Late Intervention with anti-BRAFV600E Therapy …...curs in this model by 35 d after tumor implantation as we pre-viously showed (20). Groups 1–4 were assigned to no treatment and

Late Intervention with anti-BRAFV600E TherapyInduces Tumor Regression in an Orthotopic MouseModel of Human Anaplastic Thyroid Cancer

Matthew A. Nehs,* Carmelo Nucera,* Sushruta S. Nagarkatti, Peter M. Sadow,Dieter Morales-Garcia, Richard A. Hodin, and Sareh Parangi

Thyroid Cancer Research Laboratory (M.A.N., C.N., S.S.N., D.M.-G., R.A.H. S.P.), Endocrine Surgery Unit,Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts; Division of CancerBiology and Angiogenesis (C.N.), Department of Pathology, Beth Israel Deaconess Medical Center,Harvard Medical School, Boston, Massachusetts, and Department of Pathology (P.M.S.), MassachusettsGeneral Hospital, Harvard Medical School, Boston, Massachusetts

Human anaplastic thyroid cancer (ATC) is a lethal disease with an advanced clinical presentationand median survival of 3 months. The BRAFV600E oncoprotein is a potent transforming factor thatcauses human thyroid cancer cell progression in vitro and in vivo; therefore, we sought to targetthis oncoprotein in a late intervention model of ATC in vivo. We used the human ATC cell line 8505c,which harbors the BRAFV600E and TP53R248G mutations. Immunocompromised mice were random-ized to receive the selective anti-BRAFV600E inhibitor, PLX4720, or vehicle by oral gavage 28 d aftertumor implantation, 1 wk before all animals typically die due to widespread metastatic lung diseaseand neck compressive symptoms in this model. Mice were euthanized weekly to evaluate tumorvolume and metastases. Control mice showed progressive tumor growth and lung metastases by35 d after tumor implantation. At that time, all control mice had large tumors, were cachectic, andwere euthanized due to their tumor-related weight loss. PLX4720-treated mice, however, showeda significant decrease in tumor volume and lung metastases in addition to a reversal of tumor-related weight loss. Mouse survival was extended to 49 d in PLX4720-treated animals. PLX4720treatment inhibited cell cycle progression from 28 d to 49 d in vivo. PLX4720 induces striking tumorregression and reversal of cachexia in an in vivo model of advanced thyroid cancer that harbors theBRAFV600E mutation. (Endocrinology 153: 985–994, 2012)

Despite tremendous advances in our understanding ofthe genetic and molecular events that characterize ag-

gressive thyroid cancers, anaplastic thyroid cancer (ATC)remains a lethal disease with a median survival of less than6 months (1, 2). ATC frequently presents as a large andrapidly expanding neck mass (3), and at the time of pre-sentation, the tumor has often invaded surrounding struc-tures and has metastasized to distant locations. It is forthese reasons that surgical therapies consistently fail, andtracheostomy is often necessary as a palliative measure toimprove patient comfort for local-regional control in thesetting of a massive tumor that invades or obstructs the

airway (4). These tumors tend to be resistant to standardchemotherapy, external beam radiation, and radioiodine(131I) due to loss of the sodium iodide symporter throughmalignant dedifferentiation (5). Because all of these treat-ments for ATC have marginal efficacy, new treatments areneeded that use rationally designed molecularly-targetedagents and demonstrate efficacy for tumors at this latestage of presentation (6).

B-RafV600E is an oncoprotein that is implicated in thepathogenesis and progression of ATC (7–9) and is prev-alent in several other human cancers, such as papillarythyroid cancer (PTC), melanoma, and colon cancer.

ISSN Print 0013-7227 ISSN Online 1945-7170Printed in U.S.A.Copyright © 2012 by The Endocrine Societydoi: 10.1210/en.2011-1519 Received July 18, 2011. Accepted November 30, 2011.First Published Online December 27, 2011

* M.A.N. and C.N. contributed equally to this study.

Abbreviations: ATC, Anaplastic thyroid cancer; ECM, extracellular matrix; GFP, green flu-orescent protein; IHC, immunohistochemical; PTC, papillary thyroid cancer; SCID, severecombined immunodeficient; TSP-1, thrombospondin-1; TUNEL, terminal deoxynucleotidyltransferase dUTP nick end labeling.

T H Y R O I D - T R H - T S H

Endocrinology, February 2012, 153(2):985–994 endo.endojournals.org 985

B-RafV600E oncoprotein is a potent transforming factorthat causes human thyroid cancer cell progression in vitroand in vivo (9). This mutation is present in approximately25%–38% of ATCs (10, 11). The mutant B-RafV600E pro-tein results from a transversion (T1799A) in exon 15 of theB-Raf gene, which causes a valine-for-glutamate substi-tution at residue 600 of the protein. This mutation inducesa conformational change in the protein that constitutivelyactivates the MAPK pathway (i.e. ERK1/2) (12).

PLX4720 is an ATP analog that selectively inhibitsB-RafV600E by stabilizing it in an inactive conformation(13). This highly selective molecule binds with an IC50 of13 nM, and thus its off-target binding is minimized.PLX4720 has undergone considerable in vitro and in vivotesting in animal models of melanoma and colon cancer(13–15). In addition, a recent clinical trial (16) using arelated compound (PLX4032) in patients with melanomahas shown considerable efficacy and may lead to furthertrials with other cancers that harbor the B-RafV600E mu-tation, such as colorectal and thyroid cancers.

The utility of molecular targeting with PLX4720 forthyroid cancer has been tested in vitro by Nucera et al. (9)and others (17) showing strong inhibition of B-RafV600E inthyroid cancer cell lines with this mutation. We have re-cently demonstrated that mRNA and protein expressionlevels of human thrombospondin-1 (TSP-1), a key regu-lator of tumor invasion and extracellular matrix (ECM)remodeling, were significantly increased in B-RafV600E

-positive human PTC compared with those with wild-typeB-Raf or normal thyroid tissue. Additionally, we foundthat either TSP-1 or B-RafV600E mRNA knockdown in8505c ATC cells by short hairpin RNA (9), or targetingB-RafV600E with PLX4720 (18), significantly inhibitedproliferation, migration, and invasion of these thyroid tu-mor cells.

