Serum NSE is Early Marker of Transformed Neuroendocrine Tumor After EGFR-TKI Treatment of Lung Adenocarcinoma.

Background: Transformation to a lung neuroendocrine tumor (LNET) is a mechanism of resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI). Serum neuron-specific enolase (NSE) is a useful marker in the detection of LNET. Therefore, we explored the clinical significance of serum NSE levels in the detection of transformed neuroendocrine tumors after EGFR-TKI therapy.
Methods: We report a cohort of 5 cases in our treatment group. The characteristics of the patients, pathological diagnoses, immunohistochemistry with molecular detection, laboratory examination, and treatment histories are analyzed. The tumor markers of serum NSE were analyzed. Additionally, we reviewed the publications reporting the tumor markers before and after LNET transformation during EGFR-TKI therapy.
Results: Most patients are female (3/5), aged <60 years old (4/5), nonsmokers (4/5) and harbor the EGFR 19 exon deletion (4/5). The median time of LNET transformation was 19 months (range: 12-31 months). The clinical characteristics were similar to those reported in previous studies. Laboratory examination revealed an increased NSE level before the LNET is defined. Sixteen publications were reviewed. Of those, 86.67% (13/15) publications showed an increased level of NSE when the LNET transformation was defined.
Conclusion: Adenocarcinoma tumors in non-smokers, young patients harboring the EGFR 19 exon deletion tended to transform to LNETs after EGFR-TKI therapy. Combining our findings and a review of the literature, we suggest that serum NSE may be a useful tumor marker to predict neuroendocrine tumor transformation.

Introduction

Lung cancer is the leading cause of cancer-related death in China and worldwide.1,2 Based on its biological behavior and morphological characteristics, lung cancer is classified into two histological subgroups: non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC).3,4 Epidermal growth factor receptor (EGFR) gene mutations are observed in approximately 50% of patients with advanced adenocarcinoma (ADC) in Asian populations.5 The first-line treatment for patients with advanced NSCLC with EGFR mutations is the EGFR tyrosine kinase inhibitor (TKI), such as gefitinib or osimertinib.6–9

However, most patients who initially respond to EGFR-TKI therapy subsequently report disease progression. Several studies have reported the histological transformation from NSCLC to lung neuroendocrine tumors (LNETs) as a mechanism of EGFR-TKI resistance.10,11 LNETs are tumor originating from neuroendocrine cells of the lung.12 The transformed LNETs can be SCLC or large cell neuroendocrine carcinoma. During target therapy, about 5% of patients with NSCLC tumors transform into LNETs, at which point treatment plans must change. SCLC and large cell neuroendocrine carcinoma are considered the most aggressive types of malignant lung cancer. The treatments for SCLC and NSCLC are different. Despite a high rate of positive responses to first-line treatments, these cases show rapid growth and metastasis and acquire multidrug resistance. Thus, it is important to identify transformed LNETs.

However, despite the significance in detecting cancer transformation during EGFR-TKI therapy, rebiopsies are not routinely performed in lung cancer patients who show resistance to TKI treatment. Previous studies have reported that the rebiopsy rate is about 50%.13,14 Monitoring tumor markers may be a potential alternative method. Serum neuron-specific enolase (NSE) is a good marker for SCLC.15 NSE is a key enzyme in glycolysis. In tumors derived from neuroectoderm or neuroendocrine tissues, higher levels of NSE can be detected in serum.16 Previous studies believe that the increased NSE at initial diagnosis is an important negative prognostic factor for worse response to chemotherapy, radiotherapy, or immunotherapy.17–19 However, to date, only some cases have reported the role of tumor markers (including NSE) in the detection of LNET transformation, although to date no systematic analysis has been performed evaluating its clinical role.

Herein, we report 5 cases of LNET transformed from pulmonary ADC after EGFR-TKI therapy in a single institute, in addition to a review of the previous literature. We found elevated NSE levels before transformed LNETs were clinically defined.

