Noninvasive imaging of cell proliferation following mitogenic extracellular kinase inhibition by PD0325901

The mitogen-activated protein kinase (MAPK) pathway involving RAS, RAF, mitogenic extracellular kinase (MEK1/2), and extracellular signal-regulated kinase (ERK) regulate cell proliferation, survival, and differentiation in response to extracellular signals. The pathway is activated in several cancers with overall mutation rates of ∼15% to 30% for RAS (1) and ∼7% for BRAF (2). In human melanoma, BRAF mutations occur in ∼66% of cases (2), with the most common mutation being a glutamic acid for valine substitution at position 600 (V600E). This substitution, by mimicking phosphorylation of the activation segment, activates BRAF ∼500-fold and stimulates constitutive MAPK activity (3). In colorectal cancer, K-RAS mutations occur in ∼27% of cases (4). The MAPK pathway is therefore a key target for therapeutic intervention. MEK1/2 has no reported oncogenic properties but serves as a central axis for this pathway.

A number of drugs, including CI-1040, PD0325901, and AZD6244 have been developed to target MEK1/2 (5–9). Activating mutations in BRAF and RAS have been shown to sensitize cells to these MEK1/2 inhibitors (6, 9), but clinical response is not limited to tumors harboring such mutations (5). Furthermore, all these inhibitors have been shown to reduce levels of phosphorylated ERK1/2 (pERK1/2) in cells in vitro or in tumor biopsies (5–9), although the best clinical response observed has been stable disease (5, 7, 8). These findings indicate that reduction in pERK1/2 does not necessarily predict growth inhibition. Cell proliferation has been proposed as a useful biomarker of MEK1/2 inhibition (10). Clearly, assessment of cell proliferation by imaging will be more desirable than multiple biopsy methods. In the present study, we examined the relationship between imaging variables derived from 3′-deoxy-3′-[¹⁸F]fluorothymidine-positron emission tomography ([¹⁸F]FLT-PET) and cognate biochemical effects of the MEK1/2 inhibitor PD0325901 (9). [¹⁸F]FLT-PET measures thymidine kinase 1 (TK1) activity in vivo (11, 12). A decrease in the activities of G₁ phase cyclins and retinoblastoma hypophosphorylation regulate TK1 mRNA and protein, leading to decreases in [¹⁸F]FLT uptake (13). Efficacious doses of cytotoxic drugs and gene therapies also induce reductions in tumor [¹⁸F]FLT uptake (14, 15). Indeed, Solit and coworkers have recently shown the advantage of [¹⁸F]FLT-PET over [¹⁸F]fluorodeoxyglucose-PET for monitoring the therapeutic response to PD0325901 (16). However, dynamic imaging was not done in that study; therefore, the value of different imaging variables and their relationship(s) to proliferation signals were not established. Prior to in vivo studies with PD0325901, we established the time course of drug effect in a sensitive cell line (SKMEL-28) in vitro. The potential of different [¹⁸F]FLT-PET variables as downstream pharmacodynamic imaging end points of growth inhibition following MEK1/2 inhibition was then assessed in two human tumor xenograft models—SKMEL-28 melanoma and HCT116 colorectal cancer.

[¹⁸F]FLT was synthesized by radiofluorination of the 3-N-boc-5′-O-dimethoxytrityl-3′-O-nosyl-thymidine precursor using a GE TracerLab MX synthesis module (GE Healthcare UK Limited) based on a previously described method (17). All samples had >98% radiochemical purity as determined by high-performance liquid chromatography with radiochemical detection, and the specific radioactivity was 120 to 150 GBq/μmol at the end of the synthesis. PD0325901 (N-[(R)-2,3-dihydroxy-propoxy]-3,4-difluoro-2-[2-fluoro-4-iodo-phenylamino]-benzamide) was obtained from the Division of Signal Transduction Therapy, University of Dundee, Dundee, Scotland.