We have also recently shown that oral administrationof PLX4720 within 1 wk of orthotopic ATC implantationcan result in delayed tumor growth and decreased tumorvolume in an early intervention in vivo model (12). Al-though B-RafV600E inhibition did decrease tumor growthin our previous prevention trial, it is also clinically relevantto know whether molecular targeting at a later interven-tion time point would also be efficacious given the uni-formly late clinical presentation of patients with ATC. Nostudy to date has tested a selective B-RafV600E inhibitor ina late-intervention orthotopic animal model of ATC,which is a model that reproduces the advanced clinicalpresentation in humans with this type of cancer. Thereforein this study, we began PLX4720 treatment in mice afterlarge orthotopic ATC and widespread lung metastaseswere already established. Bioimaging and histology haveconfirmed the rapid timeline for disease progression and

metastasis in this animal model of human ATC, a situationthat closely resembles the rapidly fatal disease seen in hu-mans. Our results demonstrate, for the first time, that anti-B-RafV600E therapy with orally delivered PLX4720 causesorthotopic ATC regression and reverses cachexia, evenwhen administered just 1 wk before death using a late-intervention model of ATC in vivo.

Materials and Methods

Human ATC cell lineThe human ATC 8505c cell line harboring B-RafV600E (19)

was purchased from DSMZ (German collection of microorgan-isms and cell culture). The 8505c cells were engineered to expressgreen fluorescent protein (GFP) as described by Nucera et al.(20). Cells were grown in RPMI 1640 with penicillin/strepto-mycin/amphtoericin with 10% fetal bovine serum.

Tumor implantationThe animal work was done in the animal facility at Massa-

chusetts General Hospital (Boston, MA) in accordance with fed-eral, local, and institutional guidelines. We used an orthotopicmouse model of ATC as previously described and validated byNucera et al. (9. 20). Briefly, 48 female 6-wk-old severe com-bined immunodeficient (SCID) mice (Charles River Laborato-ries, Wilmington, MA) were anesthetized using ketamine andxylazine. For each mouse, the neck was shaved and preparedwith betadine. The skin and sc tissues were incised with scissors,and the salivary glands were reflected superiorly. The centralcomponent of the neck was exposed, and the overlying strapmuscles were bluntly dissected away from the right thyroid lobeusing forceps. Once the right thyroid was exposed, 5 � 105

8505c cells resuspended in 10 �l of serum-free RPMI were in-jected into the thyroid using a Hamilton syringe (Fisher Scien-tific, Pittsburgh, PA) attached to a 27-gauge needle. After theinjection, the salivary glands were repositioned, and the incisionwas closed using nylon sutures. Triple antibiotic ointment wasapplied to the wound, and the mice were placed under a warminglamp while they recovered from anesthesia. There were no sur-gical or anesthetic complications.

PLX4720 preparation and administrationWe dissolved PLX4720 in the vehicle, dimethyl sulfoxide,

initially at a concentration of 120 mg/ml (final concentration: 30mg/kg body weight/d). It was then suspended in a 1% solution ofcarboxymethycellulose (Sigma Chemical Co., St. Louis, MO)daily at the time of administration. Mice were gavaged with thedrug suspension using a 24-gauge blunt-tipped gavage needle.

Group stratification and randomizationIn detail, 48 mice were randomly divided into eight equal

groups of six mice for the purpose of establishing a time courseof tumor progression and the response to late therapeutic inter-vention with PLX4720 (Fig. 1). Late therapeutic interventionwas defined as initiating treatment 1 wk (at 28 d) before natural(not per protocol) death (mice showed physical criteria to bekilled, i.e. advanced thyroid tumors that caused distention and

986 Nehs et al. Anti-BRAFV600E Therapy and Thyroid Tumor Regression Endocrinology, February 2012, 153(2):985–994

pressure of the trachea and esophagus), which reproducibly oc-curs in this model by 35 d after tumor implantation as we pre-viously showed (20).

Groups 1–4 were assigned to no treatment and were killedafter 7, 14, 21, and 28 d, respectively, to allow a detailed radio-logical, gross, and histological analysis of tumor progression andmetastasis, knowing that these large groups serve as a timelineproxy for all the groups before the initiation of treatment.

Group 5 was a control group treated with vehicle on d 28–d35 and then killed at 35 d given the large tumor burden andcachexia. Finally, groups 6, 7, and 8 had treatment withPLX4720 initiated at 28 d after tumor implantation at 30 mg/kg/d PLX4720 for 1, 2, and 3 wk, respectively. Each group waskilled after the completion of its treatment.

Because it has been previously demonstrated that mice in thismodel have a large and dramatic tumor burden and cachexia at35 d after tumor injection (9), we chose to begin treatment withPLX4720 1 wk earlier (28 d after injection) to accurately simu-late the advanced disease presentation seen in humans. To char-acterize the growth pattern of this model, six mice were killedweekly, their tumors were measured, lungs were evaluated formetastases, and tissues were analyzed by histology. Body weight(grams) of all remaining mice was measured weekly. Tumor sizewas measured using an electronic caliper (Fisher Scientific,Hampton, NH), and tumor volume was calculated as (1/2) �length � width � height. Lungs were also collected and analyzedfor metastasis by histology and GFP imaging as described below.Mice were killed when they had lost more than 25% of theirbaseline weight or had signs of tracheal compressive symptoms.As expected in this study, all control mice displayed features ofadvanced disease (e.g. compression of the trachea and esophagusor severe cachexia, by 35 d). We report “no surviving mice” afterthe 35-d time point for control mice, and this reflects the fact thatthese mice were intentionally killed because of these physicalcriteria to minimize pain and suffering for the humane treatmentof the animals. These physical criteria were established a priorifor both groups at the beginning of the study.

BioimagingTo detect GFP signal in the primary tumor and lungs, we used

a Multi-spectral Fluorescence Scanner (CRi Maestro 500,Woburn, MA) that was able to detect and image fluorophoresbetween 450 and 900 nm. The tumor and lung specimens wereimaged with fluorescence microscopy immediately after nec-ropsy (wavelength absorption, 488 nm) (Leica Microsystems,Bannockburn, IL).

