Companion diagnostic assays for PD-1/PD-L1 checkpoint inhibitors in NSCLC

ABSTRACT

The immune checkpoint inhibitors pembrolizumab and nivolumab together with their diagnostic assays have recently been granted market authorization for treatment of advanced non-small-cell lung cancer in the USA. The two assays, PD-L1 IHC 22C3 pharmDx and PD-L1 IHC 28-8 pharmDx (both by Dako, Glostrup, Denmark), are the first PD-L1 IHC assays to obtain regulatory approval through the Premarket Approval process. This approval is supported by recent clinical studies that have shown a positive correlation between PD-L1 expression and the outcome following treatment with different PD-1/PD-L1 checkpoint inhibitors. These diagnostic assays are able to identify the group of non-small-cell lung cancer patients who will benefit most from treatment with the immune checkpoint inhibitors. However, so far, it is only the PD-L1 IHC 22C3 pharmDx assay, which is linked to the use of pembrolizumab, that has obtained regulatory status as a companion diagnostic.

KEYWORDS:

• Companion diagnostics
• complementary diagnostics
• PD-L1
• NSCLC
• atezolizumab
• nivolumab
• pembrolizumab
• personalized medicine
• precision medicine

We have recently seen the regulatory approval of a new class of anticancer drugs; the immune checkpoint inhibitors pembrolizumab (Keytruda, Merck Sharp & Dohme) and nivolumab (Opdivo, Bristol-Myers Squibb). These compounds are antibodies that target the programmed cell death protein (PD-1) that plays a prominent role in modulating the T-cell activity through the interaction with its ligands, PD-L1 and PD-L2.[1] In 2014 both pembrolizumab and nivolumab obtained regulatory approval for treatment of patients with advanced melanoma; however, without a companion diagnostic linked to their use. The reason for this was that PD-L1 expression did not seem to be useful in the selection of melanoma patients, as clinical studies showed similar treatment outcomes in the patient irrespective of a positive or negative PD-L1 expression status.[2,3] Despite some uncertainty as well as data inconsistency with respect to the value of PD-L1 as a predictive biomarker for the immune checkpoint inhibitors in melanoma, the situation seems to be different for patients with non-small cell lung cancer (NSCLC). In October 2015, the US FDA approved two different immunohistochemical (IHC) assays that detected PD-L1 expression in formalin-fixed, paraffin-embedded NSCLC tissue. The two assays, PD-L1 IHC 22C3 pharmDx (Dako) and PD-L1 IHC 28-8 pharmDx (Dako), are linked to the use of pembrolizumab (Keytruda, Merck Sharp & Dohme) and nivolumab (Opdivo, Bristol-Myers Squibb), respectively. The decision made by the US FDA is also supported by recent published clinical data from three larger studies with different immune checkpoint inhibitors.[4–6]

The first larger data set to demonstrate a positive correlation between PD-L1 expression and the treatment outcome in patients with advanced NSCLC came from the KEYNOTE-001 study.[4] It was a large-scale phase I study aimed to evaluate the safety and efficacy of the PD-1 checkpoint inhibitor pembrolizumab and to clinically validate the PD-L1 IHC 22C3 pharmDx assay at the same time. The study enrolled a little less than 500 patients of whom approximately one-third were assigned to a training group and two-thirds to a validation group. The data from the training group were used to define the clinical cutoff for the PD-L1 assay. Based on receiver operating characteristic analysis a proportion score of ≥50% was selected, which corresponds to a PD-L1 membrane expression in at least 50% of the tumor cells. When the treatment outcome data from the validation group were analyzed, applying the selected cutoff value, the results showed that patients with a proportion score ≥50% had a higher response rate and longer progression-free and overall survival than patients with a proportion score <50%. The data from the KEYNOTE-001 study showed a clear link between PD-L1 expression and the efficacy of pembrolizumab, which was also reflected in the indication for use statement as approved by the FDA. Here, it is stated that the PD-L1 IHC 22C3 pharmDx assay is indicated as an aid in identifying NSCLC patients for treatment with pembrolizumab.[7] This type of wording is similar to several other companion diagnostic assays for drugs like crizotinib (Xalkori, Pfizer), vemurafenib (Zelboraf, Roche/Genentech) and trastuzumab (Herceptin, Roche/Genentech).[8]

