Repotrectinib Exhibits Potent Antitumor Activity in Treatment-Naïve and Solvent-Front–Mutant ROS1-Rearranged Non–Small Cell Lung Cancer

This article is featured in Highlights of This Issue, p. 3071

Chromosomal rearrangements in ros proto-oncogene1 (ROS1) gene defines a distinct molecular subset in non–small cell lung cancer (NSCLC; refs. 1–3). The development of ROS1 targeted agents and the approval of crizotinib as a standard first-line therapy has transformed the course of disease for ROS1-driven NSCLC. Although ROS1 tyrosine kinase inhibitors (TKI) markedly improve clinical outcomes, patients inevitably relapse within a few years, and subsequent therapy overcoming acquired resistance remains limited (3, 4). Several major resistance mechanisms to ROS1-TKIs have been identified, including secondary mutations within the ROS1 kinase domain and activation of bypass signaling pathways (1, 5). Approximately 50% to 60% crizotinib-resistant mutations are found within the ROS1 kinase, of which ROS1-G2032R (ROS1^G2032R) solvent-front mutation (SFM) is the most common and recalcitrant-resistant mutation. These reports highlight the importance of developing novel ROS1 inhibitors with potent activity against G2032R (1, 6).

The brain is considered a major pharmacologic sanctuary because of the blood–brain barrier (BBB; refs. 7, 8). Indeed, approximately 50% of patients treated with ROS1-TKIs experience disease progression due to central nervous system (CNS) metastases (7, 9). The high incidence of CNS progression and poor prognosis in ROS1-rearranged NSCLC indicates an unmet clinical need for novel ROS1-TKIs with improved efficacy against brain lesions.

Repotrectinib (TPX-0005) is a novel next-generation ROS1/TRK/ALK-TKI specifically designed to overcome refractory SFMs, which commonly emerge in patients with ROS1/NTRK/ALK-rearranged malignancies who have relapsed on currently available TKIs (10). Repotrectinib for TKI-refractory patients is currently under phase I/II clinical trial (NCT03093116).

Existing preclinical studies in ROS1-rearranged NSCLC use a limited number of commercial or genetically engineered cell lines. This fails to fully encompass the genetic complexities observed in patients, leading to discrepancies between preclinical and clinical drug responses. Here, we evaluated the antitumor activity of repotrectinib in ROS1⁺ NSCLC patient-derived cell lines (PDC) and patient-derived xenograft (PDX) models. Moreover, we examined the intracranial antitumor activity of repotrectinib in a brain-metastasis mouse model developed through intracranial PDC implantation.

All patient samples were obtained from patients with ROS1 rearrangements before and after treatment with ROS1-TKIs at Yonsei University Severance Hospital (Seoul, Republic of Korea). All patients provided written informed consent. The study protocol was approved by the Institutional Review Board of Severance Hospital (IRB no.: 4-2016-0788). Patients treated with repotrectinib are enrolled in clinical trial NCT-03093116. This study was conducted according to the principles set out in the World Medical Association Declaration of Helsinki and the United States Department of Health and Human Services Belmont Report.

PDCs were established from malignant effusions of patients with advanced lung adenocarcinoma (11). Briefly, patient pleural effusion samples were centrifuged at 500 × g for 10 minutes at 25°C and resuspended in PBS. Cells were separated with Ficoll-Paque PLUS solution, following manufacturer's protocol. The interface containing the mononuclear cells was washed twice in HBSS and plated on collage IV-coated plates in RPMI medium supplemented with 10% FBS. To determine whether PDCs maintained patient characteristics, sanger sequencing and whole-exome sequencing were performed. FACS staining of EpCAM confirmed PDCs with over 99% cancer purity.

The following ROS1-rearranged PDCs were used: YU1078 (CD74-ROS1) and YU1079 (CD74-ROS1 G2032R) (Table 1). Sanger-sequencing-identified fusion partners in PDCs were identical to those identified in corresponding patient biopsies.

