1 BACKGROUND

Primary lung cancer (hereafter refers to as lung cancer) that begins in the lungs is one of the most common malignant tumors in China [1] and worldwide [2]. According to the World Health Organization, the Global Cancer Observatory (GLOBOCAN) 2020 and the National Cancer Center database, 815,600 new lung cancer patients and 714,700 deaths caused by lung cancer were seen in 2020 in China, ranking first among all malignant tumors causing mortality [3]. More than 80% of lung cancer patients have non-small cell lung cancer (NSCLC) [1, 2].

In the past decade, discovering predictive biomarkers has paved the way for new treatment opportunities using targeted therapy and immunotherapy in the NSCLC [4-6]. Anaplastic lymphoma kinase (ALK) gene rearrangement is the third most common driving mutation in the development of NSCLC after mutations in epidermal growth factor receptor (EGFR) and Kirsten rat sarcoma viral oncogene homolog (KRAS) [7, 8], and previous studies from China [9] and America [10] showing that ALK gene rearrangement occurs in 3%-7% of NSCLC cases, with a high incidence among adenocarcinoma and non-smoking patients. The echinoderm microtubule-associated protein-like 4 (EML4)-ALK fusion is the primary type of ALK rearrangement. In addition to EML4, other genes fused with the ALK gene in lung cancer include kinesin family 5B (KIF5B), kinesin light chain (KLC), trafficking from ER to golgi regulator (TFG), and others [11-13].

Currently, the molecular target drugs in the market (e.g., crizotinib and alectinib) for lung cancer with ALK gene rearrangement and other genetic variations have significantly improved the patients' outcome [14, 15], even for squamous cell carcinoma [16]. Therefore, detecting the ALK gene rearrangements in an accurate, standardized, and cost-effective manner is critical for selecting drugs to manage lung cancer. Presently, China's National Medical Products Administration (NMPA) has approved ALK gene rearrangement diagnostics in four technical platforms, such as immunohistochemistry (IHC)-VENTANA⁃D5F3, fluorescence in-situ hybridization (FISH), real-time polymerase chain reaction (RT-PCR), and next-generation sequencing (NGS). The FISH technique requires a small amount of tumor sample, but the procedure is technologically demanding and has atypical negative signals. RT-PCR has high specificity and sensitivity but can only detect currently known variants. IHC-VENTANA-D5F3 is simple, inexpensive, and can be performed in almost all locations, but the criteria for positive response are not applicable in all cases, leading to ambiguous results and the need for other platforms. NGS can detect most variants, but it is expensive and requires a test run. To address this situation, experts from the RATICAL study and the Chinese Society of Clinical Oncology issued guidelines for the ALK fusion test in NSCLC in 2021 to adopt in clinical practice [17].

However, selecting an appropriate diagnostic method with good quality control for detecting ALK rearrangements has been challenging, impacting the clinical benefit. To gain knowledge on the current status and quality control of ALK gene test in lung adenocarcinoma, we planned a nationwide multicenter clinical study (the RATICAL study) including NSCLC patients who underwent ALK gene test from 30 medical centers. This study aimed to investigate the real-world test patterns of ALK gene rearrangements in advanced NSCLC, including the detection methods and quality control, and to investigate the positive rates, the associated clinicopathological characteristics, and the presence of other mutations.

2 METHODS

3 RESULTS

3.2 Patient characteristics

A total of 25,616 patients were screened, 187 patients were excluded due to missing data and unclear ALK gene rearrangement results, and 1,740 patients were excluded because they had early-stage NSCLC. Therefore, 23,689 samples were analyzed, including 17,436 patients with adenocarcinoma (Figure 1). In the present study, 19.6% of patients were tested with samples obtained from metastatic foci and 81.7% of ALK tests were performed on histological samples in clinical practice, followed by cytological samples (15.9%). The overall ALK gene rearrangement rate was 6.7% in advanced NSCLC. ALK rearrangements were more common in adenocarcinoma (8.2%) than in squamous (1.5%) and other types (4.2%) (P < 0.001). The characteristics of the patients with NSCLC and adenocarcinoma are shown in Table 1.

