MYC and EIF3H Coamplification Significantly Improve Response and Survival of Non-small Cell Lung Cancer Patients (NSCLC) Treated with Gefitinib

Abstract

Background

We investigated the incidence of eukaryotic translation initiation factor 3 subunit H (EIF3H) and MYC amplification in non-small cell lung cancer (NSCLC) patients, and whether MYC/EIF3H increased gene copy number affected response to Epidermal Growth Factor Receptor tyrosine kinase inhibitors.

Methods

Metastatic NSCLC patients (n = 54) treated with gefitinib were analyzed for the genomic content of EIF3H and MYC genes by fluorescence in situ hybridization (FISH) using a custom-designed 3-color DNA probe set.

Result

Amplification of EIF3H (ratio EIF3H/CEP8 >2), was observed in 10 cases (18.5%), and MYC was coamplified in all. MYC amplification without coamplification of EIF3H was observed in 2 cases (3.7%). Receiver operating characteristic analysis was conducted to identify the cutoff for MYC and EIF3H copy number best discriminating sensitive and resistant populations. MYC FISH positive patients (MYC+, mean ≥2.8) had a significantly higher response rate (p = 0.003), longer time to progression (p = 0.01) and overall survival (OS: p = 0.02) than MYC− (mean <2.8). Similarly, EIF3H FISH positive patients (EIF3H+, mean ≥2.75) had a significantly higher response rate (p = 0.002), longer time to progression (p = 0.01) and OS (p = 0.01) than EIF3H− (mean <2.75).

Conclusion

Our results indicate that MYC and EIF3H are frequently coamplified in NSCLC and that a high copy number correlates with increased epidermal growth factor receptor tyrosine kinase inhibitors sensitivity.

Despite tangible progress made during the last few years, the prognosis of patients with non-small cell lung cancer (NSCLC) is still disappointing. Recent therapeutic advances include drugs targeting the Epidermal Growth Factor Receptor (EGFR), such as the monoclonal antibody cetuximab,¹ or the orally available tyrosine kinase inhibitors (TKIs) gefitinib and erlotinib.²

It is thought that EGFR-TKIs are more effective in patients with certain clinical or biologic characteristics. However, the best method for patient selection is still controversial. Available data suggest that the presence of activating EGFR mutations identifies individuals with the highest chance to have tumor shrinkage,³ while EGFR gene copy number identifies patients with the highest benefit in terms of survival.^4–7 Conversely, the presence of a KRAS mutation identifies individuals with the lowest chance of responding to EGFR-TKI,⁸ with a possible detrimental effect on survival.⁹ Nevertheless, none of the above mentioned biomarkers accounts for all sensitive or resistant cases, suggesting that other factors associated with drug sensitivity could be useful for improved patient selection.

Molecular cytogenetic analyses have indicated that gain of the long arm of chromosome 8 (8q) is a frequent event in many human malignant diseases, including lung cancer.¹⁰ MYC, localized at 8q24.1, is a well characterized oncogene involved in cell growth, differentiation, and apoptosis.¹¹ Previous studies reported that MYC amplification was associated with tumor progression and a worse prognosis.^12,13 In the study conducted by Lockwood et al.,¹⁴ MYC was the most frequent amplification observed in lung cell lines. A recent analysis, conducted in 300 patients with gastric adenocarcinoma, showed coamplification of MYC and either EGFR or HER2,¹⁵ suggesting that MYC might influence the activity of drugs targeting these receptors. Moreover, preliminary results from the National Surgical Adjuvant Breast and Bowel B-31 trial suggested that HER2-positive breast cancer patients with primary tumor MYC amplification derived more benefit from trastuzumab, a monoclonal antibody against HER2, than patients without MYC amplification.^16,17

