Phase I/II Study of the Mammalian Target of Rapamycin Inhibitor Everolimus (RAD001) in Patients with Relapsed or Refractory Hematologic Malignancies

Materials and Methods

Patients. Patients with histologically confirmed advanced, relapsed, or refractory AML or MDS, chronic lymphocytic leukemia (CLL), T-cell leukemia, myelofibrosis, or mantle cell lymphoma (MCL) were eligible for enrollment.

Treatment protocol. Patients were treated with 5 (first three patients) or 10 mg orally once daily continuously either in a fasted state or after a light fat-free meal. One treatment cycle consisted of 28 days of therapy. If emesis occurred after ingesting everolimus, the dose was not repeated. Novartis Pharmaceuticals (East Hanover, NJ) provided everolimus film-coated tablets at strengths of 5 mg. Treatment was stopped if any of the following events occurred: unacceptable toxicity, intercurrent illness or change in the patient's condition that rendered the patient unsuitable for further therapy in the judgment of the investigator, or patient withdrawal from the study.

Assessment of toxicity and response. All patients were evaluable for toxicity if they received at least one dose of everolimus. Adverse events were assessed at each visit and graded according to the National Cancer Institute Common Toxicity Criteria version 3.0. Hematologic dose-limiting toxicity (DLT) was defined as myelosuppression with bone marrow hypoplasia (cellularity <5%) without evidence of leukemia for >42 days. Nonhematologic DLT was defined as any grade 4 toxicity, except for the following: grade 3 nausea or vomiting, drug-related fever of any grade, alopecia, and grade 3 lipid or electrolyte abnormalities.

Sample collection. After obtaining written informed consent, whole-blood samples were collected into 10 mL sodium heparin tubes (Becton Dickinson, Sparks, MD) at baseline and 24 hours after administration of everolimus during the first cycle of therapy.

Western blot analysis. For the Western blot analyses, peripheral blood mononuclear cells were lysed in phosphoprotein lysis buffer (10 mmol/L NaF, 1 mmol/L Na₃VO₄, 150 mmol/L NaCl, 1 mmol/L MgCl₂, 1 mmol/L CaCl₂, 0.1% NaN₃, 10 mmol/L iodoacetamide, 3 mmol/L phenylmethylsulfonyl fluoride, and 1% Triton X-100) supplemented with a protease inhibitor cocktail (Roche Diagnostics Corp., Indianapolis, IN). Equal amounts of proteins were then separated on a 12% polyacrylamide gel, transferred to Hybond-P membranes (Amersham Pharmacia Biotech, Little Chalfont, United Kingdom), immunoblotted with specific antibody, and further visualized by using enhanced chemiluminescence detection system (Amersham Pharmacia Biotech). Western blots were analyzed on a Storm 860 system by using ImageQuant software (Molecular Dynamics, Sunnyvale, CA). Integrated absorbance of each lane (band obtained with the phosphorylated-specific antibody) was recorded, and protein levels were determined after normalizing for levels of corresponding total protein or glyceraldehyde-3-phosphate dehydrogenase. The antibodies for phosphorylated p70S6K (Thr³⁸⁹) and total p70S6K, phosphorylated Akt (Ser⁴⁷³)/(Thr³⁰⁸) and total Akt, phosphorylated S6RK and S6RK, phosphorylated 4E-BP1 (Thr⁷⁰)/(Thr³⁷/Thr⁴⁶) and 4E-BP1, and phosphorylated extracellular signal-regulated kinase (ERK; p44/42) were obtained from Cell Signaling Technology, Inc. (Danvers, MA) The antibody for ERK was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and the antibody for glyceraldehyde-3-phosphate dehydrogenase was from Chemicon International (Temecula, CA).

