Trastuzumab Emtansine (T-DM1) in Patients with Previously Treated HER2-Overexpressing Metastatic Non–Small Cell Lung Cancer: Efficacy, Safety, and Biomarkers

The development of targeted therapy for non–small cell lung cancer (NSCLC) with specific molecular alterations such as anaplastic lymphoma kinase (ALK) rearrangements or EGFR mutations represents a tremendous advance (1, 2). Current research is focused on the identification and development of targeted therapy for additional molecular subtypes. HER2 is overexpressed on the surface of multiple tumor cell types, including NSCLC (3–5). Survival data meta-analyses show HER2 overexpression is associated with poor prognosis in lung cancer (3, 6). There are currently no standard therapies targeting the HER2 pathway in NSCLC, whereas HER2-targeted therapies are standard for breast and gastric cancer. In breast and gastric cancer, HER2 overexpression generally occurs in the context of gene amplification, whereas discordance between IHC overexpression and gene amplification is observed in NSCLC (7−10). Instead, increased HER2 expression in NSCLC may result from upregulated transcriptional/posttranscriptional mechanisms (7, 9).

The reported prevalence of “HER2-positive” NSCLC varies because previous studies have used different definitions and testing methods (ISH and IHC), with HER2 positivity defined by IHC 2+/3+ staining reported in 13% to 20%, IHC 3+ staining in 2% to 6%, and HER2 gene amplification by ISH in 2% to 4% (8). HER2 gene mutation, another HER2 alteration in NSCLC, has a frequency of 1% to 4% (8). HER2 gene mutations and amplifications show limited co-occurrence in NSCLC (8, 10). The absence of correlation between HER2 overexpression, amplification, or mutation suggests 3 biologically distinct NSCLC subtypes, leaving the question of which subtypes will be most effectively treated with HER2-targeted therapy.

Despite observation of these 3 alterations in NSCLC, the role of these abnormalities as therapeutic biomarkers remains undefined. Few data sets report on the use of HER2-targeted therapies in patients with HER2-positive (11, 12) or -mutated (13–16) NSCLC. These studies were small, and some included patients treated concurrently with chemotherapy and a HER2-targeting agent. Of note, most trials do not assess presence of all 3 types of HER2 alterations, and tend to use significantly divergent definitions, which impairs the interpretation of the predictive value of specific HER2 alterations in NSCLC.

Trastuzumab emtansine (T-DM1) is an antibody–drug conjugate composed of trastuzumab joined via a stable linker to DM1, a cytotoxic microtubule-inhibitory agent (17). T-DM1 targets the delivery of DM1 to HER2-positive cancer cells, maximizing the therapeutic index of DM1 and minimizing off-target effects. T-DM1 has been approved for the treatment of previously treated HER2-positive metastatic breast cancer. In preclinical studies, T-DM1 demonstrated potent in vitro growth inhibition of HER2 IHC 3+ (Calu-3, H2170) and IHC 1+ (H1781) NSCLC cells (18). T-DM1 also showed robust antitumor activity in Calu-3 HER2 IHC 3+ and H1781 HER2 IHC 1+ mouse xenograft tumor models.

Altogether, this evidence provided rationale to conduct this first phase II study of T-DM1 in patients with HER2-overexpressing advanced NSCLC prior to considering a more definitive trial and to investigate potential predictive biomarkers. Given the T-DM1 mechanism of action, we selected patients with moderate-to-high HER2 overexpression, defined as IHC 2+ or 3+ only. Specifically, HER2 overexpression can occur due to increased HER2 transcription, independent of HER2 gene amplification or mutation (7−10). To date, there are no clinical data suggesting T-DM1, in contrast to trastuzumab or a HER2 tyrosine kinase inhibitor (TKI), would have a stronger effect in HER2-mutated or HER2-amplified NSCLC. Although amplification and mutations are considered potential driver oncogenes, we hypothesize that targeting HER2 overexpression will address a potentially larger patient population that may be more relevant taking into account the mechanism of action of T-DM1, targeting the HER2 extracellular component.

