Author Disclosure: K.A. Cengel, None; J. Lee, None; A. Workman, None; E. Arguiri, None; C.C. Solomides, None; K. Hill,None; M. Christofidou Solomidou, None.

2655 Association Between the Incidence of Synchronous and Metachronous Metastases: Analysis of SEER Data

J. J. Meyer1, Y. Anacak2, L. B. Marks1

1Duke University Medical Center, Durham, NC, 2Ege University Medical School, Izmir, Turkey

Purpose/Objective(s): Cancer patients are usually staged based on the presence of radiographically/clinically detectableregional and/or distant disease. It is understood that current staging is inexact, and clinically-staged M0 (cM0) patients may havemicroscopic metastatses (i.e. are cM0pM1) that later cause relapse and death. Since the clinical tools used to stage patients arefairly similar for different tumors, the ratio of the rates of metachronous to synchronous metastases should be similar fordifferent tumors (hypothesis 1). Improvements in diagnostic tools should have caused the ratio of metachronous-to-synchronousmetastases to have decreased over time (hypothesis 2). Finally, the fraction of patients with either metachronous or synchronousmetastases should have declined over time due to increased screening and earlier diagnoses (hypothesis 3). To test thesehypothesis, we analyzed data from the SEER database from 1973–1998.

Materials/Methods: Data for 19 solid tumors were extracted from the SEER database. The rate of synchronous metastses(cM1) is reported in this database. The rate of metachronous metastases was estimated from the 5 year-mortality for the cM0patients (i.e., death rate at 5 years for cM0 patients/rate of cM0 at presentation). A graph of metachronous vs synchronousmetastasis rates was generated and a linear fit was tested to assess if their ratio was consistent between cancer sites (hypothesisnumber 1). This process was repeated for data from 1973 and 1998 to assess if improvements in staging studies altered this ratio(hypothesis number 2), and to assess if the sum of the synchronous and metachronous rates had changed over time (hypothesisnumber 3).

Results: For the 1973 and 1998 data, a linear relationship is seen between the rates of metachronous and synchronousmetastases, with modestly strong correlation coefficients (R2 � 0.77 and 0.71, respectively) (see Figure). Over time, the datapoints have shifted downward and to the left, suggesting a decrease in the number of patients with metastases, consistent withhypothesis number 3. However, the slopes of the lines for 1973 and 1998 are simlar, suggesting that changes in staging methodshave not significantly altered the ratio of metachronous/synchronous metastases (converse to hypothesis number 2) with time.

Conclusions: There is a good correlation between the rates of synchronous and metachronous metastases across tumors. Thus,the rate of anticipated metachronous metastases can be estimated from the rate of clinically-evident metastases at presentation.Changes in screening/staging of disease over time have also reduced the overall fraction of patients with metastases.

Author Disclosure: J.J. Meyer, None; Y. Anacak, None; L.B. Marks, None.

2656 Chk1 Phosphorylates Poly-A Binding Protein, Triggering Translocation to the Nucleus

S. R. Floyd1, J. R. Stehn2, E. Wilker3, M. B. Yaffe4

1Harvard Radiation Oncology Program, Boston, MA, 2Childrens’ Hospital at Westmead, Sydney, Australia, 3MassachusettsInsitute of Technology, Cambridge, MA, 4Massachusetts Institute of Technology, Cambridge, MA

Background: Noxious stimuli that result in DNA damage, such as UV radiation, produce multiple and sometimes unexpecteddownstream events in the cell. These downstream events involve alterations of many cellular functions, most likely to enablethe cell to respond to DNA damage in an efficient manner. A number of DNA damage response kinases have been identifiedthat seem to coordinate the various alterations in cellular function that facilitate this goal of an efficient DNA damage response.Exploring the function of these kinases and the downstream events that they regulate will enhance our understanding of howcells respond to DNA damage, and potentially give insight into how to productively alter this response for improved cancertherapy.

Previous work from our lab utilized a proteomic screening approach to identify novel cellular targets downstream from theChk1 DNA damage kinase (manuscript in preparation). Using this approach we identified Poly-A Binding Protein 1 (PABP1)as a novel and unexpected target of Chk1. PABP1 binds the poly-A region of mRNA transcripts, and is thought to facilitateprotein translation. Further characterization of the action of Chk1 on PABP1 can elucidate the influence of DNA damage onthe cellular protein translation machinery.

Purpose/Objective(s): To further characterize the phosphorylation of PABP1 by Chk1 in order to gain insight into the cellularfunction of this DNA damage downstream event.

Materials/Methods: We used known substrate specificity motif data for Chk1 to identify potential phosphorylation sites. Wethen created phospho-specific antibodies, and utilized 2D gel electrophoresis and Western blotting to demonstrate phosphor-

S574 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 3, Supplement, 2006

ylation following UV irradiation. We then utilized immunofluorescence microscopy and sub-cellular fractionation to investigatethe functional consequences of this phosphorylation event.

Results: PABP1 contains three potential Chk1 phosphorylation sites based on sequence motifs: serine 388, serine 470 andserine 478. Mutagenesis of these sites demonstrated that serine 470 was a likely Chk1 phosphorylation site. We developed aphospho-specific serine 470 antibody for PABP1. This antibody demonstrated in-vivo phosphorylation of PABP1 in 2D gelelectrophoresis Western blots after stimulation with UV radiation, demonstrating that PABP1 is phosphorylated on serine 470in response to UV induced DNA damage. Immunofluorescence demonstrated that PABP1 shifted from the cytoplasm to thenucleus following UV irradiation. Sub-cellular fractionation confirmed this cellular localization shift. Pre-treatment withcaffeine, an inhibitor of the ATM and ATR kinases that are upstream activators of Chk1, abrogated this effect.