Quantitative real-time reverse transcription-PCR. Total RNAs were prepared using Trizol reagent as described by the manufacturer (Life Technologies, Inc., Gaithersburg, MD). Total RNA (1 μg) was reverse transcribed by avian myeloblastosis virus reverse transcriptase (Roche Diagnostics) under standard conditions. Duplicate samples of 1 μL of each cDNA were amplified by PCR in the ABI Prism 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA). The amplification reaction mixture (25 μL) contained cDNAs, forward primers, reverse primers, probes, and Taqman Universal PCR Master Mix (PE Applied Biosystems). β₂-microglobulin was coamplified as an internal control to normalize for variable amounts of cDNA in each sample. The thermocycler variables were as follows: 50°C for 2 minutes, 95°C for 10 minutes, 40 cycles of 95°C for 15 seconds, and 60°C for 1 minute. Taqman gene expression assay for cyclin D1 (assay Hs00277039_m1), cyclin D2 (assay Hs00277041_m1), and Glut1 (assay Hs00197884_m1) was purchased from PE Applied Biosystems and used as directed by the manufacturer. Results were collected and analyzed to determine the PCR cycle number that generated the first fluorescence signal above a threshold (threshold cycle, C_t; 10 SD above the mean fluorescence generated during the baseline cycles). The abundance of each transcript of interest relative to that of β₂-microglobulin was calculated as follows: relative expression = 100 × 2^−ΔCt, where ΔC_t is the mean C_t of the transcript of interest less the mean C_t of the transcript for β₂-microglobulin.

Statistical analysis. At least three patients had to be observed for at least 4 weeks during their first cycle of treatment with 5 mg daily without a DLT. With each DLT, three additional assessable patients had to be accrued, and further escalation could occur only if no more DLTs were observed. In the absence of such toxicity, the highest dose level (i.e., 10 mg daily) was to be used for phase II testing.

For the phase II part of the study, the primary objective was to estimate the overall response rate (responses were defined as per standard criteria; refs. 18–21). As everolimus has a noncytotoxic mechanism of action, a response rate of 10% in a very poor prognosis group of patients was considered of sufficient interest to warrant additional investigation. A maximum total of 25 evaluable patients were to be entered in each group of patients with advanced or refractory disease: (a) AML or MDS, (b) myelofibrosis, and (c) CLL, T-cell leukemia, or MCL. This sample size would yield an 82% posterior credibility interval for probability of response of width ∼0.16. Interim analyses were to be done in each group after 14 patients had been evaluated for response. Thus, in each patient cohort, the trial would be terminated after the first 14 patients if no responses were observed. As a further safeguard to prevent exposure of patients to an agent with minimal activity, the overall response rate was reviewed periodically. If overall response rate was <1 in 26 patients, 2 in 42 patients, or 3 in 55 patients, this would indicate that the probability of an overall response rate of <10% among this patient population was >90%. The study was closed to accrual after 26 evaluable patients were enrolled.

Results

Study group. A total of 27 patients with AML, MDS, CLL, MCL, myelofibrosis, natural killer cell/T-cell cell leukemia, and T-cell prolymphocytic leukemia was entered onto the study between April 2004 and May 2005. Patient characteristics are listed in Table 1 . Fourteen patients had received three or more prior treatment regimens for their disease, including 2 patients who had received an allogeneic stem cell transplant. Two patients had not received prior therapy for their disease: a 75-year-old patient who developed therapy-related AML 2 years after being treated for Burkitt-like non–Hodgkin's lymphoma and a 59-year-old man who developed MDS 9 years after being treated with uracil and tegafur plus leucovorin for metastatic colon cancer. A third patient had been diagnosed with essential thrombocythemia in 1992, which evolved to myelofibrosis in 1999; the patient progressed to refractory anemia with excess blasts (RAEB) in transformation while receiving therapy with imatinib for myelofibrosis. The patients with myelofibrosis, natural killer cell/T-cell leukemia, and T-cell prolymphocytic leukemia had received two or more prior treatment regimens. Three of four patients with MCL were in leukemic phase with circulating lymphoma cells. All six patients with CLL had received prior fludarabine-based therapy.

Side effects. All 27 patients were evaluable for toxicity. No DLT occurred in the first three patients treated with 5 mg daily; all subsequent patients received 10 mg daily. The frequency and grading of potentially treatment-related adverse effects are summarized in Table 2 . These included grade 1 or 2 hyperlipidemia (44%), elevation of transaminases and/or alkaline phosphatase (41%), anorexia (37%), oral aphthous ulcers (37%), diarrhea (29%), hyperglycemia (26%), and hypomagnesemia (22%). Uncomplicated grade 3 hyperglycemia occurred in six patients requiring insulin therapy; five had a history of diabetes mellitus with grade 2 hyperglycemia noted at baseline in four patients. The hyperglycemia was resolved to baseline or better with cessation of everolimus. Other grade 3 toxicities included hypophosphatemia and fatigue in two patients each and anorexia and diarrhea in one patient.