Histological and immunohistochemical (IHC)analysis

All tissue specimens were fixed with 10% buffered formalinphosphate and embedded in paraffin blocks. Histopathologyevaluation was performed by an endocrine pathologist (P.S.) onhematoxylin and eosin-stained tissue sections of the orthotopicthyroid tumors, the surrounding perithyroidal tissues, and thelungs. These were visualized with an Olympus BX 41 microscopeand the Olympus Q COLOR 5 photo camera (Olympus Corp.,Lake Success, NY). Sections (4-�m thick) of formalin-fixed tis-sues were used for IHC procedures. After baking overnight at 37C, deparaffinization with xylene/ethanol and rehydration wereperformed. IHC analysis was performed using the phospho-ERK1/2 (Cell Signaling Technology, Danvers, MA), Ki67 (DakoEnvision system, DAKO Corp., Carpinteria, CA), cyclins E2(CCNE2) (Abcam, Cambridge, MA), caspase-3 and cleavedcaspase-3 (Cell Signaling Technology), Cyclin D1, Cyclin B1,p21, and p27/kip1 (Abcam) primary antibodies. The sections,treated with pressure cooker for antigen retrieval (Biocare Med-ical, Concord, CA), were incubated at 123 C in citrate buffer(Dako Target Retrieval Solution, S1699; DAKO Corp.), cooledand washed with PBS. Antigen retrieval was performed for 60min at room temperature. The primary antibody was detectedusing a biotin-free secondary antibody (K4011) (Dako Envisionsystem) and incubated for 30 min. All incubations were carriedout in a humid chamber at room temperature. Slides were rinsedwith PBS between incubations. Sections were developed using3,3�-diaminobenzidine (Sigma Chemical Co.) as a substrate andwere counterstained with Mayer’s hematoxylin.

The IHC markers phospho-ERK1/2, CCNE2, p27/kip1, cas-capase-3, and cleaved caspase-3 were assessed semiquantita-tively using the following scoring method: 0 negative, 1 �1–10% positive cells (low expression), 2 � 11–50% positive cells(moderate), and 3� more than 50% positive cells (highexpression).

Proliferative index and apoptosis assay in vivoImmunostaining for nuclear Ki67 was evaluated by an endo-

crine pathologist (P.S.) in the PLX4720-treated and controlgroups. The proportion of nuclei that stained positive for Ki67relative to background nuclei was counted for each tumor andaveraged. Data are reported as mean percent for positive Ki67staining. The apoptosis assay was performed using terminal de-oxynucleotidyl transferase dUTP nick end labeling (TUNEL) as-say according to manufacturer’s recommendation (ApopTagperoxidase in situ detection kit) (Millipore Corp., Bedford, MA);apoptotic cells were assessed semiquantitatively using the fol-lowing scoring method: 0 negative, 1 � 1–10% positive cells(low expression), 2 � 11–50% positive cells (moderate), and 3�more than 50% positive cells (high expression).

FIG. 1. Experimental design. Orthotopic anaplastic tumors wereimplanted in 48 SCID mice. Each week, six mice were killed. Theirtumors and lungs were evaluated by histology and GFP biomaging.Either vehicle or PLX4720 was begun at 28 d, and the response totherapy was evaluated weekly. “No surviving mice” reflects the factthat all mice at this time point showed physical criteria to be killed (e.g.cachexia and/or advanced thyroid tumors causing compression of thetrachea and esophagus).


Scoring for metastasesMetastatic dissemination in the lungs was assessed by GFP

imaging at necropsy. The lungs were dissected from the mouse,washed with PBS, and visualized by GFP imaging. For eachmouse, the number of metastases was counted as the number ofdistinct GFP signals per square centimeter (cm2) in an anteriorprojection of the lungs. The number of metastases found in eachmouse was averaged per group.

Statistical analysisStatistical analysis was carried out using Microsoft Excel

software. Two-tailed t-tests were used to evaluate statisticaldifferences. Data are reported as the mean value, and errorbars represent the SE of the mean for each group. Results withP � 0.05 were considered significant. *, P � 0.05; **, P �0.01; ***, P � 0.001.

Results

Late intervention with PLX4720 reverses weightloss and cachexia in an in vivo model of humanATC

Weight measurements in control mice were used to gen-erate a clear comparison group. Mice in all groups lostweight starting 14 d after tumor implantation. As we havepreviously demonstrated (20), control untreated mice ingroups 1–5 (n � 6/group) consistently lost weight, devel-oped cachexia and piloerection, and met a priori criteria tobe killed (weight loss �25% from baseline or trachealcompression) by 35 d (Fig. 2A). At the 35 d time point, thecontrol mice had lost additional weight (average weight12.5 g � 0.3 g) and had progression of cachexia as shownin Fig. 2B. No mice in the control group survived longerthan 35 d (all six had lost �25% of their baseline weight).

In the mice treated with PLX4720, however, not onlywas the progression of weight loss and cachexia stopped,but the process was reversed. Each treated mouse gainedweight within a week of PLX4720 treatment, and therewas a significant increase in weight in all treated micecompared with the control mice by 35 d after tumor im-plantation (15.9 g � 0.5 g vs. 12.5 g � 0.3 g; P � 0.005).As shown in Fig. 2A, mice in both the control andPLX4720-treated groups had a similar progression ofweight loss from d 14–d 28 (from �19 g to 15.5 g). Atypical example from each treatment group is shown inFig. 2B. The control mouse (from group 5) had a baselineweight of 18.7 g. Its weight had declined to 16.6 g by d 28and had progressed to severe cachexia (11.2 g) by d 35.Conversely, the mouse from group 8 began the experimentwith a baseline of 18.3 g and had progressed to a weightof 14.4 g on d 28. After 3 wk of daily PLX4720 treatment,there was a progressive weight gain up to 17 g by 49 d anda clinical resolution of the cachectic phenotype.

