does it signify?

Answer 17

Recall that Dq is a function of the extrapolation number, and that a larger shoulder = larger n. Well, a larger shoulder also means a larger Dq. The shoulder reflects repair capacity (larger shoulder = higher dose to overwhelm repair); larger Dq = larger repair capacity

Question 18

Where on the curve is the SHMT model (un)realistic?

Answer 18

Models higher dose better than lower dose

Question 19

How do you solve: “What is the dose needed to kill 99% of tumor cells?”

Answer 19

It is asking for SF=0.01

By LQ model: SF(D) = e^-(aD+bD2)

Question 20

How do you solve: “What is the dose needed to give 99% tumor control with a starting population of 10^9 cells?”

Answer 20

It is asking for TCP = 0.99 –> SF = 0.01/10^9 = 10^-11

Question 21

What is the LQ equation for survival?

Answer 21

SF(D) = e^-(aD+bD2)

Question 22

What does a low a/b ratio indicate?

Answer 22

High repair capacity; relatively resistant at small fractions and sensitive at large fractions

Question 23

What does a high a/b ratio indicate?

Answer 23

Low repair capacity: sensitive at small fractions and resistant at large fractions

Question 24

What is the a/b ratio of CNS?

Answer 24

1-2

Question 25

What tumors have an a/b ratio close to 1.5-4?

Answer 25

Prostate and Breast

Question 26

What is the equation for BED?

Answer 26

BED(a/b/) = n * d * [1+ d/(a/b/)]

for ‘n’ fractions of ‘d’ dose

Question 27

What is the equation for EQD2?

Answer 27

EQD2 = n * d * [(a/b)+d/(a/b)+2]

Question 28

What is the Thames H-factor?

Answer 28

It corrects for incomplete repair, applied to BID or TID regimens. The dose per fraction is multipled by (1+H)

Ex: For H2 correction factor of 0.2, 1.5 Gy given BID is equivalent to what QD fraction dose?
= 1.5 * (1+0.2)
= 1.5 * 1.2
= 1.8 QD fraction

Thus, 45 Gy given in 30 1.5 BID fractions = 30 x 1.8 = 54 Gy

Question 29

Where on the curve is the LQ model (un)realistic?

Answer 29

Fits data well at low doses, poorly at high doses where there may be additional mechanisms for cell survival