One patient with refractory anemia with ringed sideroblasts developed a biopsy-proven grade 3 cutaneous leukocytoclastic vasculitis (LCV) after 120 days of therapy, which was believed to be related to everolimus. Everolimus was discontinued. She did not have any systemic manifestations of vasculitis nor malignancy. Systemic immunosuppressive therapy, including corticosteroids, was not used, and no new lesions have appeared. She required surgical debridement and skin grafting.

Thirty-eight infectious episodes were noted in 21 (78%) patients at some time during the therapy. These included fever of unknown origin in 6 (22%) patients and documented infections in 20 (74%) patients: 2 patients with sinusitis, 2 patients with bronchitis, 2 patients with bacteremia involving Enterococcus faecalis and Staphylococcus aureus, four urinary tract infections in 4 patients (one of whom developed a vaginal yeast infection and another Clostridium difficile–positive diarrhea while on antibiotics), 14 pneumonias in 12 patients (one of whom developed oral thrush while on antibiotics), and 1 patient each with cellulitis, folliculitis, paronychia, and presumed herpetic stomatitis/mucositis. Fourteen (37%) of the 38 infectious episodes that occurred in 9 patients required either hospitalization or prolonged hospitalization.

One patient with RAEB-1 developed a prolonged Escherichia coli urinary tract infection during the eleventh cycle of therapy, which was complicated by recurrent hematuria, requiring platelet transfusions, antibiotics, and hospitalization. Therefore, everolimus was held during the twelfth cycle of therapy. Cytoscopy with bladder biopsy revealed a marked chronic inflammatory infiltrate, vascular congestion, and hemorrhage with no leukemic cells present. The hematuria was resolved 37 days after discontinuation of everolimus. This has been observed with prolonged symptomatic urinary tract infections and was not believed to be associated with administration of everolimus.

Of the 12 patients with pneumonia, only 2 had positive cultures (Aspergillus species and coagulase-negative Staphylococcus and Aspergillus niger, Klebsiella species, and methicillin-resistant Staphylococcus aureus). Two episodes of pneumonia occurred in one patient with myelofibrosis who was subsequently diagnosed with biopsy-proven bronchiolitis obliterans with organizing pneumonia, which was treated successfully with steroids. Of the remaining 11 patients with pneumonia, 1 patient had a chronic esophageal stricture, requiring dilations every 2 months, due to prior radiotherapy for Burkitt-like lymphoma and was believed to have an aspiration pneumonia involving the right lower lobe, 4 had grade 3 or 4 neutropenia, and 2 had CLL with hypogammaglobulinemia. Two patients died while on study: one patient with natural killer cell/T-cell leukemia from neutropenic sepsis and one with CLL from uncontrolled gastrointestinal bleeding secondary to arteriovenous malformations.

Response to treatment. The median number of cycles administered was 1 (range, 0-11) with a median of 50 doses given (range, 4-318). Nine patients missed a median of two doses of everolimus (range, 1-6), two patients missed three doses, and one patient missed six doses at some time during the therapy. No patients required dose reduction; only one patient had dose escalation from 5 to 10 mg daily. One patient with AML received hydroxyurea for 4 days during the first cycle of therapy to control blast counts, was subsequently taken off study without completing one cycle of therapy because of intercurrent illness, and was replaced. Therefore, 26 patients were evaluable for response.

One objective response (hematologic improvement) was observed. A patient with RAEB-1 (patient 5) who was pancytopenic and red cell transfusion dependent had a major hematologic improvement in platelet count from baseline values of 22 to 26 × 10⁹/L to a maximum of 102 × 10⁹/L. Time to response was 50 days. Relapse occurred 104 days after achievement of major hematologic improvement in platelet count. Duration of minor and major platelet responses was 210 and 177 days, respectively. Fifty days after starting therapy with everolimus, this patient also had an improvement in the absolute neutrophil count by >0.5 × 10⁹/L from baseline values of 1.3 to 1.4 × 10⁹/L to a maximum of 3.0 × 10⁹/L.

A second patient with therapy-related RAEB-1 (patient 3) had an improvement in platelet count from baseline values of 11 to 14 × 10⁹/L to a maximum of 34 × 10⁹/L; the improvement was noted 113 days after the dose of everolimus was increased from 5 mg daily to 10 mg daily. However, duration of response was only 25 days and, therefore, did not fulfill the criteria for minor hematologic improvement in platelet count. Response was lost during the episodes of recurrent hematuria. While off treatment and after resolution of the hematuria, this patient has had a sustained increase in platelet counts ranging from 29 to 42 × 10⁹/L of >2-month duration. The patient with refractory anemia with ringed sideroblasts (patient 7), who developed a cutaneous LCV, had decreased red cell transfusion requirements, with only three transfusions administered over a 9-month period since discontinuing everolimus. The maximum interval between transfusions was 116 days compared with a baseline of every 21 to 26 days before treatment with everolimus.