ROS1-rearranged PDX model YHIM1047 was established as previously described (Table 1; ref. 12).To generate PDC-derived tumor xenograft models, cells (YU1078 and YU1079, 5 × 10⁶ in 100 μL) were implanted subcutaneously into the flanks of 6-week-old female nu/nu mice. Animals were randomly divided (N = 5 per group) when tumor volume reached 150 to 200 mm³. Each group received one of the following treatments: once-daily crizotinib (50 mg/kg, every day), ceritinib (25 mg/kg, every day), entrectinib (30 mg/kg, every day), lorlatinib (30 mg/kg, every day), cabozantinib (30 mg/kg, every day), and twice-daily repotrectinib (15 mg/kg, twice a day). Tumor volumes (0.532 × length × width) were measured with an electronic caliper. Percentage change in tumor volume was calculated as follows: (V_t − V₀)/V₀ × 100. Tumor growth inhibition (TGI) was calculated with two formulas according to Drilon and colleagues (10). All mice were euthanized via CO₂ inhalation at the end of the experiment. Animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) and Animal Research Committee at Yonsei University College of Medicine.

Repotrectinib (TP Therapeutics) was kindly provided by the manufacturer and crizotinib, ceritinib, entrectinib, lorlatinib, and cabozantinib were purchased from Selleckchem.

About 2,500 to 3,000 cells were seeded in 96-well plates in growth media and incubated overnight at 37°C before adding serially diluted crizotinib, ceritinib, lorlatinib, cabozantinib, repotrectinib, and appropriate controls. After treating the drugs, cells were incubated at 37°C for 72 hours before performing CellTiter Glo (Promega) following the manufacturer's protocol. Dose–response curves and IC₅₀ values were calculated using GraphPad Prism.

Colony-forming assays were performed by seeding the cells in six-well plates in growth media with appropriate drug-containing media replaced every 3 days. Plates were fixed in 4% PFA and stained in crystal violet for 1 hour after 14 days of drug treatment.

Immunoblots were performed to determine relative phosphorylation of kinases and total protein levels of interest. Protein samples were separated by SPS-PAGE and transferred to nitrocellulose membrane for immunoblotting using the corresponding antibodies. P-ROS1 (T2274), ROS1, p-ERK (T202/Y204), ERK, p-AKT (S473), AKT, p-STAT3 (T705), STAT3, and Ki67 were all purchased from Cell Signaling Technologies. β-Actin was purchased from Sigma.

Intracranial tumor models were generated through implanting YU1078-luc human NSCLC PDC into brains of female BALB/c nude mice. After anesthetization, mice were immobilized with a stereotactic apparatus for incision of the skull's right hemisphere. A 0.5 mm burr hole was drilled through the right frontal lobe to implant a guide screw. Next, 5 × 10⁵ cancer cells (diluted in 5 μL PBS) were stereotactically injected for 5 minutes into the right frontal lobe via the guided screw and the incision was closed with surgical glue. At 13-day post-implantation, when photon flux measured through IVIS reached 1 × 10⁶ photons/second, mice were treated with a daily oral dose of repotrectinib (15 mg/kg, twice a day), entrectinib (30 mg/kg, every day), lorlatinib (30 mg/kg, every day), and vehicle. Brain metastatic-tumor growth was measured using an IVIS Spectrum Xenogen (Caliper Life Sciences).

IHC was performed on 4-μm-thick formalin-fixed, paraffin-embedded (FFPE) tissue sections. Slides were baked, deparaffinized in xylene, passed through graded alcohols, and then antigen retrieved with 1 mmol/L EDTA, pH 8.0 at 125°C for 30 seconds. Slides were pretreated with Peroxidase Block (Dako USA) for 5 minutes, and then washed in 50 mmol/L Tris-Cl, pH 7.4. Slides were blocked using normal goat serum (Dako USA), and subsequently incubated with anti-phosphSTAT3 mAb (Tyr705, D3A2, 1:100; CST)/anti-Ki67 antibody (D2H10; 1:100) for 1 hour. Slides were then washed in 50 mmol/L Tris-Cl, pH7.4 and treated with Signalstain boost IHC detection reagent [horseradish peroxidase (HRP), rabbit, #8114; CST] for 30 minutes. After further washing, immunoperoxidase staining was developed using a 3,3′diaminobenzidine (DAB) chromogen (Dako) for 5 minutes. Slides were counterstained with hematoxylin, dehydrated in graded alcohol and xylene, mounted, and coverslipped.