Late intervention with PLX4720 induces tumorregression and reduces tumor volume andinvasiveness in an in vivo model of human ATC

Nucera et al. (21) and Salerno et al. (17) have previouslyshown that PLX4720 is able to suppress B-RafV600E signalingcascade (i.e. down-regulation of ERK1/2 phosphorylation lev-els) in human thyroid cancer cells harboring the B-RafV600E

mutation (e.g. 8505c human thyroid cancer cells). Therefore,we administered the anti-B-Raf V600E compound, PLX4720in a late-intervention model of human ATC.

A tumor volume growth histogram demonstrates pro-gressive growth until the tumor burden was fatal by 35 d(Fig. 3A). At 7 d after injection, small tumors measuring anaverage of 1 mm3 were present in the lateral aspect of theinjected thyroid lobe; by 14 d after tumor implantation,there was an average tumor volume of 14 mm3 withoutevidence of tracheal compression or displacement. At 21 d,the tumor mass (average 44 mm3) had expanded to cause aslight compression of the trachea resulting in a leftward de-viation. By 28 d, the control tumors had grown in the planebetween the tracheaandesophagusandextended tobilateralinvolvement (average 57 mm3).

FIG. 2. PLX4720 extends survival and reverses cachexia in orthotopicmice with human ATC. A, Mouse weight over time. B, Untreated micedeveloped weight loss, piloerection, and cachexia by 28 d. Vehicle-treated mice progressed to severe cachexia, whereas PLX4720-treatedmice had reversal of weight loss, cachexia, and piloerection over the3-wk course of therapy. *, P � 0.05.


The mice that received vehicle from d 28 – d 35 had anaverage tumor volume of 115 mm3 at 35 d and showedcircumferential tumor involvement around the tracheawith invasion into the esophagus. All of these mice werekilled because of their severe tumor burden by 35 d after

tumor implantation. All mice had large tumors by 28 dafter implantation (average 57 mm3).

At 35 d after tumor implantation, those animals treatedwith 7 d of daily oral PLX4720 showed dramatically de-creased tumor volume compared with controls (average

FIG. 4. PLX4720 induces regression of lung metastases. Number of metastases/cm2 as detected by GFP imaging. A, Vehicle-treated mice hadexponential growth in the number of metastases, whereas PLX4720-treated mice had a significant reduction in the number of metastases by 49 d(B); *, P � 0.05. C, Anterior projections of the lungs with GFP imaging and gross photos.

FIG. 3. PLX4720 induces tumor regression. A, Vehicle-treated mice had an exponential tumor growth up to 115 mm3 � 14 mm3 by 35 d. Nomouse in this group survived beyond 35 d. B, PLX4720-treated mice had tumor regression down to 14 mm3 � 1 mm3 by 49 d. C, Tumorprogression and regression (gross photos and GFP images). *, P � 0.05.


25 mm3 for PLX4720 and 115 mm3 for controls; P �0.001). With continued treatment, tumor size continuedto decrease (18 mm3 at 42 d and 14 mm3 at 49 d) as shownin Fig. 3B. Thus, 3 wk of treatment with PLX4720 reducedmean tumor volume by 75% from 57 mm3 at 28 d to 14mm3 at 49 d. Vehicle-treated mice had large, invasive tu-mors that progressively grew circumferentially around thetrachea as shown in Fig. 3C. Although this invasion waspresent after 28 d of tumor growth, the mice that weretreated with 3 wk of PLX4720 had regression of invasionin addition to significantly smaller tumors (Fig. 3C).

PLX4720 induces regression of lung metastases inan in vivo model of human ATC

Untreated mice developed a progressive rise in the num-ber of lung metastases as evaluated by GFP imaging atnecropsy up to d 28 of the experiment (Fig. 4A). Meta-static disease to the lung was an early event, with miceshowing evidence of a small number of GFP-positive tu-mor colonies in the lung even as early as 1 wk after tumor

implantation (Fig. 4C). The six vehicle-treated mice hadon average 383 GFP metastases per cm2 of lung by 28 dafter tumor implantation, and the lung metastatic burdenincreased significantly to an average of 624 GFP metas-tasis/cm2 of lung (P � 0.001) by 35 d after tumor implan-tation, and no mouse in the vehicle-treated group survivedbeyond this time point (Fig. 4C). Mice treated with 1 wkof PLX4720, however, had a stabilization of the averagenumber of lung metastases at 35 d (370 GFP metastasis/cm2), and this was significantly fewer than vehicle-treatedmice at 35 d as shown in Fig. 4B (624 � 92 in controls vs.370 � 43 GFP metastases/cm2 in PLX4720-treated mice;P � 0.05). Importantly, with 3 wk of PLX4720 treatment,the mice in groups 7 and 8 (42-/49-d time point) had evenfurther significant regression in the number of lung me-tastases compared with the 28-d untreated mice in group4 (down to 223 � 51 GFP metastases/cm2; P � 0.05. Theprogression and regression of lung metastases are shownby GFP imaging in Fig. 4C.

PLX4720 inhibits tumorproliferation and cell cycleprogression in an in vivomodel of human ATC

Control tumors at 35 d demonstratedextensive tracheal and esophageal inva-sion, and large number of lung microme-tastases (Fig. 5, A–C). Mice treated with3 wk of PLX4720, however, showedsmaller tumors, less invasion, and de-crease in the number of lung metastases(Fig. 5, D–F). Tumor cell proliferationrates in vivo were measured in bothgroups using nuclear Ki67 expression byIHC analysis (Fig. 5G). Control tumor-bearing mice treated with vehicle had anaverage Ki67 expression (proliferationindex) of 71.7% � 6% at 35 d aftertumor implantation. Late therapeuticintervention with PLX4720 starting at28 d after tumor implantation resulted ina significant decrease in the tumor pro-liferation rate (proliferation index) up to49 d in the orthotopic tumors down to15.0% � 2.9% expression as shown inFig.5(P�0.01).Moreover,wefoundthatnuclear phospho-ERK1/2 protein levelswere significantly decreased in thePLX4720-treated tumors (3�) comparedwith controls (1�) (Fig. 5H) (P � 0.01).

