ASCO 2017 Annual Meeting Highlights

Published: June 14, 2017

The Society for Immunotherapy of Cancer (SITC) is pleased to present highlights of the latest advances in immunotherapy emerging from the American Society of Clinical Oncology (ASCO) Annual Meeting 2017, held in Chicago, Illinois on June 2-6, 2017.

Please click on the heading below to read the corresponding highlights:


AE=adverse event

BID=twice daily

BTK=Bruton’s tyrosine kinase

CAR=chimeric antigen receptor

CR=complete response

DCR=disease control rate


MEK= Mitogen-activated protein kinase kinase

MRD=minimal residual disease

OS= overall survival

PFS= progression-free survival

PR=partial response


SD= stable disease

TCR=T cell receptor

Combination ublituximab + ibrutinib increases ORR compared to ibrutinib alone in patients with genetically high-risk chronic lymphocytic leukemia

Describing the current treatment landscape for patients with genetically high-risk chronic lymphocytic leukemia (CLL), Dr. Jeff Porter Sharman (Willamette Valley Cancer Institute and Research Center/US Oncology Research) explained that despite the availability of ibrutinib, patients with high-risk molecular features rarely achieve remission. Dr Sharman outlined the phase 3 GENUINE trial, which evaluated the safety and efficacy of anti-CD20 antibody ublituximab (UTX) added to ibrutinib (IBT), versus IBT alone, in pre-treated patients with CLL characterized by deletions of 17p or 11g, or a P53 mutation (Abstract 7504). Following enrolment difficulties that necessitated a modified trial design, ORR remained a primary endpoint but PFS was re-categorized as a secondary endpoint. Patients (n = 126) were randomized 1:1 to UTX+IBT (900mg IV on days 1, 8, and 15 during cycle 1, day 1 during cycles 2 to 6, then every 3 months if disease was stable) or IBT monotherapy (420 mg QD). Treatment arms were well matched (median age 67 yrs, approx. 70% male; 60% ECOG 0/1, median of 3 prior treatments), the only notable difference being a higher percentage of patients with bulky disease in the combination arm (45% vs. 26%). The study met its primary endpoint, demonstrating that combination therapy with UTX+IBT significantly increased ORR compared with IBT alone (78% vs. 45%; p<0.001). When data were re-analyzed to account for patients with PR and lymphocytosis (due to previous treatment with BTK inhibitors), ORR increased to 83% in the UTX+IBT group, and 59% with IBT alone, which remained statistically significant. In all, 7% of UTX+IBT patients achieved CR and 71% PR; the IBT group had no CRs and a PR rate of 45%. The proportion of patients with MRD-negative status (19% vs. 2%; p<0.01) and reduction in nodal size >75% (66% vs. 52%) were both higher with UTX+IBT than IBT alone. The trend towards longer PFS in the combination arm was not statistically significant. With the exception of infusion-related reactions, the addition of UTX to IBT did not alter the AE profile.


PD-L1 blockade may augment response to CD19-targeted CAR T cell therapy in relapsed/refractory pediatric acute lymphoblastic leukemia (ALL)

In a session focused on immune-based strategies beyond PD-1 blockade, Dr. Shannon Maude (Children’s Hospital of Philadelphia, PA) presented data demonstrating the benefit of adding pembrolizumab to augment response to CAR T cell therapy in relapsed childhood acute lymphoblastic leukemia (ALL) (Abstract 103). In a trial designed to evaluate whether pembrolizumab could overcome T cell exhaustion to induce persistent response in patients with waning disease control (NCT02374333), Dr. Maude’s group treated six patients with relapsed/refractory ALL following chemotherapy, with a humanized CD19-directed CAR T cell (HuCART19), monitoring for MRD, B cell aplasia and HuCART19 persistence. Patients with poor persistence after initial HuCART19 infusion had the option of reinfusion with subsequent addition of pembrolizumab for continued poor T cell persistence. Since CD19-directed T cells eliminate healthy B cells, B cell recovery was used as a marker of declining CD19 T cell activity. Among the five patients who received HuCART19, there was 1 partial response, 3 cases of poor persistence, and 1 non-response. In the 3 patients with poor persistence, manifest by early B cell recovery, subsequent re-infusion with addition of pembrolizumab every 3 weeks resulted in prolonged persistence and continued B cell aplasia in 2 patients, and non-response with CD19+ minimal residual disease in 1 patient. In one patient with widespread lymphomatous disease, CART19 infusion led to good CART19 proliferation with complete bone marrow response at 28 days. Subsequent treatment with pembrolizumab for 2-3 weeks significantly reduced PET-avid lesions. Summarizing, Dr. Maude emphasized that poor CAR T cell persistence is associated with risk of relapse, explaining that immunogenicity may not fully explain why some patients respond to retreatment despite poor T cell persistence.  This small case series demonstrated that PD-1 blockade can safely be used to augment CD19 CAR T cell infusion and may generate prolonged CAR T cell persistence and objective clinical response.