Materials and Methods

Patients and Follow Up

From 2015–2020, we reviewed the pathological records of LNET cases, and identified 5 cases of ADC who received EGFR-TKI therapy that had transformed into lung LNETs. Pathology samples used to diagnose ADC and LNET were reviewed by at least two experienced pathologists. Immunohistochemistry markers, including TTF-1, p40, CD56, and CK, were used to distinguish subtypes if hematoxylin and eosin (H&E) staining was not sufficient. All patients were treated in the Pulmonary and Critical Care Medicine or Thoracic Oncology Departments, Shanghai Chest Hospital, Shanghai, China. Clinical and follow-up data were obtained for retrospective analysis, including age, sex, smoking history, clinical stage, and treatment. During therapy, patients were advised to see a doctor every two months to evaluate their condition. Chest computed tomography (CT) scans, B-mode ultrasonography (abdomen, supraclavicular lymph node), and samples for tumor markers were obtained when patients visited the hospital. A rebiopsy was carried out if the target lesion was diagnosed as progressive disease.

Pathology Diagnosis and EGFR Mutation Test

All samples were fixed in 4% formaldehyde, and 5 mm sections were stained with H&E. Sections were examined with an Olympus microscope. If pathologists were unable to distinguish the subtype of lung cancer, then immunohistochemistry was used. Immunohistochemical staining of CD56, TTF-1, CK, and p40, was performed. Pathologists examined each section using a double-blind procedure.

Before testing for EGFR mutations, tumor cells were evaluated. A minimum of 100 cancer cells were used. Two kits were used to detect EGFR mutations. The first was the EGFR mutation kit (Amoydx, Xiamen, China), and the second was the capture-based targeted sequencing panel (Burning Rock Biotech, Guangzhou, China). The tests were carried out according to the manufacturer’s instructions.

NSE Detection

A 3mL sample of venous blood was collected from each patient, and the serum was obtained after centrifugation. The serum NSE level was detected by Roche Elecsys 2010 automatic electrochemiluminescence immunoanalyzer. In the event of rupture of red blood cells, which will lead to a false increase in NSE, samples with evidence of hemolysis were discarded and a new blood sample was obtained.

Review of the Literature

All studies before 1 October 2021, describing the LNET transformation and tumor markers were reviewed and analyzed. Detailed data, including author, year published, the pathology diagnosis before and after transformation, and LNET specific tumor markers (NSE, progastrin gastrin releasing peptide [ProGRP]) before and after transformation.

Ethical Statement

This study was conducted following the Helsinki Declaration, and was approved by The Ethics Committee of Shanghai Chest Hospital (KS1752). All included patients signed informed consent forms of providing the case details and medical images. The authors are responsible for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Results

Five patients were included in our study: two males and three females. The median age was 47 years. Most patients were non-smokers who harbored an EGFR 19 exon deletion (4/5, 80%). Four patients translated to SCLC after therapy, and one patient transformed to lung LNET. One patient received five biopsies, the pathology types were ADC-SCLC-NSCLC-ADC-LNET, respectively. Three patients received only EGFR-TKI prior to LNET transformation, while two patients received both EGFR-TKI and chemotherapy. The response to EGFR-TKI was a partial response (PR) of four patient. Tumors from four patients were found to be transformed to LNET at the second biopsy, and the fifth at the third biopsy. ADC transformed to SCLC in four patients, and poorly differentiated LNETs were present in two patients. The median time of LNET transformation was 19 months (range: 12–31 months; Tables 1 and 2, and Figure 1). The immunohistochemistry of the transformed LNETs showed that CD56 was positive in each case, and TTF-1 was also expressed (Figure 2). Four patients detected EGFR in transformed LNETs. The transformed LNETs retained the initial EGFR mutation in those four cases. In the fifth biopsy from case 3, and the second biopsy from case 4, next-generation sequencing (NGS) was performed on the transformed LNET. In addition to the EGFR mutation, MYC amplification, and co-mutations of TP53 and BRCA2 were also detected (Table 1).