Data on nivolumab from the CheckMate 057 study have been published very recently, and this showed superiority of the PD-1 checkpoint inhibitor over docetaxel in a group of patients with advanced non-squamous NSCLC.[5] Although the benefit of nivolumab was demonstrated in an unselected patient population, a retrospective analysis of the tumor specimens showed a somewhat greater efficacy in the patients whose tumors expressed PD-L1 compared to the group with no or very low expression. The predictive value of PD-L1, using the PD-L1 IHC 28-8 pharmDx assay, was seen across all the efficacy end points of the CheckMate 057 study at an expression level ≥1%. In the group of patients with PD-L1 expression <1% the efficacy, seems to be similar for the two treatment arms. The retrospective testing for PD-L1 expression was not performed on all patient specimens from the CheckMate 057 study but on a subset of 78%, which must be taken into consideration when the data are interpreted due to the risk of selection bias. Looking at the approved indication for use statement for PD-L1 IHC 28-8 pharmDx, it also seems like the US FDA might have had some reservations in relation to the strength of the clinical data submitted for review to support the assay. The wording in the Premarket Approval states that PD-L1 expression as detected by the PD-L1 IHC 28-8 pharmDx assay in non-squamous NSCLC may be associated with enhanced survival from nivolumab.[9] So currently, this assay has not been approved as a companion diagnostics but rather as a kind of ‘complementary test’.

At the same time as the data from the CheckMate 057 study were published, data on a not yet approved PD-L1 checkpoint inhibitor, atezolizumab (Roche/Genentech), were presented at the European Cancer Congress in Vienna.[6] The data came from the POPLAR study where a mixed population of patients with advanced squamous and non-squamous NSCLC was randomized to receive either atezolizumab or docetaxel. In this study PD-L1 expression was analyzed prospectively by IHC using the SP142 antibody assay (Ventana). In contrast to the assays used in the KEYNOTE-001 and CheckMate 057 studies, the SP142 antibody assay not only assessed the PD-L1 expression in the tumor cells but also on the tumor-infiltrating immune cells. Although a different immune checkpoint inhibitor and a different companion diagnostic assay, the results of the POPLAR study seem similar to the results of the CheckMate 057 study. Also here the primary analysis demonstrated superiority of atezolizumab over docetaxel in the unselected intent to treat population independently of the PD-L1 expression, but similar to the CheckMate 057 study a greater benefit was demonstrated in the group of patients with the highest expression of the biomarker. In the group of patients with no or very low expression of PD-L1 the treatment efficacy of atezolizumab seems to be similar to docetaxel.

The results of the KEYNOTE-001, CheckMate 057 and POPLAR studies demonstrate that the efficacy of the PD-1/PD-L1 checkpoint inhibitors in patients with advanced NSCLC is correlated to the level of PD-L1 expression. That PD-L1 expression matters in relation to the efficacy of PD-1 checkpoint inhibitors in NSCLC was also the opinion of the US FDA when they approved pembrolizumab with its companion diagnostic assay. Despite the fact that PD-L1 expression seems to select a group of NSCLC patients that achieves a favorable outcome following treatment with the different the PD-1/PD-L1 checkpoint inhibitors, there is still room for improvement. The different clinical studies also showed that some patients with low or no expression of PD-L1 also responded to the treatment. Here, it is important to remember that PD-L1 is an inducible and dynamic biomarker and it needs to be considered differently compared to other well-known oncological biomarkers like the human epidermal growth factor receptor 2 (HER2) and the estrogen receptor (ER) in breast cancer.[1,10] The PD-1/PD-L1 checkpoint inhibitors are important contributions in finding more effective treatments against cancer, and it is likewise important that we have companion diagnostics available with a high clinical sensitivity and specificity for these compounds.[11] We need to continue our efforts to find biomarkers that exhibit these characteristics and here analysis of mismatch-repair deficiency and mutation burden might be a supplement to the current companion diagnostics.[12–14]

Financial & competing interests disclosure

The author works as a consultant for Dako/Agilent (Glostrup, Denmark) and Euro Diagnostica (Malmö, Sweden) and has given lectures at meetings sponsored by AstraZeneca (London, UK), Merck Sharp & Dohme Corp (Kenilworth, NJ, USA) and Roche (Basel, Switzerland). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.