[Show abstract][Hide abstract] ABSTRACT: Radiation therapy is a staple treatment for pancreatic cancer. However, owing to the intrinsic radioresistance of pancreatic cancer cells, radiation therapy often fails to increase survival of pancreatic cancer patients. Radiation impedes cancer cells by inducing DNA damage, which can activate cell cycle checkpoints. Normal cells possess both a G1 and G2 checkpoint. However, cancer cells are often defective in G1 checkpoint due to mutations/alterations in key regulators of this checkpoint. Accordingly, our results show that normal pancreatic ductal cells respond to ionizing radiation (IR) with activation of both checkpoints whereas pancreatic cancer cells respond to IR with G2/M arrest only. Overexpression/hyperactivation of Rac1 GTPase is detected in the majority of pancreatic cancers. Rac1 plays important roles in survival and Ras-mediated transformation. Here, we show that Rac1 also plays a critical role in the response of pancreatic cancer cells to IR. Inhibition of Rac1 using specific inhibitor and dominant negative Rac1 mutant not only abrogates IR-induced G2 checkpoint activation, but also increases radiosensitivity of pancreatic cancer cells through induction of apoptosis. These results implicate Rac1 signaling in the survival of pancreatic cancer cells following IR, raising the possibility that this pathway contributes to the intrinsic radioresistance of pancreatic cancer. INTRODUCTION Radiation therapy is a staple treatment for pancreatic cancer patients, especially for those with locally advanced or borderline resectable disease [1, 2]. However, for the vast majority of pancreatic cancer patients, radiation therapy still fails to significantly increase survival or improve quality of life . This lack of efficacy is primarily due to the intrinsic radioresistance of pancreatic cancer cells, as only 12–40% of pancreatic tumors exhibit an objective response to radiation therapy [4–22]. Thus, there is an urgent need to develop strategies for increasing the efficacy of radiation therapy in pancreatic cancer patients. The ionizing radiation (IR) delivered by radiation therapy impedes cancer cells mainly by the production of DNA damage. In response to IR-induced DNA damage, human cells will engage several protective mechanisms that promote DNA repair and survival . Among these is the activation of cell cycle checkpoints that block cell cycle progression to allow time for DNA repair . Depending on the phase of the cell cycle at which the damage is sensed, cells can activate either a G1 or G2 checkpoint, to respectively block the cell cycle at the G1/S
[Show abstract][Hide abstract] ABSTRACT: Purpose/Objective(s)
Different prescription (Rx) isodose lines are used in stereotactic radiosurgery (SRS) and stereotactic radiation therapy (SRT). This study aimed at investigating the dosimetric effect of Rx lines selection on normal tissues for Linac-based SRS and SRT.
Eight SRS or SRT patients (11 tumors) previously treated in our institution with dynamic conformal arcs and a Rx line at 90% were randomly selected. For each patient, alternative plans with 80%, 70%, 60%, and 50% Rx lines were retrospectively generated, yielding a total of 40 plans for the 8 patients. For each plan, PTV coverage was assessed and deemed acceptable based on our clinical standard. Two absolute doses corresponding to the 50% and 30% dose for the original 90% Rx line plan were calculated and denoted as D50 and D30. The volumes of D50 and D30 were calculated for all plans. R50 and R30 were computed as the volume ratios of D50s and D30s of alternative plans (50-80% Rx line) to D50 and D30 of clinical plans (90%), and used to quantify normal tissue sparing.
The average PTV volume for those 11 tumors is 7cc (range from 0.4cc to 37.3 cc). The conformity indexes of those 40 plans range from 1.1 to 1.6. R50 (mean ± standard deviation) are 0.685±0.200, 0.621±0.148, 0.641±0.125,0.790±0.126,1±0 for Rx line plans of 50, 60, 70, 80 and 90%, respectively. R30 are 0.770±0.205, 0.702±0.163, 0.711±0.135, 0.816±0.109, 1±0 for the same cases. The lowest R50 and R30 are achieved when the Rx isodose lines are 70% or 60%, indicating sharper dose falloff and better normal tissue sparing with such Rx line selections. At the same time, decreasing Rx lines inherently result in increasing hot spot within the target (RTOG dose homogeneity index increases from 1.11±0.02 for the 90% Rx case to 1.99±0.07 for the 50% Rx case). Thus, part of PTV is covered by higher dose when the Rx lines decreases, resulting in higher tumor control probability (TCP) when the PTV dose coverage is similar. In other words, compared with plans with 90% Rx line, Rx isodose lines of 60 or 70% yield plans with potentially higher TCP and lower normal tissue complication probabilities (NTCP). ANOVA (Analysis of variance) analysis finds the R50 and R30 differences between the alternative plans (50-80% Rx lines) and clinical plan (90% Rx line) are significant (p<0.001).