In this multicenter, single-arm, clinical trial (trial registration NCT02289833), eligible patients were aged ≥18 years with HER2-positive (IHC 2+ or 3+) locally advanced or metastatic NSCLC previously treated with ≥1 prior platinum-based chemotherapeutic regimen. Patients with EGFR-mutated or ALK gene–rearranged NSCLC were eligible if they had also experienced disease progression following treatment with prior targeted therapy or if they were intolerant to such treatment.

Archived tumor specimens from previously collected tissue were centrally and prospectively tested for HER2 status (Ventana Pathway HER2 (4B5) IHC assay; Ventana Medical Systems, Inc.). HER2 overexpression was evaluated by IHC, the gold standard for HER2 assessment in breast and gastric cancer. If archival tissue was unavailable for HER2 testing, patients could have a newly collected biopsy specimen tested. Based on results from central testing, patients with a HER2 status of IHC 2+ (defined as weak-to-moderate complete, basolateral, or lateral membranous reactivity in ≥10% of tumor cells) or IHC 3+ (defined as strong complete, basolateral, or lateral membranous reactivity in ≥10% of tumor cells) were eligible.

A retrospective exploratory biomarker analysis was conducted if sufficient tissue was available following IHC testing. In this analysis, HER2 gene amplification (HER2/CEP17 gene ratio ≥2) was also assessed by ISH (similar to breast and gastric cancers). HER2 mRNA expression levels were measured by quantitative reverse transcriptase PCR (Roche Molecular Diagnostics) and evaluated in subgroups defined using the median mRNA expression distribution values (>median vs. ≤median). Pending tissue availability, HER2 mutation status and amplification was also assessed using next-generation sequencing (NGS; Foundation Medicine Inc.), where amplification was defined by copy number ≥5 in a diploid model.

Eligible patients also had an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, measurable disease (per RECIST v1.1), adequate organ function, and left ventricular ejection fraction ≥50% (echocardiogram or multiple-gated acquisition scan). Patients with untreated or symptomatic brain metastases (baseline brain imaging required but repeat imaging was not required), or who had received anticancer or investigational therapy within 21 days of the start of study treatment were excluded.

Patients were able to enroll in the screening portion of the trial and underwent IHC testing prior to disease progression on current therapy to determine if they were potentially eligible for the trial. Eligible patients were subdivided according to HER2 IHC status (2+ or 3+). Overenrollment was allowed to the IHC 2+ cohort to accommodate patients who had initiated the screening process and met the eligibility criteria. All patients received T-DM1 (3.6 mg/kg intravenously every 3 weeks) until investigator-assessed disease progression, unmanageable toxicity, or study termination. Survival status was assessed every 3 months following discontinuation due to disease progression. Tumor assessments by radiographic evaluation were performed every 6 weeks.

The trial protocol was approved by the Institutional Review Boards at each participating center. The study was conducted in accordance with the Declaration of Helsinki, Good Clinical Practice guidelines, and applicable local laws. All patients provided written informed consent.

The primary efficacy endpoint was confirmed investigator-assessed objective response rate (ORR) using RECIST v1.1, defined as a complete response (CR) or partial response (PR) determined on 2 consecutive assessments ≥4 weeks apart (19).

Secondary endpoints were investigator-assessed progression-free survival (PFS), duration of response (DoR; time from initial documented response to documented disease progression or death from any cause), clinical benefit rate [CBR; proportion of patients with CR or PR or stable disease (SD) determined by the investigator at 6 months), overall survival (OS), and safety. Patients were monitored continuously for adverse events (AEs).

The primary efficacy objective was investigator-assessed ORR. Given the exploratory nature of this study, a sample size of 20 patients per cohort was selected to estimate 2-sided 95% Clopper–Pearson confidence intervals (CIs) for ORR in each cohort and overall. The data cutoff for the ORR analysis occurred when all enrolled patients were expected to have completed ≥4.5 months of follow-up (when 3 postbaseline tumor assessments were expected to be performed). Efficacy and safety were evaluated in patients who received ≥1 dose of T-DM1. DoR and Kaplan–Meier estimates of PFS and OS are reported for all evaluable patients and separately for the 2 IHC cohorts. For PFS and DoR, data for patients without disease progression or death were censored at the time of the last tumor assessment. For OS, data from patients without a death event were censored on the date last known to be alive. Exploratory biomarker data were analyzed using descriptive statistics.