We have also analyzed protein ex-pression of different molecules (cyclin

FIG. 5. Histology of orthotopic human ATC. Control tumors at 35 d demonstratedextrathyroidal extension (e.g. tracheal and esophageal invasion) (panels A and B), and lungmetastases (asterisk, panel C). Mice treated with 3 wk of PLX4720, however, showed smallerand circumferential tumors (arrow, panels D and E), and decreases in lung metastases(asterisk, panel F). Ki67 immunostaining was significantly decreased with PLX4720 treatment(71.6% � 6% in controls compared with 15% � 2.9% in PLX4720-treated mice, P � 0.01)(panel G). Phospho-ERK1/2 immunostaining was significantly decreased with PLX4720treatment (�80% in controls compared with 10–20% in PLX4720-treated mice, P � 0.01)(panel H).


D1, cyclin E2, cyclin B1, p27/kip1, and p21) involved incell cycle checkpoints for the regulation of cell prolifera-tion/DNA synthesis. Importantly, we have found thatPLX4720 effectively targeted B-RafV600E in vivo and in-hibited cell cycle progression in vivo by down-regulatingprotein levels of CCNE2 (Fig. 6A) and up-regulating cellcycle inhibitor p27/kip1(Fig. 6B). Additionally, we foundthat protein levels of both cleaved caspase-3 (an activatedform of caspase-3) (low expression) and total caspase-3(moderate/high expression) (Fig. 7, A and B) were not

significantly different in the PLX4720-treated orthotopicATC compared with untreated ATC; this result was alsoconfirmed by TUNEL assay that showed a slight increaseof apoptotic cells in the PLX4720-treated group at 42 dand less at 49 d compared with controls (Fig. 7C).

Discussion

ATC is unresponsive to all current treatment regimens.Rational drug design targeting this cancer looks at altering

FIG. 6. Immunoexpression of Cyclin E2 and p27/kip1. A, Control tumors show strong nuclear Cyclin E2 (CCNE2) protein levels (2�) at 28 d and35 d. Orthotopic tumors treated with 3 wk of PLX4720 show a significant decrease of nuclear CCNE2 protein levels (moderate, 2�) at 35 d and42 d, which progressively decrease up to 49 d (week, 1�). All magnifications, �400. B, Control tumors show low nuclear p27/kip1 protein levels(2�) at 28 d and 35 d. Orthotopic tumors treated with 3 wk of PLX4720 show a significant increase of p27/kip1 protein levels (strong, 3�) at 35 dand 42 d, which progressively increase up to 49 d (strong, 3�). All magnifications, �400.

FIG. 7. Immunoexpression of total caspase-3 and cleaved caspase-3. A, Control tumors and PLX4720-treated tumors show strong (3�)cytoplasmic total caspase 3 protein levels (2�) at 28, 35, 42, and 49 d. All magnifications, �400. B, Control tumors show moderate cytoplasmiccleaved caspase 3 protein levels (2�) at 28 d and 35 d. Orthotopic tumors treated with 3 wk of PLX4720 show a slight but not significant increaseof cytoplasmic cleaved caspase 3 protein levels at 35 d (2�) and 42 d (3�), and moderate at 49 d (2�). C, TUNEL assay using peroxidase system.Slight increase of nuclear staining (apoptotic cells) in the PLX4720-treated orthotopic tumors at 42 d compared with vehicle (control)-treatedtumors. All magnifications, �400.


the level of a known mutant oncoprotein implicated in theaggressive course of this lethal disease. Molecular targetsfor ATC and recurrent PTC are mostly centered aroundthe RAS/RAF/MAPK (mitogen-activated ERK1/2) kinasesignaling pathway given the prominence of this pathwayas an oncogenic event in PTC and ATC progression (22–25). In 2002, Davies et al. (26) identified an oncogenewidespread among human cancers, mutant V600E B-Raf,or B-RafV600E. B-RafV600E is expressed in a variety of humancancers including melanoma, colorectal, glioma, sarcoma,and thyroid cancer (8, 12). Interestingly, the B-RafV600E mu-tation is seen exclusively in PTC and ATC but not in otherhistotypes, such as medullary or follicular thyroid cancers.Histopathological correlation studies in patients with PTCdemonstrate an increase in the B-RafV600E mutation fre-quency in tumors of patients with extrathyroidal extension,thyroid capsular invasion, and lymph node metastases (27,28). Confirmatory animal studies have also shown that thy-roid specific expression of B-RafV600E in transgenic miceleads to poorly differentiated PTC (29).

Recent studies have looked for a distinct molecular ex-planation for how the B-RafV600E mutation leads to amore aggressive histopathological phenotype. Recently,Nucera et al. (9) have shown that B-RafV600E-positivePTCs have a distinct genetic profile with altered gene ex-pression in critical gene sets important in tumor invasionand ECM remodeling. The presence of the B-RafV600E mu-tation promotes ATC invasion via TSP-1 in vitro and invivo, a key regulator of ECM remodeling. Additionally,knockdown of B-RafV600E in vivo results in decreased tu-mor volume and lung metastases. Nucera et al. (9) havepreviously shown that oral administration PLX4720within 1 wk (early intervention) in an orthotopic mousemodel of human ATC can result in significant growth in-hibition of tumor volume.

The current study demonstrates the remarkable effi-cacy of this compound to induce significant regression inan orthotopic mouse model of ATC even when adminis-tered at a very late therapeutic intervention stage. Al-though previous studies have investigated the mechanisticrole of anti-B-RafV600E therapies in melanoma and colo-rectal cells (13, 30–32), and in thyroid cancer cells models(9, 17, 21), no previous study has investigated the effect ona late intervention model of ATC that reproduces the ad-vanced disease seen in patients with this malignancy.

In this study, the timeframe of tumor growth and me-tastasis of ATC 8505c cells was characterized, and theresponse to a late intervention treatment was evaluated.Nucera et al. (20) have demonstrated that this orthotopicmodel of human ATC is characterized by the induction ofintra-thyroidal tumors that grow and extend into sur-rounding structures, metastasize to the lungs, and result in

death by 35 d after the implantation of the tumors. Inter-estingly, we have also found that metastasis is a very earlyevent in this model (Fig. 4C), occurring by 7 d after tumorimplantation when the primary tumor size is only 1.5 mm3

(Fig. 3C). This finding is consistent with the aggressive andlethal disease seen in humans and also highlights previousreports on the important role of the B-RafV600E mutationin the early invasion and metastasis in this human thyroidcancer model (9). Moreover, Knauf et al. (33) have shownthat tumor initiation by oncogenic B-Raf renders thyroidcells susceptible to TGF�-induced epithelial-mesenchy-mal transition, through a MAPK-dependent process.

