The results of our study indicated that the linperlisib-GEMOX combination led to an ORR of 53.8% in patients with heavily pretreated DLBCL. The overall safety profile was deemed manageable, with particular attention to the occurrence of diarrhea and pneumonia, both of which were effectively managed. These findings highlight the potential of this combination therapy as a promising new treatment strategy for R/R DLBCL.

PI3K is integral to the regulation of B-cell proliferation and survival. The PI3Kδ isoform, in particular, serves as a pivotal node in the signaling pathways governing B-cell growth and maintenance. Its dysregulated activation is a critical factor in the malignant transformation of B cells [15, 16]. Additionally, the interlinkage between B-cell receptor signaling and PI3Kδ-mediated pathways, as well as other survival networks such as the JAK-STAT pathway, underscores the potential for combined therapeutic effects in B-cell malignancies [17, 18]. Wright et al. [19] proposed the seven subtypes of DLBCL, among which multiple genetic subtypes, including MCD, BN2, ST2, and EZB, exhibit activation of the PI3K signaling pathway. This suggests that these subtypes may benefit from PI3K-targeted therapies.

Despite the potential of PI3K inhibitors, their use as monotherapy in treating DLBCL has shown limited success. For example, the PI3K inhibitor copanlisib resulted in an ORR of 25% in patients with R/R DLBCL, with five out of 40 patients achieving CR [20]. A phase II trial evaluating the efficacy of parsaclisib in DLBCL also reported a modest ORR of 25.5%, leading to early termination due to the lack of sufficient efficacy in interim analyses [21]. In a phase Ib trial of linperlisib, including 24 DLBCL patients, the ORR was 29.2% [22]. In light of these findings, we conducted a phase Ib/II clinical trial to assess the combination of linperlisib and GEMOX in patients with R/R DLBCL. The results demonstrated an ORR of 53.8% in this cohort, with a median PFS of 5.4 months. By the data cutoff date, three patients were still undergoing treatment. This trial included a significant proportion of heavily pretreated patients, many of whom had primary refractory disease, a group typically associated with poor prognosis. In the SCHOLAR-1 study, primary refractory patients had a median OS of only 6.1 months [7]. In contrast, in our study, the ORR for primary refractory DLBCL patients was 43.3%, with a median DOR of 5.4 months, and a median OS of 17.2 months. Notably, the combination of linperlisib and GEMOX improved outcomes in R/R DLBCL without a substantial increase in treatment-related toxicities. These results suggest that linperlisib combined with chemotherapy may provide meaningful therapeutic benefits for patients with R/R DLBCL, warranting further exploration.

Several studies have evaluated the efficacy of GEMOX-based regimens in patients with R/R DLBCL. A phase II trial investigated GEMOX combined with rituximab (GEMOX-R) in 49 patients and reported an ORR of 61%, with a CR rate of 44%, a median PFS of 5 months, and a median OS of 11 months [23]. In another study involving 32 patients treated with GEMOX-R, the ORR was 43%, with a CR rate of 34%, a 1-year PFS rate of 29%, and a 1-year OS rate of 41%, with a median OS of 9.1 months [24]. These data suggest that while GEMOX-R shows efficacy, outcomes remain suboptimal for patients with poor prognostic factors, particularly those who are primary refractory or non-candidates for ASCT. In comparison, the current study demonstrated an ORR of 53.8% and a median PFS of 5.4 months in heavily pretreated patients receiving linperlisib plus GEMOX, with a 1-year OS rate of 65.5%. Notably, this cohort included a significant proportion of patients with primary refractory disease, a population associated with worse outcomes. These findings indicate that the addition of the PI3Kδ inhibitor linperlisib to GEMOX may enhance treatment efficacy. However, a future clinical trial combining linperlisib with GEMOX-R could provide additional insights and potentially establish a more effective salvage regimen for R/R DLBCL.

