Localized Hyperthermia Combined with Intratumoral Dendritic Cells Induces Systemic Antitumor Immunity

Research Article

Arunika Mukhopadhaya,1 Joseph Mendecki,1 Xinyuan Dong,1 Laibin Liu,1 Shalom Kalnicki,1 Madhur Garg,1 Alan Alfieri,1 and Chandan Guha1,2,3

Departments of 1Radiation Oncology, 2Montefiore Medical Center and Albert Einstein Cancer Center, and 3Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, New York

Abstract specific antigen (PSA), which is a surrogate for future clinical recurrence. This translates to roughly 100,000 patients per year Prostate adenocarcinoma, treated with localized tumor hyperthermia (LTH), can potentially serve as a source of facing the possibility of recurrent prostate cancer after initial local tumor antigen, where dying apoptotic/necrotic cells release treatment (1). Persistence of local disease after radiation is a tumor peptides slowly over time. In addition, LTH-treated significant issue in the management of this disease, as there cells can release heat shock proteins that can chaperone remains a continuing potential for symptomatic local recurrence or antigenic peptides to antigen-presenting cells, such as metastatic seeding. Currently, there is no curative treatment for dendritic cells. We attempted to discern whether sequential recurrent and metastatic prostate cancer. LTH and intratumoral dendritic cell and/or systemic granu- Hyperthermia has been used effectively as a radiation sensitizer in the treatment of locally advanced and recurrent solid tumors, locyte macrophage colony-stimulating factor (GM-CSF) would including breast, head and neck, and prostate cancer (2–4). activate antitumor immune response in a syngeneic murine Although hyperthermia has been evaluated as a definitive model of prostate cancer (RM-1). Palpable RM-1 tumors, treatment in combination withexternal beam radiation therapy grown in the distal appendage of C57BL/6 male mice, were for locally advanced prostate cancer, the inability to achieve high subjected to LTH (43.7°Cfor1h) 2, separated by 5 days. temperature uniformly within the tumor limits its application (2). Following the second LTH treatment, animals received either It is possible to expose parts of the tumor to high temperatures PBS or dendritic cells (2 106) intratumorally (every 3 days (>40jC) using hyperthermia, which is sufficient to induce cell for three injections). Separate cohorts also received i.v. deathin prostate cancer cells by bothnonapoptotic (<43 jC) and injection of recombinant adenovirus-expressing murine GM- apoptotic mechanisms (>43jC; ref. 5). Despite these problems, local CSF (AdGMCSF), 1 day after LTH. Control animals received tumor hyperthermia (LTH) could be considered as a palliative AdenoLacZ or AdenoGFP. Intratumoral dendritic cell injec- treatment in the management of locally advanced and recurrent tion induced tumor-specific T-helper cell activity (IFN; prostate cancer. ELISPOTS) and CTL activity, which was further augmented LTH increases the release of heat-shock proteins (HSP) from by AdGMCSF, indicating amplification of tumor-specific TH1 dying tumor cells, whereby HSPs in the extracellular milieu can immunity. The combination of LTH, AdGMCSF, and intra- act simultaneously as a source of antigen due to their ability to tumoral dendritic cell injection resulted in significant tumor chaperone intracellular peptides and as a maturation signal for growth delays when compared with animal cohorts that dendritic cells (6). The maturation step enables dendritic cells to received LTH alone. These results support an in situ cross-present tumor peptides that are released from LTH-treated autovaccination strategy where systemic administration of tumor cells to class I–dependent CD8+ T cells. Thus, ex vivo GM-CSF and/or intratumoral injection of autologous dendritic culturing of dendritic cells withLTH-treated melanoma tumor cells, when combined with LTH, could be an effective cells enhanced cross-priming and activation of melanoma- treatment for local and systemic recurrence of prostate specific CTLs (7). Furthermore, it was shown that heat treatment cancer. [Cancer Res 2007;67(16):7798–806] of tumor cells up-regulates bothHSP and tumor antigen expression and proteosomal degradation, thus augmenting T-cell Introduction cross-priming. Prostate cancer is the most common malignancy in American Various immunotherapeutic approaches are being evaluated for men and the second most common cause of death. Although patients withmetastatic prostate cancers (8–10) by inducing surgery and radiation therapy are curative treatment options for systemic immunity to antigens suchas PSA expressed by prostate early, organ-confined disease, treatment of locally advanced cancer. Prostate cancer cells are inefficient in their antigen- prostate cancer has been suboptimal. Radiation therapy cures presenting capacity, as they frequently do not express molecules few patients. Frequently, adjuvant hormonal therapy is used to important for antigen processing and presentation, suchas the improve the efficacy of radiation therapy. However, a subpopula- antigen transporter gene product, TAP-2, and class I MHC tion of androgen-independent cells will not respond to hormonal molecules (11, 12). In addition, they lack T-cell costimulatory therapy and f25% to 60% of patients will show elevated prostate- molecules (CD80 and CD86) and, therefore, could potentially induce anergy of T cells. Thus, the central defect in inducing tumor-specific immunity against prostate cancer is inefficient antigen presentation Requests for reprints: Chandan Guha, Department of Radiation Oncology, by tumor cells. Various studies have shown that the potency of Golding 103, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY whole-cell tumor vaccines were further enhanced by cytokines, 10467. Phone: 718-920-4361; E-mail: [email protected]. I2007 American Association for Cancer Research. especially granulocyte macrophage colony-stimulating factor doi:10.1158/0008-5472.CAN-07-0203 (GM-CSF) in murine tumor models (13) and in patients with

