This prospective clinical trial provides meaningful insight into the activity of vactosertib and imatinib combination therapy in patients with desmoid tumors. To the best of our knowledge, this is the first study to report a novel targeting agent, a TGF-β inhibitor, in desmoid tumors. Further investigation is needed to confirm the clinical benefits of incorporating this therapy into standard treatment regimens for this rare tumor.
A wide spectrum of treatments are available for desmoid tumors. However, due to the rarity of desmoid tumors, a few prospective trials have addressed the chemotherapeutic options. Because of the toxic side effects of cytotoxic drugs such as liposomal doxorubicin, vinblastine, and methotrexate [10, 13, 28], potential alternative non-cytotoxic agents are highly desired. Tyrosine kinase inhibitors, such as imatinib, sunitinib, and sorafenib, have achieved partial success. As a first-generation TKI, imatinib has modest efficacy (~ 10% ORR), with favorable toxicity profiles [14, 16, 29]. In DeFi trial, a γ-secretase inhibitor showed significant efficacy results in 71% risk reduction of disease progression or death than placebo [19]. Multitargeted TKIs such as sorafenib (ALLIANCE A091105 trial) and pazopanib (DESMOPAZ trial) have meaningful activity (33% ORR to sorafenib, 27% ORR to pazopanib) and 80+% progression free rates after 1 year. In our study, the notable efficacy (25.6% of ORR and 81% progression free rate after 1 year) of vactosertib and imatinib was consistent with previously studied multitargeted TKIs. In addition, it is notable that the radiological response continuously improved, with a median of 7.5 months to best response. Compared to ALLIANCE A091105 (36%) [18] and DeFi (61%) trial [19], more than 70% of participants in our study had already been treated with systemic therapy, with 40% of heavily treated patients (≥ 2 prior). Furthermore, patients with large tumors were enrolled in our study, with a 10 cm median total size of target lesions, compared to 8.4 cm in the previous sorafenib study. In addition, synergistic efficacy was also seen in imatinib-refractory patients (n = 4), with one cPR and two patients achieving long-term SD.
In this trial, the TGF-β inhibitor vactosertib, in combination with imatinib, was generally well tolerated. Mild constitutional symptoms, including myalgia and fatigue, were the major adverse events. Grade 3–4 neutropenia or anemia occurred in ~ 20% of patients but were manageable without clinically significant complications. In the ALLIANCE A091105 trial, lower doses of sorafenib were administered at 400 mg daily, which is 50% of the licensed dose, while the DESMOPAS trial used the standard dose of pazopanib (800 mg daily). Dose modification and discontinuation due to toxicities (29.5% and 7.4% respectively for vactosertib, 25.9% and 3.7% respectively for imatinib) were much lower for our study treatment than the ALLIANCE A091105 (31% dose modification and 20% discontinuation) and DESMOPAS trials (73% dose reduction and 8% discontinuation). Compared to pazopanib and sorafenib, less toxicities may enable to achieve long term stabilization via less frequent dose reduction and/or withdrawal. Furthermore, the possibility of interaction between vactosertib and imatinib was evaluated in our pharmacokinetic study. Although the pharmacokinetics of vactosertib was affected by imatinib, its clinical significance may be limited considering the toxicity profiles in the present study. Additionally, the PFS of treatment-naïve patients in our study is very promising, not only as a salvage therapy, but also as a front-line therapy alternative. Our clinical findings support the need for future randomized trials that compare our combination therapy with other agents in a front-line setting.
Despite being the recommended standard treatment for many cancers, the underlying mechanisms of multi-tyrosine kinase inhibitors remain unknown. In the DESMOPAZ study, PDGF receptor-like protein and thrombospondin-4 were highly expressed. Thrombospondin-4, a proangiogenic factor, stimulates tumor growth via TGF-β signaling [30]. In addition, TGF-β and connective tissue growth factor, downstream effectors of Wnt/β-catenin signaling, are highly expressed in desmoid tumors [31, 32]. Serine protease inhibitors, including plasminogen activator inhibitor-1, were upregulated in desmoid tumors [33] and increased tumor growth and invasion in a TGF-β-dependent manner [33]. Therefore, TGF-β may be a potential therapeutic target for desmoid tumors, and combined treatment with TGF-β and imatinib shows efficacy and synergy compared to monotherapy in preclinical models19, [34]. Indeed, our clinical study validated that combination treatment was effective and well-tolerated for desmoid tumors.
In our exploratory analysis, the TGF-β signaling enrichment score was not statistically significant. However, various canonical and non-canonical pathways related to TGF-β signaling were significantly enriched in responders. Upon DNA damage, ATM activates TGF-β signaling, which subsequently maintains genomic stability through DNA repair pathways [35, 36]. TGF-β and unfolded protein response pathways are associated with the secretion of extracellular matrix proteins and epithelial-mesenchymal transition [37, 38]. Meanwhile, considering the immunosuppressive role of TGF-β signaling within the tumor microenvironment [39], tumors enriched with immune responses may not depend on TGF-β signaling and are not responsive to TGF-β inhibitors, as shown in our study. Therefore, further investigation of TGF-β inhibitors with correlative markers in large cohorts might help elucidate the efficacy and mechanisms underlying clinical responses.
This study has several limitations. The study design did not include comparison of the direct effects of the combination regimen with other agents. Additionally, progression according to RECIST was not mandatory for inclusion in our study. However, more than two-thirds of the participants in our study had previously been treated and enrolled after disease progression according to the RECIST criteria. Furthermore, although randomized trials for rare diseases are generally difficult, we were able to rapidly compile the necessary clinical data based on a multicenter study. This was made possible by the high unmet need for novel therapeutic options for desmoid tumors. In our study, the majority of patients had been treated for more than 1 year, but the appropriate duration of combination treatment remains to be determined. Four patients categorized as long-term responders are in the resting period, and long-term follow-up results for those may elucidate more information on appropriate treatment durations.
In conclusion, the combination of TGF-β inhibitor and imatinib showed promising efficacy with favorable safety profile in the desmoid tumors. Based on the promising response rates, further investigation with a randomized trial is needed to confirm these findings. This novel therapeutic approach may provide a silver lining for this rare and disabling disease.