Given that our previous studies have shown that anti-B-RafV600E therapy inhibits the early growth of 8505c invivo (9), we anticipated that tumor size, mouse weight,and metastases would stabilize with PLX4720. Surpris-ingly, however, we found that the mice recovered 86% ofweight that they lost from baseline and had clinical reso-lution of cachexia (piloerection, muscle atrophy, weightloss). The vehicle-treated mice progressed to severe ca-chexia, whereas the PLX4720-treated mice demonstratedreversal of cachexia and weight gain. The underlying causeof thisweight gain is incompletelyunderstood,but itmighthave resulted from the inhibition of the systemic burden ofmetastases, from the relief of esophageal compression bythe large thyroid tumors, or a combination of these twoprocesses. Late therapeutic intervention with PLX4720treatment not only reversed weight loss, decreased tumorvolume, tumor cell proliferation and number of lung me-tastasis dramatically but also extended the mice’s lifespanby 40% compared with controls. All mice were killed ac-cording to a predetermined experimental plan by 49 dafter tumor implantation, partially due to limited drugsupply. Many of these had the equivalent tumor burden ofa control mouse at 2 wk after implantation and thus mayhave lived several weeks longer if they had not been killed.

The tumors themselves showed a surprising degree ofregression after only 1 wk of PLX4720 therapy, and thiseffect continued further with tumor volume decreasingdown to 9 mm3 with 3 wk of PLX4720 treatment. Primarytumor regression caused by PLX4720 correlated with sig-nificant decreases in tumor cell proliferation rates (Ki67nuclear expression), and in phospho-ERK1/2 (42 d treated��35 d treated, vs. 35 d untreated) protein levels down-stream of B-RafV600E kinase activity. Furthermore, thenumber of pulmonary metastases was significantly de-creased after 3 wk of PLX4720 treatment (223/cm2), eventhough treatment was initiated at such a late stage. Over-all, our results do show that even when treating mice withinitially advanced tumors for 21 d (from 28 d to 49 d) withPLX4720 we saw no increase in tumor size. Additionally,we have found that when the same 8505c human meta-


static thyroid cancer cells are engineered with short hair-pin RNA to stably knock-down B-RafV600E and then or-thotopically injected into the thyroid of SCID mice, therewas a significant delay in thyroid tumor growth and in-hibition of lung metastasis at 35 d compared with controls(9). We have previously shown that PLX4720 arrests8505c cells in G1 (32), and here we found that cyclinE2 was down-regulated in the orthotopic ATC afterPLX4720 treatment. Cyclin E2 belongs to the family ofE-type cyclins that operate during the G1/S phase progres-sion in mammalian cells. It has a catalytic partner, cyclin-dependent kinase 2, and activates its associated kinaseactivity at similar times during cell cycle progression. Cy-clin E2 controls S-phase (DNA synthesis) entry of differentcell types in a tissue-specific fashion and cell cycle pro-gression (34). Additionally, p27kip1 (but not p21), astrong inhibitor of cell proliferation, was up-regulatedwith PLX4720 treatment in vivo from 35 d up to 49 d.Although PLX4720 significantly inhibited cell cycle pro-gression interfering with G1/S checkpoints (causing G1 ar-rest and inhibition of S-phase/DNA synthesis), overall thenumber of apoptotic cells assessed by cleaved caspase-3protein expression levels and TUNEL assay were not sig-nificantly different in the PLX4720-treated orthotopicATC compared with controls.

Although development of this drug for use in someforms of thyroid cancer is promising and should be pur-sued, lessons from melanoma treatment trials in patientswith advanced melanoma show that in some cases humancancer cells could develop escape mechanisms to PLX4720.Further studies into molecular and stromal targets of theB-RafV600E mutation will allow rational design of serial orconcurrent therapeutics (18, 35). Although resistance totherapy with this drug remains a possibility for prolongedhuman use, the dramatic results of our study are encour-aging for the treatment of B-RafV600E-positive ATC, forwhich no effective therapy currently exists.

In conclusion, the anti-B-RafV600E compound PLX4720 in-duces striking tumor regression, reverses cachexia, andextends survival in a late-intervention orthotopic model ofATC. Our results demonstrate that even late therapeuticintervention directed at the proper molecular step in thy-roid tumor progression may be effective in halting andreversing the process. Translating this therapeutic strategyto patients presenting with large ATC that harbor theB-RafV600E mutation may cause a reduction in primarytumor volume or metastatic burden, thus allowing time toproceed with additional therapies such as surgical deb-ulking or radiation. Additionally, interesting recent invitro data suggest potential sodium iodide symporter reg-ulation in certain thyroid cancer cell lines by mutantB-RafV600E. If down-regulation of B-Raf with anti-B-

RafV600E therapy also causes sodium iodide symporter up-regulation, then theoretically some patients may be givenanti-B-RafV600E therapy both to reduce tumor size and met-astatic burden and possibly to allow redifferentiation, thusmaking radioactive iodine administration possible in somepatients for control of additional metastatic burden (36, 37).Theseresults, incombinationwithotherpreclinicalstudiesofPLX4720, provide justification for a human clinical trial ofthis drug for ATC that harbor B-RafV600E. By implementingmolecularly targeted agents that inhibit this cancer’s drivingmutations, we may expand the armamentarium of treat-ments for this otherwise incurable disease.

Acknowledgments

We thank Dr. Yutaka Kawakami (Keio University, Tokyo, Ja-pan) for providing HIV-U6 vectors. We also thank Dr. GideonBollag and Dr. Paul Lin of Plexxikon (Berkeley, CA) for technicalassistance with PLX4720 administration.

This work was funded by National Institutes of HealthGrants 1R01CA149738-01 (to S.P.) and 5T32 DK007754-10(to R.H.). C.N. is an M.D. and recipient of Ph.D. fellowshipsupported from the Italian Ministry of Education, Universitiesand Research (MIUR) and carried out at Harvard MedicalSchool; he was also funded through A. Gemelli Medical School(Catholic University, Rome, Italy).