By decreasing the Rx isodose line, OARs and normal tissues receive less radiation dose. While an 80% Rx line also results in normal tissue sparing compared with 90% Rx line, greater sparing will be achieved with 60 or 70% Rx line, in which case dose heterogeneity will increase, but this could potentially yield an increased TCP while at the same time lowering NTCP. On the other hand, the gain of increased TCP and decreased NTCP needs to be weighted over possible tumor necrosis due the larger maximum dose within the target.
International journal of radiation oncology, biology, physics 09/2014; 90(1):S907. · 4.59 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Objectives: A new dosimetric variable, dose dropping speed (DDS), was
proposed and used to evaluate normal tissue sparing among stereotactic
radiosurgery (SRS) plans with different prescription isodose lines. Methods:
Forty plans were generated for 8 intracranial SRS cases, prescribing to isodose
levels (IDLs) ranging from 50% to 90% in 10% increments. Whilst maintaining
similar coverage and conformity, plans at different IDLs were evaluated in
terms of normal tissue sparing using the proposed DDS. The DDS was defined as
the greater decay coefficient in a double exponential decay fit of the dose
drop-off outside the PTV, which models the steep portion of the drop-off.
Provided that the prescription dose covers the whole PTV, a greater DDS
indicates better normal tissue sparing. Results: Among all plans, the DDS was
found the lowest for the prescription at 90% IDL and the highest for the
prescription at 60% or 70%. Beam profile slope change in penumbra and its field
size dependence were explored and given as the physical basis of the findings.
Conclusions: A variable was proposed for SRS plan quality evaluation. Using
this measure, prescriptions at 60% and 70% IDLs were found to provide best
normal tissue sparing. Advances in knowledge: A new variable was proposed based
on which normal tissue sparing was quantitatively evaluated, comparing
different prescription IDLs in SRS.
British Journal of Radiology 08/2014; · 1.22 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Purpose:
The dose dropping speed (DDS) in normal tissue was proposed to assess plan quality. It was tested in stereotactic radiosurgery (SRS) cases.
The DDS was tested with a total 40 plans for 8 SRS cases. Dynamic conformal arc plans were generated for each case, with prescription dose to isodose levels (IDL) ranging from 50% to 90% with 10% increment to cover the PTV. Forty non-overlapping rind structures of 1mm thickness were created layer by layer from each PTV surface. The average dose in each rind was calculated and fitted with a double exponential function (DEF) of the distance from the PTV surface, which models the steep and moderate portions of the average dose curve in normal tissue. The parameter charactering the steep portion of the average dose curve in the DEF quantifies the DDS in the normal tissue receiving high dose. Provided that the prescription dose covers the whole PTV, a greater DDS indicates better normal tissue sparing.
Among all plans, the DDS was found to be the smallest for prescription of 90% IDL and reached a higher plateau region for prescription to 60% or 70%. The ratio of the highest slope to that of 90% prescription was 1.68±0.35. Analyzed with ANOVA, the DDS difference between the plans with prescription to 60% or 70% IDLS and those with prescription to 90% IDL were found significant (p<0.01). This can be attributed to the field size and penumbra slope change with different prescription levels. The trend became less pronounced with increasing PTV size due to less change of field profile for larger field size.
A method was proposed to estimate dose dropping speed as a measure of plan quality. Based on this measure, prescriptions to 60% and 70% IDLs were found to provide best normal tissue sparing.
Medical Physics 07/2014; 41(7):157. · 2.91 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Human liver has an unusual sensitivity to radiation that limits its use in cancer therapy or in preconditioning for hepatocyte transplantation. Because the characteristic veno-occlusive lesions of radiation-induced liver disease do not occur in rodents, there has been no experimental model to investigate the limits of safe radiation therapy or explore the pathogenesis of hepatic veno-occlusive disease.