Pembrolizumab shows promise in liposarcoma and undifferentiated pleiomorphic sarcoma; pre-treatment cytotoxic T cell tumor infiltration and PD-L1+ status enhance outcomes

Melissa Burgess (University of Pittsburgh Physicians) presented data from SARC028, a multicenter phase 2 study of single agent pembrolizumab in advanced sarcoma, a disease for which few treatments have generated durable responses (Abstract 11008).  Supplementing results delivered at ASCO 2016, Dr. Burgess presented updated efficacy data and recently analyzed PD-L1 tumor expression (22C3 clone) and immune infiltrates by multispectral immunohistochemistry analysis (Vectra). Patients with soft tissue sarcoma [STS: leiomyosarcoma (LSM), liposarcoma (LPS), synovial sarcoma (SS), undifferentiated pleiomorphic sarcoma (UPS), n= 10 in each group; median age 53 yrs; 64% male] or bone sarcoma (BS: osteosarcoma (OS) n=22; Ewing’s sarcoma (ES) n = 13; chondrosarcoma (CS); n = 5; median age 33 yrs; 62% male), were heavily pretreated (>80% had received ≥ 2 therapies). In the STS group, pembrolizumab 200mg IV every 3 weeks led to RECIST 1.1 CR in 1 (2.5%) patient with UPS, and PR in 6 patients (15%; 3 UPS, 2 LPS, 1 SS). Median PFS and OS in patients with STS were 18 weeks and 49 weeks, respectively. In patients with BS, the same pembrolizumab dose regimen led to PR in 2 patients (1 OS, 1 CS), with no CRs; median PFS and OS were 8 weeks and 52 weeks, respectively. Treatment was generally well tolerated with AEs as anticipated based on existing experience with pembrolizumab. Only 3/70 tumor samples, all in UPS, were positive for PD-L1 (>1%), of which 2 samples were evaluable and showed 1 CR and 1 PR. Treatment responders had a greater number of infiltrating CD3+ and CD8+ T cells pre-treatment compared to non-responders (n =28 paired samples). Although sample size was relatively small, these data demonstrate promising activity for pembrolizumab in certain sarcoma subtypes, particularly UPS and LPS, with increased likelihood of response in patients with cytotoxic T cell tumor infiltration prior to starting treatment, and in patients with PD-L1+ UPS. An expansion study in UPS and LPS is planned to build on these findings.


Anti-glucocorticoid-induced tumor necrosis factor receptor (GITR) agonist induces encouraging clinical response in combination with a PD-1 inhibitor (nivolumab), in patients with advanced solid tumors

