Factores pronósticos en el cáncer de pulmón no célula pequeña en estadios iniciales

Abstract

Lung cancer (LC) is a devastating disease and a major therapeutic burden with poor survival rates. Nowadays, it is the second most common tumour worldwide and is responsible for the highest number of cancer deaths, with an estimated 1.8 million deaths per year. In fact, in Spain it was the main cause of death by tumour in 2021, as in previous years, with an estimation of 22,438 deaths in both men and women. This Doctoral Thesis focuses on Non-Small-Cell Lung Cancer (NSCLC), which accounts for 75-80% of all LC diagnoses. Most NSCLC patients are still diagnosed in advanced stages of the disease, and, in most cases, only 20-30% of the tumours are identified in early stages with radical local treatment options. Therefore, surgery remains the standard treatment in clinically operable patients, with the aim of achieving a complete resection (R0) of the tumour lesion. Indeed, just the patients whose lesion is potentially resectable should be considered for curative surgery, as there is no evidence of benefit from tumour volume reduction surgery. Despite this curative intent at the early stage of the disease, treatment failure and patient mortality rates remain high, because of distant recurrence. Patient general condition at diagnosis and TNM stage constitute the two basic pillars on which the prognosis and the indication for adjuvant treatment of an interventional NSCLC case are based. Currently, the standard treatment for these patients consists of a two-drug adjuvant chemotherapy regimen (with cisplatin), which results in an overall improvement in absolute survival of 4-5% at 5 years. In recent years, targeted therapies, and immunotherapy (IT) have emerged as an important mean of the disease management for patients with NSCLC, as significant therapeutic achievements have been reached, leading to an increase in overall survival (OS) in this setting. Significant efforts are currently underway to replicate these accomplishments in early-stage NSCLC. However, advances in molecular and clinical understanding of this condition do not yet provide measurable data to predict disease progression and eligibility for adjuvant treatment at diagnosis in an early operable stage. Nevertheless, the concept of neo and adjuvant therapy in LC has been revolutionised exponentially in these years with the incorporation of IT. In this sense, adjuvant therapy has also been enriched with the administration of targeted therapies complementary to the use of standard chemotherapy. Therefore, it has been proved necessary to better understand the prognostic and predictive factors in early-stage NSCLC to have an impact on the evolution of this pathology. Consequently, knowing the clinical, histological, molecular, and therapeutic reality of these patients from a scientific point of view will allow us to apply the new potential advances that are expected in the coming years and will help us to understand the real scope of these innovations in our healthcare practice. In this Doctoral Thesis we review the prognostic impact of the main factors that might influence the evolution of patients with LC. Some of these factors are dependent on the patient, while others are related with the surgical technique or the histopathology of the tumour itself. We found that, despite the current prognostic and predictive evidence, some histopathological, molecular and tumour microenvironment (TME) factors are not yet used in standard clinical practice to make decisions about changes in adjuvant treatment or follow-up in patients with NSCLC who have undergone surgery. Our working hypothesis is that, apart from TNM stage, the tumour histological features, TME and molecular alterations may influence disease-free survival (DFS) and OS of patients with NSCLC, not only in advanced stages of the disease, but also in early stages, and thus, it is important to include these factors in the management of earlystage NSCLC. To address this hypothesis, we included in our study patients who underwent surgery for NSCLC in the University Hospital of Jaén in the period between 2010 and 2013 (four full years) with the following observational, clinical and research objectives: • Describe the histological, clinical and stage characteristics of the patients operated on in this period of time. • Assess the proportion of early-stage diagnoses in our hospital. • Verify the prognostic capacity of TNM in our setting. • Analyse the existence of new TME prognostic factors: tumour-infiltrating lymphocytes (TIL), PD-L1 and transforming growth factor-β (TGF-β) expression, as well as their prognostic implications. • Identify EGFR mutations in all non-smokers, ex-smokers for more than 15 years and in all non-squamous patients (as would be done in advanced NSCLC), to determine the frequency of these mutations and their prognostic implication. • Detect and analyse the types of recurrence and their relationship with OS and DFS. For these purposes, the study population analysed consisted of 94 NSCLC patients evaluated in the Medical Oncology Unit of the University Hospital of Jaén (HUJ) who underwent surgery between 2010-2013, that is, who had resectable tumours and, therefore, were in the initial stages of NSCLC. In a second phase of the study, from the 94 patients who underwent surgery for pathological stages I to IIIA NSCLC at our hospital between 2010 and 2013, 55 samples available for analysis were included to evaluate the presence of TILs and the expression of PD-L1 and TGF. Evolutionary data have been gathered until June 30th, 2018, with a median follow-up period of 5 years from diagnosis. Not only the date of death was recorded, but also the time of recurrence and whether it was local or distant. Second tumour samples have also been collected in those cases in which they have been presented. Our results showed that after a median follow-up of 61.5 months (1-99 months), 55.3% of the study population was still alive. The median expected survival time since surgery was 86 months. Of the total population, 39.4% of patients suffered a recurrence, while the rest had no relapse. In our series, early recurrence has been shown to be a risk factor for death. Thus, patients with early recurrence were 5.78 times more likely to die than patients who did not relapse within 24 months of surgery; Hazard Ratio (HR) = 5.78, Interval Confidence (IC) of 95% (3.06; 10.91). In fact, 87.5% of patients who relapsed within the first two years after surgery died. Interestingly, of all the relapses