We performed a dose-escalation study in a primate, the cynomolgus monkey, using hypofractionated stereotactic body radiotherapy in 13 animals.
At doses ≥40 Gy, animals developed a systemic syndrome resembling human radiation-induced liver disease, consisting of decreased albumin, elevated alkaline phosphatase, loss of appetite, ascites, and normal bilirubin. Higher radiation doses were lethal, causing severe disease that required euthanasia approximately 10 weeks after radiation. Even at lower doses in which radiation-induced liver disease was mild or nonexistent, latent and significant injury to hepatocytes was demonstrated by asialoglycoprotein-mediated functional imaging. These monkeys developed hepatic failure with encephalopathy when they received parenteral nutrition containing high concentrations of glucose. Histologically, livers showed central obstruction via an unusual intimal swelling that progressed to central fibrosis.
The cynomolgus monkey, as the first animal model of human veno-occlusive radiation-induced liver disease, provides a resource for characterizing the early changes and pathogenesis of ven occlusion, for establishing nonlethal therapeutic dosages, and for examining experimental therapies to minimize radiation injury.
International journal of radiation oncology, biology, physics 12/2013; · 4.59 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The immunogenicity of a cellular immunotherapy using genetically modified vaccines to express α(1,3)galactosyl epitopes (αGal) was evaluated in advanced prostate cancer (PC) patients. In this dose escalation phase I study, we report safety, feasibility, and immunologic data of an immunotherapy composed of 2 human PC cell lines engineered to express αGal epitopes (HyperAcute-Prostate, HAP, NewLink Genetics). Eight patients received up to 12 biweekly vaccinations with HAP. Enrolled patients (aged range, 53-85 y) had American Joint Committee on Cancer stage IV, any T, any N, M1, Eastern Cooperative Oncology Group PS≤2, at least 1 prior hormonal treatment and <3 prior chemotherapies, adequate bone marrow and organ function, and albumin ≥3.0 g/dL. Serum IgG antibodies to synthetic peptides overexpressed in PC were determined by enzyme-linked immunosorbent assay. Results indicate that HAP immunotherapy induced humoral immune responses to autoantigens in 2 of 8 patients. These patients developed IgG antibody to multiple epitopes overexpressed in PC after immunization. These responding patients received higher doses of the immunotherapy suggesting a dose response. Two immunogenic proteins (prostate-specific membrane antigen, hepsin) belong to the extracellular molecules family participating in malignant cell invasion. Median overall survival for patients was 25.1 months with 1 patient surviving over 70 months with stable PSA and bone metastasis before expiring of other causes. Three of 8 patients showed PSA stabilization (>100 d). In conclusion, HAP immunotherapy induces IgG responses to epitopes from autoantigens overexpressed in PC suggesting dose-dependent effect. HAP represents a viable immunotherapy approach to induce immune responses against tumor cells and may provide clinical benefit with minimal toxicity.