SKMEL-28 and HCT116 cells were cultured in RPMI 1640 growth medium containing 10% (v/v) fetal bovine serum, 2 mmol/L of l-glutamine, 100 units/mL of penicillin, and 100 g/mL of streptomycin and grown in a 5% CO₂ incubator at 37°C. Cells in exponential growth were used for subsequent studies. To examine the effect of PD0325901 in vitro on SKMEL-28 cells, the growth medium was replaced with medium containing 1 μmol/L of PD0325901 or DMSO (the final concentration of DMSO in both cases was 0.1%). The plates were then incubated for 0.5, 2, 4, 24, or 48 hours and harvested.

The levels of pERK1/2, total ERK1/2, cyclin D1, TK1, pAKT, total AKT, cleaved caspase-3 (Asp¹⁷⁵), and β-actin (loading control) protein were determined by Western blot analysis. In addition, Western blot analyses were also done on excised frozen tumor material. For this, SKMEL-28 and HCT116 tumors (after imaging studies) were snap-frozen in liquid nitrogen and stored at −80°C until analysis. The tumors were pulverized in liquid nitrogen and homogenized in 100 μL of ice-cold triple detergent buffer [50 mmol/L Tris-HCl (pH 8.0), 150 nmol/L NaCl, 0.02% sodium azide, 0.1% SDS, 1% NP40, 0.5% sodium deoxycholate, and 10 μL/mL protease inhibitor mixture]. Lysates were clarified by centrifugation (15,000 × g for 10 min at 4°C) and the supernatants were analyzed for total protein content using a commercial bicinchoninic acid protein assay kit (Perbio Science). Aliquots of the supernatant (containing 30 μg of protein) were assayed as previously reported (13).

The primary antibodies used in this study included the following: rabbit polyclonal anti-phosphorylated p44/42 MAPK (Thr²⁰²/Tyr²⁰⁴) antibody (1:1,000), rabbit polyclonal anti-total p44/42 MAPK antibody (1:1,000 dilution), rabbit polyclonal anti–cyclin D1 antibody (1:1,000), mouse biotinylated monoclonal anti-TK1 (no. 1-8-5:4-1D11, 1:20,000 dilution; Svanova Biotech), rabbit polyclonal antiphosphorylated pAKT and rabbit polyclonal anti-total AKT (Cell Signaling Technology), cleaved caspase-3 (Asp¹⁷⁵), and mouse polyclonal anti–β-actin (1:2,000; Santa Cruz Biotechnology). The membranes were processed and bands were quantified as previously reported (13). Four tumor samples were analyzed from each of the control and treatment groups.

The effect of 1 μmol/L of PD0325901 on the transcription of a number of candidate genes was assessed in SKMEL-28 cells. Cells were washed and RNA from the cells was extracted with Trizol (Invitrogen Life Technologies) as per the manufacturer's instructions with a slight modification in that the aqueous phase was cleaned up with chloroform/isoamyl alcohol mixture (24:1).

The cDNA generated was used as a template in real-time PCR reactions with a SYBR-Green PCR master mix (Applied Biosystems). All amplification reactions were done in triplicate using 96-well optical reaction plates (MicroAmp, Applied Biosystems) on an Applied Biosystem 7500 Fast System. The results were then normalized to the housekeeping gene HPRT and expressed as a ratio. The primers were designed from the literature or from the GenBank Ensembl sequence. The sequences were assessed for species-specificity by a BLAST search.⁵

Because [¹⁸F]FLT uptake is dependent not just on TK1, but also ATP (11), we examined the effect of PD0325901 on ATP in tumor samples. ATP levels were determined by a bioluminescence assay (ENLITEN ATP assay system; Promega) as previously reported (13).

