A New Target in Lung Cancer: CD39 Blockade Shows Promise in the MATISSE Study
In this article
Study: MATISSE Phase 2 (NCT05742607) · Presenter: Prof. Dr. Fabrice Barlesi (Institut Gustave Roussy & Université Paris-Saclay)
Perioperative combinations of chemotherapy + immune checkpoint inhibitors have revolutionized the treatment of non-small cell lung cancer (NSCLC) in recent years, yet 5-year event-free survival still remains only around 50%. The MATISSE Phase 2 study presented at the AACR 2026 Annual Meeting focused on a new target to move beyond this limit: CD39, a key enzyme in the adenosine pathway. When IPH5201, a first-in-class anti-CD39 monoclonal antibody, was added to durvalumab and platinum-based chemotherapy in patients with operable stage 2-3A NSCLC, a striking increase in pathological complete response (pCR) rates was observed: in tumors with PD-L1 ≥50%, this rate reached %50. This may signal the emergence of an exciting new mechanism in oncology.
CD39 and the Adenosine Pathway: A Hidden Layer That Silences Immunity
In the era of immunotherapy, the question “why do some tumors not respond to PD-L1 inhibitors?” has become one of oncology's most critical research topics. Part of the answer lies in the adenosine pathway.
The mechanism works as follows: in healthy tissues, extracellular adenosine triphosphate (ATP) is an important signal of cell damage and activates the immune system. In the tumor microenvironment, however, CD39 first breaks this ATP down into adenosine monophosphate (AMP), after which CD73 converts the AMP into adenosine. Adenosine accumulating around the tumor has precisely the opposite effect: it silences immune cells, suppresses T-cell activation, and blunts the effectiveness of PD-L1 inhibitors.
🔬 KEY CONCEPT: WHY DOES RESISTANCE TO IMMUNOTHERAPY DEVELOP?
High levels of adenosine in the tumor microenvironment effectively tell the immune system “be quiet, there is no problem here”. CD39 controls the first step of this suppressive mechanism. This enzyme is overexpressed in NSCLC tissue — the tumor actively uses this pathway to escape immunity. Blocking CD39 with IPH5201 theoretically interrupts this “be quiet” signal and strengthens the effect of PD-L1 inhibitors.
Immune suppression in the tumor microenvironment and points for pharmacological intervention
The MATISSE Study: Design and Participants
MATISSE (NCT05742607) is a multicenter, single-arm phase 2 study sponsored by Innate Pharma in collaboration with AstraZeneca. A preliminary analysis included 40 patients from centers in Europe and the US. Participants were previously untreated patients diagnosed with resectable stage 2/3A NSCLC. The study assesses the pathological complete response (pCR) rate and safety profile as its primary endpoints.
Pathological Complete Response Rates: A Striking Trend
The study's primary results reveal a clear trend toward benefit, particularly in tumors with high PD-L1 expression:
In addition to these findings, better pCR rates were also observed in tumors with high baseline CD39 expression — suggesting that CD39 might not only be a “therapeutic target” but also a potential biomarker capable of predicting treatment response. The study will continue enrolling only patients with PD-L1 ≥1% tumors to confirm this trend.
Comparison with AEGEAN: One Step Ahead?
Because MATISSE is single-arm, it is not a direct comparative study; however, the AEGEAN study is used as a reference for an indirect comparison. AEGEAN is the phase 3 study that led to FDA approval of durvalumab in combination with platinum-based chemotherapy for resectable NSCLC. The comparison is as follows:
Pathological complete response rates by PD-L1 subgroup — an indirect comparison of different study populations
The chart reveals a striking pattern: in the overall population, the MATISSE arm's %27,5 pCR rate is clearly above AEGEAN's 17,2%. Little difference is seen in the PD-L1-negative group (8,3% vs 9%), but in the PD-L1 ≥50% subgroup, MATISSE's %50 is almost twice AEGEAN's 27,5%. This finding suggests that IPH5201 strengthens durvalumab's effect, particularly in PD-L1-positive tumors.
But an important caution: MATISSE is a single-arm phase 2 study reporting preliminary results from 40 patients. The comparison with AEGEAN is indirect — the patient populations, time periods, and center profiles differ. A randomized phase 3 study is required to confirm this trend.
NeoCOAST-2 and Competition in the Adenosine Pathway: CD39 or CD73?
Presented around the same time as MATISSE, the NeoCOAST-2 study, targeting CD73, the next enzyme in the adenosine pathway, tested oleclumabin a similar perioperative design. In the study of 74 patients, the overall pCR rate was reported as 20,3% — lower than MATISSE, but in PD-L1-negative tumors the pCR rate was %16 , higher than MATISSE's 8,3%.