Lillian Siu (Princess Margaret Cancer Center, Canada) presented data from the first clinical phase I/IIa study to investigate combination treatment with an anti-glucocorticoid-induced tumor necrosis factor receptor (GITR) agonist (BMS-986156, a fully human IgG; ‘anti-GITR’) and a PD-1 inhibitor (nivolumab), in patients with advanced solid tumors (Abstract 104). GITR is a costimulatory activating receptor that is upregulated on activated T cells and expressed at higher levels in regulatory T cells (Treg) than effector T cells (Teff). Anti-GITR binding increases Teff survival and functioning while reducing the immunosuppressive effects of Treg-mediated Teff suppression, and converting Tregs to other immune cells. High levels of GITR expression on intratumoral Tregs may lead to their preferential depletion through mechanisms including antibody-dependent cell-mediated cytotoxicity (ADCC). Based on evidence of synergistic activity between anti-GITR and anti-PD-1 in a mouse model, Dr Siu’s group compared escalating doses of anti-GITR alone (10-800mg IV, Q2W) and in combination with nivolumab (anti-GITR 30-800mg IV Q2W plus nivolumab 240mg IV Q2W) in 66 patients with advanced solid tumors (NCT02598960). Baseline demographics and treatment exposure were similar between groups (median age 55-60 years; >30% of patients had received ≥ 3 prior therapies; treatment duration approx. 7-18 weeks). Incidence of treatment-related AEs was somewhat higher in the combination anti-GITR+nivolumab arm (26/37; 70%) than the anti-GITR monotherapy arm (17/29; 59%) but there was no dose-limiting toxicity and only 1 grade 3/4 treatment-related AE (combination arm), indicating a similar safety profile to nivolumab monotherapy. Anti-GITR with or without nivolumab demonstrated linear pharmacokinetics with dose-related exposure and low incidence of immunogenicity. The experimental combination also increased proliferation of Ki67+ NK and CD8+ effector cells in peripheral blood at 8 days, and increased proliferation and activation of C8+ effector and central memory cells. One patient with nasopharyngeal carcinoma who had received 3 previous lines of therapy achieved PR with 43% maximum reduction in tumor burden. These preliminary data suggest the potential for clinical benefit even in patients with disease progression during or after anti-PD-1 therapy, and warrant continued evaluation of anti-GITR in combination with anti-PD-1 therapy.

Combination nivolumab+ipilimumab shows high antitumor activity when given first-line in patients with melanoma brain metastases

Georgina Long (Melanoma Institute Australia/University of Sydney, Australia) opened her presentation on the open-label phase II Anti-PD1 Brain Collaboration (ABC) trial by explaining the unmet need for treatment in the 20-25% of patients with stage IV melanoma brain metastases, the majority of whom do not respond to available systemic treatment (Abstract 9508). The ABC trial randomized patients with asymptomatic melanoma brain metastases and no previous checkpoint inhibitor or local brain therapy (BRAF or MEK therapy was allowed) to one of three trial cohorts. Cohorts A (n=33) and B (n=27) were asymptomatic with no prior local brain metastases whereas cohort C (n=16) included patients who were neurologically symptomatic and/or had leptomeningeal disease, with brain metastases that were unresponsive to local therapy. Patients in cohort A received combination therapy consisting of nivolumab (1mg/kg) + ipilimumab (3mg/kg) Q3W x4, followed by nivolumab (3mg/kg) Q2W whereas cohorts B and C both received nivolumab monotherapy (3mg/kg) Q2W. Among patients in cohort C, 50% had >4 brain metastases, 75% had received prior BRAF or MEK inhibitors and 81% were BRAF-positive. ORR (RECIST v. 1.1) was 42% (15% CR, 27% PR), 20% (12% CR, 8% PR) and 6% (0 CR, 6% PR) in cohorts A, B and C, respectively, and median duration of response was not reached in any arm. Combination nivolumab + ipilimumab increased median PFS (4.8 months) and intracranial 6-month PFS rate (46%) compared with nivolumab alone (2.7 and 2.5 months, and 28% and 13%, in cohorts B and C, respectively). Treatment-related AEs occurred in 96% of combination therapy patients compared to 68% and 56% of patients in cohorts B and C; of these, 46%, 24% and 19% were Grade 3/4, respectively. Summarizing, Dr. Long emphasized that combination nivolumab + ipilimumab demonstrated high antitumor activity when given first line, and was much less effective following BRAF or MEK therapy, supporting this combination as first-line therapy in patients with melanoma brain metastases. The efficacy of nivolumab monotherapy was low following multimodality treatment or in patients with leptomeningeal melanoma. Based on these results, the investigators are planning a trial of radiotherapy in patients with PD as best response following treatment with combination nivolumab + ipilimumab.