The eukaryotic translation initiation factor 3 (EIF3) is a large multiprotein complex originally identified as a factor that binds the 40S ribosomal subunit and prevents its association with the 60S subunit. EIF3 consists of 13 putative subunits known as EIF3A-EIF3M.¹⁸ Various EIF3 subunits have been found to have altered expression in malignant tumors. In humans, overexpression of EIF3H has been found in prostate, breast, and liver cancer.^19–22 In many cases, this seems to be due to amplification of the 8q23 region which includes the EIF3H gene at 8q23.3. EIF3H amplification has been observed in 20% of untreated primary breast cancer, 30% of hormone refractory prostate carcinoma¹⁹ and 26% of hepatocellular carcinoma.²⁰ High level amplification of the EIF3H gene has also been associated with advanced stage and poor prognosis prostate cancer.²¹ At the cellular level, EIF3H overexpression increases proliferation, growth and survival.²³ Recent studies have demonstrated that EIF3H and MYC are coamplified in prostate cancer, suggesting that there may be cooperation between MYC and EIF3 to further up-regulate translation initiation.²⁴

Whether MYC and EIF3H are coamplified in lung cancer, and whether this could affect anti-EGFR responses, is unknown. Based on these premises, we evaluated MYC and EIF3H gene status in patients with NSCLC treated with the EGFR-TKI gefitinib.

PATIENTS AND METHODS

Patient Selection

The present retrospective study was conducted in a cohort of NSCLC patients previously evaluated for EGFR and HER2 gene copy number.^4,25 Among the 102 patients included in the original study, 54 had tumor tissue available and were selected for the present analysis. No additional clinical or biologic characteristic was used for patient selection. All patients had histologically confirmed NSCLC with measurable, locally advanced or metastatic disease, progressing or relapsing after chemotherapy, or with medical contraindications for chemotherapy. Patients were classified as never smokers (<100 cigarettes per lifetime), former smokers (quit smoking more than 6 months before starting gefitinib therapy), or current smokers (quit smoking less than 6 months before starting gefitinib therapy or active smokers). Patients received gefitinib (250 mg/d) and were evaluated for response after 2 months according to the RECIST criteria.²⁶ Tumor response was assessed by computer tomography scan, with a confirmatory evaluation repeated in patients with complete response (CR), partial response, and stable disease at least 4 weeks after the initial determination of response. The study population included 14 (26.4%) patients positive for EGFR by fluorescence in situ hybridization (FISH). Eleven patients (22.4%) had an EGFR mutation in exon 19 or 21 and 10 patients (21.7%) had a KRAS mutation. All patients received at least one previous chemotherapy, including platinum in 80.4% of cases. The study was approved by the local Ethics Committee and was conducted in accordance with ethical principles stated in the most recent version of the Declaration of Helsinki.

Tissue Preparation and FISH Analyses

Sections from paraffin-embedded tissue blocks containing representative malignant cells obtained at time of diagnosis were used for this analysis. Histopathological classification was determined on hematoxylin-eosin stained sections based on the World Health Organization criteria.²⁷

DNA insert of the BAC clone NO 764P21, encompassing sequences of the EIF3H gene, was amplified and labeled by nick-translation using a Nick Translation Kit (Vysis/Abbott Molecular Catalog#32-801300) and SpectrumGreen conjugated dUTPs according to the manufacturer’s instructions. The labeled DNA was precipitated in the presence of human Cot1 DNA. The chromosomal mapping and the hybridization efficiency of the new reagent was tested in specimens with normal karyotypes. Three-color FISH assays were performed on blank 4 µm, paraffin-embedded lung cancer tissue sections, using the SpectrumGreen labeled EIF3H probe, the LSI C-MYC SpectrumOrange and the CEP-8 SpectrumAqua commercial probes (Abbott Molecular) according to standard protocol in the laboratory.