TABLE 1. Clinicopathological characteristics of the patients with advanced NSCLC and advanced adenocarcinoma

	Advanced NSCLC (n = 23,689)			Advanced adenocarcinoma (n = 17,436)
Variable	ALK-positive	ALK-negative	P	ALK-positive	ALK-negative	P
Total, n (%)	1,590 (6.7)	22,099 (93.3)		1,422 (8.2)	16,014 (91.8)
Sex, n (%)			<0.001			<0.001
Male	757 (5.2)	13,725 (94.8)		658 (7.1)	8,669 (92.9)
Female	833 (9.0)	8,374 (91.0)		764 (9.4)	7,345 (90.6)
Age, n (%)			<0.001			<0.001
≤ 35	110 (31.9)	235 (68.1)		102 (34.6)	193 (65.4)
36-65	1,217 (8.4)	13,320 (91.6)		1,090 (9.8)	10,001 (90.2)
> 65	263 (3.0)	8,544 (97.0)		230 (3.8)	5,820 (96.2)
Smokers, n (%)			<0.001			<0.001
Current	280 (4.2)	6,381 (95.8)		238 (5.6)	3,985 (94.4)
Never	853 (8.3)	9,439 (91.7)		782 (9.3)	7,650 (90.7)
Unknown	457 (6.8)	6,279 (93.2)		402 (8.4)	4,379 (91.6)
Tumor cellularity, n (%)			0.024			0.033
≤ 30	227 (6.7)	3,181 (93.3)		215 (7.9)	2,494 (92.1)
31-50	214 (7.2)	2,756 (92.8)		196 (8.3)	2,155 (91.7)
> 50	325 (8.3)	3,590 (91.7)		295 (9.8)	2,720 (90.2)
Tumor location, n (%)			<0.001			<0.001
Right upper lobe	173 (4.4)	3,760 (95.6)		155 (5.2)	2,807 (94.8)
Right lower lobe	196 (7.8)	2,317 (92.2)		177 (9.5)	1,696 (90.5)
Right middle lobe	132 (9.8)	1,213 (90.2)		121 (11.9)	898 (88.1)
Left upper lobe	226 (6.8)	3,103 (93.2)		204 (8.4)	2,230 (91.6)
Left lower lobe	179 (8.7)	1,887 (91.3)		160 (10.5)	1,371 (89.5)
Other	684 (6.5)	9,819 (93.5)		605 (7.9)	7,012 (92.1)
Sampling lesion, n (%)			<0.001			<0.001
Primary	1,043 (5.6)	17,443 (94.4)		934 (7.1)	12,247 (92.9)
Metastatic	498 (10.7)	4,151 (89.3)		452 (11.5)	3,469 (88.5)
Unknown	49 (8.8)	505 (91.2)		36 (10.8)	298 (89.2)
Pathology, n (%)			<0.001
Adenocarcinoma	1,422 (8.2)	16,014 (91.8)
Squamous	50 (1.5)	3,392 (98.5)
Other	118 (4.2)	2,693 (95.8)
TNM stage, n (%)			0.001			0.003
III	331 (5.7)	5,512 (94.3)		293 (7.8)	3,463 (92.2)
IV	874 (6.9)	11,752 (93.1)		780 (7.8)	9,240 (92.2)
Unknown	385 (7.4)	4,835 (92.6)		349 (9.5)	3,311 (90.5)
Sample type, n (%)			0.003			0.003
Histological specimen	1,340 (6.9)	18,017 (93.1)		1,204 (8.3)	13,224 (91.7)
Cytological specimen	226 (6.0)	3,536 (94.0)		196 (7.7)	2,337 (92.3)
Body fluid specimens	15 (3.2)	448 (96.8)		13 (3.4)	375 (96.6)
Other	9 (8.4)	98 (91.6)		9 (10.3)	78 (89.7)

• Abbreviation: NSCLC, non-small cell lung cancer.

