JS04.6.A The landscape of tumor cell states and spatial organization in H3-K27M mutant diffuse midline glioma across age and location

Background

Histone 3 lysine27-to-methionine mutations (H3-K27M) frequently occur in childhood diffuse midline gliomas (DMGs) of the pons, thalamus and spinal cord, presumed to be driven by the specific spatiotemporal context of these midline locations during postnatal development. While most common in the pons and at mid-childhood ages, the same oncohistone mutation is recurrently detected in adult DMGs and throughout different midline regions. The potential heterogeneity of tumors at different ages and in different anatomical locations of the midline are vastly understudied.

Material and Methods

Through dissecting the transcriptomic, epigenomic and spatial architectures of a comprehensive cohort of patient H3-K27M DMGs - spanning the age range from 2-68 years and locations from spinal cord to thalamus - at single cell resolution, we delineate how age- and location-dependent contexts shape glioma cell-intrinsic and -extrinsic features in light of the shared driver mutation.

Results

We identify that oligodendrocyte precursor (OPC)-like cells constitute the stem-like compartment in H3-K27M DMGs across all clinico-anatomical groups, however, depending on location, display varying levels of maturity resembling less differentiated pre-OPCs or more mature OPCs further differentiated along the oligodendroglial lineage. We further demonstrate increased mesenchymal cell states in adult tumors, which we link to age-related differences in glioma-associated immune cell compartments. We for the first time resolve the spatial organization of H3-K27M DMG cell types in intact patient tissues, identifying a local niche of the oligodendroglial lineage.

Conclusion

Our study provides a powerful resource for rational modeling and therapeutic frameworks taking into account determinants of age and location in this lethal glioma group.

Final results of a phase I trial with the novel anthracycline derivative RTA 744 in patients with primary brain tumors

2062

Background: RTA 744 is an anthracycline derivative that was shown preclinically to cross the blood-brain barrier, not be a substrate for P-gp or MRP, and improve survival in an orthotopic murine model of glioblastoma. A trial of RTA 744 was conducted in patients with progressive high-grade gliomas. Methods: RTA 744 was administered as a 2-hour intravenous infusion on each of the first 3 days of a 21-day cycle. Five dose levels were tested until an MTD was reached. Pharmacokinetic samples were taken at multiple time points on days 1–5 of cycle 1. Tumor activity was assessed according to the MacDonald criteria. Results: As of December 2006, RTA 744 has been administered to 20 patients at doses ranging from 1.2 to 9.6 mg/m2/day. The MTD was determined as 7.5 mg/m2/day, and the DLTs were neutropenia and thrombocytopenia. No neurotoxicity, cardiotoxicity, alopecia, or drug-related nausea and vomiting were reported. The pharmacokinetic profile indicates dose proportionality, with a mean plasma half-life of 35 hours. Clear evidence of anti-tumor activity has been reported in 7 of 20 patients. A Complete Response was observed in a patient with GBM after seven cycles at 2.4 mg/m2/day, the patient continues greater than 5 months post-RTA744 treatment. Additionally, a Partial Response (81% tumor reduction) has been observed in a patient with AO after two cycles at 7.5 mg/m2/day who continues on treatment at this time. A Minor Response was observed at 2.4 mg/m2/day, and several patients have experienced Stable Disease of several months duration. Two patients with stable disease have shown radiographic evidence of necrosis. Median time to tumor progression was 6 weeks (range 2 to >50). Median number of cycles was 2 (range 1 to 7). The study is now enrolling additional patients at the MTD to further profile the drug’s safety and activity. Additionally, a once-weekly dosing schedule is being tested in a satellite cohort. Conclusions: RTA 744 is well tolerated when administered at doses up to and including 7.5 mg/m2/day for the first three days of a 21-day cycle. Based on clear evidence of drug activity, pivotal trials of RTA 744 in high-grade gliomas are planned as well as trials in patients with brain metastases.

[Table: see text]

A phase I study of temozolomide (TMZ) and the farnesyltransferase inhibitor (FTI), lonafarnib (Sarazar, SCH66336) in recurrent glioblastoma

1556

Background: Farnesylation is essential for the post-translational modification of several proteins that play a role in cell proliferation and growth. FTIs demonstrate antiproliferative effects in a variety of tumor cell lines and xenograft models and enhanced efficacy by combining lonafarnib with TMZ compared with TMZ alone. Clinical studies show that dose-intense TMZ may have better efficacy then conventional dosing (Wick, Neurology 2004). Objective: Determine the maximum tolerated dose of lonafarnib when combined with a dose-intensified schedule of TMZ. Eligibility criteria: Histologically proven GBM; evidence of relaspe or progression; up to 2 prior chemotherapy regimens, prior TMZ permitted; KPS ≥ 60. No enzyme-inducing anticonvulsants. Methods: TMZ was given at 150 mg/m2 days 1–7 and 15–21; lonafarnib escalated by cohort: 100 mg BID, 150 mg BID, 200 mg BID on days 8–14 and 22–28 of a 28-day cycle. Response: Toxicity measured after cycle 1 using NCI CTCAE v3. Patient (pt)accrual used a 3 + 3 design. Response was evaluated q2 cycles. Standard response criteria (MacDonald) were used. Results: 15 patients were accrued. Median age 47 yrs; median KPS 90. M:F ratio was 11:4. No DLTs in any dose cohort, a pt (cohort 1) died day 29 of unrelated cardiac disease. 4 pts demonstrated partial response, all with prior TMZ failure, 3 pts showed prolonged stable disease. 6-month PFS rate estimated at 33%, median PFS 14 weeks. Treatment was well tolerated: 1 episode each of grade 3 diarrhea, esophagitis, fatigue and hypokalemia; 3 episodes of grade 3/4 leukopenia, 2 episodes of grade 3/4 neutropenia, and 2 episodes of grade 4 thrombocytopenia. Seven pts had grade 3/4 lymphopenia, some with persistant lymphopenia from prior TMZ. 1 pt had grade 4 pneumonitis of uncertain etiology. Conclusions: The alternating week schedule of TMZ/lonafarnib was well tolerated. Preliminary results suggest anti-tumor activity even in pts who had failed prior TMZ. It is not certain whether this activity is due to the dose-intensified TMZ or the treatment combination. Studies are planned to evaluate potential predictors of response (i.e MGMT gene promoter methylation) and expand the trial to obtain pharmacokinetic and additional efficacy data.