Although no patients with CLL patients achieved a complete or partial response, three patients had a 27% to 34% reduction in adenopathy, documented radiographically, after two to three cycles of therapy. A fourth patient had a 65% reduction in peripheral lymphadenopathy, documented clinically, after two cycles of therapy; this patient subsequently died from uncontrolled gastrointestinal bleeding due to arteriovenous malformations.

Fifteen patients subsequently received other therapies for their underlying disease (median, 1; range, 1-4), including hydroxyurea (n = 2), induction chemotherapy with cytarabine and an anthracycline or purine analogue (n = 3), fludarabine-based regimen (n = 2), azacytidine (n = 1), vincristine and decadron (n = 1), hyperCVAD and alemtuzumab (n = 1), allogeneic stem cell transplant (n = 1), and other investigational agents (n = 8). After a median follow-up of 18 weeks (range, 3 to 68+) from the time of study entry, 16 patients have died, including the 2 patients who died while on study.

Effects of everolimus on mTOR signaling pathways. The ability of everolimus to modulate mTOR signaling was examined in samples from nine patients by comparing profiles of phosphorylated mTOR target proteins before and 24 hours after everolimus administration. To investigate everolimus-specific effects on mTOR signaling in vivo, phosphorylation of the 70-kDa 40S ribosomal protein kinase (p70S6K) and 4E-BP1 was examined. Both phosphorylated p70S6K and 4E-BP1 in Thr⁷⁰ were simultaneously inhibited after 24 hours of everolimus administration in samples from four patients (patients 2, 3, 7, and 9; Table 3 ). 4E-BP1 exhibited a decrease in Thr⁷⁰ phosphorylation in two additional samples (patients 5 and 6). Interestingly, the phosphorylation of another rapamycin-sensitive residue (Thr³⁷/Thr⁴⁶) was affected only in the sample from patient 3, indicating that mTOR-insensitive phosphorylation of this site can occur as previously suggested (22). No significant changes in mTOR signaling were observed in samples from four other patients. Notably, patients 3 (CLL), 5 (RAEB-1), 6 (RAEB-1), and 7 (refractory anemia with ringed sideroblasts) displayed some evidence of clinical responses to everolimus as detailed above, and patient 5 had objective response (hematologic improvement, see above). Examples of Western blot analyses for samples 5 and 3 are represented in Fig. 2 .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Peripheral blood mononuclear cells from the responding patient with RAEB-1 (sample #5) and from a patient with CLL (sample #3) were subjected to immunoblot analyses of phosphorylated 4E-BP1 (p4E-BP1), phosphorylated p70S6K (pp70S6K), total 4E-BP1, phosphorylated Akt (pAkt), total Akt, phosphorylated ERK (pERK), total ERK, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and quantitiative real-time PCR analysis of Glut1 and cyclin D1 transcripts. The abundance of transcripts of either cyclin D1 or Glut1 relative to that of β₂-microglobulin (B2M) was assessed by real-time PCR as described in Materials and Methods. Error bars, half the difference between the maximum and minimum values that arose on substituting ΔC_t − SD or ΔC_t + SD, respectively, for ΔC_t in the following formula: relative expression = 100 × 2^−ΔCt.

Akt and mTOR are linked to each other via positive and negative regulatory circuits, and in model systems, tumors exhibiting activation of Akt are hypersensitive to mTOR inhibitors (23). We therefore analyzed effects of everolimus in the context of Akt signaling. At baseline, Akt was phosphorylated on Thr³⁰⁸ in all seven of seven samples tested and on Ser⁴⁷³ in eight of nine samples. Unexpectedly, we found inhibition of phosphorylation of Akt at Thr³⁰⁸ in four of six (patients 2, 3, 5, and 7) and at Ser⁴⁷³ in five of eight (patients 2, 3, 5, 6, and 9) patient samples (see examples on Fig. 2). Interestingly, Akt was down-regulated only in samples where inhibition of mTOR signaling was observed, suggesting that mTOR inhibition in vivo may inhibit Akt signaling. Activation (phosphorylation) of ERK was documented in samples from eight of nine patients, and unanticipated decrease in phosphorylated ERK levels was observed in samples from two patients (patients 3 and 8).