occurred in our data series, 64.8% were accumulated within the first 24 months. According to these findings, surgery in LC is not curative in all cases, although it has a clear influence on survival. Since around half of the patients who undergo surgery are at high risk of death within 5 years, the development of affordable and reproducible biomarkers has become essential to predict the adjuvant therapeutic efficacy and recurrence rate for these patients, and ultimately, to identify prognostic and relevant factors where action can be taken to increase survival. In this line, we aimed to evaluate the association of clinicopathologic characteristics and the prognostic value of PDL-1 expression and TILs, and explored the immune microenvironment by assessing their relationship, since they have been suggested as clinically applicable predictive biomarkers in surgically resectable NSCLC. Furthermore, we studied TGF-β expression and proposed a more simple and efficient way for its assessment (nuclear staining), since this molecular target has been outlined as a potential prognostic biomarker that may open a new line of research pointed towards TGF-β inhibitor therapies in NSCLC, in combination with ICIs in those patients with a worse prognosis. To assess PD-L1 expression, we used the tumour proportion score (TPS) described in the PD-L1 interpretation manual IHC 22C3 pharmDx-NSCLC; the presence of TIL was analysed by light microscopy by selecting the representative tumour infiltrating area stained with haematoxylin and eosin and in which there were sufficient viable tumour cells without necrosis. The density of intratumoral lymphocytic infiltration was assessed at low magnification (10×) and classified into two categories, 'intense' and 'not intense'. High intensity was defined as intense infiltration equivalent to the density observed in a metastatic lymph node (Brambilla et al., 2016). Regarding TGF-β analysis, apart from the known method with semi-quantitative immunohistochemistry, nuclear staining was assessed as an additional factor, as other studies on pancreatic adenocarcinoma (ADC) considered nuclear staining as a criterion to define TGF-β expression in tumour cells analogous to the interpretation of other biomarkers, such as anti-IDH1-R132H antibody in glial tumours. Nuclear staining analysis of TGF-β was performed considering its unequivocal presence in more than 1% of viable cancer cells as tumour TGF-β positivity. Data from these analyses (Cardenas-Quesada et al., 2022) reflected that neither PDL1 nor TIL appeared as possible risk factors for OS or DFS, even when adjusted for sex, age, and stage. However, when analysing TGF-β nuclear staining, according to the dichotomous variable absent (n = 25) and present (n = 30), statistical significance was found in the DFS curves in favour of absent nuclear staining, with a p-value of 0.045. This result was not observed in OS; nevertheless, significance was maintained when adjusting for age and sex [p-value: 0.044, HRa: 2.832, HRa IC: (1.029-7.794)], and remained very close to significance when adjusting for staging [p-value 0.064, HRa: 2.597, HRa IC: (0.946 - 7.126)]. Therefore, the presence of nuclear staining may be considered as a possible risk factor for relapse in patients operated on for early-stage NSCLC, as the risk is more than twice as high in patients with nuclear staining present than in those in whom it is not observed. EGFR mutations were also analysed in all nonsmokers, ex-smokers for more than 15 years and in all non-squamous LC patients (as would be done in advanced NSCLC). As a result, 58.5% of the population (55 patients out of 94) met these clinical criteria to test if they had any type of EGFR mutation. This study of EGFR mutations was carried out using the Therascreen Kit technique, which analyses mutations in exons 18, 19, 20 and 21 of the EGFR gene by means of real-time PCR, a technique commonly used in clinical practice. Our findings revealed that only 5 patients had a positive result for EGFR mutation. Of these five patients, two were women and three were men, none of them were smokers. All were diagnosed of adenocarcinoma, three of them of the acinar subtype, and all underwent stage I surgery. Two of them had distant recurrence, but none of them had died at the follow-up cut-off date. When analysing OS within the group of 55 selected patients, statistically significant differences were observed in favour of the mutated patients (p-value = 0.028). To date, it remains a challenge to delay or prevent tumour relapse in patients with NSCLC. Although oncological IT with immune checkpoint inhibitors has revolutionised the treatment of this disease, we still lack accurate prognostic and predictive biomarkers to aid in clinical decision making. Our preliminary results suggested that TGF-β staining may predict poor prognosis in early-stage NSCLC patients; since when considering the presence of unequivocal nuclear staining in >1% of viable tumour cells (present or absent) as a possible additional risk factor, DSF curves revealed a clear significance in favour of absent TGF-β nuclear staining. These findings appeared to be promising and may open a new line of research. As a result of its ability to promote tumour development, TGF-β and its signalling pathway offer potential opportunities for targeted therapy. As mentioned in the published article, several agents targeting various components of this pathway have been studied or are being developed and evaluated in clinical trials. However, little is still known about TGF-β expression in general and in NSCLC in particular. In conclusion, as occurs with histological subclassification, none of these TME factors are used in clinical practice for decision-making in early-stage NSCLC: surgical treatment, adjuvant, or follow-up after surgery. However, our review and analysis reinforce the importance of discovering new prognostic factors useful to improve the evolution of LC and, therefore, advocate for including these features in anatomical pathology reports of biopsies and surgical specimens. The scientific evidence with target treatments and IT is leading us to do the same with molecular alterations (EGFR for now) and PD-L1 expression. Current scientific progress would require an effort from all professionals involved in this pathology to be able to early diagnose and ensure a sufficient tumour sample for the analysis of all histological, molecular and TME characteristics that allow for a more personalised treatment and follow-up of each of our patients with resectable LC.