Commenting on the two studies, thoracic oncologist Tina Cascone from MD Anderson Cancer Center highlighted an important point: “We are working with small numbers, but targeting different components of the adenosine pathway may provide different benefits depending on PD-L1 expression.” CD39 and CD73 blockade may therefore be complementary, rather than alternatives to each other. Indeed, blocking both together stands out as a strategy that could be considered in the future.
- Target: CD39 enzyme (the first step in the pathway)
- Increases ATP accumulation (immune-stimulating)
- Overall pCR: 27,5%
- PD-L1 ≥50%: 50% pCR (strong response)
- PD-L1 negative: 8,3% pCR
- Developer: Innate Pharma + AstraZeneca
- Target: CD73 enzyme (the final step in the pathway)
- Directly inhibits adenosine formation
- Overall pCR: 20,3% (n=74)
- PD-L1 negative: 16% pCR (higher than MATISSE)
- Lower than MATISSE in PD-L1-positive subgroups
- Developer: AstraZeneca
Safety Profile: No New Signal
No new safety signals were observed in MATISSE. Adverse events related to IPH5201 were reported in approximately three quarters of patients. Events occurring in 10% or more were asthenia (weakness), arthralgia (joint pain), hypothyroidism, and thrombocytopenia (a reduced platelet count). This profile does not differ dramatically from the known adverse-effect profile of durvalumab and platinum-based chemotherapy — suggesting a reasonable safety window for CD39 blockade.
Expert Commentary: The Future of the Field
Barlesi gave a clear message about the study's future direction: MATISSE opens the way for CD39 and, more broadly, adenosine pathway inhibition in early-stage non-small cell lung cancer. Improvements in pathological complete response rates are observed according to tumor PD-L1 expression. To confirm this trend, MATISSE will continue enrolling patients with PD-L1 ≥1%.
Implications for Practice in Turkey and Future Possibilities
Perioperative durvalumab + platinum-based chemotherapy (the AEGEAN regimen) is increasingly becoming standard in the treatment of resectable NSCLC in Turkey. However, reimbursement coverage and variation in practice between centers still pose significant barriers. MATISSE's results will not directly affect Turkish practice today — because IPH5201 is not yet approved and remains at the clinical trial stage.
Nevertheless, three important strategic messages emerge. First, major thoracic oncology centers in Turkey should prepare to participate in future phase 3 studies — important both for scientific contribution and for our patients' early access to innovative treatments. Second, pathology infrastructure should be strengthened for routine assessment of PD-L1 and CD39 expression — CD39 immunohistochemistry in particular may become a treatment-guiding biomarker in the future. Third, we should anticipate the growing importance of the adenosine pathway in oncology practice and update educational materials for both clinicians and patients.
🎓 DROZDOGAN Academy — Commentary and Discussion
CD39 BLOCKADE · ADENOSINE PATHWAY · LUNG CANCER · MATISSE PHASE 2 · AACR 2026
The Importance of This Study
MATISSE seeks a concrete answer to a question that remains unresolved in the immunotherapy era: “Why do some NSCLC tumors not respond to PD-L1 inhibitors as well as expected?” Preclinical evidence that the adenosine pathway plays a central role in this vulnerability had accumulated for a long time; MATISSE is among the first major studies to test this hypothesis in humans with resectable disease. A first-in-class molecularly targeted treatment raising the pCR rate from 27,5% to 50% in the PD-L1 ≥50% subgroup — if confirmed in phase 3 — would represent a new layer of NSCLC treatment. We are seeing an important mechanistic principle translated into human data, not only for lung cancer but also for other solid tumors.
Strengths
A strength of the study is that it includes not only clinical efficacy data but also comprehensive biomarker analyses. Alongside subgroup analyses by PD-L1 expression, the positive correlation between baseline CD39 expression and pCR reveals the potential for a predictive biomarker indicating who may benefit most from treatment. Moreover, the safety profile not differing dramatically from the underlying chemotherapy-immunotherapy combination is an important finding supporting progression to phase 3 studies. Presentation alongside NeoCOAST-2 offers an elegant research framework suggesting that different points of the adenosine pathway may be effective in different patient groups. The multicenter international collaboration based at Gustave Roussy and led by Barlesi contributes to the quality of the data.