Cancer Res 2007; 67: (16). August 15, 2007 7798 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Immunomodulation of Hyperthermia melanoma (14), lung cancer (15), and prostate cancer (9). GM-CSF expressing murine GM-CSF, h-galactosidase, and green fluorescent protein increases circulating dendritic cells in the blood, which are available (GFP) were obtained from the Gene Therapy Core of the Albert Einstein to process and present antigens that are released from irradiated College of Medicine (AECOM). tumor cells in the whole-cell vaccine. Because dendritic cells are Animal Model thought to be the major antigen-presenting cells involved in Five- to 6-week-old male C57Bl/6 mice (Jackson Laboratory) were triggering primary T-cell responses in vivo, dendritic cell–based maintained in the animal facility, and all animal studies were done under tumor vaccination approaches are being increasingly tested (16). the guidelines and protocols of the Institutional Animal Care and Use These protocols use defined tumor antigens and peptides, whole- Committee of AECOM. RM-1 tumor cells (105) were inoculated s.c. into the cell extracts, apoptotic and/or necrotic tumor cells, and tumor- dorsum of the foot of male C57Bl/6 mice. Tumor dimensions were derived DNA and RNA (reviewed in ref. 16), and have been shown to measured by vernier caliper on alternate days and the average tumor induce significant antitumor immune response in animal tumor volume was determined by 4/3 3.14 (L/2 W/2 H/2), where L is length, W is width, and H is height. Palpable tumors, with z5mmin models. Sipuleucel-T (Provenge; APC8015; Dendreon Corp.), a novel diameter, were used for these studies. immunotherapeutic agent, which includes autologous dendritic cells pulsed ex vivo witha recombinant fusion protein (PA2024) In vitro Studies with RM-1 Cells consisting of GM-CSF and prostatic acid phosphatase, has shown Hyperthermia treatment. Exponentially growing RM-1 cells (106) were significant activity in two phase II trials in men with androgen- immersed into a temperature-controlled Haake water bathmaintained at j j j dependent biochemically relapsed prostate cancer with a decrease 42 C, 43.7 C, and 45 C for 1 h. Because optimal cell killing and release of j in PSA and prolongation in PSA doubling time (17). Data from two HSP70 was seen with43.7 C, further LTH studies were done at this temperature. phase III trials in men with asymptomatic, metastatic hormone- Apoptosis/necrosis detection assay. One day after hyperthermia refractory prostate cancer showed an improved median overall treatment, cells were assayed for necrosis or apoptosis by Annexin survival in men who received sipuleucel-T compared with placebo V–FITC/propidium iodide detection kit (BD Biosciences) according to the (17, 18). These approaches, however, require additional ex vivo manufacturer’s protocol. Briefly, 104 to 105 cells were analyzed by FACScan manipulation to grow tumor cells and/or to prepare tumor flow cytometry (Becton Dickinson, Pharmigen) and analyzed using the antigenic extracts from resected tumor specimens and the results FLOWJO program. from early clinical trials have pointed to a need for additional ELISA for HSP70 release. Release of HSP70 by hyperthermia-treated improvement of dendritic cell–based vaccines before this therapy cells was assayed using an HSP70 ELISA kit (StressGen Biotechnologies), could be added to existing cancer strategies. As an alternative according to the manufacturer’s instructions. approach, intratumoral injection of immature (19) or cytokine- Culture of immature dendritic cell. Bone marrow cells from the tibia and femur of C57Bl/6 mice were harvested following an established transduced (20–22) dendritic cells has been shown to induce strong protocol (25) and the cells were plated as 106/mL of culture medium antitumor immune response in various models of murine and consisting of RPMI 1640 supplemented with10% heat-inactivated FBS, human tumors. This approach depends on the capture and 100 Ag/mL streptomycin, 100 units/mL penicillin (Life Technologies, Inc.), processing of tumor-derived antigens by intratumoral dendritic and recombinant murine GM-CSF, 10 ng/mL (specific activity, 5 106 cells for presentation to T lymphocytes. Although the exact units/mg), and interleukin-4 (IL-4; 10 ng/mL, specific activity, 2.8 108 mechanisms of tumor antigen acquisition by dendritic cell are still units/mg; Peprotech). Immature dendritic cells were cultured for 5 days unclear, available data suggest a role for HSPs released from dying withcytokine supplementation on alternate days. 6 malignant cells and for the internalization of tumor-derived Engulfment assay of dendritic cell. RM-1 cells (10 ) and bone marrow– 7 apoptotic bodies. derived immature dendritic cells (10 ) were incubated withcarboxyfluor- Because LTH induces a temperature-dependent apoptotic or escein diacetate, succinimidyl ester (CFDA-SE; Invitrogen-Molecular Probes) and PKH 26 (Sigma Chemical) for labeling according to the necrotic cell death in tumors, which is associated with the release manufacturer’s protocol. One day after heat treatment, CFDA-SE–stained of HSPs, intratumoral injection of dendritic cell could induce a RM-1 cells were cultured with10 7 PKH26-stained resting dendritic cells for strong tumor-specific immune response. We hypothesized that 4 hand observed under confocal microscopy (Leica AOBS). LTH-treated prostate tumor cells could potentially serve as a Dendritic cell maturation assay. Bone marrow–derived dendritic cells source of tumor antigens in vivo, where dying tumor cells would were purified withbeads coated withanti-CD11c monoclonal antibodies, release tumor antigens and HSP slowly over time, thus providing a conjugated to phycoerythrin (Stemcell Technologies) from 5-day cultures of depot of tumor-derived antigenic peptides and ‘‘danger’’ signal for bone marrow cells withGM-CSF and IL-4. One day after heattreatment, 10 6 7 dendritic cell maturation. To examine our hypothesis, we evaluated RM-1 cells were cocultured with10 dendritic cells for 1 to 2 days, followed our comprehensive in situ tumor vaccination strategy using LTH by harvesting and staining with FITC-conjugated monoclonal antibody followed by intratumoral dendritic cell injection and GM-CSF against CD80 (BD PharMingen). The dendritic cells were analyzed by FACScan flow cytometry. cytokine therapy in a nonimmunogenic mouse prostate cancer Reverse transcriptase-PCR of CCR7. Autologous dendritic cells from model for the induction of systemic tumor-specific immune 5-day cultures of bone marrow cells were cocultured withheat-treated response and therapeutic efficacy. RM-1 cells for 24 and 48 h. After incubation, cells were harvested and total RNA was extracted withTRIzol (Invitrogen) and reverse transcribed using Materials and Methods the Superscript First Strand Synthesis System (Invitrogen). CCR7 expression was then evaluated by reverse transcription-PCR (RT-PCR). Cell Lines, Medium, Cytokines, and Other Reagents Mouse prostate cancer cell line, RM-1 (a gift from Dr. T.C. Thompson, In vivo Experiments Baylor College of Medicine, Houston TX) originally isolated from a ras + LTH treatment. Palpable RM-1 tumors (>21–25 mm3) were treated with myc-transformed mouse prostate carcinoma (23, 24) were cultured in LTH (43.7jC) for 1 h, using a temperature-controlled (Haake Circulator) DMEM, supplemented with10% heat-inactivated fetal bovine serum (FBS; water bath. Animals were anesthetized with i.p. ketamine and xylazine Atlanta Biologicals), 0.1 mmol/L nonessential amino acids, 1 Amol/L (7:1 mg/mL for 100 AL/mouse) and immobilized in a lucite jig from which sodium pyruvate, 100 Ag/mL streptomycin, and 100 units/mL penicillin the legs projected through. The foot bearing tumor was immersed into (Life Technologies Invitrogen Corporation). Recombinant adenoviruses the water heated at 43.7jC and the tumor-bearing legs were maintained www.aacrjournals.org 7799 Cancer Res 2007; 67: (16). August 15, 2007

