Posted on

Cbd oil for cll leukemia

Cannabinoid Receptors Are Overexpressed in CLL but of Limited Potential for Therapeutic Exploitation

Conceived and designed the experiments: KV. Performed the experiments: PF EAP TL CP. Analyzed the data: PF EAP TL CP KV. Contributed reagents/materials/analysis tools: MG PS UJ. Wrote the paper: PF EAP KV. Clinical information: MG PS UJ.

This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Associated Data

S1 Fig: No differences in survival of CLL patients in relation to CNR2 mRNA expression. (A) Mean overall survival (OS) for high expressing patients was 196 months vs. 230 months for low expressing patients (N = 107; p = 0.763). (B) Mean treatment free survival (TFS) in CNR2 high and low mRNA expressers was 100 months vs. 135 months in high and low expression groups, respectively (N = 107; p = 0.2290). One hundred and seven patients were included in the analysis, median mRNA expression of CNR2 (3.77) was used as cut-off.

S2 Fig: Cytotoxic effect of cannabinoids in combination with fludarabine. CLL primary cells (N = 5) were incubated in triplicates in co-culture with M2-10B4 mouse fibroblasts and incubated for 30 minutes with increasing concentrations of cannabinoids before fludarabine (5 μM) was added. Viability was determined after 48h. Incubations with vehicle served as control. For comparison, cells were incubated with fludarabine alone, with AMD3100 alone, and with AMD3100 in combination with fludarabine (N = 6). Mean values and standard deviations are shown. Hatched lines mark experimental blocks. The synergistic effect of the combination 40 μM ACEA with 5μM fludarabine was significantly different from the effect of 40μM ACEA alone. *p = 0.047. Abbreviation: CNB, (-)-cannabidiol.

S3 Fig: Cytotoxic effect of cannabinoids under serum-reduced conditions. PBMC of 5 CLL patients were incubated in triplicates for 48h in increasing compound concentrations at 1%, 2.5%, 5%, and 10% serum containing medium in suspension and in co-culture with M2-10B4 mouse fibroblasts before viability was measured. Mean values and standard deviations are shown. (A) (R)-(+)-methanandamide in suspension and (B) in co-culture. (C) (-)-cannabidiol in suspension and (D) in co-culture. (E) ACEA in suspension and (F) in co-culture. (G) JWH133 in suspension and (H) in co-culture. (I) AM251 in suspension and (J) in co-culture. (K) AM630 in suspension and (L) in co-culture. Note different scales on x- and y-axes.

S4 Fig: Cytotoxicity of cannabinoids in relation to CNR1 mRNA expression. PBMC from CLL patients were incubated in triplicates in increasing compound concentrations in suspension and co-culture with M2-10B4 mouse fibroblast cells for 48h before viability was measured. (A) (R)-(+)-methanandamide (N = 10). (B) (-)-cannabidiol (N = 18). (C) ACEA (N = 16). (D) JWH133 (N = 16). (E) AM251 (N = 16). (F) AM630 (N = 16). The x-axis shows the measured mRNA expression for each CLL sample tested (healthy CD19 sorted cells set as 1) from highest (left) to lowest (right) expression. Absent values may indicate that i) sample was not tested, or ii) IC50 could not be calculated, or iii) 50% viability reduction could not be achieved. Note different scales on Y-axis for A and D.

S5 Fig: Cytotoxicity of cannabinoids in relation to CNR2 mRNA expression. PBMC from CLL patients were incubated in triplicates in increasing compound concentrations in suspension and co-culture with M2-10B4 mouse fibroblast cells for 48h before viability was measured. (A) (R)-(+)-methanandamide (N = 10). (B) (-)-cannabidiol (N = 18). (C) ACEA (N = 16). (D) JWH133 (N = 16). (E) AM251 (N = 16). (F) AM630 (N = 16). The x-axis shows the measured mRNA expression for each CLL sample tested (healthy CD19 sorted cells set as 1) from highest (left) to lowest (right) expression. Absent values may indicate that i) sample was not tested, or ii) IC50 could not be calculated, or iii) 50% viability reduction could not be achieved. Note different scales on Y-axis for A and D.

