Introduction

Lung cancer is the most common cause of cancer-related deaths and the second most frequently diagnosed cancer worldwide [1]. As in the rest of the world, lung cancer is a major health burden in Austria comprising 19% of all cancer deaths. With 2402 male and 1472 female deaths in 2017, it is the most frequent cause of cancer-related deaths in males and the second most frequent cause in females. Furthermore, the lung cancer incidence has shown a steady increase over the last five decades, comprising 11% of all newly diagnosed cancers [2]. Diagnosis and staging of lung cancer is a complex multistep process and the management of these patients requires a multidisciplinary and multimodal approach to determine the optimal treatment strategy, which includes surgery, radiotherapy, immunotherapy, chemotherapy and targeted approaches for tumors with driver mutations as well as individual combinations of all of the above [3].

The identification of driver mutations in a subset of tumors has altered the therapeutic landscape and substantially improved survival outcomes in lung cancer [4]. Therefore, efficient and effective molecular testing of lung cancer specimens is crucial. Due to the increasing number of molecular biomarkers, multiplexed assays such as next generation sequencing (NGS) are used to detect a multitude of drugable mutations in a single workflow [4].

Real-world data on routine practice of lung cancer management in Austria is scarce [2]. Gaining information on the diverse and complex population of lung cancer patients as a whole, however, is crucial as the enrolment of patients in clinical cancer trials has been shown to be highly selective including patients several years younger, with superior performance status and less comorbidities at the start of first line therapy than patients in the clinical setting [5,6,7]. Generalizability of recommendations based on clinical trials is therefore limited, as the results of clinical trials may reflect only outcomes under best-case scenarios. This results in a potential gap of information on patient outcomes especially among older patients or patients with comorbidities. The collection of real-world data complements the results from clinical trials by enabling us to gain insights into the incidence, characteristics and treatment reality of lung cancer patients and to improve healthcare policy decisions [7].

The Landsteiner lung cancer research platform (LALUCA) is a multicenter, clinical registry with prospective data obtained from lung cancer patients, with the aim of documenting data on molecular testing, diagnostic and treatment modalities as well as the course of disease. This article describes in detail the objectives, the methodology used and the implementation of the LALUCA registry.

Patients, material and methods

Population and registry area

The LALUCA registry is based in Vienna, the capital city of Austria. Vienna has a population of approximately 2 million inhabitants, who have free access to high-quality healthcare [8]. Between 2018 and 2020, 1028 cases of tracheal and lung cancer on average have been reported for Vienna to the Austrian National Cancer Registry [9]. The Departments of Respiratory and Critical Care Medicine of the Klinik Floridsdorf and the 2nd Department of Internal Medicine with Pulmonology of the Klinik Ottakring facilitate diagnosis and treatment for a large number of lung cancer patients in Vienna. Within this patient cohort, the LALUCA registry is recruiting patients for study participation.

Objectives

The purpose of this project is to set up a clinical platform to collect representative real-world data on lung cancer management in Austria. We aim to evaluate clinical characteristics, diagnostic and treatment modalities, as well as outcome of unselected patients in real-life practice. In contrast to data of clinical trials, this registry is able to show data on the whole lung cancer patient population including information on older and comorbid patients. This registry focuses specifically on the collection of representative data on molecular biomarker testing and NGS data of patients before the start of first-line treatment, including information on rare mutations and clinical consequences of NGS testing. Another equally relevant objective of this registry is to assess indicators of the quality of care in patients with lung cancer such as timeliness of diagnosis and first treatment application as well as frequencies, modalities and sequences of imaging and treatment. In this respective, in an ongoing investigation we assess the implementation of biomarker testing by analyzing testing rates, timeliness and areas of improvement in molecular testing. By evaluating the incorporation of guideline recommendations, we hope to reveal challenges in molecular testing and highlight areas of improvement. Moreover, investigations of specific therapeutic settings are planned to analyze outcomes in a real-world patient cohort, thereby complementing results from cancer trials and adding information for the development of guidelines and policy decisions.