[Table: see text]

Taurine inhibits rat substantia nigra pars reticulata neurons by activation of GABA- and glycine-linked chloride conductance

Whole-cell recordings from rat midbrain slices showed that bath application of taurine (30 microM-10 mM) reversibly hyperpolarized and inhibited the firing of substantia nigra pars reticulata neurons. These effects were mediated by an increase in membrane chloride conductance. Taurine cross-desensitized with GABA and glycine. Strychnine competitively antagonized the actions of glycine and taurine but not that of GABA. On the other hand, the responses to taurine and glycine were not sensitive to bicuculline. These findings suggest that taurine inhibits SNR neurons by opening GABA and glycine-linked chloride channels.

Platelet aggregation in laminar flow. II. Shear rate and ADP-dependent effects of acetylsalicylic acid and indomethacin

The relative contributions of physiologic activators and shear rates (G) on the formation and stability of platelet aggregates arising from platelets treated with widely clinically used non-steroidal anti-inflammatory drugs such as acetylsalicylic acid (ASA) need to be characterized for conditions of low G believed to be important in vivo. In Part I of these studies it was found that platelets in human citrated platelet-rich plasma (PRP) undergo ADP-induced aggregation in laminar flow by apparently two distinct processes: a low ADP-dependent process, generating large visible aggregates stably only at low shear rates, and a higher ADP-threshold-dependent process yielding more complete and shear-resistant large aggregate formation. Similar studies of both macroaggregation (TA) and actual % aggregation (%PA) were made with ADP injections into PRP in the cone-in-plate device as a function of G and pre-treatments with ASA (200 microM) and indomethacin (10 microM). These drugs reduced %PA in part and yielded only reversible macro-aggregation for studies conducted at the higher [ADP] (4-6 microM) and at high G (greater than or approximately 120-150 sec-1), with no significant effects at low G (less than or approximately 20 sec-1). The results are discussed in terms of a platelet stickiness factor 's', ascribed to changes in fibrinogen binding sites reported for these drug actions, and related to in vivo flow and variable platelet activation conditions.

Platelet aggregation in laminar flow. I. Adenosine diphosphate concentration, time and shear rate dependence

The relative contributions of physiologic activators and shear rates (G) on the formation and stability of platelet aggregates for in vitro laminar flow need to be characterized for conditions of low G believed to be important in vivo. Platelet aggregation was therefore studied in human citrated platelet-rich plasma (PRP) prepared at 37 degrees C for G = 0-190 sec-1 in laminar flow in a cone-in-plate device as a function of adenosine diphosphate (ADP) activation. Macroaggregation (TA) was assessed by visual observation and light transmission (%T), as routinely done in aggregometry, while actual percent aggregation (PA) was determined by microscopy of sub-samples from the decrease in single platelet concentration. Unactivated PRP never yielded shear-induced aggregation while samples pre-mixed with low [ADP] (less than or approximately 0.7 microM) led to large visible aggregate formation stable only for G less than or approximately 30-60 sec-1, apparently arising from pre-existing microaggregates. Direct ADP injections into the flow device showed that at low [ADP] (less than 1 microM), maximal values for both PA and TA were reached when step-wise increases in G reached less than or approximately 30 sec-1, with TA decreasing at higher G; these values were only some 10-20% of maximal values achieved with high (ADP] (greater than or approximately 4 microM), when PA was again maximal at G greater than or approximately 30 sec-1, but TA expressed as a rate of aggregation continued to increase with G up to 150 sec-1. The results are discussed in terms of the temporal shear-rate and ADP-dependent dynamics of a platelet stickiness factor, 's'.

Kinetics of ADP-induced human platelet shape change: apparent positive cooperativity

The kinetics of ADP-induced human platelet shape change have been examined. Initial velocities of platelet shape change were estimated by two methods: (1) the slope of the initial decrease in light transmission through stirred, citrated platelet-rich plasma, and (2) direct examination of platelet morphologies by phase-contrast microscopy. In both cases, a value of the Hill coefficient, NH, significantly greater than 1 is obtained (2.0 +/- 0.2 and 1.8 +/- 0.2, respectively). The observed elevated value of NH is not due to a substantial fraction of the ADP being platelet bound, the presence of factors in the plasma, platelet heterogeneity, or the influence of the rate of platelet shape change reversion. Our observations suggest that ADP-induced platelet shape change may be a positively cooperative or "threshold" type response.