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Cancer Research in that position for 1 h. A second LTH treatment was delivered, 5 days after substrate 3-amino-9-ethylcarbazole and analyzed using the ELISPOT the first LTH treatment. analyzer. Mean numbers of spot-forming cells were calculated from the Intratumoral dendritic cell and recombinant adenovirus injection. triplicate assays. The results are presented as means F SDs for three Ten cohorts were used for these studies as follows: (a) untreated; (b) separate experiments. adenovirus-expressing GM-CSF (AdGMCSF) alone; (c) intratumoral dendritic cell; (d) intratumoral dendritic cell + AdGFP; (e) intratumoral dendritic CTL Assay cell + AdGMCSF; (f ) LTH alone; (g) LTH + control AdGFP or AdLacZ [8 109 For CTL assay, splenocytes were prepared after the same protocol as plaque-forming unit (pfu)]; (h) LTH + AdGMCSF (8 109 pfu); (i)LTH+ described in the ELISPOT assay and stimulated with LTH-treated RM-1 cells intratumoral dendritic cell; and (j) LTH + AdGMCSF (8 109 pfu) + in 5-day cultures. The cells were then analyzed for cytotoxic activity using a intratumoral dendritic cell. On day 1 after the first LTH treatment, nonradioactive CTL assay [lactate dehydrogenase (LDH) release assay] recombinant adenoviruses (8 109 pfu in 100 AL PBS) containing either method (CytoTox 96 Assay-Promega), according to the manufacturer’s murine GM-CSF or control GFP was i.v. injected via tail vein. On day 1 after protocol. Briefly, the assay was done in a 96-well round-bottomed plate the second LTH treatment, 2 106 bone marrow–derived dendritic cells using RM-1 cells as the target cells. CTL assays were done at lymphocyte were injected intratumorally in separate cohorts of AdGMCSF- and AdGFP- effector/target (E/T) ratios of 5:1 and 10:1. The results were expressed treated animals. A total of three dendritic cell injections were administered according to the following formula: on 3-day intervals. Cell death assay after LTH in vivo. One day after LTH, tumor cells were % Specific analysis ¼ðexperimental release harvested from tumors grown in LTH-treated and untreated mice and cells À Þ=ð were prepared for Annexin V–FITC/propidium iodide staining according to spontaneous release maximum release the manufacturer’s protocol (BD Biosciences) and analyzed by Cell Quest À spontaneous releaseÞ program. ELISPOT assay. Splenocytes were harvested from C57Bl/6 mouse spleens 21 days after the first LTH treatment according to standard procedures (25). The ELISPOT assays were done following established Results protocols (14, 26). Briefly, 96-well nitrocellulose-based microtiter plates To evaluate the effect of LTH treatment and dendritic cell–based were coated overnight at room temperature with anti-IFNg monoclonal immunotherapy, we chose the RM-1 murine prostate cancer model. antibodies (BD Biosciences-PharMingen), diluted in PBS. After the plate The RM-1 cells were originally derived from a mouse prostate was washed with PBS, all wells were blocked with PBS + 1% bovine serum reconstitution model where the ras and myc oncogenes were j 6 A albumin for 2 hat 37 C. The splenocytes (10 /50 L/well) from mice of retrovirally introduced into the urogenital sinus (23). The mouse various cohorts were mixed with LTH-treated RM-1 cells (105/50 AL/well) prostate reconstitution model consistently produces androgen- and cocultured in RPMI 1640 with10% FCS for 30 to 40 hat 37 jCin5% independent, poorly differentiated prostate adenocarcinoma and CO2. Following this incubation, the wells were washed in PBS-Tween 20 and incubated withbiotinylated anti-IFN g monoclonal antibodies for the derivative cell lines, such as RM-1 prostate cancer cells, have 2 hat 37 jC, followed by incubation withstreptavidin–horseradish been used to examine the efficacy of a variety of gene, peroxidase conjugate for 1.5 h. Spots representing a single cytokine chemotherapeutic, immunotherapeutic, and hormonal manipula- secreted by individual cells were developed by using the peroxidase tion strategies (27, 28).

Figure 1. Induction of cell death 24 h after LTH in vitro (A and B) and in vivo (C and D). RM-1 cells or tumor-bearing foot pads were treated either with LTH (43.7jCfor1h;B) or incubated in control temperature (37jCfor1h;A). Twenty-four hours after treatment, tumor cells were harvested and stained for fluorescence-activated cell sorting with propidium iodide (PI) and Annexin V–FITC. Flowcytometry (representative of three experiments) of Annexin V–FITC and propidium iodide–stained RM-1 cells shown below. Note the presence of early apoptotic (5.5%, Annexin+/propidium iodideÀ), late apoptotic (31.4%, Annexin+/ propidium iodide+), and necrotic (21.7%, AnnexinÀ/propidium iodide+) RM-1 cell population after heat treatment in vitro (B), whereas in vivo (C, D) LTH treatment primarily induced apoptosis (93.1%, Annexin+/propidium iodide+; D).