Study design

The LALUCA registry is registered at ClinicalTrials.gov (NCT04733430) as an observational, prospective, multicenter clinical registry. Overseen by the Karl Landsteiner Institute for Lung Research and Pulmonary Oncology, the LALUCA registry is run by a multidisciplinary team, including pulmonologists, pathologists, oncologists, thoracic surgeons, radio-oncologists, clinical scientists, system engineers and several data entry specialists. The registry is designed as a custom web-based platform created by celsius37TM (celsius37.com AG, Mannheim, Germany) using the push method of data extraction and input by trained staff. Data entry personnel (DEP), each working with a separate user account, extract detailed information from electronic health records (EHR) and transform running text into specific variables that are manually documented into standardized electronic case report forms (eCRF). The eCRF is filled out on the LALUCA website in accordance with good clinical practice (GCP) guidelines, as well as with additionally created LALUCA guidelines ensuring the accuracy, legibility, completeness and timeliness of data documentation.

Recruitment

All patients, who are diagnosed with lung cancer at one of the two centers, are offered participation. Patients are eligible for inclusion if they have a histologically or cytologically confirmed lung cancer, are over the age of 18 years, and sign a consent form. Additionally, patients need to be able to understand and complete the patient reported outcomes assessment instruments. Recruitment for patients started in November 2020 and is ongoing.

Ethics and sponsors

The study was reviewed and approved by the ethics committee of Vienna (EK-20-061-VK). This project is supported by the following industry partners: Amgen, AstraZeneca, Bristol Myers Squibb, Böhringer, Eli Lilly Oncology, Gilead, Janssen, Merck Group, Merk Sharp&Dohme, Novartis, Pfizer, Roche, Sanofi, and Takeda Oncology. Sponsors had no influence on design, methodology, or parameter selection of the registry.

Data collection and measurements

After patients have signed the informed consent form, the LALUCA team is given access to their medical health records relevant to the management of lung cancer. Additionally, patients are required to fill out a case history questionnaire. To enable DEP to gather all relevant information for the eCRF in the LALUCA registry, specifically tailored forms are available to the physicians in either digital or physical form for patient documentation. The collection of patient information starts at the first consultation and includes the patient’s self-reported questionnaires, tailored forms from the physicians, as well as imaging records, pathological, histological and cytological reports from the EHR. All obtained data variables organized by category are shown in the Table 1. As an example for the documentation of specific elements on the platform, collected data items for molecular pathology are displayed in Table 2.

Table 1 Data categories and variables of the LALUCA registry. ECOG eastern cooperative oncology group, FEV forced expiratory volume, DLCO diffusion capacity for carbon monoxide, FVC forced vital capacity, pTNM pathological tumor-node-metastasis, PD-L1 programmed death-ligand 1, TPS tumor proportion score, IHC immunhistochemistry, PCR polymerase chain reaction, FISH fluorescence in situ hybridization, NGS next generation sequencing, EGFR epidermal growth factor receptor, VAMLA,video-assisted mediastinoscopic lymphadenectomy, IP investigational product
Full size table
Table 2 Example of the complete documentation of biomarker diagnostics. PD-L1 programmed death ligand 1, ALK anaplastic lymphoma kinase gene, ROS1 ros proto-oncogene 1, BRAF b-raf proto-oncogene, EGFR epidermal growth factor receptor, MET MET proto-oncogene, RET ret proto-oncogene, NTRK neurotrophic tropomyosin receptor kinase, KRAS Kristen rat sarcoma virus, HER2 human epidermal growth factor receptor
Full size table

Data quality control

To guarantee a sufficiently high quality of the data, various measures were implemented to ensure data accuracy, completeness, capture, data standardization and timeliness [10]. First, all DEP are trained by senior staff members and participate in a monthly training seminar to update their knowledge and to be equipped with all tools necessary to abstract relevant information from the EHR and enter it into the eCRFs. To enhance data accuracy and to reduce the possibility of interpretative errors, the LALUCA guidelines were created containing all the data items, codes and definitions assessed in the registry and therefore serve as a data dictionary during the input process. All DEP have access to the LALUCA guidelines to allow a uniform interpretation of EHR and input into the data registry. Second, the tailored forms filled out by physicians serve the purpose of inputting complete information into the LALUCA registry. Third, input variables are mostly standardized and seldom allow free text. Finally, quality assurance monitoring is carried out by a clinical monitor throughout the whole study and data quality checks are performed periodically using custom-made Python scripts including completeness and extensive plausibility checks, as well as regular statistical reports of the data sets. In cases of discrepancies or missing values, the study sites are contacted to make any necessary changes to the data.

