TNBC has special biological behavior and clinicopathological features, and the prognosis is poorer than that of other types. Meanwhile, the treatment of mTNBC is even more challenging. Traditional TNBC treatment options include only chemotherapy and no other treatments. However, in recent years, an increasing number of studies and experiments have prove that immune checkpoint inhibitors (ICIs) are effective for TNBC, especially for PD-1/PD-L1 positive patients.
We included five studies with a total of 3104 patients with mTNBC and showed that ICIs plus chemotherapy significantly improved PFS not only in PD-L1-positive patients but also in the ITT population. At the same time, in the population with high PD-L1 expression, the addition of ICIs to therapy improved both PFS and OS. However, ICIs combined with chemotherapy did not improve OS in PD-L1 positive patients or in the ITT population. For grade 3-5 AEs, our meta-analysis showed that there was no significant difference in the ICIs group compared with the chemotherapy alone group.
The IMpassion130 [15,22] study was the first to show positive results in the 2018 European Society for Medical Oncology (ESMO) advanced TNBC Phase III clinical trial and was also published in the New England Medical Published in the journal Science. The two primary endpoints were PFS and OS. The results of the first reported study[15] showed that in the ITT population, PFS was 7.2 months in the combination treatment group and 5.5 months in the control group (HR=0.8, 95% CI: 0.69-0.92; P=0.002). In the PD-L1positive population, the benefit was further expanded, and the PFS in the combination therapy group was 2.5 months longer than that in the control group (7.5 vs 5.0 months, HR=0.62, 95% CI: 0.49-0.78; P<0.001). Later data reported [23,24] suggested that ICIs combined with chemotherapy have better PFS, while there was no significant difference in OS. Later, the IMpassion131 [17] study showed that combining ICIs with chemotherapy did not improve PFS or OS versus chemotherapy alone. We believe that these factors may be responsible for the different results between IMpassion130 and IMpassion131. First, the chemotherapeutic drug in IMpassion131 was paclitaxel, while in IMpassion130, it was nab-paclitaxel. At the same time, paclitaxel repuires more use of dexamethasone, which may reduce the efficacy. Second, the ratios of the experimental group and the control group were different between 1:1 (IMpassion130) and 1:2 (IMpassion131),respectively. The IMpassion132 study [28], a multinational double-blind, placebo-controlled, two-arm, randomized phase III trial comparing atezolizumab plus chemotherapyversus placebo plus chemotherapy in early-stage relapsed TNBC, has not been published so far.
The KEYNOTE-355 [16] study demonstrated that pembrolizumab in combination with chemotherapy can significantly and clinically improve PFS compared to the placebo plus chemotherapy group among patients with mTNBC with CPS of 10 or more. These findings suggest that the addition of pembrolizumab to standard chemotherapy is the first-line treatment for mTNBC. The KEYNOTE-119 [26] study showed that pembrolizumab did not significantly improve OS compared with chemotherapy in previously treated mTNBC patients, and in patients treated with pembrolizumab, higher PD-L1 expression (CPS≥20) was associated with a higher long median OS.
The IMpassion130 [15,22], KEYNOTE-119 [26] and KEYNOTE-355[16] studies suggested that different PD-L1 expression has different sensitivities toICI treatment and that overexpression can better improve PFS/OS. Several RCTs have been performed in mTNBC patients treated with ICIs. In the era of monotherapy, the results suggest that these patients can compare the associated clinical benefits of immunotherapy with chemotherapy alone.
In addition, KEYNOTE-012 [29],a nonrandomized phase Ib trial of the single-agent pembrolizumab given to patients with advanced PD-L1 positive advanced TNBC, showed a single-agent pembrolizumab-related ORR (18.5%) for capecitabine that was reported twice as much (9%).
Thus far, chemotherapy combined with ICIs can also improved prognosis in early TNBC. Several RCTs have been conducted in patients with early TNBC [11,30,31], and ICIs plus chemotherapy have a higher PFS/OS than chemotherapy alone. This is especially true in patients who achieve pathological complete response (pCR) after neoadjuvant chemotherapy. KEYNOTE -522[32], a multicenter, randomized, double-blind, placebo-controlled phase III clinical study. A total of 1174 patients with stage II-III TNBC were randomized 2:1 to immunotherapy or placebo plus chemotherapy. Compared with the placebo combined chemotherapy group (201 patients), the pembrolizumab combined chemotherapy group (401 patients) increased pCR by 13.6% (64.8% vs 51.2%, P<0.001); among them, the PD-L1 positive population (CPS ≥1%) and pCR were 68.9% and 54.9% in the two groups and 45.3% and 30.3% in the PD-L1-negative population, respectively[32].
Based on the results of the NeoTRIPaPDL1 study, atezolizumab in combination with chemotherapy did not improve PFS in patients with mTNBC[33]. The ENHANCE 1 study showed that eribulin plus pembrolizumab was generally well tolerated and promised antitumor activity in mTNBC[34].
As the most important immune checkpoint drugs, PD-1/PD-L1 inhibitors have transformed the treatment practice of TNBC, including neoadjuvant therapy and advanced first-line therapy. However, the existing treatment mode still needs to be combined with chemotherapy, and there is still a long way to go to find "precise biomarkers" for predicting immunotherapy and finally achieve "de-chemotherapy".
In the analysis of AEs, their results showed that the addition of ICIs increased the incidence of AEs, while our AE analysis showed different results.
Another meta-analysis [35] included four RCTs; their sample size was smaller than ours (2482 vs 3104), and the result of the PFS in ITT or PD-L1 positive analysis was consistent with ours (HR, 0.82; 95% CI, 0.74–0.90; P <.001/HR, 0.69; 95% CI, 0.59– 0.80; P <.001). Nevertheless,subgroup analysis only analyzed PFS without OS in PD-L1 positive patients, whereas we had a different result.