All experiments were done by licensed investigators in accordance with the United Kingdom Home Office's regulations under the Animal (Scientific Procedures) Act 1986 and were within guidelines set out by the United Kingdom Coordinating Committee for Cancer Research's Ad hoc Committee on the Welfare of Animals in Experimental Neoplasia (18). We used two xenograft models to assess the biological activity of PD0325901, i.e., human melanoma SKMEL-28 and human colon carcinoma HCT116 xenografts. Six- to 8-week-old BALB/c nu/nu mice were obtained from Harlan United Kingdom, Ltd., and tumors were induced by s.c. inoculation of 5 × 10⁶ SKMEL-28 cells (BRAF mutant) or HCT116 cells (K-RAS mutant) on the back of the mice. SKMEL-28 cells were prepared as a suspension in 25% of growth medium in basement membrane matrix (BD Matrigel; BD Bioscience); HCT116 cells were prepared in PBS. Tumor dimensions were measured continuously using a caliper and tumor volumes were calculated by the equation: volume = (π/6) × a × b × c, where a, b, and c represent three orthogonal axes of the tumor. Mice were used when their tumors reached ∼100 mm³. Size-matched tumor-bearing mice were randomized to receive 25 mg/kg (0.005 mL/g mouse) of PD0325901 or vehicle (0.5% hydroxypropyl methylcellulose plus 0.2% Tween 80) daily by oral gavage. [¹⁸F]FLT-PET scanning was done at ∼25 hours after commencing treatment (i.e., after two doses with the second dose given 1 hour before scanning) or after 10 daily treatments (again with the last dose given 1 hour before scanning). After imaging, one part of the tumor was snap-frozen in liquid nitrogen and stored at −80°C; the other part was fixed in formal saline.

Dynamic [¹⁸F]FLT-PET studies were carried out on a dedicated small animal PET scanner, quad-HIDAC (Oxford Positron Systems). The features of this instrument have been previously described (19). Scanning was done using a previously reported protocol (13). Dynamic scans were acquired in list-mode format over a 60-min period.

Cumulative images comprising of 30 to 60 min of the dynamic data were used for visualization of radiotracer uptake and to draw regions of interest. Regions of interest were defined on five tumor and five heart-slices (each 0.5 mm in thickness). Dynamic data from these slices were averaged for each tissue and at each of the 19 time points to obtain time versus radioactivity curves for these tissues. Tumor radioactivity was corrected for physical decay and normalized to that of heart to obtain a standardized uptake value (NUV). The NUV at 60 min postinjection (NUV₆₀) was used for comparisons; to improve signal-to-noise ratio, NUV_max was also calculated in a similar manner as NUV₆₀ except that maximal voxel radioactivities for each tumors' region of interest were used. The area under the NUV curve (AUC) was calculated as the integral of NUV from 0 to 60 min. The fractional retention of tracer (FRT) at 60 min relative to that at 1.5 min was also calculated (13).

PD0325901- and vehicle-treated tumors (n = 4 each) were excised after imaging, fixed in formalin, embedded in paraffin, and cut into 5.0-μm sections. The degree of tumor proliferation was assessed in tumor sections as previously described (13). For each tumor section, the total number of Ki67-positive brown-stained cells and hematoxylin-stained cells were counted in 16 randomly selected fields of view using an Olympus BX51 microscope (Olympus UK, Ltd.) at ×400 magnification. The Ki67 labeling index (LI_Ki67) was calculated using the equation LI_Ki67 = [Ki67-positive cells / (total cells)] × 100%.

