Historically, chemotherapy was the only systemic therapy for advanced NSCLC. Over the last decades, effective targeted therapy has been developed. EGFR is the prime target since 20 years now [1], and is followed by an increasing number of targetable rare oncogenic driver mutations. This decade, the phase I trial of nivolumab – including NSCLC patients – started recruitment on July 1, 2011 [2]. Since then, immunotherapy with immune checkpoint inhibition (ICI) has become the third pillar in our therapeutic armentarium, with a very rapid development for relapsed, and thereafter for untreated advanced NSCLC. Most patients with non-oncogene addicted advanced NSCLC will nowadays have both chemotherapy and ICI, either in a concurrent or sequential way. This raises the question whether their interaction is synergistic or additive, favoring the former versus the latter approach. We will look at this from the perspective of adverse events, mechanistic background, and clinical trial outcomes. There is not much overlap between the typical and common adverse events (AEs) of chemotherapy or ICI. The former are mostly bone marrow inhibition, nausea/vomiting, mucositis, hair loss and neuropathy. ICIs are well tolerated, with no or mild AEs in most patients, but may induce immune related AEs mostly in skin or endocrine organs, liver, lungs or kidneys. Hence, there is little difference in overall adverse event profiles when the chemotherapy+ICI (chemo+ICI) and chemotherapy+placebo (chemo+PL) arms from randomized trials are compared. In the Keynote-189 trial - an example with the anti-PD-1 agent pembrolizumab – overall grade 3 to 5 AEs were present in 67.2% and 65.8% of patients with chemo+ICI and chemo+PL, respectively [3]. Grade 3 to 5 immune-related AEs, as rated by investigators blinded to the assigned therapy, were 8.9% for chemo+ICI, while 4.5% for chemo+PL. AEs leading to discontinuation of all treatment were 13.8% and 7.9%, respectively. Hence, irAEs acted additive to the overall AE profile. The notable exception in this trial was renal toxicity. There was a low frequency of auto-immune nephritis (1.7%), but there was a 5.2% incidence of acute kidney injury in the chemo+ICI arm, compared to 0.5% in the control arm. In the IMpower-133 trial – an example with the anti-PD-L1 agent atezolizumab – overall grade 3 to 5 AEs were present in 67.2% and 63.8% of patients with chemo+ICI and chemo+PL, respectively [4]. All grade immune-related AEs, as rated by investigators blinded to the assigned therapy, were 39.9% for chemo+ICI, and 24.5% for chemo+PL. So, here again a rather additive pattern in AE profile is put forward. Chemotherapy has long time been regarded as immunosuppressive and incompatible with immunotherapy. To improve on the rather low response rate with ICI alone, recent trials focused on the combination of chemo+ICI, trying to exploit the immune modulatory (synergistic) effects of chemotherapy both on the tumor cells and immune cells. The details of this interaction are beyond the scope of this abstract, but a central one is immunogenic cell death (ICD) by chemotherapy. In contrast with necrotic/apoptotic cell death, ICD is characterized by immune-promoting features on dendritic cells and macrophages, by means of inducing calreticulin expression on tumor cells, release of adenosine triphosphate in the extracellular space and of high mobility group box 1 protein from the nucleus of the cancer cell. ICD, however, has been associated with only a limited number of chemotherapeutic agents used in clinical practice, such as doxorubicin, mitoxantrone, oxaliplatin and cyclophosphamide [5]. None of these agents is part of the modern chemotherapeutic armentarium for NSCLC. Recent phase III trials with chemo+ICI versus chemo+PL demonstrated a convincing benefit in response rate, progression-free survival (PFS) and overall survival (OS) with the combination (e.g. [3]). Of note, there are no randomized data comparing the combination of chemo+ICI in a concurrent versus a sequential way. Ideally, such a trial should have several arms: one with the concurrent use of chemo+ICI in all PD-L1 tumors; a sequential one with pembrolizumab followed by chemotherapy in PD-L1 ≥50% tumors; and one with the sequential use of a chemotherapy and then ICI in PD-L1 <50% tumors. The primary endpoint preferably should be PFS2 in a intention-to-treat analysis. In the absence of such data, there is no definitive evidence that one or the other strategy is superior, and we can only make speculations about this question. At the meeting, we will present suggested algorithms based on the comparison of PFS from several recent trials. Especially the PFS2 analysis of the Keynote-024 [6,7] and KN-189 [3,8] trials are helpful in this respect. Even if there are many caveats with this approach, it may be helpful to guide clinical practice between concurrent and sequential treatment strategies.
Read full abstract
Sep 1, 2024
H Yoshioka + 15
Open Access