In the total NSCLC patients, the overall ALK gene rearrangement rates were higher in females (P < 0.001), ≤ 35 years of age (P < 0.001), never smokers (P < 0.001), with tumor cellularity of >50 (P = 0.024), samples from metastatic tumors (P < 0.001), and diagnosed using histological specimens (P = 0.003), similarly in the adenocarcinoma subgroup (Table 1). Supplementary Figure S2 shows the distribution of non-adenocarcinoma cases from the participating centers.

4 DISCUSSION

This retrospective study aimed to comprehensively understand the current status and quality control of ALK test for lung cancer in China (the RATICAL study). The analyses included 23,689 patients with advanced NSCLC who underwent ALK test in 30 medical centers across China. The overall ALK gene rearrangement rate was 6.7% in advanced NSCLC, with higher ALK gene rearrangement rates in patients aged ≤ 35 years, females, never smokers, and patients with adenocarcinoma. The IHC-VENTANA-D5F3 was used to test the ALK gene rearrangements most frequently. The largest number of samples tested were histological samples.

The overall ALK gene rearrangement rate was 6.7% in advanced NSCLC in the present study. This outcome was consistent with that of the previous studies from China (6.6% [9] and 4%-11.6% [19]). Notably, these rates are higher compared to those observed in Western populations, underscoring potential ethnic and regional variations in genetic susceptibility to ALK rearrangements. For example, a study reported that Asian patients exhibited a significantly higher prevalence of ALK gene rearrangement disease at 22.0%, compared to New Zealand Europeans at 4.4% [20]. This suggests that genetic factors may contribute to the higher incidence of ALK rearrangements in Asian populations. The ALK gene rearrangement rates were 31.9%, 8.4%, and 3.0% for patients aged ≤ 35, 36-65, and > 65 years in advanced NSCLC in the present study, respectively. The ALK gene rearrangement rate was 8.3% for advanced NSCLC patients who never smoked, which was higher than for patients who smoked or were smoking (4.2%). The positive detection rate among women was 9.0%, higher than 5.2% among men in the present study. These results were consistent with those from a previous study [10]. The ALK gene rearrangement rate was 8.2% for patients with advanced lung adenocarcinoma, numerically higher than that in patients with advanced NSCLC, indicating a lower positive rate in patients with non-adenocarcinoma. Despite this outcome, several squamous cell carcinoma patients with ALK gene rearrangement have benefitted from ALK inhibitors [16]. Therefore, Chinese clinical experts recommend ALK test for patients with advanced NSCLC (non-adenocarcinoma) to make better decisions on treatment options [21].

The diagnostic guidelines strongly recommend the IHC-VENTANA-D5F3, FISH, RT-PCR, and NGS methods for detecting the ALK fusion gene [17]. Each method has distinct advantages: FISH is highly specific but may have false-negative issues; IHC-VENTANA-D5F3 offers quick screening with high sensitivity; RT-PCR excels in sensitivity for known variants; and NGS not only detects known ALK rearrangements but also identifies concurrent genomic alterations that may impact therapy outcomes [17]. For instance, NGS has shown potential in predicting better disease control and longer progression-free survival in crizotinib-treated patients [22]. This highlights the importance of selecting appropriate diagnostic techniques based on clinical needs and available resources. The IHC methods are commonly used in clinical practice to determine ALK fusions because they are simple, fast, and cost-effective. Four ALK antibodies, such as IHC-VENTANA-D5F3, ALK1, 5A4, and 1A4, are presently available. Among them, ALK1 is less sensitive, 1A4 is less specific, and 5A4 has an unclear cut-off value [23]. Therefore, the IHC-VENTANA-D5F3 is the preferred choice for IHC due to its high sensitivity and specificity [24], and it was used in more than half of the samples analyzed in the present study. However, this test can give false-positive results when non-adenocarcinoma samples are used, warranting experienced pathologists for interpretation [24]. In the previously published first quality control analysis of the RATICAL study, 31 pathologists from 31 nationwide medical centers were selected to read 109 IHC sections. Only 19.4% of the pathologists were 100% accurate in the diagnosis. The false negative rate was 26.5%, positive cases were read as negative by at least one pathologist, and 41.7% of the 60 negative cases were read as positive (false positive) by at least one pathologist. The inconclusive interpretation was 31.2% [18]. These results highly emphasized the heterogeneity in viewing and interpretation among the pathologists on IHC slides, contributing to the variations in ALK gene rearrangement results observed in the present study. Consequently, guidelines have been issued in China to standardize the ALK test and recommend other retesting methods when the IHC-VENTANA-D5F3 results are inconclusive [17, 21].