Phase II InMotion150 trial of atezolizumab (anti-PD-L1) + bevacizumab (anti-VEGF) shows efficacy in metastatic renal cell carcinoma

Michael Atkins (Georgetown-Lombardi Comprehensive Cancer Center, Washington D.C) provided updated data from the phase II InMotion150 trial, which aimed to boost antitumor immunity by combining VEGF and PD-L1 blockade in patients with locally advanced or metastatic renal cell carcinoma (mRCC) (Abstract 4505). Patients (n = 305 treatment-naïve patients with mRCC) were randomized 1:1:1 to first-line treatment with sunitinib (50mg, 4 weeks on, 2 weeks off), atezolizumab (1200mg IV Q3W) or atezolizumab (1200mg IV)+bevacizumab (15 mg/kg) Q3W. Patients with biopsy-proven progressive disease on monotherapy could cross over into the combination arm. Data previously reported at the 2017 Genitourinary Cancers Symposium showed a strong trend towards improved PFS for the combination compared to either monotherapy arm for patients whose tumors contained ≥ 1% of immune cells expressing PD-L1. Bevacizumab appeared to provide greatest benefit when added to atezolizumab in patients whose baseline tumor biopsies demonstrated gene signatures characteristic of both high T effector cells and myeloid inflammation. In patients who crossed over to receive atezolizumab + bevacizumab (n = 95), Dr. Akins reported a 26% ORR (all partial), 24% post-atezolizumab and 28% post-sunitinib, with 72% of responses ongoing (median duration of response not reached). Median PFS was 8.8 months in crossover patients (12.6 post-atezolizumab, 8.3 post-sunitinib); patients who did not respond to first-line therapy had considerably shorter median PFS (7.6 months) than first-line responders (20.1 months). Safety of the combination treatment was consistent with safety profiles of each individual treatment. The clinical efficacy of atezolizumab + bevacizumab will be further evaluated in the phase III InMotion 151 trial.


Durable remission demonstrated with BCMA-specific CAR T cell therapy in patients with relapsed/refractory multiple myeloma

Wanhong Zhao (Second Affiliated Hospital of Xi’an Jiaotong University, China), presented data from the first 35 patients in an ongoing phase I trial of a proprietary B cell maturation protein (BCMA)-targeting CAR T cell in relapsed/refractory multiple myeloma (Abstract LBA3001). The trial enrolled heavily pre-treated adults (median age 55 yrs, 71% Durie-Salmon stage IIIA, 100% ≥3 prior treatment regimens including bortezomib) with biopsy-confirmed multiple myeloma per updated International Myeloma Working Group (IMWG) criteria, and BCMA expression > 50% on malignant plasma cells (NCT0303090659). Patients received a median of 4.7 (0.6 ~ 7.0) × 10e6/kg CAR T cells, in 3 split doses over one week, after a conditioning regimen of cyclophosphamide (300 mg/m2 dx3). Evidence of activity appeared as rapidly as 10 days after the first injection. ORR was 100%, and within 2 months 33 patients (94%) had evidence of clinical remission (CR, very good PR, or PR). Median follow-up was 208 days. Of the 19 patients followed for ≥ 4 months, 14 achieved CR, 1 PR, and 4 others have achieved very good partial remission. There was no evidence of relapse among patients who achieved CR; of the 5 patients who have been followed for 12-14 months, all remained in CR with no MRD. One patient with a very good partial remission experienced disease progression: an extramedullary lesion that reappeared 3 months after complete disappearance on CT scan. In all, 85% of patients experienced a cytokine release syndrome that was transient and manageable with supportive care in all cases. There were no neurologic side effects reported. Based on these very encouraging results, the researchers plan to enroll 100 patients in the trial in China, and open a similar trial in the U.S. in 2018.


Novel carcinoembryonic antigen (CEA) CD3 T-cell bispecific antibody demonstrates preliminary efficacy alone and in combination with the PD-L1 inhibitor atezolizumab, in patients with metastatic colorectal cancer (mCRC)