Safety was assessed through summaries of AEs, deaths, and changes in laboratory results. All AEs occurring on or after first study treatment were summarized by mapped term, appropriate thesaurus levels, and National Cancer Institute Common Terminology Criteria for Adverse Events v4.0 toxicity grade. Serious AEs were listed separately and summarized.

Of 370 screened patients, 133 had IHC 2+ or 3+ tumors (Supplementary Fig. S1). Of these, 49 patients from 5 centers in the United States and 13 European centers were eligible and enrolled between December 15, 2014, and June 21, 2016; the remaining 84 patients did not meet eligibility criteria or get elect for alternative treatment. Of the 49 enrolled patients, HER2 status was based on archival samples for 41 patients and freshly obtained samples (collected within 8 weeks before start of study treatment) for 6 patients; the sample collection date was unknown for 2 patients. Median age was 61 years (range, 36–80), and the majority of patients were male, previous smokers, and had adenocarcinoma (Table 1). All but 1 patient previously received platinum-based chemotherapy (98%; 48/49); information on prior and follow-up treatments is shown in Supplementary Table S1. The IHC 2+ cohort was overenrolled and included 29 patients, and the IHC 3+ cohort included 20 patients.

At the cutoff date (May 29, 2017), 15 of 49 (31%) patients remained on study for survival follow-up, and 1 in the IHC 3+ cohort remained on treatment; 3 were lost to follow-up, 30 died (29 due to disease progression; 1 died of an unknown cause as reported in death registry records), and 1 discontinued the study due to clinical progression.

Median follow-up was 23.1 months (range, 0.9–26.7) and 18.4 months (1.0–25.1) in the IHC 2+ and 3+ cohorts, respectively. In the IHC 2+ cohort, there were no treatment responses; 8 of 29 (28%) patients had SD, 16of 29 (55%) had progressive disease (PD), and response was not evaluable for 5 of 29 (17%; Fig. 1A and B). In the IHC 3+ cohort, ORR was 20% (95% CI, 5.7%–43.7%); 4 of 20 (20%) patients had a PR, 4 of 20 (20%) had SD, 11 of 20 (55%) had PD, and 1 (5%) died prior to first scheduled tumor assessment (response missing; Fig. 1C). CBR was 7% (2/29, 95% CI, 1%–23%) and 30% (6/20; 95% CI, 12%–54%) in the IHC 2+ and 3+ cohorts, respectively. DoR for the 4 IHC 3+ responders was 2.9, 7.3, 8.3, and 10.8 months (censored for 1 patient still on treatment; Fig. 1D).

In the IHC 2+ and 3+ cohorts, median PFS was 2.6 months (95% CI, 1.4–2.8) and 2.7 months (95% CI, 1.4–8.3), respectively; median OS was 12.2 months (95% CI, 3.8–23.3) and 15.3 months (95% CI, 4.1–NE), respectively (Fig. 2).

Median duration of T-DM1 treatment was 3.6 months (range, 0–24.8 months; Supplementary Fig. S2). AEs, regardless of causality or attribution, are reported (Table 2). Forty-five patients (92%) reported an AE (any grade). Ten patients reported grade 3 AEs (regardless of relationship to treatment), and fatigue (n = 2) was the only grade 3 AE reported in more than 1 patient. One grade 4 seizure was reported in a patient with prior history of brain metastases receiving concurrent treatment for seizures. There were no deaths due to AEs. Two patients withdrew from study treatment due to AEs (grade 2 infusion-related reaction and grade 3 influenza). Of AEs of particular interest in T-DM1–treated patients, 1 event each of grade 3 thrombocytopenia and infusion-related reaction/hypersensitivity occurred.

All responding patients were in the IHC 3+ cohort (Table 3). A higher proportion of IHC 3+ patients had HER2-amplified tumors, higher HER2 gene copy number, and HER2 mRNA expression >median compared with the IHC 2+ cohort.

Biomarker characteristics of the responders and those with SD >6 months are shown in Table 4. Of 4 responding patients, 3 had HER2-amplified tumors (ISH, HER2/CEP17 gene ratio ≥2). Three responders had central NGS data; 2 of these patients had HER2 amplification (NGS, copy number ≥5; Tables 3 and 4). The other responder had locally-determined NGS results (same platform used for central assessment) showing an equivocal test result for HER2 amplification. Two responders had HER2 mutations including 1 HER2 gene rearrangement (unknown functional status).

