In recent years, the therapeutic landscape for NSCLC has undergone rapid transformation due to a large number of new, innovative and well-tolerated approaches, particularly in the field of immunotherapy and targeted therapies (8, 26, 27). These achievements have not only influenced the therapy regime for patients with advanced tumors, but are also used as adjuvant and neoadjuvant approaches in early stages with favorable chances of a long-term cure (28–30). Efforts have been ongoing to enhance the rate of early-stage initial diagnoses, gain insight into the tumor biology aggressiveness, and identify individuals who may benefit from a more intensive therapy regime or closer follow-up care. These endeavors have led to repeated attempts to discover suitable liquid biopsy-based biomarkers having diagnostic and prognostic significance in patients with NSCLC (31, 32).
The circulating proteins CYFRA (cytokeratin fragment) and CEA (carcinoembryonic antigen, glycoprotein), which are among the earliest biomarker candidates for NSCLC, have been examined in several prior studies with the outcome that elevated plasma levels of these circulating proteins are showing a trend towards a poorer prognosis (33–38). Furthermore, the presence of circulating tumor cells (CTCs), which play a role in the formation and development of metastases, has been associated with poorer prognosis in patients with NSCLC (39–41). Most recently, a large study has delivered the first promising results regarding the potential of small RNAs in liquid biopsies for tumor detection, albeit with a low sensitivity in early stages (42). Despite these encouraging findings regarding suitable biomarker candidates, none of the candidates have yet been established as an integral part of the diagnostic or therapeutic process in the existing guidelines for the treatment of thoracic cancers. In the context of the upcoming introduction of lung cancer screening in Germany using low-dose CT, there is a significant clinical need for more specific biomarkers that could facilitate the interpretation of inconclusive results and better inform clinical decision-making.
In our study, we investigated the ability of circulating eHsp70 to predict disease stage, aggressiveness and prognosis of patients with lung cancer across different stages of disease. With the established Hsp70-exo sandwich ELISA which uses the mHsp70-specific monoclonal antibodies cmHsp70.1 (aa 451–461) and cmHsp70.2 (aa 614–623), it is possible to quantify both free Hsp70 (released predominantly by dying cells) and Hsp70 bound to lipid microvesicles (primarily derived from viable cells) in the blood (21). The results of previous studies demonstrating significantly higher eHsp70 levels in the circulation of cancer patients compared to healthy controls (21, 25, 43) are consistent with our present findings in patients with NSCLC across all stages and patients with lung metastases of different extrathoracic primary tumors.
As a molecular chaperone, cytosolic Hsp70 maintains protein homeostasis by supporting protein folding, unfolding and transport in all nucleated cell types under physiological conditions (44). In tumor cells, Hsp70 is frequently overexpressed and can contribute to a more aggressive and therapy-resistant tumor phenotype by interfering with apoptosis and proliferation pathways (45–47). Many studies have shown that apart from its cytosolic localization, Hsp70 can be presented on the plasma membrane of tumor cells via a mechanism which is enabled by a tumor-specific lipid composition (48). Furthermore, an elevated mHsp70 expression is associated with higher tumor stages and a more aggressive behavior in various female cancers, prostate cancer and brain tumors (25, 49, 50). Tumor cells expressing mHsp70 actively release eHsp70 into the extracellular space (17). Significantly increased eHsp70 levels have also been shown in patients with NSCLC compared to healthy donors (43, 51). Herein, we found a progressive increase in circulating eHsp70 levels starting from healthy donors, to low-grade NSCLC through locally advanced and metastatic NSCLC. The ROC analysis generated an AUC value of 0.83, with a specificity of 0.78 and a sensitivity of 0.73 at an eHsp70 cut-off value of 49.48 ng/mL for the presence of locally advanced or metastatic NSCLC compared to the healthy cohort. These findings further support the crucial function of eHsp70 in mediating tumor aggressiveness in lung cancer and the significant expression of Hsp70 in advanced tumors.