Josep Tabernero (Vall d’Hebron Institute of Oncology and CIBERONC, Spain) reported preliminary data from two ongoing phase I studies of a novel carcinoembryonic antigen (CEA) CD3 T cell bispecific antibody, alone and in combination with the PD-L1 inhibitor atezolizumab, in patients with metastatic colorectal cancer (mCRC) (Abstract 3002). The flexible bispecific antibody (CEA-TCB) simultaneously binds to CD3 on T lymphocytes and CEA on tumor cells, activating T cells in close physical proximity to the tumor cells. CEA-TCB was given weekly in both trials; atezolizumab was administered IV Q3W. In the phase 1 dose escalation, patients who received CEA-TCB monotherapy (IV QW; n = 59 patients ≥40mg CEA-TCB) or CEA-TCB +atezolizumab 1200mg IV Q3W (n = 33 patients ≥40mg CEA-TCB), >90% in both trials developed AEs although most were transient Grade 1/2 events. AEs ≥ Grade 3 were reported in approx. 38% of patients for both monotherapy and the combination at CEA-TCB doses ≥40mg , most commonly infusion reactions (24% vs. 12% with monotherapy vs. combination therapy, respectively) and diarrhea (7% vs. 18%). In all, 31 patients with metastatic CRC received CEA-TCB monotherapy at doses ≥60 mg and 36 patients received the combination with CEA-TCB doses ranging from 5-160 mg. Both patient cohorts were heavily pre-treated (40-50% prior adjuvant therapy, 55-70% ≥ 3 prior therapies) and ≥90% in both studies had microsatellite stable (MSS) disease. In the monotherapy group receiving CEA-TCB at doses ≥60 mg, treatment produced 6% PR and 39% SD, per RECIST 1.1.  Several patients with best response as SD experienced tumor regression. Among the 36 patients receiving the combination, investigators observed 14% PR and 47% SD. However, in the 11 evaluable patients receiving the combination at CEA-TCB doses of 80 or 160 mg, 8 patients experienced reduction in target lesions, including 2 partial responses, and 6 (55%) are progression-free at 8-24 weeks. The results suggest promising activity of this combination in MSS mCRC.


Epacadostat, in combination with immune checkpoint inhibitors, shows deep and durable responses in treatment-naïve patients with melanoma, and patients with squamous cell carcinoma of the head and neck

Raymond Perez (University of Kansas, KS) and Omid Hamid (The Angeles Clinic and Research Institute, CA) presented preliminary data from two clinical trials evaluating the selective oral indoleamine 2,3 dioxygenase-1 (IDO-1) inhibitor, epacadostat, in combination with anti-PD-1 antibodies nivolumab (Dr. Perez, Abstract 3003) or pembrolizumab (Dr. Hamid, Abstract 6010). IDO-1 is an IFNγ-induced intracellular enzyme that catalyzes the degradation of tryptophan in the kynurenine pathway, supporting an immunosuppressive tumor microenvironment. As such, epacadostat may reinstate immune surveillance and, in combination with immune checkpoint inhibition, improve patient outcomes. The phase I/II ECHO-204 and ECHO-202 (KEYNOTE-037) trials are investigating epacadostat (25-300mg PO BID) in combination with nivolumab (phase 2 dose: 240mg IV Q2W) and pembrolizumab (200mg Q3W), respectively, in advanced solid tumors. Patients previously treated with anti-PD-1 or anti-CTLA-4 therapies were excluded (except first-line anti-CTLA-4 for metastatic melanoma in ECHO-204). Baseline data varied across disease and dose cohorts. Analysis of phase II ECHO-204 data revealed an ORR of 63% (5% CR, 58% PR) and DCR 88% in 40 treatment-naïve melanoma patients receiving epacadostat 100mg or 300mg, in combination with nivolumab, with responses ongoing and median duration of follow-up 16+ weeks. Among 31 patients receiving the same treatment combination for SCCHN, ORR was 23% (3% CR, 19% PR) with DCR 61%, irrespective of HPV status or level of PD-L1 expression, with responses ongoing and median duration of follow-up 24+ weeks. The combination did not show efficacy in unselected patients with ovarian cancer or colorectal cancer. Data from ECHO-202/KEYNOTE-037, revealed ORR of 34% (8% CR, 26% PR) and DCR 61% in 38 patients receiving epacadostat and pembrolizumab for metastatic or recurrent SCCHN, again showing no effect of PD-L1 expression or HPV status. In this study, 10/13 (77%) responses were ongoing with median duration of follow-up 18.4+ weeks. In both ECHO studies, the most common treatment-related AEs were fatigue and rash, and AEs were generally manageable with supportive care. Safety and tolerability were consistent with previous experience of the checkpoint inhibitors as monotherapy. Based on these deep and durable responses, and acceptable tolerability, combination therapy with epacadostat and checkpoint inhibitors will move into phase 3 trials in treatment-naive patients with melanoma and patients with SCCHN. 