Deparaffinized sections were fixed in 3% to 4% paraformaldehyde in PBS (pH 7.4) for 15 min at room temperature and permeabilized for 20 min with PBS containing 0.25% Triton X-100 at room temperature, and then washed thrice with PBS for 5 min. After washing, the sections were blocked with 1% bovine serum albumin in PBS containing 0.25% Triton X-100 for 30 min at room temperature. Subsequently, they were incubated with a mixture of pERK1/2 antibodies (1:100; Cell Signaling) and Ki67 antibodies (1:100; Dako) diluted in 1% bovine serum albumin in PBS containing 0.25% Triton X-100 for 16 h at 4°C. After washing (five times) with PBS, the sections were incubated for 1 hour at room temperature with a mixture of Alexa Fluor 488–conjugated goat anti-rabbit antibodies (Invitrogen) diluted 1:400 and Alexa Fluor 594–conjugated goat anti-mouse antibodies (Invitrogen) diluted 1:400 in 1% bovine serum albumin. The sections were then washed five times with PBS and mounted with gel mount containing 4′,6-diamidino-2-phenylindole (Invitrogen). Specimens were examined for fluorescence with an Olympus BX51 microscope (Olympus); the results were compared with H&E staining.

Data were expressed as mean ± SE. Statistical analyses were done using the software GraphPad Prism, version 3.03 (GraphPad). Group comparisons were assessed using the nonparametric Mann-Whitney test. Correlations between tumor radiotracer levels and LI_Ki67 were determined (Pearson correlation). Two-tailed P ≤ 0.05 were considered significant.

The phosphorylation of ERK1/2 protein (44/42 kDa) was detected by Western blotting. A time-dependent reduction in pERK1/2 was induced by 1 μmol/L of PD0325901 (Fig. 1A). Maximum and significant reduction of pERK1/2 was observed at 4 hours compared with vehicle control (P = 0.03) after treating cells with PD0325901, and there was partial recovery of the protein at 24 and 48 hours after treatment (P = 0.05). The observed change in pERK1/2 expression was not accompanied by changes in total ERK1/2 expression, indicating that the reduction in ERK1/2 phosphorylation by PD0325901 was not due to a depletion of the protein.

In most cell types, sustained activation of MAPK signaling pathway leads to activation of the cyclin D–cdk4 complex to modulate cell proliferation (9, 20, 21). PD0325901 (1 μmol/L) reduced cyclin D1 protein expression (Fig. 1B) at 24 and 48 hours after treatment compared with vehicle control (P = 0.03). Compared with pERK1/2 (Fig. 1B), the recovery of cyclin D1 was slower. Like cyclin D1 protein, TK1 protein levels decreased with PD0325901 treatment in a time-dependent manner (Fig. 1C); the decrease in TK1 expression following drug treatment was also slower than those for pERK1/2, with maximum reduction at 48 hours posttreatment (P = 0.03). The drug caused a small but significant induction of caspase-3 cleavage at 24 and 48 hours compared with vehicle-treated cells (Fig. 1D). Lastly, as a negative control, we examined the effect of PD0325901 on pAKT levels. No consistent time-related alterations in pAKT levels were observed (Fig. 1E) at the drug concentration used (1 μmol/L).

The mRNA levels of TK1, cyclin D1, and ELK in SKMEL-28 cells were all significantly down-regulated within hours following treatment with 1 μmol/L of PD0325901 compared with vehicle control (Fig. 2A–C). Significant inhibition of cyclin D1 and TK1 mRNA was detectable at the earliest time point (4 hours; P = 0.001 and 0.03, respectively) and persisted for the duration of the study (48 hours). The nuclear transcription factor ELK, which is a direct downstream target of pERK1/2 (22, 23) was also down-regulated at 4 hours in response to PD0325901. The recovery kinetics of ELK were, however, more rapid than those for cyclin D1 and TK1. Of note, PD0325901 also induced a significant G₁ arrest at the lowest dose concentration studied (0.01 μmol/L): 81.6 ± 0.8% of cells were in G₁ phase compared with 56.8 ± 4.4% for vehicle control (P = 0.03; Supplementary Fig. S1).⁶