Follow-up

After enrolment, patient data are assessed every 6 months for a follow-up status and an update of ongoing treatment. In cases of loss to follow-up, the insurance provider is contacted for information about the patient’s survival status.

Results

Implementation

The development and implementation of the LALUCA registry is an effort of the Karl Landsteiner Institute for Lung Research and Pulmonary Oncology. The need for a high-quality clinical registry, especially with respect to the collection of representative real-world data on molecular biomarkers in unselected patients, was recognized and a prospective multicenter clinical registry developed. A catalogue consisting of 17 categories with 180 variables was collected based on numerous discussion rounds with specialists from pulmonology, pathology, oncology, thoracic surgery and radio-oncology. After the implementation of a web-based platform according to the custom designed variable catalogue in the first phase of the project, a test run of the system was conducted allowing system optimization and a successful launch of the registry.

Growing participation

During the implementation phase of the registry from November 2020 to December 2020, 50 patients were included. The total number of patients increased rapidly over time and the project is ongoing with a recruitment rate of about 400 patients annually. Patient enrolment across both research sites is displayed in Fig. 1. In 2023 the number of patients included in the registry rose to a total of 1200 patients, who are regularly followed up creating ever growing real-world data.

Fig. 1
figure 1

Evolution of the patient participation in the LALUCA registry

Full size image

Preliminary results

Of 1200 patients included in the registry, 159 patients were excluded from the analysis due to the findings of either synchronous primary lung carcinoma or the tumor being a recurrence, 46% of the patients were female and 54% male. The majority of patients (70%) were 61–80 years old and either current (51%) or former smokers (38%). With 49%, most patients were diagnosed at stage IVA/B (Fig. 2).

Fig. 2
figure 2

Staging at initial diagnosis. ind. indetermined, TNM tumor-node-metastasis

Full size image

In the registry adenocarcinoma was the most prevalent histological subtype with 54.3%, followed by squamous cell carcinoma with 20.5%. The prevalence of all histologic subtypes is shown in Table 3.

Table 3 Prevalence of histological subtypes
Full size table

Project continuation and outlook

The end of 2023 not only marks the inclusion of a total of 1200 patients in the LALUCA registry but also the transition to an improved version with updated variables, upgraded eCRFs and data transfer to an improved IT infrastructure. We used what we have learned so far to improve and simplify data input and management by further standardization and user-friendly drop-down options. Furthermore, to accommodate the increasing load of data, additional positions for DEP were created. Additionally, collaborations with other Austrian sites are in the planning stage.

Discussion

The LALUCA registry continuously recruits unselected lung cancer patients from two high volume sites in Vienna, Austria, since its implementation in 2020. Using a custom web-based platform, comprehensive and detailed information on molecular testing, diagnostic and treatment modalities as well as the course of disease is collected in a real-world clinical dataset.

To date, several registries have been implemented to record and evaluate the burden of cancer in the Austrian population. On a national level, the Austrian national cancer registry serves as a data collection tool for epidemiological purposes. It collects data on incidence, prevalence, mortality, and survival probability of all types of cancer based on the entire population of Austria since its implementation in 1983 [11]. Following the development of the Austrian national cancer registry, regional registries have emerged in Carinthia, Salzburg, Tyrol and Vorarlberg, which collect and provide epidemiological data for the national cancer registry, while also using their data to improve the quality of care in their Federal State [12,13,14,15]. In recent years, hospital-based clinical quality data registers have evolved in Lower and Upper Austria, pooling data from 24 and 11 hospitals, respectively. As Lower and Upper Austria are the Federal States with the second and third highest population in Austria, these registries gather data from areas populated by 1.7 and 1.5 million inhabitants, respectively [8]. Thus, both registries, combining web-based software tools with teams for quality management, provide detailed and comprehensive data on cancer patients in their respective Federal State [16, 17]. The Federal State of Vienna encompasses a population as large as 2 million inhabitants [8]. Between 2018 and 2020 the incidence rates of tracheal and lung cancer were the highest in Vienna with a total of 1028 cases compared to 964 and 728 cases in Lower and Upper Austria, respectively [9]. Within this patient collective, the LALUCA registry currently includes approximately 400 new patients on a yearly basis.