Table 1

Pathological Characteristics of the Initial ADC and Transformed Neuroendocrine Tumors

No	Age	Sex	Smoking History	Biopsy	Biopsy Result						Other
					Pathology		Sample Type	TTF-1	CD56	EGFR
1	55	F	No	First biopsy	ADC	Mediastinal lymph node	Needle biopsy	NA	NA	19del
				Second biopsy	ADC	Pleura	Needle biopsy	(+)	(-)	19delT790M
				Third biopsy	SCLC	Primary cancer	Needle biopsy	(+)	(+)	19del
2	46	M	Former	First biopsy	ADC	Primary cancer	Needle biopsy			19del
				Second biopsy	SCLC	Pleural effusion	Pleural effusion	(+)	(+/-)
3	47	M	No	First biopsy	ADC	Primary cancer	Needle biopsy	(+)	(-)	L858R
				Second biopsy	SCLC	Primary cancer	Needle biopsy	Part (+)	(+)	NA
				Third biopsy	NSCLC	Primary cancer	Tracheoscopic biopsy	Part (+)	NA	L858R
				Forth biopsy	ADC	Primary cancer	Tracheoscopic biopsy	(+)	(-)	L858R
				Fifth biopsy	LNET	Primary cancer	Tracheoscopic biopsy	Part (+)	(+)	L858R	TP53 mutation
4	76	F	No	First biopsy	ADC	Pleural effusion	Pleural effusion	(+)	(-)	19del
				Liquid biopsy						T790M
				Second biopsy	LNET	Primary cancer	Needle biopsy	(+)	(+)	19del	MYC amplificationBRCA2 mutation
5	47	F	No	First biopsy	ADC	Primary cancer	Surgery resected	NA	NA	19del
				Second biopsy	SCLC	Supraclavicular lymph node	Needle biopsy	Part (+)	(+)	19del

Abbreviations: ADC, adenocarcinoma; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; LNET, lung neuroendocrine tumor; EGFR, epidermal growth factor receptor.

Table 2

Treatment Before and After Transformation

No	Initial Tumor	Initial Stage	Treatment Before Transformation				Transformed Tumor	Transformed Stage	Treatment After Transformation
			TKI	Chemotherapy	Response	PFS			TKI	Chemotherapy
1	ADC	IV	Gefitinib,	TS-1	PR	12	SCLC	IV	-	EC
			Osimertinib			7
2	ADC	IV	Erlotinib	AP	PR	23	SCLC	IV	-	EC
			Osimertinib			1
3	ADC	IV	Gefitinib	-	PR	19	SCLC	IV	Gefitinib	EC, AP
4	ADC	IV	Icotinib,	-	PR	10	LNET	IV	-	AP, EC
			Osimertinib			5
5	ADC	IV	Gefitinib	-	SD	12	SCLC	IV	-	EC

Abbreviations: ADC, adenocarcinoma; SCLC, small cell lung cancer; LNET, lung neuroendocrine tumor; TKI, tyrosine kinase inhibitor; PFS, progression free survival; PR, partial response; SD, stable disease; TS-1, tegafur.

Figure 1

Treatment history, pathology, and other characteristics. Five cases received comprehensive treatment based on epidermal growth factor receptor-tyrosine kinase inhibitor targeted therapy, and the LNET transformation was confirmed in the rebiopsy specimen.

Figure 2

Representative pictures before and after transformation. The hematoxylin-eosin and immunohistochemistry staining showed that adenocarcinoma was transformed to lung neuroendocrine tumor. CD56, a neuroendocrine marker, was negative before transformation and strongly positive after transformation.

Detailed treatment histories are described for each case in the .

We analyzed tumor markers of NSE during EGFR-TKI treatment. As shown in Figure 3, at the start of treatment with EGFR-TKIs, NSE was present at normal levels, but gradually increased during treatment to above the normal range in the final laboratory examination before transformed LNETs were detected.

Figure 3

NSE detection of each patient. NSE levels gradually increased during epidermal growth factor receptor-tyrosine kinase inhibitor therapy and exceeded the reference level before the clinical definition of lung neuroendocrine tumor.

