Tumors Models, and

Response of Tumors

Martin BrownApril 23, 2012

Hanahan and Weinberg, Cell 2011

Cancers comprise more than tumor cells. Do the stromal cells affect the response to radiotherapy?

Capillary

Endothelial cell

Endothelial cells should be a prime target in radiotherapy – Julie Denekamp

•Endothelial cells in tumors are rapidly dividing so will die rapidly after irradiation

•One endothelial cell supports 2000 tumor cells!

Tumor curability by IR does not depend on the sensitivity of the host (Budach et al, 1993)

TD50 assay for in vivo response of tumors

(Hewitt and Wilson, 1959)

Cell survival curve of leukemia cells irradiated in situ and in vitro (Hewitt & Wilson,1959)

Lung colony assay (Hill and Bush)

Tumor Growth Delay Assay

Tumor growth delay following different radiation doses.

Demonstration of Tumor Radiosensitization using TCD50 assay (Sheldon et al, 1974)

In vivo/in vitro (excision) assay

This is the fastest, cheapest and most accurate assay of the response of the tumor cells. However, it makes the assumption that the response of the cells is not altered by taking them out of the tumor.

Demonstration of “hypoxic fraction” in mouse tumors (Rockwell & Kallman, 1972)

Spheroids: an in vitro tumor model.

Mimics several aspects of tumors

-chronic hypoxia adjacent to necrosis

-non dividing cells

-cell to cell contact

-limited diffusion of drugs

Tumor Hypoxia results from Differences in the Vasculature between Tumor and Normal Tissues

Blood vessels Necrosis

HIF-1 staining

Chronic Hypoxia (HIF-1 staining) in Human Head & Neck Cancers(Aebersold et al,

2001)

Oxygenation of human tumors is usually measured with an oxygen electrode (“Eppendorf”)

Oxygen tensions of Normal Tissue and Nodes of Head and Neck Cancers

0

4

8

12

16

20

1000 10 20 30 40 50 60 70 80 90

0

4

8

12

16

20

0 10 20 30 40 50 60 70 80 90 100

Oxygen Partial Pressure (mm Hg)% o

f M

easu

rem

en

ts

Normal Subcutaneous Tissue

Tumor

Median pO2

0.0

0.1

0.2

0.3

10 20 30 40 50 60 700 80

Median pO2 (mmHg)

Median pO2 Values for Tumor and Normal Tissues for 62 Head & Neck Cancer Patients (Le et al 2004)

Median pO2 values for Tumor and Normal

Proportionof

Values

Tumor Median= 11.8 mmHg

Normal Median= 51.9 mmHg

Effect of Tumor Hypoxia on Local Control of Head and Neck Cancer (Brizel et al, 1999)

63 patients treated.Similar differencesfor survival and DFS.

Oxygenation of Prostate Carcinoma by Eppendorf Electrode

Movsas et al, 2001 Parker et al, 2004

Med = 2.4mmHg Med = 4.5 mmHg

Effect of tumor hypoxia on outcome of prostate ca to RT (Movsas et al, 2002)

pO2 = 5.3 mmHg

pO2 = 1.0 mmHg

Biochemical failure by PSA

Median Oxygen Levels of Human Tumors

(Brown & Wilson Nat. Rev. Cancer 4: 437 2004)

0.1 1 10 1001.0

1.5

2.0

2.5

3.0

OE

R(p

)

Oxygen Partial Pressure (p) (mm Hg)

It is incorrect to think that tumors are composed of hypoxic and aerobic cells. In fact

the majority are at intermediate pO2

Koch et al,1984

Whillans and Hunt,1982

Hypoxic AerobicIntermediate

0 10 20 30 40 50 60-10

-9

-8

-7

-6

-5

-4

-3

-2

-1

10

10

10

10

10

10

10

10

10

10

100

No Hypoxic Cells

Sur

vivi

ng F

ract

ion

Total Dose (2 Gy Fractions)

10% hypoxic cells with full reoxygenation (binary)

10% hypoxic cells with no

reoxygenation

When the cells at intermediate hypoxia are considered the predicted survival to 2 Gy

fractions is dramatically altered

10% hypoxic cells and intermediate p02 cells with full reoxygenation

Wouters and Brown,1997

Consequences of Tumor Hypoxia for Cancer Treatment

• Hypoxic cells are resistant to killing by IR

– Extent of hypoxia affects response to radiotherapy

• Hypoxia is associated with slowing of proliferation:

– Leads to resistance to most anticancer drugs.