[Show abstract][Hide abstract] ABSTRACT: grade 3 and 6% had grade 4 toxicity. Grade 3 and higher toxicities are summarized in the table. All patients completed chemotherapy except 8% missed and 12% needed dose reduction for the last cycle. A total of 14% patients needed hospitalization and 12% needed transfusion. Overall, severe myelotoxicity decreased from 60% to 33% for cervical cancer. Grade 2 or higher myelotoxicity decreased from 45% to 36% for rectal cancer. The toxicity rates for anal cancer are comparable at 55% to the previously reported rate of 51%. Conclusions: The rates of myelotoxicity have decreased with the utiliza-tion of the current dose constraints in IMRT planning for cervical and rectal cancer however neutropenia rates continue to be high for anal cancer, which may be attributed to the toxic effects of chemotherapy with mitomycin and 5-FU used concurrently. Further research is needed to validate the findings of this study in larger population in multi institutional setting. Purpose/Objective(s): Gastrointestinal (GI) melanoma is a rare disease. Whether it is different from skin melanoma remains unclear. The objective of this study is to compare the overall survival (OS), cancer specific survival (CSS) and prognostic factors of GI melanoma to those of skin melanoma using the Surveillance, Epidemiology, and End Results (SEER) registry. Materials/Methods: Patients diagnosed with invasive GI and Skin mela-noma between 1973 and 2008 were identified from the SEER database. Prognostic factors analyzed included age (18-60/61-100), gender, race (White/nonwhite), marital status, stage (localized/regional/distant), year of diagnosis (1973-1987/1988-1997/1998-2008), and type of treatment (radiation therapy (RT)/surgery). OS and CSS were evaluated using the Kaplan-Meier method. Cox proportional hazards regression analysis examined what variables are prognostic of survival. Results: A total of 406 patients with GI melanoma and 173622 patients with skin melanoma were identified. The median age was 69 and 57 for the GI and skin groups, respectively. The GI group tends to be older and with more advanced stage than the skin group. Among the GI group, 85% underwent surgery and 18% received RT while 95% of skin melanoma patients underwent surgery and 2% received RT. The GI group had a median OS and CSS of 15 and 16 months, respectively while the skin group had a median OS of 283 months and did not reach a median CSS. Cox analysis showed that the skin melanoma had significantly lower risk of total mortality and cancer mortality compared to GI melanoma (Hazard Ratio (HR) 0.40, p < 0.0001) and (HR 0.34, p < 0.0001). Factors significantly associated with improved OS and CSS in the skin group included: age 60, female gender, non-white race, early stage, being married, more current diagnosis, undergoing surgery and not receiving RT. In contrast, factors associated with improved OS and CSS in the GI group included: age 60, early stage, non-white race and undergoing surgery. Subgroup analysis on patients who underwent surgery showed that lymph node status is the only prognostic factor for GI melanoma, while all of the previously identified prognostic factors except for race are associated with OS and CSS in the skin group. Conclusions: Outcomes of patients with GI melanoma are significantly inferior to patients with skin melanoma. The melanomas in these two sites also have different prognostic factors. Based on these findings, future studies should explore the reasons behind these differences, which may enable one to identify patients who may have optimal outcomes.
International Journal of Radiation OncologyBiologyPhysics 11/2012; · 4.52 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The purpose of this study was to evaluate the standard outpatient dose of 131-Iodine tositumomab (75 cGy) combined with high-dose carmustine, etoposide, cytarabine, and melphalan (BEAM) followed by autologous stem cell rescue for the treatment of chemotherapy-sensitive relapsed or refractory, or high-risk first complete remission (CR) patients with diffuse large B cell non-Hodgkin's lymphoma (DLBCL). Forty patients with chemotherapy-sensitive persistent or relapsed or high/intermediate or high international prognostic index DLCBL were treated in a phase II trial combining 75 cGy 131-Iodine tositumomab with high-dose BEAM followed by autologous stem cell transplantation. The CR rate after transplantation was 78%, and the overall response rate was 80%. Short-term and long-term toxicities were similar to historical control patients treated with BEAM alone. With a median follow-up of 6 years (range, 3-10 years), the 5-year overall survival (OS) was 72% (95% confidence interval [CI], 55%-83%), and the 5-year progression-free survival (PFS) rate was 70% (95% CI, 53%-82%). The PFS and OS were encouraging in this group of chemotherapy-sensitive persistent, relapsed, or high-risk patients with DLBCL. A follow-up phase III trial with 131-Iodine tositumomab/BEAM vs rituximab/BEAM was planned based on this information.