Address all correspondence and requests for reprints to: SarehParangi, M.D., Associate Professor of Surgery, Harvard MedicalSchool, Massachusetts General Hospital, Endocrine SurgeryUnit, Wang ACC 460, 15 Parkman Street, Boston, Massachu-setts 02115. E-mail: [email protected].

Disclosure Summary: All authors declare no financial con-flicts of interest.

References

1. Neff RL, Farrar WB, Kloos RT, Burman KD 2008 Anaplastic thy-roid cancer. Endocrinol Metab Clin North Am 37:525–538

2. Are C, Shaha AR 2006 Anaplastic thyroid carcinoma: biology,pathogenesis, prognostic factors, and treatment approaches. AnnSurg Oncol 13:453–464

3. Broome JT, Gauger PG, Miller BS, Doherty GM 2009 Anaplasticthyroid cancer manifesting as new-onset Horner syndrome. EndocrPract 15:563–566

4. Pasieka JL 2003 Anaplastic thyroid cancer. Curr Opin Oncol 15:78–835. Braga-Basaria M, Ringel MD 2003 Clinical review 158: Beyond

radioiodine: a review of potential new therapeutic approaches forthyroid cancer. J Clin Endocrinol Metab 88:1947–1960

6. Wiseman SM, Griffith OL, Deen S, Rajput A, Masoudi H, Gilks B,Goldstein L, Gown A, Jones SJ 2007 Identification of molecular mark-ers altered during transformation of differentiated into anaplastic thy-roid carcinoma. Arch Surg 142:717–727; discussion 727–729

7. Nikiforov YE 2004 Genetic alterations involved in the transitionfrom well-differentiated to poorly differentiated and anaplastic thy-roid carcinomas. Endocr Pathol 15:319–327


8. XingM2007BRAFmutationinpapillarythyroidcancer:pathogenicrole,molecular bases, and clinical implications. Endocr Rev 28:742–762

9. Nucera C, Porrello A, Antonello ZA, Mekel M, Nehs MA, GiordanoTJ, Gerald D, Benjamin LE, Priolo C, Puxeddu E, Finn S, Jarzab B,Hodin RA, Pontecorvi A, Nose V, Lawler J, Parangi S 2010B-Raf(V600E) and thrombospondin-1 promote thyroid cancer pro-gression. Proc Natl Acad Sci USA 107:10649–10654

10. Smallridge RC, Marlow LA, Copland JA 2009 Anaplastic thyroidcancer: molecular pathogenesis and emerging therapies. EndocrRelat Cancer 16:17–44

11. Ricarte-Filho JC, Ryder M, Chitale DA, Rivera M, Heguy A,Ladanyi M, Janakiraman M, Solit D, Knauf JA, Tuttle RM, Ghos-sein RA, Fagin JA 2009 Mutational profile of advanced primary andmetastatic radioactive iodine-refractory thyroid cancers reveals dis-tinct pathogenetic roles for BRAF, PIK3CA, and AKT1. Cancer Res69:4885–4893

12. Nucera C, Goldfarb M, Hodin R, Parangi S 2009 Role ofB-Raf(V600E) in differentiated thyroid cancer and preclinical vali-dation of compounds against B-Raf(V600E). Biochim Biophys Acta1795:152–161

13. Tsai J, Lee JT, Wang W, Zhang J, Cho H, Mamo S, Bremer R,Gillette S, Kong J, Haass NK, Sproesser K, Li L, Smalley KS, FongD, Zhu YL, Marimuthu A, Nguyen H, Lam B, Liu J, Cheung I, RiceJ, Suzuki Y, Luu C, Settachatgul C, Shellooe R, Cantwell J, Kim SH,Schlessinger J, Zhang KY, West BL, Powell B, Habets G, Zhang C,Ibrahim PN, Hirth P, Artis DR, Herlyn M, Bollag G 2008 Discoveryof a selective inhibitor of oncogenic B-Raf kinase with potent anti-melanoma activity. Proc Natl Acad Sci USA 105:3041–3046

14. Packer LM, East P, Reis-Filho JS, Marais R 2009 Identification ofdirect transcriptional targets of (V600E)BRAF/MEK signalling inmelanoma. Pigment Cell Melanoma Res 22:785–798

15. Emery CM, Vijayendran KG, Zipser MC, Sawyer AM, Niu L, KimJJ, Hatton C, Chopra R, Oberholzer PA, Karpova MB, MacConaillLE, Zhang J, Gray NS, Sellers WR, Dummer R, Garraway LA 2009MEK1 mutations confer resistance to MEK and B-RAF inhibition.Proc Natl Acad Sci USA 106:20411–20416

16. Flaherty KT, Puzanov I, Kim KB, Ribas A, McArthur GA, SosmanJA, O’Dwyer PJ, Lee RJ, Grippo JF, Nolop K, Chapman PB 2010Inhibition of mutated, activated BRAF in metastatic melanoma.N Engl J Med 363:809–819

17. Salerno P, De Falco V, Tamburrino A, Nappi TC, Vecchio G, SchweppeRE, Bollag G, Santoro M, Salvatore G 2010 Cytostatic activity of aden-osine triphosphate-competitive kinase inhibitors in BRAF mutant thyroidcarcinoma cells. J Clin Endocrinol Metab 95:450–455

18. Nucera C, Lawler J, Parangi S 2011 BRAFV600E and microenvi-ronment in thyroid cancer: a functional link to drive cancer pro-gression. Cancer Res 71:2417–2422

19. Meireles AM, Preto A, Rocha AS, Rebocho AP, Maximo V, Pereira-Castro I, Moreira S, Feijao T, Botelho T, Marques R, Trovisco V,Cirnes L, Alves C, Velho S, Soares P, Sobrinho-Simoes M 2007Molecular and genotypic characterization of human thyroid follic-ular cell carcinoma-derived cell lines. Thyroid 17:707–715

20. Nucera C, Nehs MA, Mekel M, Zhang X, Hodin R, Lawler J, NoseV, Parangi S 2009 A novel orthotopic mouse model of human ana-plastic thyroid carcinoma. Thyroid 19:1077–1084