mTOR signaling has been reported to affect the expression of the target genes by different mechanisms (transcription, translation, or protein degradation; refs. 24–26). We examined the effects of mTOR inhibition on the transcription of the glucose transporter Glut1 and D-type cyclins that control the G₁-S transition in the cell cycle. In samples from six patients, in which administration of everolimus inhibited mTOR signaling, down-regulation of Glut1 mRNA (>40%) was observed in three samples (patients 2, 3, and 6) and down-regulation of either cyclin D1 or cyclin D2 in three and three samples, respectively (Table 3). In samples 3 (CLL) and 5 (RAEB-1), concomitant decreases in mRNA expression of both cyclin D1 and D2 were observed.

Discussion

Because of accumulating evidence documenting the importance of the phosphatidylinositol 3-kinase/Akt/mTOR pathway in the pathogenesis of a variety of hematologic malignancies, several therapeutic strategies are being developed to modulate this signaling pathway. The importance of mTOR is underscored by its key regulatory role in protein translation (1, 2). The current phase I/II study is the first to evaluate everolimus in patients with advanced hematologic malignancies. In addition, the effect of everolimus on multiple mTOR-regulated proteins was evaluated.

Therapy with everolimus at a dose of 10 mg orally once daily was relatively well tolerated. Toxicities observed with everolimus were consistent with those reported in previous studies using mTOR inhibitors and consisted of hyperglycemia, hyperlipidemia, anorexia, elevated liver enzymes, oral aphthous ulcers, diarrhea, hypophosphatemia, fatigue, and hypomagnesemia (27–30). The frequency of infectious episodes did not seem to be increased in the study patients, an important consideration for future studies in the hematologic malignancies. There were no documented episodes of cytomegalovirus infections or varizella zoster infections.

One patient with advanced myelofibrosis had recurrent pulmonary infiltrates while receiving therapy with everolimus. The first episode occurred 57 days after initiating therapy and resolved with antibiotics and ongoing therapy with everolimus. This was subsequently biopsy proven to be bronchiolitis obliterans with organizing pneumonia and treated successfully with steroids. Everolimus was discontinued due to the intercurrent illness. She was not rechallenged. Bronchiolitis obliterans with organizing pneumonia is usually idiopathic, although it has been associated with hematologic malignancies and cytotoxic drugs (31, 32) Nonspecific pneumonitis has been described in patients treated with temsirolimus and sirolimus (27, 33). Six of 42 patients were diagnosed with bronchiolitis obliterans with organizing pneumonia, all of whom were recipients of solid organ transplants. Approximately one third of patients experienced improvement of the infiltrates after discontinuation of the drug, and two of four patients with renal cell cancer experienced recurrent pneumonitis on rechallenge (27). Preclinical studies indicated that everolimus could prevent the development of bronchiolitis obliterans in a porcine bronchial model (34). Similarly, the combination of mycophenolate and sirolimus could attenuate the progression of bronchiolitis obliterans syndrome in lung and heart-lung transplant recipients (35). Therefore, the relationship of bronchiolitis obliterans with organizing pneumonia to everolimus is unclear.

A patient with MDS developed cutaneous LCV while receiving therapy with everolimus on this study. Causes of LCV include malignancy, including MDS, and drugs (36). LCV has been reported with sirolimus in four patients. This occurred 0.75 to 4 months after initiation of sirolimus to prevent solid organ rejection. The involved sites were skin in three cases and gastrointestinal in the fourth. Symptoms were resolved within 0.2 to 4 months after discontinuation of sirolimus. Rechallenge with sirolimus resulted in recurrence of the LCV in one case.

Only one patient with MDS achieved an objective response (i.e., major hematologic improvement in platelet count), although one other patient achieved a minor response in platelet count of 25-day duration. Although only five patients with MDS were treated with everolimus, the response rate (20%) was comparable with what has been observed with sirolimus in this patient population (16%; ref. 28). None of the nine patients with AML treated with everolimus had an objective response, in contrast to an overall response rate of 44% (44% partial response) reported with sirolimus in patients with AML (37). It is unclear whether this is due to differences in patient characteristics, duration of therapy (only 28 days with sirolimus), or significant differences in the agents (1). We observed no responses in 10 patients with AML treated with temsirolimus (38). On interim analysis, we observed that only 2 of 26 (8%) of patients with AML or MDS treated with AP23573 had a hematologic improvement (39). Response rates of 38% (3% complete response; 35% partial response) were observed in patients with relapsed or refractory MCL receiving single agent temsirolimus; median time to response was 1 month (range, 1-8 months; ref. 29). All four patients with MCL treated with everolimus had progressive disease; three of whom did not complete a course of therapy. Furthermore, three patients had circulating lymphoma cells at baseline and all four patients in the current study had refractory disease compared with 54% of those given temsirolimus.