Limitations and Open Questions
Several important limitations emerge on critical assessment. First, MATISSE is a single-arm phase 2 study with no randomized placebo/control arm. The comparison with AEGEAN is indirect and may involve systematic bias arising from different patient populations. Second, the sample of 40 patients is limited, especially for subgroup analyses — the “50% pCR” figure in the PD-L1 ≥50% group is subject to wide confidence intervals because of the small sample. Third, the primary endpoint is pCR, but what really matters in lung cancer is long-term survival and event-free survival data. While pCR is a valuable intermediate endpoint, longer follow-up is necessary for long-term benefit. Fourth, the study was funded by Innate Pharma and AstraZeneca, and the principal investigator has relationships with these companies — although this is an inevitable reality, it makes independent phase 3 confirmation even more important. Finally, why benefit remained limited in PD-L1-negative tumors — whether the adenosine pathway acts together with another immune escape mechanism in these groups — remains an important question to answer.
Implications for Clinical Practice
For clinical practice, today's message is clear: IPH5201 is not yet an approved treatment and cannot enter routine use. However, there are important implications for thoracic oncology centers in Turkey. First, the existing AEGEAN regimen (perioperative durvalumab + platinum chemotherapy) should be used as standard in resectable stage 2-3A NSCLC — but it should be remembered that more than half of patients relapse even under this regimen. Second, rigorous measurement of PD-L1 expression (using the 22C3 antibody in experienced pathology centers) is critical for predicting treatment response; adding CD39 IHC may be considered in the future. Third, referral of suitable patients to phase 3 studies should be considered. This development has considerable educational value for oncology practitioners: new targets such as the adenosine pathway, TIGIT, and LAG-3 will increasingly be discussed in the coming years as the “second generation” of immunotherapy. The most important practical message is that pathological complete response in early-stage NSCLC remains a strong predictor of long-term survival — every strategy that increases pCR matters, but whether that increase translates into survival must be shown in follow-up studies.
Questions for the Future
The critical questions for the coming period are: will MATISSE's PD-L1 ≥1% expansion cohort (and a possible phase 3 study) confirm the pCR advantage under randomized conditions? Will CD39 expression enter routine pathological assessment as a predictive biomarker, and which IHC platforms will become standardized? Could combinations targeting different points in the adenosine pathway (CD39, CD73, the A2a receptor) yield better results, and in which PD-L1 subgroups will these combinations be most beneficial? Will combinations of adenosine pathway inhibitors with other new immune checkpoint inhibitors such as TIGIT and LAG-3 be considered? Will survival results reflect the pCR advantage, or will we face the paradox “pCR is high but survival is unchanged” — does this problem, repeatedly encountered in oncology in recent years, apply to MATISSE as well? Specifically for Turkey: when will adenosine pathway inhibitors be integrated into our country's clinical trial network, and how will access to these innovations be opened for our patients? The answers will shape the next decade of perioperative treatment in NSCLC.
Scientific Sources
- Barlesi F et al. (MATISSE). Dual CD39 and PD-L1 inhibition: Interim results from the Phase 2 MATISSE trial of IPH5201 plus durvalumab and platinum-based chemotherapy in patients with resectable NSCLC. AACR Annual Meeting 2026, Abstract CT231. Presentation date: 21 April 2026.
- Medscape Medical News. CD39 Blockade Shows Potential in Early-Stage Non-Small Cell Lung Cancer. M. Alexander Otto, 22 April 2026.
- Innate Pharma. MATISSE Phase 2 Interim Results: IPH5201 in Resectable NSCLC. Press release, 17 April 2026.
- Heymach JV et al. (AEGEAN). Perioperative Durvalumab for Resectable Non-Small-Cell Lung Cancer. N Engl J Med 2023;389:1672–1684.
- NeoCOAST-2 study team. Oleclumab (anti-CD73) combined with durvalumab and chemotherapy in resectable NSCLC: perioperative results. Presented: AACR 2026.
- Allard B et al. The adenosine pathway in anti-cancer immunity: CD39 and CD73 as emerging therapeutic targets. Nat Rev Clin Oncol 2020;17:611–629.
- Vijayan D, Young A, Teng MWL, Smyth MJ. Targeting immunosuppressive adenosine in cancer. Nat Rev Cancer 2017;17:709–724.
- ClinicalTrials.gov. MATISSE: Perioperative IPH5201 with Durvalumab and Chemotherapy in Resectable NSCLC. Identifier: NCT05742607.
This article was created using various editorial tools, including artificial intelligence, as part of the preparation process. It was reviewed by human editors before publication.
Was this article helpful?
Your feedback helps us improve our content.
Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.