Comprehensive genomic profiling was performed in repeat biopsies in 2 patients who progressed after repotrectinib. FFPE slides were subjected to Illumina HiSeq based targeted sequencing capturing 171 cancer-related genes. Resultant reads were mapped to the human genome reference (hg19) using Burrows-Wheeler Alignment (BWA) followed by analysis using Genome Analysis ToolKit. Somatic mutations were called using MuTect2 and annotated with Oncotator. FoundationOne CDx (Foundation Medicine) was used to confirm the identified genomic alterations. A whole-genome shotgun library was constructed from 50 to 1,000 ng of extracted DNA and hybrid capture-selected libraries were sequenced at >500× median coverage using the Illumina HiSeq platform. All classes of cancer-related genomic alterations (e.g., base substitutions, indels, rearrangements, and copy-number alterations) were assayed.

Statistical analyses were performed in SPSS version 21 and GraphPad Prism 5. All data are expressed as means ± SD or ±SE for three or more independent replications. Between-group differences were evaluated using Student t test, Mann–Whitney test, Kruskal–Wallis with Dunn post hoc test, or ANOVA with Tukey post hoc test, as appropriate. Significance was set at P < 0.05.

In this study, we established three patient-derived preclinical models including one treatment-naive and two ROS1-TKI–resistant models from patients with ROS1⁺ NSCLC (Table 1). We examined the activities of clinically available ROS1-TKIs including crizotinib, ceritinib, entrectinib, lorlatinib, and repotrectinib in treatment-naive PDC YU1078 cells generated from a patient with ROS1-TKI–naïve NSCLC with CD74–ROS1 (35 years/female). Entrectinib (IC₅₀, 0.052 μmol/L), lorlatinib (IC₅₀, 0.009 μmol/L), and repotrectinib (IC₅₀, 0.021 μmol/L) exhibited potent growth inhibition in YU1078 cells compared with crizotinib (IC₅₀, 0.156 μmol/L) and ceritinib (IC₅₀, 0.412 μmol/L; Fig. 1A). Colony formation was significantly reduced with lorlatinib and repotrectinib treatment compared with other TKIs (Fig. 1B) and was accompanied by markedly suppressed ROS1 and ERK phosphorylation levels (Fig. 1C).

Consistent with the in vitro findings, lorlatinib and repotrectinib displayed pronounced tumor regression in subcutaneous YU1078 xenograft model with 189.2% TGI. Crizotinib and entrectinib resulted in 150.7% and 124.9% TGI, respectively, whereas ceritinib displayed 64.7% TGI (Fig. 1D). Although tumor regression was comparable between lorlatinib and repotrectinib, repotrectinib produced the strongest antitumor effect, which was most apparent after drug withdrawal. After drug withdrawal at 21 days, tumors in three of four repotrectinib-treated mice were persistently suppressed for over 80 days. One mouse exhibited slow regrowth after 60 days post-drug withdrawal. However, tumors in two of four lorlatinib-treated mice abruptly regrew 10 days after drug withdrawal (Fig. 1E). To determine whether repotrectinib is an effective subsequent treatment option after failure to prior ROS1-TKI, we administrated repotrectinib to tumor progressed on ceritinib. Subsequent repotrectinib treatment prominently suppressed tumor growth where of note, upfront repotrectinib completely induced tumor regression (Fig. 1F). Collectively, upfront and subsequent repotrectinib treatment demonstrated potent TGI based on in vitro and in vivo ROS1⁺ preclinical models.

Lorlatinib and entrectinib have been reported to penetrate the BBB and regress brain metastasis (13–15). We investigated the ability of repotrectinib to penetrate BBB using an intracranial tumor model of luciferase-transduced YU1078 cells (YU1078-luc). Compared with vehicle and entrectinib, repotrectinib treatment significantly reduced tumor load (as measured using photon flux) in the brain without body weight loss (Fig. 2A and B). In addition, all repotrectinib-treated mice survived for up to 110 days, whereas median survival of vehicle- and entrectinib-treated mice was 49 and 57 days, respectively (Fig. 2C). Lorlatinib showed comparable reduction of tumor load to that of repotrectinib (Fig. 2A and B). IHC further revealed that repotrectinib treatment completely suppressed viable brain tumors, with pronounced reduction in Ki67-positive and TUNEL-positive cells compared with vehicle and entrectinib treatment (Fig. 2D).