S6 Fig: Impact of cannabinoids on CLL cell migration in unsorted PBMC. Primary cells of 5 CLL patients were pre-incubated with cannabinoids before being transferred to transwell plates and incubated for 4h for migration. Control experiments included CXCL12 alone (control), no CXCL12 (control w/o CXCL12), incubation with vehicle (DMSO, ethanol), and incubation with the CXCR4 inhibitor AMD3100. CLL cells were incubated either with agonist (ACEA, JWH133) or antagonist (AM251, AM630) before migration. In addition, cells were treated with antagonist before agonist incubation before migration was allowed (CB1: AMS251&ACEA; CB2: AM630&JWH133). Bars represent mean values of migration indices + standard deviations, hatched lines indicate experimental blocks.* p = 0.0016; ** p

S2 Table: Compound concentrations and duration of incubations before initiation of migration experiments. (PDF)

S3 Table: Comparison of patient characteristics between CNR2 high and low mRNA expressing groups. (PDF)

All relevant data are within the paper and its Supporting Information files.

Abstract

The cannabinoid receptors 1 and 2 (CNR1&2) are overexpressed in a variety of malignant diseases and cannabinoids can have noteworthy impact on tumor cell viability and tumor growth. Patients diagnosed with chronic lymphocytic leukemia (CLL) present with very heterogeneous disease characteristics translating into highly differential risk properties. To meet the urgent need for refinement in risk stratification at diagnosis and the search for novel therapies we studied CNR expression and response to cannabinoid treatment in CLL. Expression levels of CNR1&2 were determined in 107 CLL patients by real-time PCR and analyzed with regard to prognostic markers and survival. Cell viability of primary CLL cells was determined in suspension and co-culture after incubation in increasing cannabinoid concentrations under normal and reduced serum conditions and in combination with fludarabine. Impact of cannabinoids on migration of CLL cells towards CXCL12 was determined in transwell plates. We found CNR1&2 to be overexpressed in CLL compared to healthy B-cells. Discriminating between high and low expressing subgroups, only high CNR1 expression was associated with two established high risk markers and conferred significantly shorter overall and treatment free survival. Viability of CLL primary cells was reduced in a dose dependent fashion upon incubation with cannabinoids, however, healthy cells were similarly affected. Under serum reduced conditions, no significant differences were observed within suspension and co-culture, respectively, however, the feeder layer contributed significantly to the survival of CLL cells compared to suspension culture conditions. No significant differences were observed when treating CLL cells with cannabinoids in combination with fludarabine. Interestingly, biologic activity of cannabinoids was independent of both CNR1&2 expression. Finally, we did not observe an inhibition of CXCL12-induced migration by cannabinoids. In contrast to other tumor entities, our data suggest a limited usability of cannabinoids for CLL therapy. Nonetheless, we could define CNR1 mRNA expression as novel prognostic marker.

Introduction

Cannabinoids, the active components of the hemp plant Cannabis sativa, have been used for centuries for medical and recreational purposes. These compounds exert their activity by binding the cannabinoid receptors.

Cannabinoid receptors 1 and 2 (CNR1/CNR2; CB1/2) belong to the group of G protein-coupled receptors (GPR) and are part of the endocannabinoid system. The native ligands of the two receptors, such as 2-arachidonoyl glycerol, are produced on demand and fulfill a variety of functions. Although other receptors have been described for the endocannabinoid system [1], CB1&2 still are the two receptors for which most knowledge has been gathered.

CNR1 is primarily expressed in the brain, CNR2 in cells and tissues of the immune system, but both receptors have also been found outside these main sites of expression [2–4]. The two receptors appear to mediate similar responses and to exert overlapping influences, they seem to interact with respect to inflammatory and neurologic/psychotic conditions [5–9], and are involved in migration of cells of different origin under different physiological states [10–12]. Thus, 2-arachidonoyl glycerol acts as chemo-attractant for both immature and mature B-cells via CB2 [13–16] and appears to interfere with CXCR4 expression and/or CXCL12 induced migration [14]. Similar migratory effects were reported also for other cannabinoids [17], although the extent of these effects seems to be variable [18, 19].