Initially the slides were incubated for 2 hours at 65°C, deparaffinized in Citri-Solv (Fisher) and washed in 100% ethanol for 5 minutes. The slides were sequentially incubated in 2× SSC at 75°C for 10 to 18 minutes, digested in 0.25 mg/ml Proteinase K/2× SSC at 45°C for 10 to 18 minutes, washed in 2× SSC for 5 minutes, and dehydrated in ethanol. The probe mixture was prepared using 0.5 µl of the stock reagent of each of the commercial probes and 200 ng of DNA from the EIF3H probe and applied to the selected hybridization areas, which were covered with glass coverslips and sealed with rubber cement. DNA denaturation was performed for 15 minutes at 80°C and slides were incubated at 37°C for 24 to 48 hours. Posthybridization washes were performed with 2× SSC/0.3% NP-40 at 72°C for 2 minutes. The slides were then washed in 2× SSC for 1 minute, dehydrated in ethanol and chromatin was counterstained with diamidino-2-phenylindole (0.3 µg/ml in Vectashield, Vector Laboratories, Burlingame, CA). Analysis was performed on epifluorescence microscope using single interference filter sets for green (fluorescein isothiocyanate), red (Texas red), aqua (Aqua) and blue (diamidino-2-phenylindole) as well as dual (red/green) and triple (blue, red, green) band pass filters. For documentation, images were captured using a charge- coupled device camera and merged using dedicated software.

Statistical Analysis

The primary end-point of the study was the association of MYC and EIF3H gene copy number with response to gefitinib therapy. Analysis of receiver operating characteristic (ROC) curve was carried out with the aim of determining a cutoff point for MYC and EIF3H gene copy numbers as a continuous variable.²⁸ Sensitivity and specificity were expressed in terms of percentage and the highest value has been chosen as the best cutoff point. Secondary end-points were association with time to progression (TTP) and overall survival (OS). TTP was calculated from the time of first gefitinib dose administration to time of disease progression or last disease assessment. OS was calculated from the time of first gefitinib dose administration to patient death or last contact.

Differences in response rate (RR) were compared by Fisher’s exact test or χ² test. TTP, OS, and the 95% confidence intervals were evaluated by survival analysis using Kaplan-Meier method.²⁹ TTP and OS for the groups with negative and positive biomarker were compared using the log-rank test. Statistical significance was set at <0.05 for each analysis. All statistical analyses were performed using SPSS version 11.5.1 (SPSS Italia srl, Bologna, Italy).

RESULTS

Patient Characteristics

Characteristics of the 54 patients included in the study are shown in Table 1. The median age was 61 years (range 25–79), the majority of patients were male (64.8%), and most were former or current smokers (85.2%) with a good performance status (ECOG PS 0–1: 90.7%). Adenocarcinoma was the most frequent histologic type (55.6%). In the study population, the overall response rate was 17.0%, including 1 CR and 8 partial responses (PR). The disease control rate, including CR, PR, and stable disease, was 41.5%, median TTP was 3.5 months, and median survival was 11.0 months. EGFR FISH and mutation status was known in 53 and 49 patients, respectively. EGFR was positive by FISH in 14 (26%) cases and by mutation in 11 (22%). Response rate was significantly higher in EGFR FISH positive than in negative (50% versus 5%, p = 0.001) and in patients with EGFR mutations than in EGFR wild type (63% versus 2%, p = 0.0001).

Table 1.

Patient Characteristics

Characteristics	Total	%
Total	54
Median age (year, range)	61 (25–79)
Gender
Male	35	64.8
Female	19	35.2
Smoking history
Never	8	14.8
Former	22	40.8
Current	24	44.4
Histology
Adenocarcinoma + bronchioloalveolar	35	64.8
Squamous cell carcinoma	12	22.2
Other	7	13.0
ECOG performance status
0	33	61.1
1	16	29.6
2	5	9.3
Total evaluated for EGFR FISH	53
Positive	14	26.4
Negative	39	73.6
Total evaluated for EGFR mutation	49
Exon 19	7	14.3
Exon 21	4	8.1
Wild type	38	77.6
Total evaluated for KRAS mutation	46
Mutated	10	21.7
Wild type	36	78.3
Total evaluable for response to gefitinib therapy	53
Complete response	1	1.9
Partial response	8	15.1
Stable disease	13	24.5
Progressive disease	31	58.5
Time to progression (mo)	3.5
Median survival (mo)	11.0

EGFR, epidermal growth factor receptor; FISH, fluorescent in situ hybridization; ECOG, Eastern Cooperative Oncology Group.