Clinical potency of repotrectinib was demonstrated in a TKI-naïve ROS1⁺ patient treated with repotrectinib. A 69-year-old female CD74-ROS1–rearranged patient was presented with liver and multiple brain metastases. After receiving 40 mg repotrectinib once daily on the phase I repotrectinib trial, the patient achieved confirmed PR (−80%) based on RECIST1.1, currently ongoing at 20 months (Fig. 3A). The phase I repotrectinib trial also included a ROS1⁺ lung cancer patient who had recurred under ceritinib. Pre-repotrectinib brain imaging revealed multiple asymptomatic brain metastases. This patient tolerated repotrectinib treatment (40–240 mg, every day) and achieved confirmed RECIST1.1-based PR (−40%) that is ongoing at 16 months (Fig. 3B). As a clinical proof of concept, the potent intracranial activity of repotrectinib has been observed in a TKI-naïve patient and a patient who had progressed on prior TKI enrolled in an ongoing phase I/II clinical trial. Multiple metastatic brain lesions disappeared after 2-month treatment and were maintained over 15 months in each patient.

We identified the ROS1^G2032R mutation in YU1079, which was serially established in the same patient as YU1078 but after progressing on crizotinib treatment. On the basis of recent studies examining lorlatinib and cabozantinib in Ba/F3 cells engineered with CD74–ROS1^G2032R (10, 13), we investigated these TKIs in our YU1079 with CD74-ROS1^G2032R. Cabozantinib (IC₅₀, 0.111 μmol/L) and repotrectinib (IC₅₀, 0.097 μmol/L) effectively inhibited YU1079 ROS1^G2032R growth, concomitant with marked reduction in ROS1 and downstream signal phosphorylation (Fig. 4A–C). However, lorlatinib (IC₅₀, 1.509 μmol/L) was relatively ineffective in inhibiting YU1079 ROS1^G2032R growth, possibly due to a weak inhibitory effect on phospho-ROS1, ERK, and STAT3, along with persistent phospho-AKT (Fig. 4C).

The in vivo efficacy of these TKIs was further evaluated in YU1079 ROS1^G2032R PDX model. Consistent with the in vitro assays, cabozantinib and repotrectinib significantly led to 102.1% and 90.9% TGI, respectively, compared with 22.7% TGI for lorlatinib (Fig. 4D). Considering severe treatment-related toxicities observed in patients receiving cabozantinib (1, 16, 17), we focused on comparing repotrectinib with clinically available entrectinib in the PDX model YHIM1047 ROS1^G2032R, established from a CD74-ROS1 patient with NSCLC who progressed on entrectinib. As expected from the patient's clinical response, entrectinib failed to inhibit tumor growth whereas repotrectinib treatment induced 142.8% TGI (Fig. 4E). IHC revealed marked reduction of phosphorylated STAT3 and Ki67 in repotrectinib-treated YHIM1047 ROS1^G2032R tumors compared with entrectinib and vehicle (Fig. 4F).

The clinical activity of repotrectinib against ROS1 SFM was seen in a 49-year-old female ROS1-rearranged patient who progressed after 44 months of crizotinib treatment with an identified CD74-ROS1 G2032R mutation. The patient received 160 mg twice a day repotrectinib which induced a confirmed PR (−80%) based on RECIST1.1 currently ongoing at 6 months (Fig. 4G). Taken together, these results suggest that repotrectinib is a promising therapeutic strategy for ROS1-rearranged NSCLC harboring ROS1^G2032R.