Overexpression of both receptors has been found both in solid tumors and hematologic malignancies and cannabinoids were shown to inhibit cell migration, angiogenesis, to reduce proliferation and viability, to induce apoptosis in vitro and reduce tumor burden in vivo [4, 20–32]. The sensitivity of mantle cell lymphoma (MCL), chronic lymphocytic leukemia (CLL), and Hodgkin lymphoma (HL) cell lines to cannabinoids was linked to the overexpression of CNR1 and/or CNR2 [23, 33, 34]. While some of these reports used relatively selective agonists like ACEA (CB1), JWH133, or JWH015 (both CB2) [25, 30, 35–37], in the majority of studies cannabinoids were tested which appear to display broader activity on G-coupled protein receptors [23, 28, 31, 33, 38]. Thus cannabidiol acts as CB1 antagonist, CB2 inverse agonist, GPR55 antagonist, and agonist for the VR1 vanilloid and the μ-opioid receptors [39, 40], (R)-(+)-methanandamide as CB1 agonist but also displays activity at vanilloid receptors and other G-protein coupled receptors and ion channels [1, 41].

In solid tumors, expression of the two cannabinoid receptors has been linked to patient outcome. In hepatocellular carcinoma and mobile tongue squamous cell carcinoma, both CB1 and CB2 overexpression was associated with good prognosis [42, 43]. In contrast, CB1 expression was reported to be a marker of bad prognosis in prostate and colorectal cancer [44–46], while CB2 was shown to be a poor prognostic marker in colon cancer [32] and was linked to poorer survival in HER2 positive breast cancer and squamous cell carcinoma of head and neck [47, 48]. Whether increased expression of one of the two receptors or both has clinical implications in hematologic malignancies appears to be variable [49–51].

The development of targeting drugs in recent years has greatly improved therapeutic options in CLL [52–54]. However, it is not known how long targeting molecules will display their potential before patients develop resistances and/or progress. In this line, several reports already discussed genetic changes developing during treatment with such compounds [55, 56]. CLL, like other malignancies, consists of a pool of malignant clones [57–59], which develop and evolve during disease course. Changes in this clonal landscape may occur during treatment and/or due to the acquisition of resistance mutations. Therefore, there still is an urgent need for agents which can be used for combination therapy as well as supportive regimens to increase treatment options and to improve patient care.

Based on the reported aberrant expression of cannabinoid receptors in neoplasms, we studied the expression of the two receptors in CLL patients analyzing it in relation to clinical parameters to determine their usability for prognosis. Additionally, considering the versatile aspects of cannabinoid actions, we evaluated the potential of cannabinoids for use in CLL therapy.

Materials and Methods

Patient material

Peripheral blood samples were collected from 107 consecutive patients diagnosed with CLL at the Division of Hematology and Hemostaseology of the General Hospital in Vienna, Austria. All patients and the four healthy volunteers included in the study signed informed consent according to the Declaration of Helsinki. The study was approved by the Ethics Committee of the Medical University of Vienna (approval Nr: 1011/2012). Clinical characteristics for the patients used in mRNA expression analysis are listed in S1 Table.

Peripheral blood mononuclear cells (PBMC) were isolated by Ficoll (Biocoll, Biochrom, Berlin, Germany) separation following standard procedures, cells were stored viable in liquid nitrogen. For protein analyses, 1×10 7 PBMCs were centrifuged, the supernatant was discarded and cell pellets were stored at -80°C. For RNA extraction, primary cells were stored at -80°C in TRIzol (Life Technologies Ltd, Paisley, UK). B lymphocytes from healthy donors were isolated using the EasySep ™ Human B Cell Enrichment Kit without CD43 Depletion (STEMCELL Technologies SARL, Grenoble, France) following the manufacturer´s protocol.

Reagents

Cannabinoids used in cytotoxicity and migration experiments were ACEA (CB1 agonist), AM251 (CB1 antagonist, μ-opioid receptor antagonist, GPR55 agonist, ion channel activation), AM630 (CB2 antagonist/inverse agonist, weak partial CB1 agonist, ion channel activation), (-)-cannabidiol (CNB) (GPR55 & weak CB1 antagonist, CB2 inverse agonist, weak agonist at VR1 vanilloid receptors, and modulator at opioid receptors), JWH133 (CB2 antagonist), and R-(+)-methanandamide (RM) (CB1 agonist, activity against GPR and ion channels) purchased from TOCRIS Bioscience (Bristol, UK). Fludarabine (2-Fluoroadenine-9-β-D-arabinofuranoside, SIGMA-ALDRICH, Vienna, Austria) served as control in cytotoxicity experiments, CXCL12/SDF-1α (R&D SYSTEMS, Abingdon, UK) and CXCR4 antagonist AMD3100 octahydrochloride (TOCRIS Bioscience, Bristol, UK) were controls in migration experiments.