A heatmap of the top 20 genetic alterations found across all patients was prepared for those with central NGS results (n = 24, Supplementary Fig. S3). Of these 24 patients, HER2 mutations were detected in 4 patients (no responders), and HER2 amplification was detected in 5 patients (2 responders).

In NSCLC, activation of HER2 occurs through various mechanisms including protein overexpression, gene amplification, or mutation, and is considered an oncogenic driver. In our study of T-DM1 in patients with HER2-overexpressing advanced NSCLC, objective responses were observed in patients with IHC 3+ tumors. No responses were observed in the IHC 2+ cohort, regardless of HER2 amplification status. However, 4 additional patients achieved SD >6 months, resulting in CBRs of 7% and 30% in the IHC 2+ and 3+ cohorts, respectively.

Several small studies have also investigated the use of HER2-targeted therapies in lung cancer (Supplementary Table S2). A phase II trial investigated treatment with T-DM1 in NSCLC characterized by overexpressed, amplified, or mutated HER2 and was stopped early due to limited efficacy (20). Among 15 patients with HER2-positive NSCLC treated with T-DM1 (5 IHC 3+, 3 IHC 2+/FISH-positive, 7 with HER2 mutations), there was only 1 PR in a patient with HER2 mutation (ORR, 6.7%; 90% CI, 0.2%–32.0%). The small number of patients enrolled in the trial, and in the various molecular subsets limits the interpretation of these results. Analyses from the NCT02675829 basket trial further support the potential role of TDM-1 in advanced NSCLC characterized by HER2 oncogenic alterations. This trial investigates efficacy of T-DM1 in various HER2-amplified solid tumors and HER2-mutant lung cancers (21, 22). In the cohort of patients with advanced HER2-mutant lung adenocarcinoma, ORR was 44% (8/18; 95% CI, 22%–69%), median PFS was 5 months (95% CI, 3–9 months), and median DoR was 4 months (95% CI, 2–9 months). In the cohort of patients with HER2-amplified advanced lung cancer, ORR was 50% (3/6; ref. 22).

Other limited datasets have reported activity of HER2-targeted therapy in NSCLC. In a retrospective analysis of patients with a HER2 mutation who received 22 individual anti-HER2 treatments, there were 11 PRs, leading to an ORR of 50% (11/22; ref. 13). The MyPathway basket trial (NCT02091141) is evaluating dual inhibition of the HER2 pathway with pertuzumab plus trastuzumab in patients with HER2-overexpressing, -amplified, or -mutated tumors, including lung cancer (23). Based on interim data for NSCLC, ORR was 13% (2/16; 95% CI, 2%–38%) in the HER2-amplified cohort and 21% (3/14; 95% CI, 5–51%) in the HER2-mutated cohort.

In a phase II study, patients with HER2-mutant or HER2-amplified lung cancer received treatment with dacomitinib, a pan-HER TKI, and the ORR was 12% (3/26; 95% CI, 2%–30%) for HER2-mutant disease and 0% (0/4; 95% CI, 0%–60%) for HER2-amplified disease (14). Neratinib, another pan-HER TKI, has been investigated in HER2-mutated lung cancer. In a cohort of 26 lung cancer patients with HER2 or HER3 mutations from the phase II SUMMIT basket trial, only 1 patient with a HER2 exon 20 mutation (L755S) achieved a response (ORR 3.8% at 8 weeks; ref. 24). Median PFS was 5.5 months. A phase I study of neratinib plus temsirolimus, an mTOR inhibitor, observed responses in 2 of 7 patients with HER2-mutated NSCLC; however, this was at the cost of significant treatment-related toxicity (25). Pyrotinib, a TKI targeting HER1 and HER2 receptors, has also been studied, and preliminary results from a phase II study found an ORR of 55% (6/11) among patients with previously treated HER2-mutant advanced NSCLC (26). Studies of afatinib, an ErbB family blocker, have showed varied results. In a phase II study of patients with heavily pretreated lung adenocarcinoma, a cohort of 7 patients with HER2 mutation received afatinib monotherapy, and none experienced an objective response (27). Another phase II study of afatinib demonstrated an ORR of 7.7% (1/13) among pretreated patients with NSCLC harboring HER2 exon 20 mutations (28). Finally, a retrospective review of patients with metastatic HER2-mutant lung cancer treated with afatinib from 7 institutions found an ORR of 11% (3/27; ref. 29). Of note, a study investigating response to immune checkpoint blockade in HER2-mutated advanced lung cancer found an ORR of 12% (3/26) and a median PFS of 1.9 months (95% CI, 1.5-4.0; ref. 30).