21. Nucera C, Nehs MA, Nagarkatti SS, Sadow PM, Mekel M, FischerAH, Lin PS, Bollag GE, Lawler J, Hodin RA, Parangi S 2011 Tar-geting BRAFV600E with PLX4720 displays potent antimigratoryand anti-invasive activity in preclinical models of human thyroidcancer. Oncologist 16:296–309

22. Gupta-Abramson V, Troxel AB, Nellore A, Puttaswamy K,Redlinger M, Ransone K, Mandel SJ, Flaherty KT, Loevner LA,O’Dwyer PJ, Brose MS 2008 Phase II trial of sorafenib in advancedthyroid cancer. J Clin Oncol 26:4714–4719

23. Sherman SI, Wirth LJ, Droz JP, Hofmann M, Bastholt L, MartinsRG, Licitra L, Eschenberg MJ, Sun YN, Juan T, Stepan DE, Schlum-berger MJ, Motesanib Thyroid Cancer Study G 2008 Motesanib

diphosphate in progressive differentiated thyroid cancer. N EnglJ Med 359:31–42

24. Kloos RT, Ringel MD, Knopp MV, Hall NC, King M, Stevens R,Liang J, Wakely Jr PE, Vasko VV, Saji M, Rittenberry J, Wei L,Arbogast D, Collamore M, Wright JJ, Grever M, Shah MH 2009Phase II trial of sorafenib in metastatic thyroid cancer. J Clin Oncol27:1675–1684

25. Cohen EE, Rosen LS, Vokes EE, Kies MS, Forastiere AA, WordenFP, Kane MA, Sherman E, Kim S, Bycott P, Tortorici M, ShalinskyDR, Liau KF, Cohen RB 2008 Axitinib is an active treatment for allhistologic subtypes of advanced thyroid cancer: results from a phaseII study. J Clin Oncol 26:4708–4713

26. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S, TeagueJ, Woffendin H, Garnett MJ, Bottomley W, Davis N, Dicks E, EwingR, Floyd Y, Gray K, Hall S, Hawes R, Hughes J, Kosmidou V,Menzies A, Mould C, Parker A, Stevens C, Watt S, Hooper S, et al.2002 Mutations of the BRAF gene in human cancer. Nature417:949–954

27. Xing M, Clark D, Guan H, Ji M, Dackiw A, Carson KA, Kim M,Tufaro A, Ladenson P, Zeiger M, Tufano R 2009 BRAF mutationtesting of thyroid fine-needle aspiration biopsy specimens for pre-operative risk stratification in papillary thyroid cancer. J Clin Oncol27:2977–2982

28. Frasca F, Nucera C, Pellegriti G, Gangemi P, Attard M, Stella M,Loda M, Vella V, Giordano C, Trimarchi F, Mazzon E, Belfiore A,Vigneri R 2008 BRAF(V600E) mutation and the biology of papillarythyroid cancer. Endocr Relat Cancer 15:191–205

29. Knauf JA, Ma X, Smith EP, Zhang L, Mitsutake N, Liao XH, Refet-off S, Nikiforov YE, Fagin JA 2005 Targeted expression ofBRAFV600E in thyroid cells of transgenic mice results in papillarythyroid cancers that undergo dedifferentiation. Cancer Res 65:4238–4245

30. Hatzivassiliou G, Song K, Yen I, Brandhuber BJ, Anderson DJ, Al-varado R, Ludlam MJ, Stokoe D, Gloor SL, Vigers G, Morales T,Aliagas I, Liu B, Sideris S, Hoeflich KP, Jaiswal BS, Seshagiri S,Koeppen H, Belvin M, Friedman LS, Malek S 2010 RAF inhibitorsprime wild-type RAF to activate the MAPK pathway and enhancegrowth. Nature 464:431–435

31. Heidorn SJ, Milagre C, Whittaker S, Nourry A, Niculescu-Duvas I,Dhomen N, Hussain J, Reis-Filho JS, Springer CJ, Pritchard C,Marais R 2010 Kinase-dead BRAF and oncogenic RAS cooperate todrive tumor progression through CRAF. Cell 140:209–221

32. Poulikakos PI, Zhang C, Bollag G, Shokat KM, Rosen N 2010 RAFinhibitors transactivate RAF dimers and ERK signalling in cells withwild-type BRAF. Nature 464:427–430

33. Knauf JA, Sartor MA, Medvedovic M, Lundsmith E, Ryder M, Sal-zano M, Nikiforov YE, Giordano TJ, Ghossein RA, Fagin JA 2011Progression of BRAF-induced thyroid cancer is associated with ep-ithelial-mesenchymal transition requiring concomitant MAP kinaseand TGF� signaling. Oncogene 30:3153–3162

34. Geng Y, Yu Q, Whoriskey W, Dick F, Tsai KY, Ford HL, Biswas DK,Pardee AB, Amati B, Jacks T, Richardson A, Dyson N, Sicinski P2001 Expression of cyclins E1 and E2 during mouse developmentand in neoplasia. Proc Natl Acad Sci USA 98:13138–13143

35. Nucera C, Lawler J, Hodin R, Parangi S 2010 The BRAFV600Emutation: what is it really orchestrating in thyroid cancer? Onco-target 1:751–756

36. Durante C, Puxeddu E, Ferretti E, Morisi R, Moretti S, Bruno R,Barbi F, Avenia N, Scipioni A, Verrienti A, Tosi E, Cavaliere A,Gulino A, Filetti S, Russo D 2007 BRAF mutations in papillarythyroid carcinomas inhibit genes involved in iodine metabolism.J Clin Endocrinol Metab 92:2840–2843

37. Riesco-Eizaguirre G, Rodrıguez I, De la Vieja A, Costamagna E,Carrasco N, Nistal M, Santisteban P 2009 The BRAFV600E onco-gene induces transforming growth factor � secretion leading to so-dium iodide symporter repression and increased malignancy in thy-roid cancer. Cancer Res 69:8317–8325


Documents

Late Intervention with anti-BRAFV600E Therapy …...curs in this model by 35 d after tumor implantation as we pre-viously showed (20). Groups 1–4 were assigned to no treatment and