Primary antibodies

The following cannabinoid receptor directed antibodies were tested:

Anti-CB1 antibodies: Cat.No. PA1-745, Thermo Fisher Scientific Inc., Waltham, MA, USA; Cat.No. GTX100517, GeneTex Inc., Irvine, CA, USA; Cat.No. AF1185a, Abgent, San Diego, CA, USA.

Anti-CB2 antibodies: Cat.No. AF1575a, monoclonal, Abgent, San Diego, CA, USA; Cat.No. AF1186a, polyclonal, Abgent, San Diego, CA, USA; Cat.No. PA1-744, Thermo Fisher Scientific Inc., Waltham, MA, USA; Cat.No. GTX100391, GeneTex Inc., Irvine, CA, USA.

Cat.No. CB1001, anti-GAPDH monoclonal antibody, Calbiochem/EMD Millipore, Merck, Darmstadt, Germany.

Recombinant proteins for Western blots: CB1 and CB2 human recombinant proteins (Cat.Nos. H00001268-G01 and H00001269-G01; Abnova, Taipei City, Taiwan).

Secondary antibodies

IRDye 680 conjugated goat anti-mouse polyclonal IgG (H+L) (Cat.No. 926–32220), and IRDye 800CW conjugated goat anti-rabbit polyclonal IgG (H+L) (Cat.No. 926–32211), both purchased from LI-COR (Bad Homburg, Germany).

RNA extraction and cDNA synthesis

RNA was extracted from samples of CLL patients and healthy donors using TRIzol, RNA was dissolved in 10 μl DEPC water, and the amount of isolated RNA measured. Two μg of RNA were used for cDNA synthesis (all products from Promega Corporation, Madison, USA), cDNA was stored at -20°C until real time PCR.

Real time PCR

Real time PCR was carried out using TaqMan Gene Expression Assays on demand for Cannabinoid receptors 1 and 2 (Cat.Nos. Hs00275634_m1 and Hs00361490_m1, Life Technologies Ltd, Paisley, UK) according to the manufacturer´s protocol. Applied Biosystems ® Human ACTB (Cat.No. 4326315E) served as housekeeping gene. Samples were run in duplicates on an ABI Prism 7000 Sequence Detector and analyzed using the SDS Software. For calculation of mRNA expression, the ΔΔCt-method was used [60], for which CD19 sorted pooled healthy B cells were set as 1.

Protein isolation and Western Blot

Proteins from patient samples were prepared using RIPA Buffer, concentration was determined using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific Inc., Rockford, USA). Sixty μg of protein per lane were separated by 10% SDS PAGE and transferred to PVDF Immobilon-FL Transfermembrane (pore size 0.45 μM; Merck Millipore, Billerica, US). Membranes were incubated for 16 hours at 4°C with the primary antibody, were washed 3 times for 5 minutes with PBS (PAA laboratories, Pasching, Austria) + 0.1% Tween20 (Sigma Aldrich, St. Louis, MO, USA), and subsequently incubated 45 minutes at room temperature (RT) in the dark with the secondary antibody before detection on an Odyssey Imager (LI-COR, Bad Homburg, Germany)

Cell culture

All cells were cultured under standard conditions (95% humidity, 5% CO2, 37°C). M2-10B4 (purchased from the American Type Culture Collection; http://www.lgcstandards-atcc.org) were kept in RPMI1640 containing 10% fetal calve serum (FCS) and 1% Penicillin/Streptomycin (PS) as were primary CLL cells and healthy PBMC. All reagents were purchased from Life Technologies (Carlsbad, CA, USA).