Finally, we investigated genetic mutation status in biopsies of 2 patients who progressed on repotrectinib in clinical trial using targeted sequencing. Patient A, a 46-year-old male with a 20 pack-year smoking history, was diagnosed with adenocarcinoma. The patient underwent a first biopsy before treatment, followed by a second biopsy after progression on repotrectinib treatment for about 10 months (Table 1). A variant of the ROS1 translocation, CD74-ROS1 fusion, was detected by targeted sequencing. Mutations in the ROS1 kinase domain were not detected. Post-repotrectinib tumor biopsy had mutations in CCND3 (E253Q), TP53 (H178Q&H179Y), and SMAD4 (E538*), which were absent in the corresponding baseline biopsy (Supplementary Table S1). These mutations occurred at different variant allele frequencies (VAF), 7%, 53%, and 47%, respectively. Patient B was a 38-year-old female with adenocarcinoma with a 10 pack-year smoking history. The patient was biopsied following progression on repotrectinib after 10 months of treatment (Table 1). ROS1 fusion partner was identified as SLC34A2. We identified CEBPA (196_197insHP, VAF 41%), RB1 (H555R, VAF 45%), ERBB2 (R143Q, VAF 51%), and TP53 (E171G, VAF 9%) mutations in the post-repotrectinib tumor biopsy. We were unable to demonstrate the absence of these mutations in the baseline biopsy due to lack of available tumor tissue. Altogether, these findings suggest the possibility of non-ROS1 dominant mechanism as acquired resistance to repotrectinib. Further studies are necessary to investigate the functional role of these mutations to elucidate acquired resistance mechanism to repotrectinib.

In this study, we demonstrated that repotrectinib is a potent next-generation TKI against ROS1 and ROS1^G2032R mutation with remarkable activity in the CNS. Our findings highlight the potential of repotrectinib as a first-line therapy in ROS1⁺ NSCLC as well as those with G2032R.

Crizotinib is the FDA-approved standard care in ROS1-rearranged NSCLC with entrectinib emerging as a promising ROS1-TKI recently approved by the FDA. Ceritinib and lorlatinib have both previously showed clinical activity in treatment-naïve ROS1⁺ NSCLC (8, 18). In particular, lorlatinib is the only recommended TKI after progression to crizotinib under National Comprehensive Cancer Network (NCCN) guidelines (19). To our knowledge, repotrectinib exhibited the most potent in vitro and in vivo activity compared with other ROS1-TKIs in treatment-naïve YU1078 model (Fig. 1A–D). Despite comparable activity between repotrectinib and lorlatinib, repotrectinib was able to persistently and strongly suppress tumor recurrence in YU1078-derived xenograft model for more than 80 days even after drug withdrawal (Fig. 1E). Consistent with previous reports (10), our findings suggest that repotrectinib is a potent ROS1-TKI in the first-line setting.

CNS metastases remain a major cause of morbidity and mortality in ROS1-driven NSCLC. Recent reports highlighted that 36% of ROS1⁺ patients experience CNS metastasis at baseline or 50% after crizotinib treatment similar to ALK+ NSCLC (9). The low brain-to-plasma ratio of crizotinib suggests that frequent CNS metastasis during crizotinib treatment may be due to poor BBB penetration and brain exposure (6). For the first time, we demonstrated superior CNS penetration of repotrectinib compared with entrectinib in YU1078-luc intracranial tumor model with significant reduction of metastatic brain lesion and extensive mouse survival over 100 days (Fig. 2). This was clinically supported by two CD74-ROS1 NSCLC patient response data where repotrectinib rapidly regressed tumor and the metastatic brain lesion (Fig. 3). Therefore, upfront repotrectinib therapy could effectively reduce dismal prognosis due to CNS progression, which might overcome current limitations of first-line crizotinib.

Acquired resistance to targeted therapy is a major obstacle for achieving durable response in oncogene-driven NSCLC. Approximately 50% to 60% of acquired resistance to crizotinib is caused by on-target mutations, of which G2032R is the most recalcitrant and commonly observed mutation in ROS1⁺ patients. Ceritinib, entrectinib, and lorlatinib did not show robust activity against G2032R, as reported previously (10, 13). Cabozantinib exhibited comparable activity to repotrectinib but toxicity observed in clinical trials should be managed further (16). Recently, a novel ROS1/NTRK inhibitor DS-6051b was reported to have activity against G2032R in preclinical models but requires validation in the clinical setting (20). Consistent with previous reports demonstrating the activity of repotrectinib in BaF3 CD74–ROS1^G2032R cell line and xenograft models (10), we demonstrated prominent antitumor activity of repotrectinib in crizotinib-refractory YU1079 ROS1^G2032R and entrectinib-refractory YHIM1047 ROS1^G2032R. This was supported by pronounced tumor regression in patient with CD74-ROS1 G2032R NSCLC (Fig. 4). Recently, preliminary results of TRIDENT1 trial reported meaningful clinical activity and favorable safety profiles of repotrectinib (21, 22). The confirmed objective response rate (ORR) was 91% in 11 patients with TKI-naïve ROS1⁺ NSCLC and 39% in 18 TKI-pretreated patients after repotrectinib. ROS1^G2032R patients demonstrated ORR 43% (n = 3/7). Most adverse events were manageable and grade 1 to 2. Collectively, repotrectinib is an effective therapy against ROS1⁺ NSCLC patients who are TKI-naïve or TKI-refractory harboring G2032R.