Conclusions: UV irradiation triggers phosphorylation of PABP1 by Chk1 at serine 470, and causes a subcellular localizationshift from the cytoplasm to the nucleus. PABP1 is assumed to carry out its function of binding mRNA transcripts and facilitatingtranslation in the cytoplasm, where the bulk of protein translation takes place. This unexpected finding of phosphorylationinduced relocation to the nucleus could be an example of alteration of normal cellular processes following DNA damage, inthat this relocation could significantly alter protein translation.

Author Disclosure: S.R. Floyd, None; J.R. Stehn, None; E. Wilker, None; M.B. Yaffe, None.

2657 Radiobiological Considerations in Designing Hypofractionated Radiation Treatments for LiverIrradiation

A. Tai, B. Erickson, K. Khater, X. A. Li

Medical College of Wisconsin, Milwaukee, WI

Purpose/Objective(s): RTOG is initiating a new hypofractionation regimen (RTOG 0438) to treat patients with livermetastases. Recent publications demonstrated that 3D conformal radiation therapy (3DCRT) with high dose per fraction (4.6Gy/fx) leads to encouraging outcomes and an acceptable probability of radiation induced liver disease (RILD) for patients withprimary liver carcinoma. The purpose of this work is to optimize dose-fractionation regimens based on normal tissuecomplication probability (NTCP) and survival rates for liver irradiation, considering the recent clinical data.

Materials/Methods: A set of plausible radiobiological parameters for primary liver tumors were obtained from a model fittingto a series of clinical survival data. These parameters are alpha/beta � 12.8 � 1.0 Gy, alpha � 0.013 � 0.002 Gy-1, thepotential doubling time: 123 � 9 days, and colonogenic cell number: 1302 � 47. In addition, a new set of Lyman modelparameters, TD50�40.5 Gy, m�0.28 and n�1.1 (Set 1), were extracted from the recent clinical data of hypofractionated3DCRT. These parameters become: TD50�39.8 Gy, m�0.12 and n�0.97 (Set 2), based on a conventional 3DCRT of 1.5 Gyper fraction. Using all of these parameters, we computed a biological equivalent dose (BED) and devised a series of dosefractionation regimens that lead to equal or improved survival rates and controlled NTCP.

Results: Tab. 1 presents a series of dose fractionation regimens for primary liver tumor. The NTCP were calculated for twoeffective volume (Veff) bins. The maximum tolerable doses (MTD) with the corresponding fraction size and the survival rates(SR) at one year, calculated for NTCP�10% and Veff�40%, are also listed in the table. The dose/fx, marked by *, are thoserecommended in RTOG 0438. Other regimens produce nearly the same BED for tumor as that for the proven dose fractionationregimen (1.5 Gy/fx and 10 fractions/wk). The table shows (1) the RTOG 3.5 and 4 Gy regimens have substantially lower BEDs,(2) Total doses for all RTOG regimens can be escalated to achieve higher BED and SR, with NTCP � 10%, and (3) Regimenswith dose per fraction between 3.5 to 5 Gy lead to improved one-year survival rates (around 90%).

Tab. 1 NTCP are calculated using parameters of Set 1 with exception of 2 and 3 Gy per fraction of which NTCP are calculatedusing parameters of Set 2. An alpha/beta ratio of 2 Gy is assumed for normal liver in order to normalize each prescribed doseby the LQ model to 4.6 Gy/fx or 1.5 Gy/fx at which the Lyman model parameters were obtained.

Conclusions: We have calculated a series of dose fractionation regimens for primary liver tumor. These data may be used todesign clinical trials aimed to improve survival for liver tumor patents.

Author Disclosure: A. Tai, None; B. Erickson, None; K. Khater, None; X.A. Li, None.

Dose/fx(Gy) Fraction/wk

TotalFractions

Pres. Dose(Gy)

TreatmentTime(day)

BED forTumor(Gy)

Veff(%)�30-35NTCP(%)

Veff(%)�35-40NTCP(%)

MTD(Gy)

SR(1y) (%)at MTD

2 5 33 66 45 56.8 0-0.4 0.4-3.3 71.8 753 3 21 63 47 57.4 0.7-6.5 6.5-28 57.4 663.5* 5 10 35 12 39.4 0.2-0.3 0.3-0.4 84.9 914 3 13 52 29 55.7 0.7-1.2 1.2-2.1 77.9 884* 5 10 40 12 47.3 0.3-0.5 0.5-0.8 77.9 904.5* 5 10 45 12 55.6 0.6-1.0 1.0-1.6 71.7 885* 5 10 50 12 64.3 1.0-1.8 1.8-3.2 66.7 885 2 10 50 33 55.2 1.0-1.8 1.8-3.2 66.7 835.8 2 8 46.4 26 56.2 1.1-2.1 2.1-3.6 59.9 839.4 1 4 37.6 21 56.1 2.1-4.0 4.0-7.1 41.0 75

21.3 1 1 21.3 1 56.3 3.6-7.1 7.1-13 20.1 62

S575Proceedings of the 48th Annual ASTRO Meeting