For both tumor types, tumor volumes were not significantly reduced at day 1 following PD0325901 (25 mg/kg) treatment (Fig. 3A). In SKMEL-28, tumor volume changed from 168.0 ± 38.5 mm³ (pretreatment group) to 181.2 ± 28.8 mm³ and 248.0 ± 36.9 mm³ at days 1 and 10, respectively, following treatment with vehicle. In contrast, tumor shrinkage was observed after daily treatment with PD0325901 (25 mg/kg) to 135.8 ± 19.0 and 43.2 ± 16.3 mm³ (P = 0.05), at days 1 and 10, respectively. HCT116 tumor volumes showed a rapid increase in the vehicle-treated group from 106.670 ± 15.70 mm³ (pretreatment group) to 136.7 ± 38.6 mm³ and 542.3 ± 115.0 mm³ at days 1 and 10, respectively. In contrast, tumor stasis was seen with PD0325901 treatment; tumor volumes were 82.9 ± 18.9 and 128.5 ± 29.3 mm³ (P = 0.05) at days 1 and 10, respectively. No significant changes in mouse body weights were observed with drug treatment (data not shown).

Typical (0.5 mm) transverse PET image slices demonstrating the accumulation of [¹⁸F]FLT in tumor regions are shown in Fig. 3B. Qualitative reductions in radiotracer localization were observed in the tumors of PD0325901-treated mice. Whereas NUV increased over time in vehicle-treated tumors, PD0325901 induced significant reductions in [¹⁸F]FLT uptake compared with vehicle treatment (Fig. 3C and D). Summary data are presented in Table 1 and show that although the NUV₆₀, NUV_max, AUC, and FRT were unchanged in the vehicle control groups, all the variables were significantly lower in the PD0325901-treated groups compared with corresponding vehicle-treated groups. FRT in HCT116 tumor decreased but the decrease was not statistically significant.

Significant changes in LI_Ki67 were seen after 1 and 10 days of drug treatment in SKMEL-28 xenografts (P = 0.03 for both; Fig. 4A and B). Similarly, LI_Ki67 was lower in HCT116 tumors after 10 days of drug treatment (P = 0.05; Fig. 4C). In this tumor, changes in LI_Ki67 at day 1 were not statistically significant. All tumor data (SKMEL-28 and HCT116) were combined for the assessment of the degree of association between tumor [¹⁸F]FLT kinetic variables and cell proliferation. The correlation coefficient (r) between LI_Ki67 and NUV₆₀ was 0.630 (P = 0.001), and that for LI_Ki67 versus FRT was 0.596 (P = 0.003); LI_Ki67 versus AUC was also significant (r = 0.647; P = 0.002). There was also a significant correlation between [¹⁸F]FLT kinetic variables and tumor TK1 expression: r was 0.518 (P = 0.002), 0.423 (P = 0.015), and 0.480 (P = 0.005) for NUV₆₀, AUC, and FRT, respectively.

Significant reductions in pERK1/2, cyclin D1, and TK1 protein levels were observed after drug treatment in vivo compared with vehicle controls (Supplementary Table S2). ATP levels were also significantly lower after 1 and 10 days of drug treatment compared with controls (Supplementary Table S2).⁶

The coexpression of Ki67 and pERK1/2 in SKMEL-28 tumor sections is shown in Fig. 5. Whereas Ki67-positivity (red) was confined to the nucleus pERK1/2-positivity (green) was both nuclear and cytoplasmic. Coexpression of pERK1/2-positive cells and Ki67-positive cells was indicated by the yellow fluorescence. In vehicle-treated tumors, almost all the pERK1/2-positive cells were also Ki67-positive (Fig. 5A). The coexpression of pERK1/2 and Ki67 was detectable after 1 day of PD0325901 treatment but the total number of (yellow) cells within the tumor mass had decreased (Fig. 5B). At 10 days after PD0325901 treatment, Ki67 nuclear staining was barely detectable. In contrast, there were pockets of cells with pERK1/2 cytoplasmic staining; no pERK1/2 nuclear staining was detectable (Fig. 5C).