Biology of blood and marrow transplantation: journal of the American Society for Blood and Marrow Transplantation 08/2012; · 3.15 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Purpose: To demonstrate that standardization in documentation format can significantly reduce manual data entry error in patient chart. Method and Materials: Due to lack of direct data link between CT on rail imaging registration software and patient R&V system, therapists have to manually enter translation correction of patient position into R&V system after performing imaging registration between CT image taken just before treatment delivery and treatment planning CT image. Approximately six months after CT on rail was placed into clinical service, we started requiring therapists to use a standard format to document the shifts when they manually enter data to reduce manual data entry error rate. The therapist manual data entry errors in R&V system before (551 entries) and after the format standardization (1645 entries) are the subjects of statistical analysis. The errors are divided into two categories (recoverable and non-recoverable errors) depending on whether a human being can recover the translation shifts in lateral, longitudinal, and vertical directions from the therapist notes. Fisher's exact test is performed to test the statistical significance of therapist's data entry error reduction after our imposing the documentation format requirement. Temporal information on when the errors were made is also analyzed to find out when errors are more likely to happen. Results: Statistical analysis indicated that reductions in the numbers of recoverable, non-recoverable, and total errors after standardization in document format are all statistically significant with p values less than 0.05. Conclusion: A simple and low cost measure like standardization in document format can significantly reduce the errors operators introduce to patient R&V system when they perform manual data entry. We hope our experience will convince people to follow more disciplined documentation rule to reduce the error rate and therefore potentially can improve the quality of patient care when manual data entry is involved.
Medical Physics 06/2012; 39(6):3750. · 2.91 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Purpose: To quantify pancreatic head tumor rotation and deformation due to the respiration using 4DCT and fiducial markers. Methods: This study included seventeen pancreatic head tumor patients who were treated with gated SBRT using Novalis system in our institution. Each patient had two 5-mm-long fiducial markers placed in pancreatic head approximately 2 cm apart for the gating treatment. All patients had 4DCT scans of 3mm-slice thickness using a CT scanner (SENSATION, SIEMENS) under free breathing condition to create the internal target volume (ITV). The respiratory curve was generated through a pressure sensor placed on the patient's abdomen. The 4DCT image data sets were binned into 8 phases: 0% inhale (end of exhale), 25% inhale, 50% inhale, 75% inhale, 100% inhale (end of inhale), 75% exhale, 50% exhale, and 25% exhale. The fiducial markers were contoured on each phase of 4DCT images, and the 3D coordinates (x,y,z) of centroid of the fiducial marker were recorded. The distance between two markers was calculated for each phase, and its variation on eight phases indicated the pancreatic head deformation. The orientation change of the two markers on eight phases indicated the pancreatic head rotation. Student t-test was performed for the statistical analysis. Results: Pancreatic head rotation and deformation were observed through the breathing cycle. The largest rotation and deformation happened to the course from 0% inhale to 100% inahle. The rotation was 8.0±6.2 degree ranging from 2 to 20 degree. The deformation was 2.0±2.4 mm ranging from 0 to 6.6 mm. Both were statistically significant with p<10-4 for rotation and p=0.0033 for deformation. Conclusions: The 4DCT images showed significant pancreatic head rotation and deformation due to respiration, and both rotation and deformation were highly variable between patients. For patients receiving SBRT with tight margin, rotation and deformation need to be considered, and individualized margin may be necessary.
Medical Physics 06/2012; 39(6):3680. · 2.91 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We report a rare presentation of a 66-year-old female with diffuse metastatic adenocarcinoma of unknown primary involving liver, lymphatic system and bone metastases. The neoplastic cells were positive for CK7 and OC125, while negative for CK20, thyroid transcription factor 1, CDX2, BRST-2, chromogranin, synaptophysin, estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2/neu). Fluorescence in situ hybridization showed no amplification of the HER2/neu gene. Molecular profiling reported a breast cancer origin with a very high confidence score of 98%. The absence of immunohistochemistry staining for ER, PR, and HER2/neu further classified her cancer as triple-negative breast cancer. Additional studies revealed high expression levels of topoisomerase (Topo) I, androgen receptor, and ribonucleoside-diphosphate reductase large subunit; the results were negative for thymidylate synthase, Topo II-a and O6-methylguanine-DNA methyltransferase. The patient was initially treated with a combination regimen of cisplatin and etoposide, and she experienced a rapid resolution of cancer-related symptoms. Unfortunately, her therapy was complicated by a cerebrovascular accident (CVA), which was thought to be related to cisplatin and high serum mucin. After recovery from the CVA, the patient was successfully treated with second-line chemotherapy based on her tumor expression profile. We highlight the role of molecular profiling in the diagnosis and management of this patient and the implication of personalized chemotherapy in this challenging disease.