Cancer Res 2007; 67: (16). August 15, 2007 7800 www.aacrjournals.org

incubation for 24 h. In contrast to unheated control cells that had <8% apoptotic/necrotic cell population, 50% to 58% of heat-treated RM-1 cells exhibited either apoptosis or necrosis (Fig. 1A and B). As displayed in Fig. 1B, Annexin V–FITC/propidium iodide staining showed the presence of early apoptotic (5.53%, Annexin+/ propidium iodideÀ), late apoptotic (31.4%, Annexin+/propidium iodide+), and necrotic (21.7%, AnnexinÀ/propidium iodide+) RM-1 cell population after heat treatment. We next treated mice with 2- to 3-week-old palpable tumors withLTH, 43.7 jC for 1 h. One day after LTH, there was significant increase in late apoptotic cells (Annexin V+/propidium iodide+) in RM-1 tumors (90 F 5%), compared withcontrol animals (<1%; Fig. 1 C and D). Interestingly, in contrast to heat treatment in vitro, pure necrotic cell population was less prominent after LTH treatment in vivo. This could be due to an increase in tumor tissue temperature during our LTH treatment in vivo, when apoptosis dominates with higher temperatures (>43–44jC), as supported by previous studies in human prostate cancer cells (5). Alternatively, necrotic cells may have been discarded as debris during the harvesting of tumor cells from foot tumors. To examine whether HSP70 was released from heat-treated RM-1 cells, ELISA was done to quantitate HSP70 protein levels in the cell culture medium of heat-treated RM-1 cells and in serum of Figure 2. Engulfment of heat-treated RM-1 cells by immature dendritic cells. Confocal microscopy showing CFDA-SE–stained apoptotic RM-1 cell (A), PKH LTH-treated mice. There was significant release of HSP70 (range 6 26–stained immature dendritic cell (B), and dendritic cell engulfing RM-1 1–8 ng/mL) in culture medium of heat-treated 10 RM-1 cells and cells (nucleus stained with 4¶,6-diamidino-2-phenylindole; C and D). in serum (1–5 ng/mL) of LTH-treated mice, indicating the induction of HSP70 protein expression and subsequent release Heat treatment induces both apoptosis and necrosis and the from dying tumor cells. release of HSP70 from RM-1 cells cultured in vitro and in vivo. Immature dendritic cells engulfed hyperthermia-treated To determine the type of cell death induced by heat treatment, RM- RM-1 cells followed by maturation. Immature dendritic cells 1 cells were treated with hyperthermia (43.7jC) for 1 h, followed by have the highest capacity of engulfing apoptotic/necrotic cells and

Figure 3. Maturation of dendritic cells after coculturing with treated RM-1 cells. fluorescence-activated cell sorting (A–C) demonstrating an increase in CD80 cell surface expression in dendritic cells, 24 h after coincubation with (A) untreated, (B) heat-treated RM-1, and (C) freeze-thawed RM-1 cells; (D) RT-PCR showing an increase in expression of CCR7 mRNA in dendritic cells cocultured with LTH-treated RM-1 cells for 48 h.

www.aacrjournals.org 7801 Cancer Res 2007; 67: (16). August 15, 2007

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Cancer Research dendritic cells lose the efficiency of antigen uptake upon Systemic TH1-type immune response is induced after LTH maturation. Therefore, immature dendritic cells were purified from and intratumoral dendritic cell injection. Next, we examined bone marrow cultured for 5 days in medium containing GM-CSF whether injection of immature dendritic cells in LTH-treated RM-1 and IL-4 and used for these studies. Twenty four hours after heat tumors could stimulate T cells and induce RM-1–specific systemic treatment, CFDA-SE (green)–stained RM-1 cells were cocultured immune response in vivo. ELISPOT assays showed that splenocytes withPKH26 (red)–stained dendritic cells for 4 hfor examining from mice that received LTH + dendritic cell or LTH + AdGMCSF whether dendritic cells can engulf heat-treated RM-1 cells. contained significantly (P > 0.001) higher frequencies of IFNg- Confocal microscopy showed that green debris of RM-1 cells were secreting T cells (117 and 93 colonies per 106 cells, respectively) in the vesicles within the PKH26-stained red dendritic cells, than mice that received LTH alone (29 colonies per 106 cells). A indicating dendritic cells have phagocytosed necrotic or apoptotic combination of systemic AdGMCSF and intratumoral dendritic cell debris from heat-treated RM-1 cells (Fig. 2). When dendritic cells induced the highest frequency of IFNg-secreting splenocytes (223 were cocultured with heat-treated RM-1 cells for 24 to 48 h, flow colonies per 106 cells; Fig. 4A). When splenocytes were stimulated cytometry revealed an increase in expression of T-cell costimula- witha nonspecific, C57Bl/6-derived, heat-treated, 3LL (Lewis lung tory molecule, CD80 (B7-1), on the dendritic cell surface (Fig. 3B). adenocarcinoma) cells, the IFNg-secreting T cells were consistently Because dendritic cell maturation results in an increase in the gene below baseline values (Fig. 4A). expression for chemokine receptor, CCR7 mRNA, we did RT-PCR to Induction of CTL response. We next investigated whether the evaluate the expression of CCR7 mRNA. Figure 3D shows an combination of LTH of RM-1 primary tumor and intratumoral increase in CCR7 gene expression by 48 hof incubating dendritic dendritic cell injection induced CTLs that can lyse tumor cell cells with heat-treated RM-1 cells. Furthermore, there was minimal targets. LDH release assay showed that intratumoral dendritic cell induction of CD80 in dendritic cells that were cocultured with injection induced tumor-specific CTL activity (LDH release 17–20% cell lysate from necrotic RM-1 cells after repeated freeze-thaws versus 5% in control). Comparable levels of CTL activity was also (Fig. 3C). These experiments indicated that dendritic cells could noted in splenocytes of mice that received AdGMCSF + LTH engulf cell fragments of heat-treated RM-1 cells, followed by (aproximately 19% LDH release), indicating that systemic admin- maturation withinduction of CD80 and CCR7. istration of GM-CSF could substitute intratumoral dendritic cell

Figure 4. A, intratumoral injection of dendritic cells and/or systemic injection of AdGMCSF increase the frequency of T cells secreting IFNg in response to LTH-treated RM-1 cells. IFNg ELISPOTS demonstrating an increase in IFNg production by splenocytes in animals treated with LTH + dendritic cells (DC) F AdGMCSF. Columns, means from three to eight separate experiments of treatment cohorts; bars, SD. 3LL (Lewis lung carcinoma) are included as nonspecific stimulatory cells for each cohort 2 repeats. B, cytotoxic activity of splenocytes after combination therapy with LTH, intratumoral injection of dendritic cells, and/or systemic injection of AdGMCSF. Splenocytes were harvested 22 d after the first LTH treatment and cultured for 5 d with specific stimulation by heat-treated (43.7jC for 1 h) RM-1 cells. Cytotoxic activity was assessed by LDH release assay upon incubating the activated splenocytes with untreated RM-1 cells. The highest lytic activity was seen in splenocytes of animals that received LTH + dendritic cell + AdGMCSF. 3LL (Lewis lung carcinoma) are included as nonspecific target cells (10:1) for each cohort and cytotoxicity was V1.0.