We have shown that the reduction in cell proliferation induced by PD0325901 is detectable by [¹⁸F]FLT-PET variables NUV₆₀, NUV_max, AUC, and FRT. The MAPK signal transduction pathway is an attractive target for therapeutic intervention because several cancers harbor amplifications and mutations of genes within the pathway leading to constitutive activation of downstream effectors. In preclinical models, MEK1/2 inhibitors showed activity in tumors harboring mutations in BRAF or RAS (6, 9). In recent phase I and II clinical trials of these inhibitors, stable disease was observed as the best clinical outcome despite significant reductions in pERK1/2 levels (5, 7, 8). Smalley and coworkers (10) recently showed that drug-induced reductions in the proliferation marker Ki67 correlated better with the growth-inhibitory activity of MEK1/2 inhibitors than changes in pERK1/2 and suggested the use of this biomarker in addition to pERK1/2 in future trials of MEK1/2 inhibitors. Because a noninvasive imaging biomarker may be more desirable than a histologic biomarker such as Ki67, we investigated if and how [¹⁸F]FLT-PET imaging of cell proliferation could fulfill this role. Indeed, Solit and coworkers have recently shown the advantage of [¹⁸F]FLT-PET over [¹⁸F]fluorodeoxyglucose-PET in monitoring the response to the MEK1/2 inhibitor, PD0325901 (16).

[¹⁸F]FLT uptake into tumors is a function of the levels of the TK1 enzyme and of the cofactor ATP (11, 12). Studies in xenograft models (13–15, 19) and in human tumors (24, 25) have shown strong correlations between [¹⁸F]FLT uptake and LI_Ki67. In vitro studies in SKMEL-28 cells showed that mRNA and protein levels of TK1 were rapidly decreased after MEK1/2 inhibition by PD0325901 (Figs. 1 and 2). These findings, together with the strong G₁ arrest in treated cells, suggested that [¹⁸F]FLT might be a useful biomarker for imaging MEK1/2 inhibition by PD0325901. The time course studies after single drug dosing indicated that recovery of pERK1/2 and ELK was more rapid than that for the protein/mRNA for cyclin D1 and TK1. These findings indicate that a TK1-based imaging probe will not measure MEK1/2 activity directly but rather downstream effects of MEK1/2 inhibition, the kinetics of which may be somewhat different from direct target activity.

Treatment of SKMEL-28 and HCT116 tumors led to reductions in [¹⁸F]FLT uptake in drug-treated mice compared with vehicle controls. In contrast with previous reports suggesting that cells with K-RAS mutation are unresponsive to PD0325901 (9), we saw [¹⁸F]FLT responses in the HCT116 tumor model. In these imaging studies, baseline radiotracer uptake should be assumed to be similar to that in day 1 vehicle-treated mice. In previous [¹⁸F]FLT imaging studies in RIF-1 (14) and IGROV-1 (15), we found no differences between radiotracer uptake/biochemical variables in size-matched untreated tumors versus tumors treated with vehicle for 1 day, and therefore, as with recent studies, we compared day 1 vehicle-treated to day 1 drug-treated groups (13, 26). The lower [¹⁸F]FLT uptake in drug-treated compared with vehicle-treated tumors at day 1, therefore, supports the assertion that [¹⁸F]FLT-PET imaging can be used to measure early changes in proliferation following PD0325901 treatment. Of interest, where drug effects on proliferation are delayed, we see no early changes in tumor [¹⁸F]FLT uptake, as shown in a recent study (26). The study design reported here is different from one that will be used clinically, i.e., pretreatment scans followed by posttreatment scans in the same patient, but is useful in that it permits direct comparison of the imaging end point to tumor biochemistry in individual animal tumors.