Upfront next-generation TKI therapy prolonged survival outcome in patients with EGFR-mut and ALK-fusion NSCLC and have been approved as standard care in these patients (23, 24). Moreover, a recent study presented that sequential ALK-TKI treatment fosters development of diverse compound mutations, highly refractory to all available ALK-inhibitors (25). These data implicate that upfront use of repotrectinib could be able to prevent or delay the emergence of G2032R and subsequent compound mutations, potentially improving clinical outcomes. In line with the idea, G2032R or ROS1 compound mutation was not observed in 2 patients who progressed on first-line repotrectinib in our study (Supplementary Table S1). Therefore, optimal first-line therapy could be derived from repotrectinib-based combinations to prevent both ROS1-dependent and -independent resistance. Potent overall and intracranial activity, G2032R specificity, and favorable safety profiles suggest repotrectinib as a promising first-line TKI.

In conclusion, we demonstrated that repotrectinib is a novel, highly potent, BBB-penetrant next-generation ROS1/TRK/ALK-TKI. Our findings provide evidence for repotrectinib as an effective first-line treatment in ROS1⁺ NSCLC and after progression to prior ROS1-TKI.

A. Drilon is an employee/paid consultant for Ignyta/Genentech/Roche, LOXO/Bayer/Lilly, Takeda/Ariad/Millenium, TP Therapeutics, AstraZeneca, Pfizer, Blueprint, Helsinn, Beigene, BergenBio, Hengrui, Exelixis, Tyra, Verastem/14ner, MOREHealth, and Abbvie, and reports receiving other remuneration from GlaxoSmithKline, Teva, Taiho, PharmaMar, Foundation Medicine, Wolters Kluwer, and Merck/Puma/Merus/Boehringer Ingelheim. B. Chul Cho is an employee/paid consultant for Novartis, AstraZeneca, Boehringer-Ingelheim, Roche, Bristol-Myers Squibb, Ono, Yuhan, Pfizer, Eli Lilly, Janssen, Takeda, and MSD, and reports receiving commercial research grants from Novartis, Bayer, AstraZeneca, MOGAM Institute, Dong-A ST, Champions Oncology, Janssen, Yuhan, Ono, Dizal Pharma, and MSD, and reports receiving speakers bureau honoraria from Novartis, Bayer, AstraZeneca, MOGAM Institute, Dong-A ST, Champions Oncology, Janssen, Yuhan, Ono, Dizal Pharma, and MSD, and holds ownership interest (including patents) in TheraCanVac Inc., Gencurix Inc., and Bridgebio Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M.R. Yun, D.H. Kim, Y.W. Lee, J.J. Cui, H.R. Kim

Development of methodology: D.H. Kim, S.-G. Kang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.H. Kim, S.-Y. Kim, H.-S. Joo, H.M. Choi, A. Drilon, H.S. Shim, J.J. Cui, H.R. Kim

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.H. Kim, S.G. Heo, A.J. Schoenfeld, A. Drilon, H.S. Shim, J.J. Cui, H.R. Kim

Writing, review, and/or revision of the manuscript: D.H. Kim, A.J. Schoenfeld, A. Drilon, H.S. Shim, M.H. Hong, J.J. Cui, H.R. Kim

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C.W. Park, H.N. Kang, S.S. Lee, M.H. Hong

Study supervision: M.R. Yun, B.C. Cho

We thank all the patients who participated in this study. This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2016R1A2B3016282, to B.C. Cho) and NRF grant funded by the Korean government (NRF-2018R1D1A1B07050233, to M.R. Yun).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.