In this study, 651 patients whose sample underwent testing using the IHC-VENTANA-D5F3 protocol had a concordance rate of 99.2% in the inter-hospital agreement analysis, and 119 samples who underwent testing using other protocols (PCR, NGS, FISH, and IHC) were reviewed by the IHC-VENTANA-D5F3 and resulted in a concordance rate of 93.2%. In addition, the agreement of detection of different methods specifically for those validated cases confirmed by FISH were 98.0% for IHC-VENTANA-D5F3, 98.7% for NGS, and 91.3% for RT-PCR, showing a high concordance rate with the initial tests, but some inconsistencies remained. Indeed, the concordance rate between IHC and RT-PCR was relatively low at 91.3%. This may be attributed to the specificity of RT-PCR, which can lead to false negatives if the specific ALK fusion sequences are not present. Additionally, the potential for IHC to produce false positive results [17], contributes to these discordances. Furthermore, due to the small sample size of discordances, the detailed analysis to track possible reasons is infeasible. In a previous study of ALK test in over 3,000 NSCLC patients, FISH and IHC-VENTANA-D5F3 were used. As a result, 14 patients were read as negative by FISH and positive by IHC but further validated as ALK gene rearrangement by NGS [12]. Therefore, when inconsistencies arise with different test methods, clinicians and pathologists should discuss them to reach a consensus [17].

Regarding selecting specimens for downstream test, tumor tissue samples are highly recommended. When tumor tissue specimens do not meet the requirements or are unavailable, cytology samples can be used. In a few circumstances of advanced lung cancer, when tissue or cytology samples are objectively unavailable, blood or cerebrospinal fluid was used for the diagnosis [21]. In this study, most (more than 80%) of ALK tests were performed on histological samples in clinical practice, followed by cytological samples. Only a few evaluations were performed using body fluid samples. However, a previous study found that the ALK test on cytological specimens was reliable and sensitive [25]. In the present study, about 20% of patients were tested with samples obtained from metastatic foci. While our study observed a higher ALK gene rearrangement rate in metastatic tumors compared to primary lesions, it is important to recognize that the underlying ALK gene rearrangement status remains consistent between primary and metastatic sites. This consistency supports the reliability of using either primary or metastatic tissue samples for detecting ALK rearrangements. The literature corroborates this, showing high concordance of ALK gene rearrangements in both tumor sites, which validates the use of metastatic samples in clinical diagnostic settings when primary tumor tissue is not available [26, 27].

We also analyzed for other driver genes or biomarkers concurrently with ALK gene rearrangement detection and found a small number of cases of mutations, except for high co-positivity with PD-L1. The co-occurrence of ALK gene rearrangement and mutations in other driver genes was rare, as supported by the literature [28], except for the TP53 mutation [29], which was not assessed in the present study. RT-PCR or NGS test is recommended when ALK gene rearrangement and other driver gene mutations are considered [17]. Since co-occurrence of mutations is rare, the treatment options are uncertain [30-32]. Notably, EGFR mutation was more common in women, and KRAS mutation was more common in males, as supported by the literature [33, 34].