Cancer Res 2007; 67: (16). August 15, 2007 7802 www.aacrjournals.org

Figure 5. Tumor growth delay in LTH-treated animals. A, systemic administration of AdGMCSF enhances the tumoricidal effects of LTH (P V 0.005). B, intratumoral injection of dendritic cells augment the tumor growth inhibition achieved by combination therapy with LTH and AdGMCSF (P V 0.0001). Graph shows representative volumes (mm3) on the y-ordinate of three repeat experiments.

injection. However, the addition of systemic administration of Table 1). Because GM-CSF increases the amount of circulating AdGMCSF to intratumoral dendritic cell injection in LTH-treated dendritic cells in peripheral blood, we examined whether systemic RM-1 tumors augmented the CTL response to a maximum of 36.3% administration of a recombinant adenovirus expressing murine (range, 30–44%), indicating that systemic GM-CSF could maximize GM-CSF can enhance the tumor growth delay in LTH-treated the CTL response induced by intratumoral dendritic cell injection tumors and made comparisons to control viruses (AdLacZ or (Fig. 4B). These results show that a combination of LTH, AdGFP). These results are summarized in Table 1 and are intratumoral injection of dendritic cells and systemic administra- representative of three separate experiments. Twenty-four hours tion of AdGMCSF induces IFNg-secreting, TH1 T cells and RM-1– after the first LTH treatment, animals received either i.v. AdGMCSF specific CTLs, suggesting that LTH treatment of RM-1 tumors or control viruses, AdLacZ, or AdGFP (8 109 particles/mouse). provides a source of tumor antigens and maturation signals for Animals receiving control adenoviruses showed growth progres- dendritic cell activation, which, in turn, induces a strong tumor- sion (Fig. 5A) similar to LTH-alone group (P = 0.67). In contrast, specific systemic immune response. AdGMCSF (n = 17) significantly (P V 0.01) delayed LTH-treated Systemic AdGMCSF and/or intratumoral dendritic cell RM-1 tumor progression within the 2nd week of treatment (Fig. 5A; injection augments the tumor growth retardation of LTH. To Table 1). We then determined whether the tumor growth delay determine whether induction of tumor-specific TH1-type immune is further enhanced by combining intratumoral dendritic cell response in LTH-treated animals resulted in greater tumor control, injection withLTH + AdGMCSF treatment. Intratumoral dendritic we compared the treatment efficacy of LTH alone or in cell inoculation, administered every 3 days after the second LTH combination withdendritic cell–based immunotherapy.Two weeks treatment for three injections, augmented the tumoricidal effects after inoculation of RM-1 cells, palpable tumors were treated with of LTH (P V 0.01; Fig. 5B; Table 1). The tumor growth inhibition was LTH (43.7jC for 1 h) in two fractions separated by 5 days. LTH maximal (P V 0.0001) in animals receiving LTH + AdGMCSF + (n = 18) significantly inhibited (P < 0.001) tumor growth(Fig. 5 A; intratumoral dendritic cell injection (n = 13; Fig. 5B). Intratumoral www.aacrjournals.org 7803 Cancer Res 2007; 67: (16). August 15, 2007

Table 1. Tumor volumes in treatment cohorts

Groups Day 0 Day 18 Day 22

n Volume Volume F SE Volume F SE

Untreated control 16 25.2 795 F 64.2 841 F 7.1 AdGMCSF 8 20.1 698 F 40.5 831 F 10.4 DC F AdGFP 10 26.2 685 F 18.6 823 F 22.3 AdGMCSF + DC 10 24.8 621 F 54.6 703 F 9.2 LTH F Ad-GFP 18 21.1 238 F 9.9 612 F 30 c LTH + DC F AdGFP 11 20.8 219 F 15.8* 408 F 28 b LTH + AdGMCSF 17 22.6 111 F 18.9 314 F 41.7x c LTH + AdGMCSF + DC 13 23.9 158 F 19.5 205 F 51k

NOTE: The volumes are averaged from three experiments. *P V 0.01. cP V 0.05. bP V 0.005. xP V 0.001. kP V 0.0001.