Whereas data suggest a potential role for HER2-targeted therapy in NSCLC, further investigation of HER2 as a biological target in NSCLC is needed. The relationship between the distinct, and probably independent, features of HER2 activation, related biomarkers, and response to treatment needs further elucidation. For example, in our study, all responders had IHC 3+ tumors, 3 had HER2 amplification (by ISH and NGS), and 2 had HER2 gene mutations. By contrast, none of the 8 responders to T-DM1 in the HER2-mutant cohort of the NCT02675829 basket trial had HER2 levels beyond IHC 2+, but 1 patient had HER2 amplification (21). Activity of T-DM1 in the context of HER2 mutations remains to be biologically elucidated (21), and, given co-occurring HER2 amplification and mutation in the majority of responders and the small number of responders in our trial, we do not have sufficient sample size to assess for a potential interaction between the molecular alterations. We used IHC to identify patients with intermediate-to-high HER2 protein levels, as IHC is the standard assay used in indications where HER2-targeted therapy is established. HER2 mRNA levels were assessed as part of the exploratory biomarker evaluation; however, this did not identify a cutoff to identify patients who responded to T-DM1. Of the 4 responders, only 1 patient had HER2 mRNA levels above the median HER2 mRNA level (Table 3). HER2 IHC alone is insufficient as a predictive biomarker, and identification of additional biomarkers is required.

Our study has a few limitations. We did not test a predetermined response rate because this was an exploratory study to obtain preliminary efficacy data and investigate potential biomarkers for further rationale for development in this rare subset of NSCLC. Higher patient numbers might be needed to provide clarity on potential relationships between different HER2 biomarkers and response. In addition, only 49 of 133 prescreened patients with IHC 2+ or 3+ tumors were enrolled and started treatment with T-DM. Prescreening was allowed while patients were still on other therapy, and, at the time of disease progression, patients may have no longer met all eligibility criteria or may have elected for alternative treatment. As some sites screened patients for HER2 status prior to the central laboratory assessment, the prevalence of IHC 2+ and IHC 3+ NSCLC is likely higher than an actual population-based prevalence of HER2 IHC overexpression. In our study, most tissue came from the archival specimen (41/49; 84%). The absence of mandatory rebiopsy at study entry is a weakness—adopted for practical reasons—given that HER2 status at diagnosis and study entry may theoretically vary under selective pressure of treatments and tumor evolution, whereas the majority of patient tumor samples came from primary tumors. HER2 amplification has been hypothesized as a resistance mechanism with EGFR TKIs (31, 32), indicating potential changes in HER2 status following treatment with EGFR TKIs. Although data from NSCLC showing loss of HER2 overexpression are not available, this cannot be excluded. Sufficient tissue was available for evaluation of all mandatory biomarkers; however, our biomarker analyses were limited by sample availability. Where possible, we collected data on EGFR and ALK based on local testing, whereas broader NGS testing for other molecular alterations is not routinely performed and was available for selected patients (Supplementary Fig. S3). Finally, further dedicated studies would be strengthened by independent radiologic review.

In summary, our study indicates activity of T-DM1 in selected patients with IHC 3+ HER2-positive metastatic NSCLC. The safety profile of T-DM1 was similar to findings from prior T-DM1 clinical trials and showed that T-DM1 was also tolerable in NSCLC. Additional investigation into HER2 signaling pathway oncogenic modifications, including HER2 overexpression, amplification, or mutation, may help to refine a patient population more likely to benefit from treatment with T-DM1. Of importance, HER2 IHC—used as a single parameter—was an insufficient predictive biomarker for T-DM1 activity. Further trials are needed to refine the target population for T-DM1 as well as for other HER2-directed therapies in NSCLC.