The combination of linperlisib and GEMOX was found to be safe and well-tolerated in patients with R/R DLBCL. Compared to linperlisib monotherapy in a previous phase I study, there was no increase in the incidence of grade ≥ 3 neutropenia (46.2% vs. 44.0%) or pneumonia (12.8% vs. 16.0%). However, there was an increase in the incidence of thrombocytopenia (41.0% vs. 4.0%) and leukopenia (28.2% vs. 8.0%) [11]. The incidence of elevated ALT and AST levels related to linperlisib was observed in 51.3% and 61.5% of patients, respectively, in this study. However, grade ≥ 3 elevated ALT and AST levels were not reported in the 39 patients who received linperlisib. In comparison, 50.0% of patients (62/125) treated with idelalisib experienced elevated ALT/AST levels (grade ≥ 3 in 13%) [25]. The incidence of diarrhea/colitis was 43.6% (17/39) for any grade, with no grade ≥ 3 cases reported. The median occurrence time of diarrhea/colitis was 4 weeks, and it was mild and responded well to antidiarrheal agents. In a study involving 64 patients treated with copanlisib monotherapy, 16.4% of patients experienced diarrhea of any grade, with 1.5% reporting grade ≥ 3 diarrhea [26]. Similarly, among 146 patients with indolent non-Hodgkin lymphoma receiving idelalisib monotherapy at a dose of 150 mg, 47% had any grade of diarrhea, including terms such as colitis, enterocolitis, and gastrointestinal inflammation. Grade ≥ 3 diarrhea occurred in 14%, with 11% experiencing severe diarrhea. The median time to onset for diarrhea was 1.9 months [27]. In this study, linperlisib-related pneumonia of any grade was observed in 17.9% of patients (7/39), with 12.8% (5/39) experiencing grade ≥ 3 pneumonia. The median occurrence time of pneumonia was 27 weeks. Treatment with oxygen supplementation, anti-infective therapy, systemic corticosteroids, and drug withdrawal resulted in a favorable outcome. Grade ≥ 3 pneumonia was observed in 10.0% (45/465) of patients who received duvelisib and 13.3% (24/181) of patients who received copanlisib in the global study [27]. No interstitial lung disease was observed in our study. The incidence of grade ≥ 3 interstitial lung disease was 15.4% (2/13) in a study of 13 Chinese patients treated with copanlisib monotherapy [27]. Pneumonia was a concern with linperlisib, mainly related to its PI3Kδ target and chemotherapy. However, with cotrimoxazole prophylaxis and anti-infective therapy in this study, pneumonia improved or stabilized in most patients, and lung injury was generally reversible. The absence of grade 3 or higher events, such as AST/ALT elevations and colitis, in this study may be attributed to several factors. Structurally, linperlisib is designed to have increased selectivity for PI3Kδ while minimizing activity against PI3Kγ, potentially reducing off-target immune-related toxicities commonly associated with less selective PI3K inhibitors [12,13,14]. Additionally, linperlisib is primarily eliminated via renal excretion, with fecal excretion serving as a secondary pathway, which may further contribute to its favorable safety profile by reducing gastrointestinal and hepatic accumulation [11]. Population pharmacokinetic and exposure-response analyses of linperlisib have not demonstrated a significant correlation between systemic exposure and the incidence of grade 3 or higher diarrhea [11]. However, the relatively small sample size in this study may have limited the ability to detect rare, severe toxicities. Future studies with larger patient cohorts are essential to comprehensively evaluate the safety profile.

This study has several limitations that must be considered. First, the follow-up period was limited, and the median OS has not yet been reached, which restricts the ability to make definitive conclusions about the long-term effectiveness of the treatment Second, the single-arm design and relatively small sample size may impact the external validity of the findings. Third, tumor response was evaluated using contrast-enhanced CT or MRI based on the IWG 2007 criteria, without the incorporation of positron emission tomography-computed tomography (PET-CT) for metabolic assessment. This approach may have underestimated the ORR, as metabolic imaging provides additional sensitivity in detecting residual disease. The adoption of the Lugano 2014 criteria in future studies, which integrates metabolic evaluations, may yield more accurate response assessments.