dendritic cell injection alone or in combination withAdGMCSF temperature in vivo versus that in culture conditions. The nature of and/or AdGFP without LTH had only a marginal tumoricidal effect, cell deathinduced by LTH and thedependence on a critical indicating that LTH is necessary for volume reduction in temperature varies between cell types and tumor tissues. Li et al. established tumors (Table 1). Thus, although dendritic cell–based (5) described apoptosis as the primary mode of cell death in human immunotherapy is ineffective in local tumor control, it can enhance prostate cancer cells at higher temperatures (>43jC), whereas the tumoricidal effects of LTH and induce a systemic immune Yonezawa et al. (29) showed that apoptotic cell death was induced response against prostate cancer. in malignant fibrous histiocytoma cells at 42jC and necrosis at higher (44jC) temperatures. Sauter et al. (30, 31) showed that although dendritic cells can engulf both apoptotic and necrotic Discussion cell, the maturation signal for dendritic cells, necessary for efficient Currently, there is no satisfactory salvage treatment for locally cross-presentation of antigens to T cells, is provided by necrotic recurrent prostate cancer, after failure of radiation therapy. cells. Several groups have shown that the endogenous ‘‘danger’’ Reirradiation of prostate cancer is not possible because of limiting signal in necrotic cells, necessary for dendritic cell maturation, is toxicities of normal tissues to radiation therapy. Hyperthermia could HSP (32, 33). Our results show that HSPs are released from both be used as a salvage therapy for patients presenting with evidence LTH-treated apoptotic and necrotic RM-1 cells. Additionally, of clinical and biochemical failure, whereby it not only helps local incubation of heat-treated RM-1 cells with immature dendritic control but also provides a source of tumor antigens for intratumoral cells resulted in induction of CD80 and CCR7 molecules, indicating dendritic cells for induction of tumor-specific immune response. maturation of dendritic cells. Similar results have been reported LTH has been used in clinical trials in patients with locally advanced withtumor lysates derived from UV-irradiated apoptotic and prostate cancer (1–3) as a radiosensitizing adjuvant therapy in necrotic tumor cells (34, 35). The induction of CCR7 is a hallmark combination with radiation therapy. A number of methods, such as of dendritic cell maturation because CCR7 is the ‘‘homing receptor’’ transurethral microwave thermotherapy, radiofrequency interstitial for antigen-loaded dendritic cells to migrate into draining lymph tumor ablation, and high-frequency ultrasound have been safely nodes, where they would present the processed antigens to T cells used in the clinic to treat locally recurrent prostate cancer with LTH. for the induction of adaptive immune response (36–38). LTH can elevate tumor temperatures due to the inability to rapidly HSPs are highly conserved, abundant intracellular proteins and dissipate heat because of tortuosity of tumor blood vessels and function as molecular chaperones that guide several steps during tumor avascularity. This results in an increase in cell killing when synthesis, transportation, and degradation of proteins (39). They compared with attendant normal tissues, which have the abilities to bind to peptides that are generated by the proteosomal degradation dissipate heat by vascular dilation. of heat-denatured intracellular proteins (26). The release of HSP- Our studies show that heat treatment (42–43.7jC) induces peptide complex from dying tumor cells could induce anticancer apoptosis and necrosis of RM-1 cells resulting in release of HSPs in immune responses by targeting tumor-derived antigenic peptides to the culture medium and serum. Interestingly, we found subtle the professional antigen presenting cells, the dendritic cells, whereby differences between in vitro and in vivo effects of LTH on the dendritic cells endocytose HSP-peptide complex actively via several percentage of apoptotic or necrotic deathof RM-1 cell populations. HSP receptors (CD91 for gp96 and CD14 for HSP70; refs. 32, 40). Elevated temperatures (43.7jCfor1h)in vitro resulted in cells that Once endocytosed, the HSP-peptide complexes are routed to the were bothapoptotic and necrotic, whereas, in vivo it induced endogenous antigen pathway and cross-presented in MHC class I primarily apoptosis of RM-1 cells. The predominance of one form molecules, where it is recognized by class I–dependent CD8+ T cells. of cell deathover othercan be due to a difference in tumor tissue HSPs not only present a set of comprehensive tumor-derived peptide

Cancer Res 2007; 67: (16). August 15, 2007 7804 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Immunomodulation of Hyperthermia antigens to the dendritic cells, but also provide the maturation signal induce cell-mediated immunity in immunogenic tumors. However, for these cells. Thus, HSP proteins noncovalently bound to antigenic we now show that immunomodulation of LTH with dendritic cell– peptides are able to vaccinate against tumors and infectious agents based immunotherapy can induce cell-mediated immunity in a (39). In addition, HSP fusion proteins have been shown to induce nonimmunogenic RM-1 tumor model. This is in concordance with CTL responses (41). our earlier studies withanothernonimmunogenic tumor model Random mutations in cancer cell generate unique tumor (3LL lung adenocarcinoma), where we combined Flt3L therapy antigens, which if identified would be a logical choice for tumor withradiation therapyto induce systemic protective antitumoral vaccine specifically designed for individual patients. Because immunity (46, 47). individual mutations in cancer cells are not identified, autologous Various approaches of prostate cancer immunotherapy, such as tumor-derived HSP-peptide complexes has been proposed as a recombinant tumor antigen vaccines delivered by viral and logical personalized approach that may obviate the need to identify plasmid vectors, peptide vaccines, and HSP-peptide vaccines, are the unique antigens contained in the individual vaccine (42, 43). being tested in clinical studies. Although vaccination with defined Thus, autologous HSP vaccines are undergoing testing in clinical tumor antigens presented by dendritic cell has obvious appeal, trials and preliminary data show some promise (44). However, this natural immune variation, MHC polymorphism, and emergence of strategy requires isolation and purification of HSP-peptide com- antigen-loss variants would require an ever-changing mixture of plexes from resected tumor specimens. This report shows that potential tumor antigens in vaccine formulations. Attempts have LTH-treated primary tumor can be harnessed as an autologous been made to treat tumors withirradiated whole-cellvaccines in situ tumor vaccine, whereby the LTH-treated tumor cells provide transduced withthe GM-CSF gene in murine tumor models (13) a depot of tumor antigens in vivo and by inducing the expression and in patients withmelanoma (14), lung cancer (15), and prostate and release of HSP-peptide complex from dying tumor cells, LTH cancer (9). As mentioned before, sucha strategy requires growing treatment of prostate cancer cells could provide the very essential tumor cells ex vivo, which may not be possible in most cases of ‘‘danger’’ signals for dendritic cell maturation and induction of recurrent and metastatic prostate cancer. Instead of these whole- tumor-specific adaptive immune response. Thus, the combination cell vaccines, our in situ vaccination approachcombines local of LTH and intratumoral dendritic cell inoculation induced TH1 prostate hyperthermia to release prostate cancer antigens with type immune response withan increase in IFN g-secreting T cells dendritic cell–activating cytokine, GM-CSF, to induce a strong and CTLs resulting in enhancement of tumor growth retardation tumor-specific immune response. Immunomodulation by hyper- withthecombination treatment. Interestingly, systemic adminis- thermia with dendritic cell–stimulating cytokine therapy would be tration of AdGMCSF induced similar levels of RM-1–specific beneficial in treating locally advanced and recurrent prostate immune response and augmented tumor growthretardation, cancer and hopefully, successfully eradicate micrometastatic tumor suggesting that systemic GM-CSF administration could bypass foci. In the clinic, prostate hyperthermia can be delivered by a the requirement for intratumoral dendritic cell injection. This is variety of procedures, such as high-energy transurethral thermo- not surprising because GM-CSF increases the number of circulat- therapy (48), transperineal radiofrequency interstitial tumor ing dendritic cells in the blood, thus enabling circulating dendritic ablation (49), and high-frequency ultrasound (50). The experi- cells to infiltrate LTH-treated tumors for antigen engulfing and ments, reported here, would serve as a basis of future clinical trials processing. The combination of LTH with AdGMCSF and intra- of hyperthermia and dendritic cell–based immunotherapy in tumoral dendritic cell was found to be the most efficacious in patients withrecurrent and metastatic prostate cancer. inducing antitumor immunity and tumor growthretardation. In this study, Ad-GMCSF was used as an inexpensive and effective Acknowledgments means to administer GM-CSF in laboratory mice. However, in the Received 1/16/2007; revised 5/3/2007; accepted 6/5/2007. clinic, one could envision using s.c. injection of recombinant GM- Grant support: DOD PC020797 (C. Guha). CSF to increase the number of circulating dendritic cells. When The costs of publication of this article were defrayed in part by the payment of page feasible, autologous dendritic cells can be harvested from patients charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate thisfact. for direct intratumoral injection into LTH-treated primary tumors. This article is dedicated to the memory of Dr. Joseph Mendecki whose research Previous studies by Alfieri et al. (45) have shown that LTH can focused on the applications of hyperthermia in the treatment of cancer.