In general, the reductions in [¹⁸F]FLT kinetic variables were associated with changes in LI_Ki67 (r ≥ 0.6; P ≥ 0.003) suggesting that [¹⁸F]FLT-PET imaging may be useful as a marker of changes in cell proliferation following MEK1/2 inhibition. This is in agreement with previous studies with other drugs in which good correlations were seen between [¹⁸F]FLT uptake variables and LI_Ki67 (13–15). There were significant differences in LI_Ki67 between drug-treated animals and their respective vehicle control groups for SKMEL-28 (days 1 and 10) and HCT116 (day 10). Changes in LI_Ki67 did not, however, discriminate between tumor shrinkage in a SKMEL-28 model and modest growth inhibition in a HCT116 model, a finding that could be explained at least in part by the very different growth rates of the two tumor types. Also of note, a direct relationship between HCT116 tumor volume and [¹⁸F]FLT uptake was not seen. This finding is not dissimilar from our previous observations in mice (13–15, 19) and humans (27), and supports the notion that changes in proliferation are more sensitive and precede tumor size changes. The group-averaged PET kinetic variables may, however, have been influenced by spatial differences in radiotracer uptake, which could have introduced errors in averaging of [¹⁸F]FLT within regions of interest. For instance, although the group-averaged NUV₆₀ was unchanged between days 1 and 10 for drug-treated SKMEL-28, LI_Ki67 decreased. The reduction of NUV_max was more consistent with changes in LI_Ki67 than in NUV₆₀, presumably because of the lower signal-to-noise characteristics of the former. The reduction in PET variables was consistent with changes in TK1 enzyme and ATP levels in vivo.

An important aspect of this work was to investigate, with a dynamic imaging protocol, if different [¹⁸F]FLT-PET variables change differently with drug treatment. In general, all four [¹⁸F]FLT-PET variables decreased after drug treatment with slightly different levels of change at the two time points studied (Table 1A and B). Reductions in all four imaging variables implied that the drug-induced effects detectable by [¹⁸F]FLT-PET could not simply be attributed to modulation of, for instance, tumor perfusion which may accompany the biological activity of some kinase inhibitors. This finding also implies that the semiquantitative imaging variable—standardized uptake value—often employed in clinical PET studies may be a useful variable in the clinical pharmacodynamic assessment of MEK1/2 inhibition by PD0325901 and potentially other MEK1/2 inhibitors.

A more detailed study to link the effect of PD0325901 on pERK1/2 to changes in Ki67 showed a time-dependent reduction in both pERK1/2 and Ki67 expression by PD0325901. The presence of Ki67-negative pERK1/2 (cytoplasmic)–positive cells in tumors treated with PD0325901 for 10 days is consistent with the need for pERK1/2 to translocate to the nucleus in order to affect transcription (20–23). Analysis of tumor material by Western blotting does not discriminate cytoplasmic from nuclear pERK1/2 levels. Thus, assuming that reduction in nuclear pERK1/2 (Fig. 5C) is a major effect of this drug, small changes in pERK1/2 detectable by Western blot could be associated with large changes in Ki67. Regarding other mechanisms of drug response, an increase in sub-G₀ (apoptotic) cells and cleaved caspase-3 activities were seen in treated cells in vitro, indicating that the apoptotic pathway is active with PD0325901 treatment. The level of apoptosis was, however, distinctly low in both tumor types, indicating that apoptosis was not a major mechanism of tumor growth inhibition. An important consideration for pharmacodynamic assessment of pathway-specific drugs is that the concentration of drug required to inhibit pathway activity may be insufficient to inhibit downstream signaling irrespective of upstream mutational status. The use of end points, such as [¹⁸F]FLT-PET, that report on growth inhibition should be considered alongside end points which report on target activity.

In summary, [¹⁸F]FLT-PET is a sensitive imaging biomarker for detecting the antiproliferative effect of MEK1/2 inhibition by PD0325901; it does not directly measure MEK1/2 activity. This imaging method may be useful in the clinical development of PD0325901 and other inhibitors of the MAPK pathway.

No potential conflicts of interest were disclosed.

We thank the members of the Molecular Therapy group, Imperial College London, for critical reading and discussion of the manuscript, and the staff of the Biological Imaging Centre of Imperial College for support with imaging studies.