Outcome According to MYC Gene Copy Number

As shown in Table 2, FISH analysis for MYC was successfully performed in all patients. Median mean MYC gene copy number was 2.92 copies/cell (range 1.52–12.38). Amplification of MYC (ratio MYC/CEP8 >2), was observed in 12 cases (22.2%). In the ROC analysis, a cut-off of MYC copy number per cell at 2.8 identified an area under the curve of 0.78 (Figure 1A). Using this cutoff, 25 specimens (46.3%) were MYC FISH negative (mean per cell <2.78) and 29 (53.7%) were MYC FISH positive (mean per cell ≥2.8 copies). Among the 29 MYC FISH positive patients, 7 were EGFR FISH positive and EGFR mutated, 4 were EGFR FISH positive, 1 harbored an EGFR mutation and 5 harbored a KRAS mutation. Increased MYC gene copy number was not associated with any statistically significant clinical or biologic characteristic, even though the MYC FISH positive patients were more frequently EGFR FISH positive, EGFR mutated and never smokers. The RR significantly favored MYC FISH positive over MYC FISH negative patients (31.0% versus 0%, p = 0.003), as well as TTP (4.4 versus 2.6 months, p = 0.014) and OS (13.8 versus 6.4 months, p = 0.025), as shown in Table 2 and Figure 2. In the small subgroup of EGFR FISH positive or EGFR mutated patients (n = 17), all responders were MYC FISH positive (66.0% in EGFR positive/MYC positive versus 0% in EGFR positive/MYC negative, p = 0.04).

Table 2.

MYC/EIF3H Gene Copy Number, Patient Characteristics and Response to Gefitinib Therapy

	Total Number	MYC + (n/%)	MYC − (n/%)	p	EIF3H + (n/%)	EIF3H − (n/%)	p
Gender	54	29	25	0.18	28	26	0.13
Male	35	16/55.2	19/76.0		15/53.6	20/77.0
Female	19	13/44.8	6/24.0		13/46.4	6/23.0
Smoking	54	29	25	0.09	28	26	0.07
Never	8	7/24.1	1/4.0		7/25.0	1/3.8
Former/current	46	22/75.9	24/96.0		21/75.0	25/96.2
Histology	54	29	25	0.12	28	26	0.17
Adenocarcinomaa	35	22/75.9	13/52.0		21/75.0	14/53.8
Other	19	7/24.1	12/48.0		7/25.0	12/46.2
EGFR FISH	53	29	24	0.07	28	25	0.19
Positive	14	11/37.9	3/12.5		10/35.7	4/16.0
Negative	39	18/62.1	21/87.5		18/64.3	21/84.0
EGFR mutation	49	28	21	0.4	27	22	0.32
Mutated	11	8/28.5	3/14.3		8/29.6	3/13.6
Wild type	38	20/71.5	18/85.7		19/70.4	19/86.4
KRAS mutation	46	27	19	0.80	26	20	0.92
Mutated	10	5/18.5	5/26.3		5/19.2	5/25.0
Wild type	36	22/81.5	14/73.7		21/80.8	15/75.0
Response to gefitinib	53	29	24	0.003	28	25	0.002
Responderb	9	9/31.0	0		9/32.1	0
Non responderb	44	20/69.0	24/100.0		19/67.9	25/100.0
Response in EGFR+c	17	12	5	0.04	11	6	0.01
Responderb	8	8/66.6	0		8	0
Non responderb	9	4/33.4	5/100.0		3	6

^aAdenocarcinoma category also included bronchioloalveolar carcinoma.