References 5. Li WX, Franklin WA. Radiation- and heat-induced 10. Lodge PA, Jones LA, Bader RA, Murphy GP, Salgaller apoptosis in PC-3 prostate cancer cells. Radiat Res 1998; ML. Dendritic cell-based immunotherapy of prostate 1. Vicini FA, Horwitz EM, Gonzalez J, Martinez AA. 150:190–4. cancer: immune monitoring of a phase II clinical trial. Treatment options for localized prostate cancer based 6. Milani V, Noessner E, Ghose S, et al. Heat shock Cancer Res 2000;60:829–33. on pretreatment serum prostate specific antigen levels. protein 70: role in antigen presentation and immune 11. Sanda MG, Restifo NP, WalshJC, et al. Molecular J Urol 1997;158:319–25. stimulation. Int J Hyperthermia 2002;18:563–75. 2. Anscher MS, Samulski TV, Dodge R, Prosnitz LR, characterization of defective antigen processing in Dewhirst MW. Combined external beam irradiation and 7. Shi H, Cao T, Connolly JE, et al. J. Hyperthermia human prostate cancer. J Natl Cancer Inst 1995;87: external regional hyperthermia for locally advanced enhances CTL cross-priming. J Immunol 2006;176: 280–5. adenocarcinoma of the prostate. Int J Radiat Oncol Biol 2134–41. 12. Blades RA, Keating PJ, McWilliam LJ, George NJ, Phys 1997;37:1059–65. 8. Arlen PM, Gulley JL. Therapeutic vaccines for prostate Stern PL. Loss of HLA class I expression in prostate 3. Hurwitz MD, Kaplan ID, Svensson GK, Hynynen K, cancer: a review of clinical data. Curr Opin Investig cancer: implications for immunotherapy. Urology 1995; Hansen MS. Feasibility and patient tolerance of a novel Drugs 2005;6:592–6. 46:681–6. transrectal ultrasound hyperthermia system for treat- 9. Simons JW, Mikhak B, Chang JF, et al. Induction of 13. Dranoff G, Jaffee E, Lazenby A, et al. Vaccination ment of prostate cancer. Int J Hyperthermia 2001;17: immunity to prostate cancer antigens: results of a withirradiated tumor cells engineered to secrete 31–7. clinical trial of vaccination withirradiated autologous murine granulocyte-macrophage colony-stimulating 4. Overgaard J. The current and potential role of prostate tumor cells engineered to secrete granulocyte- factor stimulates potent, specific, and long-lasting hyperthermia in radiotherapy. Int J Radiat Oncol Biol macrophage colony-stimulating factor using ex vivo anti-tumor immunity. Proc Natl Acad Sci U S A 1993; Phys 1989;16:535–49. gene transfer. Cancer Res 1999;59:5160–8. 90:3539–43. www.aacrjournals.org 7805 Cancer Res 2007; 67: (16). August 15, 2007