[Show abstract][Hide abstract] ABSTRACT: The purpose of this work is to determine appropriate radiation therapy beam margins to account for intrafraction prostate translations for use with real-time electromagnetic position monitoring and correction strategies. Motion was measured continuously in 35 patients over 1157 fractions at 5 institutions. This data was studied using van Herk's formula of (αΣ + γσ') for situations ranging from no electromagnetic guidance to automated real-time corrections. Without electromagnetic guidance, margins of over 10 mm are necessary to ensure 95% dosimetric coverage while automated electromagnetic guidance allows the margins necessary for intrafraction translations to be reduced to submillimeter levels. Factors such as prostate deformation and rotation, which are not included in this analysis, will become the dominant concerns as margins are reduced. Continuous electromagnetic monitoring and automated correction have the potential to reduce prostate margins to 2-3 mm, while ensuring that a higher percentage of patients (99% versus 90%) receive a greater percentage (99% versus 95%) of the prescription dose.
[Show abstract][Hide abstract] ABSTRACT: To compare the dosimetric parameters of intensity-modulated radiotherapy (IMRT) and three-dimensional conformal radiotherapy (3D-CRT) in patients with intermediate-risk rhabdomyosarcoma and to analyze their effect on locoregional control and failure-free survival (FFS).
The study population consisted of 375 patients enrolled in the Children's Oncology Group protocol D9803 study, receiving IMRT or 3D-CRT. Dosimetric data were collected from 179 patients with an available composite plan. The chi-square test or Fisher's exact test was used to compare the patient characteristics and radiotherapy parameters between the two groups. The interval-to-event outcomes were estimated using the Kaplan-Meier method and compared using log-rank tests. Cox proportional hazards regression analysis was used to examine the effect of the treatment technique on FFS after adjusting for primary site and risk group.
The median follow-up time was 5.7 and 4.2 years for patients receiving 3D-CRT and IMRT, respectively. No differences in the 5-year failure of locoregional control (18% vs. 15%) or FFS (72% vs. 76%) rates were noted between the two groups. Multivariate analysis revealed no association between the two techniques and FFS. Patients with primary tumors in parameningeal sites were more likely to receive IMRT than 3D-CRT. IMRT became more common during the later years of the study. Patients receiving IMRT were more likely to receive >50 Gy, photon energy of ≤6 MV, and >5 radiation fields than those who received 3D-CRT. The coverage of the IMRT planning target volume by the prescription dose was improved compared with the coverage using 3D-CRT with similar target dose heterogeneity.
IMRT improved the target dose coverage compared with 3D-CRT, although an improvement in locoregional control or FFS could not be demonstrated in this population. Future studies comparing the integral dose to nontarget tissue and late radiation toxicity between the two groups are warranted.
International journal of radiation oncology, biology, physics 04/2011; 82(5):1764-70. · 4.59 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Pancreatic cancer does not respond to a single-agent imatinib therapy. Consequently, multimodality treatments are contemplated. Published data indicate that in colorectal cancer, imatinib and radioimmunotherapy synergize to delay tumor growth. In pancreatic cancer, the tumor response is additive. This disparity of outcomes merited further studies because interactions between these modalities depend on the imatinib-induced reduction of the tumor interstitial fluid pressure. The examination of human and murine PDGFr-β/PDGF-B pathways in SW1990 pancreatic cancer xenografts revealed that the human branch is practically dormant in untreated tumors but the insult on the stromal component produces massive responses of human cancer cells. Inhibition of the stromal PDGFr-β with imatinib activates human PDGFr-β/PDGF-B signaling loop, silent in untreated xenografts, via an apparent paracrine rescue pathway. Responses are treatment- and time-dependent. Soon after treatment, levels of human PDGFr-β, compared to untreated tumors, are 3.4×, 12.4×, and 5.7× higher in imatinib-, radioimmunotherapy + imatinib-, and radioimmunotherapy-treated tumors, respectively. A continuous 14-day irradiation of imatinib-treated xenografts reduces levels of PDGFr-β and phosphorylated PDGFr-β by 5.3× and 4×, compared to earlier times. Human PDGF-B is upregulated suggesting that the survival signaling via the autocrine pathway is also triggered after stromal injury. These findings indicate that therapies targeting pancreatic cancer stromal components may have unintended mitogenic effects and that these effects can be reversed when imatinib is used in conjunction with radioimmunotherapy.