S. Peters is a consultant/advisory board member for Roche, has received honoraria or consultation fees from Abbvie, Amgen, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Clovis, Eli Lilly, F. Hoffmann-La Roche, Illumina, Janssen, Merck Sharp and Dohme, Merck Serono, Novartis, Pfizer, Regeneron, Seattle Genetics and Takeda, given a talk in a company's organized public event for AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Eli Lilly, F. Hoffmann-La Roche, Merck Sharp and Dohme, Novartis, Pfizer, and received grants/research support as a (sub)investigator in trials sponsored by Amgen, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Clovis, F. Hoffmann-La Roche, Illumina, Merck Sharp and Dohme, Merck Serono, Novartis, and Pfizer. L. Bubendorf reports receiving commercial research grants from, holds ownership interest (including patents) in, and is a consultant/advisory board member for Roche. P. Bonomi is a consultant/advisory board member for AstraZeneca, Biodesix, Merck, Helsinn, Pfizer, and Roche Genentech. P. Garrido reports receiving speakers bureau honoraria from Roche, Pfizer, and MSD, and is a consultant/advisory board member for Roche, AstraZeneca, Abbvie, Pfizer, Bristol-Myers Squibb, MSD, and Lilly. A. Rittmeyer reports receiving speakers bureau honoraria from Roche Genentech, Bristol-Myers Squibb, and MSD, and is a consultant/advisory board member for Abbvie, AstraZeneca, Bristol-Myers Squibb, Eli Lilly, MSD, Pfizer, Boehringer Ingelheim, and Roche. L.H. Lam holds ownership interest (including patents) in Roche. M.W. Lu holds ownership interest (including patents) in Roche. T.E. Stinchcombe is a consultant/advisory board member for Roche Genentech. No potential conflicts of interest were disclosed by the other authors.

Conception and design: S. Peters, R. Stahel, L. Bubendorf, P. Bonomi, J. De Castro Carpeno, S. de Haas, L.H. Lam, T.E. Stinchcombe

Development of methodology: S. Peters, L. Bubendorf, J. De Castro Carpeno, S. de Haas, L.H. Lam, M.W. Lu, T.E. Stinchcombe

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Peters, R. Stahel, A. Villegas, D.M. Kowalski, C.S. Baik, D. Isla, J. De Castro Carpeno, P. Garrido, A. Rittmeyer, M. Tiseo, L.H. Lam, M.W. Lu, T.E. Stinchcombe

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Peters, R. Stahel, L. Bubendorf, P. Bonomi, D.M. Kowalski, C.S. Baik, J. De Castro Carpeno, A. Rittmeyer, C. Meyenberg, S. de Haas, L.H. Lam, M.W. Lu, T.E. Stinchcombe

Writing, review, and/or revision of the manuscript: S. Peters, R. Stahel, L. Bubendorf, P. Bonomi, A. Villegas, D.M. Kowalski, C.S. Baik, D. Isla, J. De Castro Carpeno, P. Garrido, A. Rittmeyer, M. Tiseo, C. Meyenberg, S. de Haas, L.H. Lam, M.W. Lu, T.E. Stinchcombe

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Peters, L.H. Lam

Study supervision: S. Peters, R. Stahel, D. Isla, J. De Castro Carpeno, S. de Haas, L.H. Lam, M.W. Lu, T.E. Stinchcombe

Other (Please enter any additional contributions here: I was member of the steering committee and directly involved in the trial, development management, oral presentation, and manuscript development, editing and writing): T.E. Stinchcombe

The authors would like to thank all the patients who participated in the trial and their families, as well as the participating study sites.

The authors are grateful for the assistance of Sven Stanzel of F. Hoffmann-La Roche, Ltd., and Yvonne G. Lin, Alan Sandler, and David Chen of Genentech, Inc.

This study was funded by F. Hoffmann La-Roche, Ltd. Support for third-party writing assistance was provided by Meredith Kalish, MD, of CodonMedical, an Ashfield Company, part of UDG Healthcare plc, and was funded by F. Hoffmann-La Roche.

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.