14. Soiffer R, Hodi FS, Haluska F, et al. Vaccination with 25. Inaba K, Swiggard WJ, Steinman R, Romani N, and kinetics of chemokine production regulate dendritic irradiated, autologous melanoma cells engineered to Schuler G. Isolation of Dendritic cells. Hoboken (NJ): cell function. Eur J Immunol 1999;29:1617–25. secrete granulocyte-macrophage colony-stimulating fac- Wiley Interscience; 2006. p. Unit 3.7. 38. Hirao M, Onai N, Hiroishi K, et al. CC chemokine tor by adenoviral-mediated gene transfer augments 26. Srivastava P. Interaction of heat shock proteins with receptor-7 on dendritic cells is induced after interaction antitumor immunity in patients withmetastatic mela- peptides and antigen presenting cells: chaperoning of withapoptotic tumor cells: critical role in migration noma. J Clin Oncol 2003;21:3343–50. the innate and adaptive immune responses. Annu Rev from the tumor site to draining lymph nodes. Cancer 15. Salgia R, LynchT, Skarin A, et al. Vaccination with Immunol 2002;20:395–425. Res 2000;60:2209–17. irradiated autologous tumor cells engineered to secrete 27. Kawakita M, Rao GS, Ritchey JK, et al. Effect of 39. Srivastava PK, Menoret A, Basu S, Binder RJ, granulocyte-macrophage colony-stimulating factor aug- canarypox virus (ALVAC)-mediated cytokine expression McQuade KL. Heat shock proteins come of age: ments antitumor immunity in some patients with on murine prostate tumor growth. J Natl Cancer Inst primitive functions acquire new roles in an adaptive metastatic non-small-cell lung carcinoma. J Clin Oncol 1997;89:428–36. world. Immunity 1998;8:657–65. 2003;21:624–30. 28. GriffithTS, Kawakita M, Tian J, et al. Inhibitionof 40. Basu S, Binder RJ, Ramalingam T, Srivastava PK. 16. Gilboa E, Nair SK, Lyerly HK. Immunotherapy of murine prostate tumor growthand activation of CD91 is a common receptor for heat shock proteins cancer withdendritic-cell-based vaccines. Cancer immunoregulatory cells withrecombinant canarypox gp96, hsp90, hsp70, and calreticulin. Immunity 2001;14: Immunol Immunother 1998;46:82–7. viruses. J Natl Cancer Inst 2001;93:998–1007. 303–13. 17. So-Rosillo R, Small EJ. Sipuleucel-T (APC8015) for 29. Yonezawa M, Otsuka T, Matsui N, et al. Hyperthermia 41. Chen CH, Wang TL, Hung CF, et al. Enhancement of prostate cancer. Expert Rev Anticancer Ther 2006;6: induces apoptosis in malignant fibrous histiocytoma DNA vaccine potency by linkage of antigen gene to an 1163–7. cells in vitro. Int J Cancer 1996;66:347–51. HSP70 gene. Cancer Res 2000;60:1035–42. 18. Small EJ, Schellhammer PF, Higano CS, et al. Placebo- 30. Sauter B, Albert ML, Francisco L, Larsson M, 42. Blachere NE, Udono H, Janetzki S, Li Z, Heike M, controlled phase III trial of immunologic therapy with Somersan S, Bhardwaj N. Consequences of cell death: Srivastava PK. Heat shock protein vaccines against sipuleucel-T (APC8015) in patients withmetastatic, exposure to necrotic tumor cells, but not primary tissue cancer. J Immunother 1993;14:352–6. asymptomatic hormone refractory prostate cancer. cells or apoptotic cells, induces the maturation of 43. Lewis JJ. Therapeutic cancer vaccines: using unique J Clin Oncol 2006;24:3089–94. immunostimulatory dendritic cells. J Exp Med 2000;191: antigens. Proc Natl Acad Sci U S A 2004;101 Suppl 2: 19. Triozzi PL, Khurram R, Aldrich WA, Walker MJ, Kim 423–34. 14653–6. JA, Jaynes S. Intratumoral injection of dendritic cells 31. Gallucci S, Lolkema M, Matzinger P. Natural 44. Srivastava PK. Therapeutic cancer vaccines. Curr derived in vitro in patients withmetastatic cancer. adjuvants: endogenous activators of dendritic cells. Opin Immunol 2006;18:201–5. Cancer 2000;89:2646–54. Nat Med 1999;5:1249–55. 45. Alfieri AA, Hahn EW, Kim JH. Role of cell-mediated 20. Nishioka Y, Hirao M, Robbins PD, Lotze MT, Tahara 32. Melcher A, Todryk S, Hardwick N, Ford M, Jacobson immunity in tumor eradication by hyperthermia. Cancer H. Induction of systemic and therapeutic antitumor M, Vile RG. Tumor immunogenicity is determined by Res 1981;41:1301–5. immunity using intratumoral injection of dendritic cells the mechanism of cell death via induction of heat shock 46. Chakravarty PK, Guha C, Alfieri A, et al. Flt3L therapy genetically modified to express interleukin 12. Cancer protein expression. Nat Med 1998;4:581–7. following localized tumor irradiation generates long- Res 1999;59:4035–41. 33. Basu S, Binder RJ, Suto R, Anderson KM, Srivastava term protective immune response in metastatic lung 21. Mazzolini G, Alfaro C, Sangro B, et al. Intratumoral PK. Necrotic but not apoptotic cell deathreleases heat cancer: its implication in designing a vaccination injection of dendritic cells engineered to secrete shock proteins, which deliver a partial maturation signal strategy. Oncology 2006;70:245–4. interleukin-12 by recombinant adenovirus in patients to dendritic cells and activate the NF-n B pathway. Int 47. Chakravarty PK, Alfieri A, Thomas EK, et al. Flt3- withmetastatic gastrointestinal carcinomas. J Clin Immunol 2000;12:1539–46. ligand administration after radiation therapy prolongs Oncol 2005;23:999–1010. 34. Larmonier N, Merino D, Nicolas A, et al. Apoptotic, survival in a murine model of metastatic lung cancer. 22. Kirk CJ, Hartigan-O’Connor D, Nickoloff BJ, et al. necrotic, or fused tumor cells: an equivalent source of Cancer Res 1999;59:6028–32. T cell-dependent antitumor immunity mediated by antigen for dendritic cell loading. Apoptosis 2006;11: 48. Corica FA, Cheng L, Ramnani D, et al. Transurethral secondary lymphoid tissue chemokine: augmentation 1513–24. hot-water balloon thermoablation for benign prostatic of dendritic cell-based immunotherapy. Cancer Res 35. Kotera Y, Shimizu K, Mule JJ. Comparative analysis of hyperplasia: patient tolerance and pathologic findings. 2001;61:2062–70. necrotic and apoptotic tumor cells as a source of Urology 2000;56:76–80. 23. Thompson TC, Southgate J, Kitchener G, Land H. antigen(s) in dendritic cell-based immunization. Cancer 49. Djavan B, Susani M, Shariat S, et al. Transperineal Multistage carcinogenesis induced by ras and myc Res 2001;61:8105–9. radiofrequency interstitial tumor ablation (RITA) of the oncogenes in a reconstituted organ. Cell 1989;56:917–30. 36. Forster R, Schubel A, Breitfeld D, et al. CCR7 prostate. TechUrol 1998;4:103–9. 24. Baley PA, Yoshida K, Qian W, Sehgal I, Thompson TC. coordinates the primary immune response by establish- 50. Sanghvi NT, Foster RS, Bihrle R, et al. Noninvasive Progression to androgen insensitivity in a novel in vitro ing functional microenvironments in secondary lym- surgery of prostate tissue by high intensity focused mouse model for prostate cancer. J Steroid Biochem Mol phoid organs. Cell 1999;99:23–33. ultrasound: an updated report. Eur J Ultrasound 1999;9: Biol 1995;52:403–13. 37. Sallusto F, Palermo B, Lenig D, et al. Distinct patterns 19–29.

Cancer Res 2007; 67: (16). August 15, 2007 7806 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Localized Hyperthermia Combined with Intratumoral Dendritic Cells Induces Systemic Antitumor Immunity

Arunika Mukhopadhaya, Joseph Mendecki, Xinyuan Dong, et al.

Cancer Res 2007;67:7798-7806.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/67/16/7798

Cited articles This article cites 49 articles, 17 of which you can access for free at: http://cancerres.aacrjournals.org/content/67/16/7798.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/67/16/7798.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/67/16/7798. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.