Before 1 October 2021, total of 16 reports, describing 16 patients, were enrolled (Table 3). Most cases were male (9/16, 56.25%), aged <60 years old (12/16, 75.00%), never smokers (10/16, 62.50%), harbored the EGFR 19del (12/16, 75.00%). Nine patients received EGFR-TKI treatment only, six patients received EGFR-TKI and chemotherapy, while the remaining patient received EGFR-TKI, chemotherapy and local radiation therapy. The median duration of response (DOR) of LNET transformation was 16.5 months (range 1–45 months). Fifteen studies reported the level of NSE during treatment, and 13 (86.67%) patients showed an increased level of NSE when LNET transformation was defined. Seven studies reported ProGRP levels, and 6 (85.71%) patients showed increased levels of ProGRP in LNET transformation.

Table 3

Review of NSE and ProGRP in Detecting Neuroendocrine Tumor Transformation After EGFR-TKI Treatment of Lung Adenocarcinoma

Author	Year	Age	Sex	Smoking	EGFR	TKI	Chemotherapy	Histopathology	EGFR	DOR	NSE	ProGRP	Ref
Zhang Yan	2013	80	M	No	19del	Icotinib	—	SCLC	19del	1	↑	NA	[20]
Watanabe Satoshi	2013	52	F	No	19del	Erlotinib	Pemetrexed+cisplatin	SCLC	19del	20	↑	↑	[21]
Xue Shuping	2015	49	F	No	19del	Icotinib		SCLC	NA	23	↑	↑	[22]
Furugen Makoto	2015	63	M	No	19del	Erlotinib		SCLC+ADC	19del	10	↑	↑	[23]
Fang Liying	2016	74	M	Yes	19del	Gefitinib		SCLC	NA	12	↑	NA	[24]
Lin Quan	2016	49	M	Yes	L858R	Gefitinib	Pemetrexed+cisplatinDocetaxel+ cisplatin	SCLC	L858R	15	↑	NA	[25]
Xu Yanjun	2017	37	F	No	19del	Gefitinib	Pemetrexed+cisplatin	SLCL	NA	7	↑	NA	[26]
Lu Hongyang	2018	55	M	Yes	L858R	Icotinib		NEC	L858R	19	↑	NA	[27]
Shiroyama Takayuki	2018	35	M	Yes	19del	Afatinib		SCLC	19del	7	—$	—$	[28]
Liu Yangyang	2018	38	M	No	L858R	Erlotinib		SLCL	NA	18	↑	NA	[29]
Iijima Yuki	2018	52	F	No	19delT790M	GefitinibErlotinibOsimertinib	Pemetrexed+carboplatinGemcitabine	SCLC	19del	45	↑	↑	[30]
Han Chaonan	2019	58	F	No	19del	Icotinib	Pemetrexed	SCLC	19del	10	↑	NA	[31]
Chen Shen	2019	49	M	No	L858RT790M	GefitinibOsimertinib	Pemetrexed+cisplatin	SCLC	NA	30.7	↑	NA	[32]
Kato Yasuhiro	2019	74	F	No	19del	Afatnib		SCLC	19del	20	NA	↑	[33]
Fiore Michele*	2019	56	F	Yes	19del	Gefitinib	Pemetrexed+cisplatin	SCLC	19del	20	↑	↑	[34]
Xie Zhanhong	2019	53	M	Yes	19del	Erlotinib		SCLC	19del	11	—	NA	[35]

Notes: *Local radiotherapy was added; $NSE and ProGRP were elevated after the first cycle of chemotherapy.

Abbreviations: EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor; DOR, duration of response; NSE, neuron-specific enolase; ProGRP, progastrin-releasing peptide; NA, not available.

Discussion and Review of the Literature

EGFR mutations are the most common mutation in NSCLC, and mutations are considered to be the drivers of NSCLC carcinogenesis.36 Half of Asian NSCLC patients, especially nonsmoking females, harbor EGFR mutations.37 EGFR-TKIs are a class of drugs that has been widely used in clinical practice to block EGFR signaling pathways and inhibit tumor cell proliferation and migration.38 Several studies have shown prolonged progression-free survival after first-line EGFR-TKI therapy,5–9 but patients gradually develop drug resistance. Different mechanisms have been described to explain this, including new EGFR mutations, amplification of the MET gene, epithelial-mesenchymal transitions, and transformation to LNETs.39 These mechanisms are also responsible for resistance to third-generation EGFR-TKIs.32 Histological transformation is found in approximately 5% of EGFR-TKI therapy-resistant patients.40 The most common type of transformation that results from EGFR-TKI therapy is ADC to SCLC, which is an aggressive cancer usually treated by chemotherapy and radiotherapy.