The LALUCA registry stands out among national and international lung cancer registries. In accordance with the current national guidelines all NSCLC patients in the LALUCA registry undergo reflex testing with an extended NGS biomarker panel; however, internationally the practice of reflex testing has been reported in only 14 of 21 European countries [18, 19]. Moreover, only a third of European laboratories use single gene testing for epidermal growth factor receptor (EGFR) mutations followed by further testing with NGS in only 45% of cases [20]. The LALUCA registry offers real-world NGS data on a wide variety of histological subtypes including squamous carcinomas and small cell lung cancer and collects crucial data for the assessment of molecular biomarkers. Furthermore, the registry has an extensive collection of surgical and oncologic treatments, which also allows the investigation of clinical consequences of NGS testing on therapeutic concepts. To the best of our knowledge, these advantages set the LALUCA registry apart from other international registries such as the Danish lung cancer registry, which cannot provide information on curative or palliative treatment intent or specify oncological treatment, or the National Lung Cancer Audit UK, which only requires EGFR mutation status, ALK and ROS1 fusion status and PD-L1 expression for minimal completion of a patient thereby limiting detailed information on molecular biomarkers in their overall dataset [21,22,23]. Additionally, our platform facilitates the collection of key data on specific supportive therapies and the investigation on changes in the diagnostic and therapeutic approach over the course of time in all types and stages of lung cancer in contrast to other international registries which only record the planned primary treatment such as the Swedish National Lung Cancer Register or only include advanced NSCLC patients such as the German CRISP registry [23, 24]. Future directions of our registry include investigations of incorporation of guideline recommendations regarding molecular testing for the assessment of quality of care as well as investigations of real-world patient outcomes in specific therapeutic settings for the complementation of trial results with data from a diverse and complex patient population.

The design of the LALUCA registry has some clear strengths: first, a multidisciplinary team of clinicians was involved in establishing the registry design and determining variables included in the dataset. The engagement of clinicians has shown improvement in the quality and usefulness of cancer registry data by ensuring the collection of relevant variables [26]. Due to the fast changes in the field of oncology, the clinicians’ perspective and a close interdisciplinary collaboration are imperative to understand which data variables are required to assess new clinical practices and to shed light on current treatment paradigms [25, 26]. Second, data entry in the LALUCA registry is performed by specially trained DEP using a data dictionary, specially curated by clinicians, for uniform interpretation of definitions and standardized input of data. Standardization and clear definition of variables reduces high interobserver variability of DEP that has been found to be an issue in other registries [27]. Especially, the involvement of clinicians in the development of data dictionaries for data standardization has been recommended to establish clear data definitions and thereby quality data [25]. Third, in the rapidly progressing field of oncology, the importance for registries to remain flexible and adaptive in order to stay relevant has been demonstrated [26, 28]. All variables included in the registry need to be carefully evaluated in order to be relevant and useful, as too many variables result in an increased workload of text mining and data entry for the registry team without adding additional benefit [28]. Accordingly, the LALUCA registry was updated to include re-evaluated variables, adding the newly relevant while omitting the obsolete. Furthermore, high data quality of the LALUCA registry is guaranteed by quality assurance monitoring and data quality checks which involve the detection and correction of missing or erroneous data.

A limitation of our registry, however, is that while the LALUCA registry recruits a representative sample of lung cancer patients, it does not cover the majority of patients diagnosed with lung cancer in the Federal State of Vienna, yet. The Steering Committee of the registry, though, has already reached out to other high volume lung cancer sites in Vienna and other cities, to further expand data collection.

In conclusion, the LALUCA registry has been designed to create high quality clinical data on lung cancer. With its primary focus on comprehensive NGS data and its high number of patients, it offers an important platform to investigate real-world practice and to evaluate the quality of care for patients with lung cancer.