Drug incubations and viability tests

All experiments were performed using medium without phenol red. Viability was determined using the CellTiter-Blue ® Viability Assay (Promega, Madison, WI, USA) following the manual. Experiments with M2-10B4 were done in triplicates and repeated twice. Drug incubations with primary cells were done in triplicates using 10–18 samples from CLL patients and 2–3 healthy donors (see respective figures for the number of samples included in each incubation). Concentrations ranged between 0 and 100 μM, the optimal range having been determined in preliminary experiments. Concentrations are indicated for each compound and experiment in the respective figures. Cells incubated with vehicle only served as controls, vehicle did not exceed 1% of experimental volume. Viability was calculated as the mean of all replicates and was normalized to vehicle control (set as 100%).

See also  Best cbd oil for sleep and fastest shipping

Primary cells from both CLL patients and healthy donors were tested in suspension culture for 48 h at 37°C at concentrations of 3×10 6 cells/ml in 96 well plates before viability was measured. For co-culture incubations of CLL cells, 2×10 5 M2-10B4 cells were seeded in 12 well plates and incubated for 24 h under standard conditions. On the next day, primary CLL cells were transferred to the wells at a concentration of 3×10 6 cells/ml. Compounds were added and plates were incubated for 48 h at 37°C. Then, CLL cells were transferred to 96 well plates before carrying out the CellTiter Blue assay.

To determine whether CB1&2 agonists and antagonists exert an additive cytotoxic effect when combined with fludarabine, CLL primary cells (n = 5) were prepared for co-culture incubation as described above. Cells were treated with increasing concentrations of ACEA and JWH133, and single concentrations of AM251, AM630, and CNB. Using the same setup, cells were pre-incubated with cannabinoids for 30 minutes before 5 μM fludarabine was added. For comparison, primary cells were incubated in co-culture with vehicle alone, fludarabine (5 μM) alone, AMD3100 (0,629 mM) alone, and were pre-treated with AMD3100 for 30 minutes before fludarabine was added. Plates were left for 48 h at 37°C before viability was determined.

To evaluate a potential effect of serum on cannabinoid toxicity, primary CLL cells from 5 patients were tested at different FCS concentrations (1, 2.5, 5, and 10%) both in suspension and in co-culture with M2-10B4. Preparation of experiments and determination of viable cells were done as described above.

Migration assays

CLL cells of 7 patients were sorted using the Human B cell Enrichment Kit (CatNo.8804-6818-74, Affymetrix eBiosciences, San Diego, CA, USA) following the manufacturer´s recommendations. Resulting CD19+CD5+ cell purity was 97.7% ± 1.04. Cells (5×10 6 cells/ml) were incubated with AM251, AM630, JWH133, ACEA, or with a combination of antagonist + agonist. Incubations with vehicle, with AMD3100, and without CXCL12 served as controls. Compound concentrations were chosen to exert minimal cytotoxic effects on cells. A list of concentrations and combinations can be found in S2 Table.

Six hundred μl RPMI1640 without phenol red were pipetted into the bottom chambers of 24-well polycarbonate membrane transwell plates (6.5 mm diameter, 5.0 μm pore size; Corning Inc., Corning, NY, USA), CXCL12 (0,1 μg/ml) or vehicle (0.1% PBS with 0.1% BSA) was added to the medium. After addition of transwell inserts, 100 μl of treated cells were transferred to the upper chamber. Migration was allowed to occur for 4h at 37°C. The contents of both the upper and the bottom wells were transferred to separate tubes, centrifuged, pellets were resuspended in PBS, and cells were counted by FACS on a BD FacsCanto II Flow Cytometer.

A similar set of experiments was done with unsorted PBMC from CLL patients (N = 5) for which cells were pre-incubated with agonist, antagonist, or the combination of both before being transferred to the transwell plates. Incubation time was 0.5 h for antagonist, 1 h for agonist, and 1.5 h for antagonist plus agonist. An overview of these combinations can also be found in S2 Table. In this setting, cells were counted in a Neubauer counting chamber.

Migration is shown as means + SD of the migration index which was calculated as the number of cells migrated in the presence of compound divided by the number of cells migrated in the presence of vehicle [61–63].