^bResponder category included patients with complete and partial response; non responder category included patients with stable and progressive disease.

^cEGFR + included patients with a positive FISH test or with an EGFR mutation.

FIGURE 1.

Receiver Operating Characteristic (ROC) curves for MYC (A) and eukaryotic translation initiation factor 3 subunit H (EIF3H) (B). The value best discriminating responders versus nonresponders to gefitinib therapy was a mean gene copy number per cell of 2.8 for MYC and 2.75 for EIF3H.

FIGURE 2.

The picture shows time to progression and survival of patients treated with gefitinib according to MYC fluorescence in situ hybridization (FISH) status. A mean gene copy number per cell ≥2.8 qualified the sample as MYC positive.

Outcome According to EIF3H Gene Copy Number

EIF3H was successfully evaluated by FISH in all cases (Table 2). Median mean EIF3H gene copy number was 2.79 copies/cell (range 1.65–5.02). Amplification of EIF3H (ratio EIF3H/CEP8 >2), was observed in 10 cases (18.5%), and MYC was coamplified in all (Figure 3). There were only two cases with MYC amplification in which EIF3H was not coamplified. Using a ROC analysis (Figure 1B), a cutoff of EIF3H copy number per cell at 2.75 identified an area under the curve of 0.84. Using this cutoff, 28 specimens (51.8%) were EIF3H FISH positive (mean per cell ≥2.75 copies) and 26 (48.2%) were EIF3H FISH negative (mean per cell <2.75). As was the case for MYC, increased EIF3H gene copy number was not associated any statistically significant clinical or biologic characteristic. Nevertheless, EIF3H FISH positive cases were more frequently never smokers and EGFR positive either by FISH or mutated. Interestingly, all EIF3H FISH positive patients were also MYC FISH positive, and this association was statistically significant (p < 0.0001). EIF3H FISH positive patients had a significantly higher RR (32.1% versus 0%, p = 0.002), significantly longer TTP (4.4 versus 2.7 months, p = 0.01) and OS (17.8 versus 6.4 months, p = 0.01) than EIF3H FISH negative, as shown in Table 2 and Figure 4. In the small cohort of patients that were EGFR FISH positive or harbored an EGFR mutation (n = 17), all responders were EIF3H FISH positive and vice versa (p = 0.01).

FIGURE 3.

Non-small cell lung cancer (NSCLC) sections hybridized with the MYC (SpectrumOrange)/eukaryotic translation initiation factor 3 subunit H (EIF3H) SpectrumGreen/CEP8 SpectrumAqua probe set showing low copy numbers (A), high copy numbers for the 3 targets (B) and coamplification of MYC and EIF3H (C). EIF3H coamplification was observed in 10 of 12 MYC amplified specimens.

FIGURE 4.

The picture shows time to progression and survival of patients treated with gefitinib according to translation initiation factor 3 subunit H (EIF3H) fluorescence in situ hybridization (FISH) status. A mean gene copy number per cell ≥2.75 qualified the sample as EIF3H positive.

DISCUSSION

The present study, conducted in a NSCLC population exposed to the EGFR-TKI gefitinib, for the first time demonstrated that MYC and EIF3H are frequently coamplified in NSCLC and this event could affect drug sensitivity. Patients with MYC or EIF3H increased gene copy number had better outcomes in terms of response, TTP and survival than individuals lacking MYC/EIF3H gene gain.

Previous studies have reported that amplification of 8q may be involved in cancer progression.¹⁰ The two genes evaluated in our study, MYC and EIF3H, have been suggested to be putative target genes in 8q23-q24.^22,30,31 These two genes are frequently coamplified in recurrent hormone-refractory prostate cancer²¹ and in breast cancer, but they were not always included in the same amplicon.^19,32 Nevertheless, Nupponen et al.,³² based on the amplicon mapping and gene expression data, reported that the strongest candidate target gene for the 8q23-q24 amplification was EIF3H. In the present study, for the first time we have shown that EIF3H and MYC are frequently coamplified in NSCLC and probably these two genes are not included in the same amplicon. Interestingly, many of the genes regulated by MYC affect ribosome biogenesis and protein translation.^33–35 Thus, coamplification of EIF3H and MYC may cooperate in enhanced protein translation, either in a general way or for a specific subset of mRNAs.