A second indication of Hsp70’s role as a marker for aggressiveness emerged from the detailed examination of the surgically treated cases and their histopathological results. Although some studies suggest an association between circulating eHsp70 levels and the tumor volume (52, 53), we found no significant association of circulating eHsp70 concentrations and the largest pathologically measured NSCLC diameter after surgical resection. However, our results revealed a clear difference (***p = 0.001) in the eHsp70 concentration between patients with pathologically proven lymph node metastases and those without. In the guidelines for the preoperative diagnosis of possible lymph node involvement and mediastinal staging in patients with NSCLC, PET-CT plays a decisive role as a non-invasive method. The reported literature states that the sensitivity of PET-CT for the detection of lymph node metastases is between 0.74 to 0.85, with a specificity ranging from 0.85 to 0.92, but with low accuracy in the detection of micrometastases in affected lymph nodes (54–58). Invasive methods such as mediastinoscopy, EBUS examination or surgical lymph node dissection can provide representative histological results (59). To our knowledge, currently no established liquid biomarker exists for the detection of lymph node metastases in the pretherapeutic setting. The ROC analysis of preoperative eHsp70 in patients undergoing surgery with curative intent yielded an AUC value of 0.82 for the presence of pathological lymph node metastases with a sensitivity of 0.8 and a specificity of 0.78 at an eHsp70 value of 40.45 ng/ml. Preoperative PET-CT results of patients with histologically confirmed lymph node metastases showed a correct clear positive finding in only 6 of 15 cases and no evidence for the presence of lymph node metastasis in 5 cases. In addition, the group with affected lymph nodes and elevated eHsp70 values exhibited a higher percentage of tumors with pathologically higher and thus more dedifferentiated grading, further suggesting a potential role of Hp70 as a marker of aggressiveness. Moreover, we could show that patients with carcinoma precursors had lower eHsp70 levels than patients with malignant tumors. In summary, we assume eHsp70 measured in the blood before start of any treatment could make a significant contribution to more precise preoperative mediastinal staging, especially in unclear cases and in combination with unclear results in PET-CT.
In addition to NSCLC eHsp70 levels were also measured in patients with lung metastases derived from a different primary tumor than NSCLC. Like patients with locally advanced, metastatic NSCLC, these patients also exhibited significantly higher eHsp70 values compared to healthy controls, thereby indicating that eHsp70 could provide a universal tumor biomarker not only for NSCLC, but also for other tumor entities.
The third interesting evidence for the potential of eHsp70 as a tumor biomarker of aggressiveness in thoracic oncology is provided by our follow-up data. Previous studies have already shown a link between elevated eHsp70 levels and a tendency to early tumor relapse or treatment failure (25, 49, 60). In our study, we could clearly demonstrate that patients who had an early NSCLC relapse in the first 6 months showed significantly higher preoperative eHsp70 values compared to patients without evidence of an early tumor relapse. Moreover, the eHsp70 values of 2 patients which were additionally determined at the time of the first follow-up after 3 months showed, that already at this early time point, the patient with recurrence showed a significantly higher value than the patient without recurrence. These observations align with our recently published study, wherein eHsp70 levels rose within three months post-surgery in early-relapse patients, whereas they remained stable in relapse-free patients (51).
The predictive value of eHsp70 for the occurrence of tumor relapse could also be found in patients with thoracic metastases derived form an extrathoracic primary tumor within the first 12 months following pulmonary metastasectomy. Measuring eHsp70 levels at multiple time points during follow-up and during adjuvant therapy could further confirm the results of this study and provide interesting and dynamic clues for possible recurrence and treatment response. The immunophenotyping of the peripheral blood lymphocytes of patients with lung cancer did not provide conclusive results as a predictive tumor biomarker for tumor aggressiveness. Irrespective of the stage and tumor aggressiveness all patients suffered from a lymphopenia and the number of immunosuppressive regulatory T cells were found to be significantly increased compared to healthy controls. An increase in CD3-/CD56+/CD94+/CD69+/NKp30+/NKp46 + NK cell subsets which correlated with a significant drop in CD4 + T helper cells, appeared to be restricted to patients with lung metastases of an extrathoracic primary tumor. Since reduced CD4 + T cell counts are associated with decreased levels of pro-inflammatory cytokines in the circulation [43], we assume that despite an increase in some NK cell subsets, these effector cells are unable to mediate protective anti-tumor immunity in advanced tumors.
Since we found no obvious association between circulation eHsp70 levels and tumor size, we hypothesize that the tumor aggressiveness (lymph node metastases, early relapse), rather than the size of the tumor, is responsible for the elevated circulating eHsp70 levels. Considering the increased detection of incidental pulmonary nodules by lung cancer screening, our easily measurable biomarker in liquid biopsies could contribute to the management of such unclear cases. In instances where elevated eHsp70 values are associated with ambiguous results, early histological confirmation or closer follow-up could be considered. On the other hand, measuring eHsp70 not only at the time of diagnosis, but also during the treatment could help to identify patients experiencing treatment failure and thus guide the selection of alternative treatment strategies or more intensive approaches. This is especially relevant given the increasing complexity and diversity of available modalities.