ALK test is available in China by IHC and FISH (these methods only target the ALK gene or protein, and therefore, not all patients have EGFR test), but also by RT-PCR and NGS assays that include ALK and multiple other genes (as listed in Tables 4–5). Therefore, the results for these genes are also available when NGS and RT-PCR assays are performed. The results were generally consistent with the mutation rates of these driver genes reported in the literature without new signals [35-37].

In managing advanced NSCLC, comprehensive molecular profiling that includes not only ALK but also ROS1, RET, NTRK, and other actionable gene fusions is critical. However, the practical limitations of conventional methods like ALK IHC necessitate the use of more advanced techniques. At our institution, we have established a versatile diagnostic model that integrates multiple test platforms to accommodate the variable quality of clinical biopsy specimens. When tissue NGS is not feasible due to sample quality issues, we employ ARMS-PCR+IHC/FISH or switch to blood ctDNA NGS test, which allows us to continue detecting driver gene variants despite the challenges [38]. This strategic approach ensures that comprehensive genomic profiling can be achieved, supporting the use of targeted therapies based on a broad spectrum of detected mutations, including common and rare gene loci. Our experiences emphasize the importance of a flexible, multi-platform test strategy to overcome the limitations of tissue sample quality and ensure that molecular diagnostics can inform effective treatment planning for all patients.

This study had some limitations. First, ALK test was based on the clinical reality of each participating hospital. Each hospital might have different test protocols to match their local equipment, hardware, software, and the pathologists' preferences. Nevertheless, the quality control principles are national and were uniformly applied by all centers. Second, it was a cross-sectional study, and the treatment of patients with ALK gene rearrangement (e.g., treatment with ALK inhibitors) was not followed. Third, missing data might have occurred due to the large sample size. Fourth, our study included data from 30 centers across the country, but the standards of ALK test and the testers' experience might have varied from center to center. Fifth, only nine of the 30 centers were performing NGS. Since biopsies provided limited specimens used primarily for histology, immunohistochemistry, and basic genetic test, insufficient material was often left for NGS. Nevertheless, multiple quality controls were performed at all centers to minimize the impact of this bias on the outcomes. Besides, while this study provides comprehensive insights into the test patterns and positivity rates of ALK rearrangements, it does not include clinical outcome data such as overall survival.

In conclusion, we investigated ALK gene rearrangement rates and test patterns in China through this real-world study. The overall ALK gene rearrangement rate among advanced NSCLC in China was 6.7%, while advanced NSCLC patients with ALK gene rearrangement had specific clinicopathological characteristics. The IHC-VENTANA-D5F3 was used most frequently for detecting the ALK gene rearrangement with high consistency. This study provides a basis for detecting ALK gene rearrangements in patients with NSCLC, aiming to optimize future ALK gene rearrangement test patterns in patients with NSCLC.

AUTHOR CONTRIBUTIONS

Lin Li: Data curation; formal analysis; writing—original draft. Lei Guo: Data curation; project administration. Wencai Li, Chunyan Wu, Yanfeng Xi, Yuan Ji, Lili Jiang, Ji Li, Jingping Yun, Gang Chen, Yuan Li, Yueping Liu, Dianbin Mu, Yuchen Han, Leina Sun, Qingxin Xia, Xiaodong Teng, Nanying Che, Wei Wu, Xueshan Qiu, Chao Liu, Xiaochu Yan, Daiqiang Li, Zhihong Zhang, Zhe Wang, Yujun Li, Zheng Wang, Lingchuan Guo, Xiu Nie, Jingshu Geng and Jianhua Zhou: Data curation; project administration; supervision. Jianming Ying: Conceptualization; methodology; funding acquisition; supervision; formal analysis; writing—review & editing.

CONFLICT OF INTEREST STATEMENT

All authors declare no competing interests.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

This retrospective study was approved by the Ethics Committee of the Cancer Hospital of the Chinese Academy of Medical Sciences and reported to the Ethics Committee of all study centers. Informed consent was obtained from all patients.

PATIENT CONSENT STATEMENT

Informed consent was obtained from the patients.