To perform this study, we developed a three-color DNA probe set to simultaneously examine the copy number of EIF3H, MYC and centromere eight sequences in lung tumors, and were able to observe amplification of both EIF3H and MYC in the same tumor cells. This finding generates the hypothesis that coamplification and overexpression of EIF3H with MYC cooperatively shifts the repertoire of actively translated mRNAs to a set of genes that drive proliferation and growth, or inhibit apoptosis.

The possibility that MYC influences the efficacy of agents targeting the EGFR family has been postulated in a study conducted in breast cancer patients.^16,17 In the National Surgical Adjuvant Breast and Bowel B-31 clinical trial, which compared treatment with chemotherapy versus chemotherapy plus trastuzumab, a total of 1549 patients were evaluated for MYC, and 30% of patients displayed gene amplification. Among patients with amplification of both HER2 and MYC, the addition of trastuzumab to chemotherapy reduced the risk of cancer recurrence by 76% and reduced the risk of death by 64%, while in individuals with HER2 amplification alone the addition of trastuzumab to chemotherapy did not reduce the risk of death.¹⁶ In our study, we observed that both MYC and EIF3H were associated with a better outcome after gefitinib therapy, and among EGFR+ patients (FISH+ and/or mutated), only individuals with increased MYC/EIF3H gene copy number had a significant tumor shrinkage. The mechanism responsible for the highest sensitivity to anti-EGFR agents in presence of MYC/EIF3H gene gain is not clear. The small number of patients included in the present study and the strong association between the clinical and biologic parameters did not allow us to assess the value of different predictors in a multivariable model. Nevertheless, preclinical data have demonstrated that drugs interfering with the EGFR family inhibit expression and activity of MYC, suggesting that MYC-dependent tumors may be more sensitive to anti-EGFR strategies.^36–38 MYC expression is known to induce apoptosis, which can be inhibited by signaling from either the RAS and AKT pathways,³⁹ antiapoptotic molecules, i.e., BCL-2,⁴⁰ or through the up-regulation of factors such as MAD-1, which directly antagonize MYC function.⁴¹ Coamplification of EIF3H may also reduce MYC-dependent apoptosis in prostate epithelial cells.⁴² If MYC up-regulation was an early event in some tumors, then growth factor signaling of EIF3H over-expression might provide a necessary antiapoptotic signal. Finally, it is not possible to exclude that other genes included in the amplicon could in some way modulate drug sensitivity.

MYC gene amplification has been evaluated in several solid tumors, including lung cancer.^13,14,43 Conversely, to our knowledge, no previous study investigated the incidence of EIF3H gene amplification in lung tumors. Previous studies reported a 13 to 23% incidence of MYC amplification in small cell lung cancer and 5 to 12% in NSCLC.¹² In the study conducted by Kubokura et al.,¹² the incidence of MYC amplification in NSCLC was up to 88%, higher than reported in other experiences including the present study probably because of the different and less stringent criteria used for defining gene amplification. Interestingly, Kubokura study showed that MYC amplification correlated with lymph node metastasis, suggesting a possible negative effect on survival. Although the absence of a control arm precludes the possibility to discriminate between the predictive versus prognostic effect of study genes, our results, together with the possible negative prognostic effect of both EIF3H and MYC^12,13,21,22 suggest that those genes could represent a favorable predictive factor for survival in patients exposed to EGFR-TKIs.

In conclusion, our results showed for the first time that MYC and EIF3H are coamplified in NSCLC and this biologic event could positively affect the response and survival of patients treated with EGFR-TKIs. These data originated from a retrospective analysis and a prospective validation is warranted.