Variability in everolimus concentrations has been observed in different ethnic groups, with lower bioavailability in African-Americans compared with Caucasians or non-African-Americans (40). In the current study, 3 of the 26 evaluable patients were African-Americans (1 CLL, 1 natural killer cell/T-cell leukemia, and 1 AML); none of whom achieved an objective response, although the patient with CLL did have a 33% reduction in lymphadenopathy after three cycles of therapy with everolimus.

Akt was activated as evidenced by phosphorylation of Ser⁴⁷³ and Thr³⁰⁸ in all but one sample tested, confirming the high frequency of Akt activation in primary leukemia cells. (1) Analysis of phosphorylation of 4E-BP1 and p70S6K, established biomarkers of mTOR blockade by rapamycin analogues, showed inhibition of phosphorylation of both proteins in four of nine samples and inhibition of phosphorylated Thr⁷⁰ in 4E-BP1 in two additional samples (i.e., inhibition in ∼66% of patients). Notably, evidence of inhibition of mTOR signaling by everolimus was documented in samples from all patients with clinical responses, with the exception of samples from three patients with CLL, where the patients did not consent to provide samples. mTOR inhibition was associated with transcriptional down-regulation of D-type cyclins (either cyclin D1 or cyclin D2) and a decrease in Glut1 mRNA levels in a subset of patients, suggesting that, in leukemic cells, everolimus attenuates transcription of these genes. However, inhibition of mTOR downstream targets did not translate into clinical responses in AML patients 2 and 9, indicating that the ability to inhibit the mTOR pathway is not always sufficient to elicit clinical responses.

Modest antitumor activity of mTOR inhibitors may also be associated with activation of alternative prosurvival signaling pathways, such as mitogen-activated protein kinase/ERK. Furthermore, ERK1/2 has been shown to phosphorylate and inactivate TSC2 and 4E-BP1 and activate p70S6K in vitro (41–43), and preclinical data indicate that inhibition of both mTOR and ERK signaling pathways is required to prevent de novo protein translation. We observed high levels of phosphorylation of ERK in the majority of samples studied, confirming our previously reported data in AML that showed phosphorylated ERK in >80% of AML patients and providing a rationale for future combination studies with Raf or mitogen-activated protein kinase/ERK kinase inhibitors (44, 45).

Recent work has shown that selective inhibition of mTOR downstream signals paradoxically increases phosphorylated Akt (46, 47), likely attributable to lack of feedback inhibition by p70S6K on the insulin-like growth factor-I receptor. We unexpectedly observed inhibition of Akt phosphorylation in five of six samples, in which everolimus modulated mTOR signaling. mTOR partitions between two scaffold proteins, the rapamycin-sensitive raptor and the rapamycin-insensitive rictor (23). Recent data show that the rictor-mTOR complex (TORC2) directly phosphorylates Akt on Ser⁴⁷³ and facilitates Thr³⁰⁸ phosphorylation by PDK1 (48). Although rapamycin does not bind to a preformed rictor-mTOR complex (49), it has been proposed that, during prolonged rapamycin treatment, the drug will sequester the newly synthesized mTOR molecules, interfere with reassembly of rictor-mTOR complex, and thereby inhibit rictor-mTOR activity. In fact, prolonged, but not acute, treatment of some human cells with rapamycin has been reported to partially inhibit Akt phosphorylation (50).

The data reported here provide the first in vivo evidence that rapamycin analogues are capable of suppressing Akt activity in hematopoietic cells, perhaps in contrast to solid tumor cell types. Although the antitumor activity of single agent mTOR inhibitors is modest in patients with hematologic malignancies, combination therapy of these agents with modulators of upstream stimulants of the pathway (e.g., vascular endothelial growth factor, Flt3, bcr-abl) is warranted. Based on data from the current study, concerns about the immunosuppressive effects of mTOR inhibitors in patients with advanced hematologic malignancies should not pose a barrier to the conduct of further studies, although this issue will require careful monitoring.