Nivolumab, with or without ipilimumab, as 2nd- or 3rd- line treatment improves disease control and overall survival in mesothelioma

Opening his presentation, Arnaud Scherpereel (University of Lille, France) explained the lack of effective second line treatments for patients with malignant pleural mesothelioma (MPM) whose disease progresses on first-line pemetrexed platinum doublet. Dr Scherpereel reported late-breaking data from the multicenter phase II MAPS-2 trial of second- or third-line nivolumab ± ipilimumab in patients with malignant pleural mesothelioma (MPM) (Abstract LBA8507). After unexpectedly rapid recruitment, patients with biopsy-proven MPM and documented disease progression on chemotherapy that included pemetrexed/platinum doublet were randomized to nivolumab (3mg/kg IV Q2W) alone (n = 63) or in combination with ipilimumab (1mg/kg IV Q6W; n = 62). Treatment-related AEs occurred more frequently in the combination therapy arm (87% vs. 78%; Grade 3/4: 18% vs. 10%); events that were significantly more frequent with combination therapy were diarrhea (20% vs. 6%) and pruritus (12% vs. 2%; both p=0.04). There were 3 deaths in the combination arm (hepatitis, encephalitis and acute renal failure). Both nivolumab monotherapy and nivolumab + ipilimumab reached the primary endpoint of DCR at 12 weeks (44.4% and 50.0%, respectively) in the first 108 evaluable patients. Both regimens also led to longer median OS than all previous reports in this setting (10.4 months and not reached with nivolumab monotherapy and nivolumab + ipilimumab, respectively). Progression-free survival was better in combination therapy patients compared to monotherapy (median PFS 5.6 vs. 4.0 months) with median follow-up of 10.4 months. In light of the favorable tolerability overall, particularly with nivolumab monotherapy, single agent nivolumab, alone or in combination with ipilimumab, may provide new second/third-line treatment options for patients with MPM. 


Diversity and composition of the gut microbiome influence response to anti-PD-1 immune checkpoint therapy in patients with metastatic melanoma

Building on evidence that the gut microbiome may influence treatment response and durability  following cancer immunotherapy, Jennifer Wargo (MD Anderson Cancer Center, TX) and colleagues evaluated bacterial signatures in oral (n=233) and fecal (n=115) samples before and after immune checkpoint therapy in patients with metastatic melanoma (Abstract 3008). Samples underwent microbiome sequencing (16S and whole genome), molecular profiling (whole exome sequencing) and immune profiling (IHC, flow cytometry, TCR sequencing and cytokine analysis) to evaluate bacterial diversity and composition. Patients who responded to anti-PD-1 therapy had greater diversity in their gut microbiome than non-responders (p<0.01), and a preponderance of Clostridia species (Ruminococcaecae and Faecalibacteria), whereas non-responders were colonized with more Bacteroidales (p<0.05). At median time to follow-up of approximately 240 days, patients with greater gut microbiome diversity had not reached median PFS whereas those with intermediate and low diversity had median PFS of 232 and 188 days, respectively (both p<0.05). Abundance of Faecalibacterium prausnitzii was associated with significantly longer PFS (p<0.05) whereas greater colonization with Bacteroidales led to significantly shorter PFS (p<0.05). Differences in the diversity and composition of the oral microbiome had no effect on outcomes. Analysis of the association between the gut microbiome and antitumor response demonstrated higher CD8+ cell density in the tumor microenvironment (TME) of responders than non-responders (p<0.05) and an abundance of Ruminococcaecae and Faecalibacteria was associated with greater density of CD8+ cells in the TME (r2=0.42; p=0.007). Concluding, Dr. Wargo raised questions that will require further research, such as whether patients should have microbiome profiling before embarking on immunotherapy, whether antibiotic use should be limited in such patients, and whether probiotics are indicated. A clinical trial of fecal microbiome transplant will open later this year to evaluate how modulation of the gut microbiome influences response to checkpoint blockade.


  • The publication yesterday of  the first clinical guidelines  by the Society for Immunotherapy of Cancer on management of side effects associated with immunotherapies really felt momentous. ...

  • It was great seeing old friends again and meeting new ones at the 32 nd Annual Meeting & Pre-Conference Programs in National Harbor! We appreciate all who joined SITC for our most successful ...

  • The cited 2017 case report states " The restoration of the cellular immune response against tumor cells through PD-1 blockade and CRLA-4 blockade  have recently been shown to be associated ...

  • This is an interesting subject to me since a few years back we published a study in the American Journal of Transplantation that showed blocking mTOR with a low dose of rapamycin enhanced ...