• Hypoxia promotes and selects for a more malignant phenotype

– Hypoxic tumors are more metastatic

Green: hypoxiaRed: proliferation (IdUrd+)

Hypoxia and proliferation in human H & N tumors (Van der Kogel et al)

Human Glioblastoma Stained with EF5 (Red) and Ki-67 (Green) (Evans and Koch,

2002)

Capillary

Distance from Capillary (µm)

Su

rviv

ing

Fra

ctio

n

O2 O2

150100500

. Radiation / Chem. DrugO2

Expect gradient of cell killing from blood vessels to necrosis

X-ray Survival of Cells in SiHa Tumors as a function of Distance from Blood Vessels

(Durand & Olive, 1997)

HypoxicAerobic

Expression of hypoxia induced proteins (eg CA IX) can be used in archival tissues as a surrogate for hypoxia.

Head and Neck

BreastAdenocarcinoma

Ovarian Adenocarcinoma

Immunohistochemical staining for CA IX

Wykoff et al, 2000

Hui et al, 2002

NPC

Can we exploit tumor hypoxia?

• A drug specifically toxic to hypoxic cells

would kill only tumor cells - and the

most resistant ones.

0 1000 2000 3000 4010 -5

10 -4

10 -3

10 -2

10 -1

100

TPZ Conc (µM)

Hypoxia

Air

00

Preferential Toxicity of Tirapazamine (TPZ) to Hypoxic Cells in vitro

Mouse SCCVII cells 1 hr exposure

100 101 102 103 10410-5

10-4

10-3

10-2

10-1

100

TPZ Conc (µM)

HCR=300

Su

rviv

ing

Fra

ctio

n

Mouse SCCVII cells 1 hr exposure

2

O

NN

N NH

O

TPZ

Capillary

Distance from Capillary (µm)

Su

rviv

ing

Fra

cti

on

O2 O2

Addition of a Hypoxic Cytotoxin to Standard Treatment Could Exploit Tumor Hypoxia

Combined

Hypoxic Cytotoxin

150100500

. Radiation / Chem. DrugO2

Complementary Killing by TPZ and IR in SiHa Tumors as a function of Distance from Blood

Vessels

(Durand & Olive, 1997)

HypoxicAerobic

Combined

Primary site failure in 92 randomized advanced H&N patients (Peters et al,2007,)

PET hypoxia status

Treatment

P-valueRT + cis /no TPZ

RT + cis/TPZ

Non-Hypoxic

2/27 (7%) 3/21 (14%) NS

Hypoxic 8/18 (44%) 0/26 (0%) 0.0002

P-value 0.008 NS

Results of multicenter randomized clinical trial of TPZ with advanced H & N cancer

693 patients without major protocol violationsAll 863 patients in 89 sites in 16 countries

Rischin et al and Peters et al, JCO, 2010

Need to select hypoxic tumors in future trials and have good QA for radiotherapy

100806040200

100

1000

10000

90% probability80% probability

% Patients with Hypoxic Tumors

Number of

Patients

Patient Numbers Needed to Detect a Change from 40 to 65% Response

Rates

SN30000 is superior to TPZ in multiple tumor models

Single doses ( 15 or 20 Gy) Fractionated 8 x 2 Gy SiHa tumor

Hicks et al, Clin Cancer Res 2010

Summary

• Sensitivity of tumor and normal cells can be assayed quantitatively in situ in experimental systems.

• Tumor hypoxia has a major negative impact on the curability of tumors by radiotherapy, and probably also chemotherapy

• Tumor hypoxia can be exploited using drugs specifically toxic to hypoxic cells.