In our study, all cases were diagnosed as ADC with sensitive EGFR mutations. Subsequent biopsies showed that TKI treatment was associated with the transformation to SCLC. A previous study reported that there is a high incidence of SCLC transformation among non-smoking, younger aged, EGFR-mutated NSCLC patients.41 Similarly, in our study the patients were mostly young (<60 years old), harbored the EGFR 19 deletion, and had never smoked. The median time for SCLC transformation was reported to be 17.8 months (range 14.3–26.2 months),11 and the transformed LNET always retained their original EGFR-activating mutations,21,32,41,42 which was consistent with our findings. Although transformed LNETs harbored the initial EGFR mutation, most patients did not continue EGFR-TKI therapy. In only one patient the histological type was NSCLC after chemotherapy for transformed SCLC, after which EGFR-TKI therapy was continued for about nine months.

Different theories have been proposed to explain the mechanism of histological transformation. One theory proposed that tumors that transform to LNET are not pure NSCLC. As the tumor is heterogeneous, samples obtained from bronchoscopy, fine-needle aspiration or pleural effusion may not provide sufficient material to detect the presence of combined histology at diagnosis.43,44 Niederst et al also proposed a cladistic evolution model: a precursor cell at baseline may transform into SCLC in response to EGFR-TKIs.10 The signaling pathway for this process may include TP53, RB1, Notch, and MYC.45–50 In our study, NGS was conducted on two patients with transformed LNETs and revealed the presence of TP53 and MYC, which was consistent with previous findings.

Although it is important to detect cancer transformation during EGFR-TKI therapy, additional biopsies are not routinely performed in lung cancer patients who show resistance to TKI treatment. Previous studies have reported that the rebiopsy rate is about 50%.13,14 Additional biopsies are not usually performed due to inaccessible tumors, deterioration of performance status, or patient refusal.51 Thus tumor markers may play a key role in monitoring cancer transformation. NSE and ProGRP are widely used to monitor SCLC treatment. NSE is a key enzyme in glycolysis. In tumors derived from neuroectoderm or neuroendocrine tissues, higher levels of NSE can be detected in serum.16 Gastrin releasing peptide has been identified in nerve fibers, brain and neuroendocrine cells of the fetal lung. It proved challenging because gastrin releasing peptide is unstable in blood, however ProGRP has been found to be more stable,52 and has been shown to be a sensitive and specific marker of SCLC.53

We reviewed and analyzed all the literature reporting LNET transformation and tumor markers. A total of sixteen reports including sixteen patients were enrolled (Table 3). The patients described in those literatures most are male, younger aged, never smoking, who harbored the EGFR 19Del mutation. The media DOR of LNET transformation was 16.5 months, which was similar to that observed in our findings. Most studies reported an increased NSE level (13/15, 86.67%) or ProGRP level (6/7, 85.71%) once LNET transformation was defined. This is consistent with our findings. In our cohort of these five cases, increased NSE levels may indicate pathological transformation. In the present study, NSE increased gradually during treatment to above the reference value prior to detection of transformed LNETs. Although NSE and ProGRP were good markers for the detection of LNET transformation, in some studies, NSE or ProGRP were normal during transformation.32,35 Therefore, our findings suggested that NSE is an accurate predictor of the transformation to LNETs. But a rebiopsy was encouraged to confirm the type of cancer.

In summary, we report 5 cases of transformation to LNETs following EGFR-TKI treatment of ADC. The clinical characteristics and pathological findings described here are consistent with previous studies. We found that the NSE levels can be good predictors of cancer transformation.

Conclusion

Serum NSE can be a useful tumor marker to predict neuroendocrine tumor transformation during EGFR-TKI therapy.