Statistical methods

CNR1 and 2 expressions are given as median, quartiles and range. The median was set as cut-off for CNR expression to distinguish CNR high and low expression groups. The impact of CNR mRNA expression on survival was tested by Kaplan-Meier plots (P values from log-ranks tests) and quantified by hazard-ratios in univariate and multivariable Cox regression analyses. Progression free survival was calculated from first treatment to progression, overall survival was measured from date of first diagnosis to follow up and treatment free survival was calculated from date of first diagnose till first treatment or follow up. Prognostic markers were compared between groups using Chi 2 -tests. Pairwise comparison of migration experiments and viability assays was done by t-test. For this, IC50 values were compared between CLL suspension and co-culture at the same compound concentration, between CLL suspension culture and healthy donors in suspension at the same compound concentration, and—for reduced serum conditions—between CLL suspension and co-culture at the same compound and serum concentrations. In addition, again for reduced serum conditions, IC50 values were compared within CLL suspension and co-culture experiments, respectively, at the same compound but at different serum concentrations. Pearson correlation and linear regression analyses were performed to determine a potential association between IC50 values and CNR mRNA expression. P-values ≤ 0,05 were considered statistical significant. Computations were performed using SPSS or GraphPad Prism.

Results

CNR mRNA and protein expression in CLL

CNR1 mRNA expression ranged from 0.00 to 140.39 with a median expression of 1.52. CNR2 mRNA had a median expression of 3.77, ranging from 0.06 to 26.54. Receptor expression of healthy, CD19 sorted B-cells was set as 1.

Using the median expression as cut-off, patients were split into high and low expressing groups for both receptors. Medians (ranges) were 0.23 (0.00–1.41) and 7.16 (1.52–140.39) for the CNR1 low and high expressing groups, and 2.35 (0.06–3.72) vs. 5.41 (3.77–26.54) for CNR2 low and high expressing groups, respectively. The number of patients in the 4 groups were 31 in the double low expressing group, 44 in the mixed group (one receptor high, the other low expressing), and 32 were double high expressers.

In order to determine whether receptor protein expression follows mRNA expression, we carried out Western blots using different commercially available antibodies (listed in Methods). In these experiments, we included recombinant proteins for both receptors as positive controls for CB1&2 expression. Despite testing different protein extraction and transfer protocols, we were not able to a) determine specific bands, and b) detect differences in receptor expression (data not shown). We concluded that the antibodies used in our experiments were unspecific, most likely showing crossreactivity with both receptors and/or detecting other proteins concomitantly. Subsequently, protein data were excluded from any analysis.

CNR1 but not CNR2 mRNA is a prognostic marker in CLL

Next, we analyzed patient data with regard to high and low mRNA expression of CNR1 and CNR2, respectively. We found that, based on the level of CNR1 expression, patient characteristics were significantly different for two prognostic markers. Thus, persons displaying high CNR1 mRNA expression were more likely to have unmutated IgHV genes (54.3 vs. 32.7%; p = 0.033), or high CD38 expression (42.6 vs. 21.6%; p = 0.026) ( Table 1 ). In addition, CNR1 high expressing patients had a significantly shorter overall and treatment free survival (OS; TFS) (OS: HR = 8.615; 95% CI 1.947–38.112; p = 0.001; TFS: HR = 2.770; 95% CI 1.603–4.785; p

Table 1

CNR1 low CNR1 high p-value
Age at diagnosis [years]
Median (Range) 59 (25–85) 63 (39–82)
Sex [%]
Female:Male 47.2:52.8 38.9:61.1
CD19+CD5+ cells [%]
Median (Range) 79.0 (28–97) 86.0 (42–97)
Mutational status * [%]
Unmutated 32.7 54.3 0.033
Mutated 67.3 45.7
CD38 [%]
Low < 30 78.4 57.4 0.026
High ≥ 30 21.6 42.6
Median (Range) 5.0 (0–91) 23.0 (0–85)
Binet at diagnosis [%]
A 88.5 81.1 0.296
B/C 11.5 18.9
Lymphocyte doubling time [%]
Low < 1 year 18.0 30.4 0.154
High ≥ 1 year 82.0 69.6
Del13q [%]
Unmutated < 5.0 46.0 50.0 0.686
Mutated ≥ 5.0 54.0 50.0
Del11q [%]
Unmutated < 8.6 84.0 71.2 0.121
Mutated ≥ 8.6 16.0 28.8
Del17p [%]
Unmutated < 10.2 90.0 92.3 0.681
Mutated ≥ 10.2 10.0 7.7
Tris12 [%]
Unmutated < 3.7 86.0 92.3 0.305
Mutated ≥ 3.7 14.0 7.7
Rearr14q [%]
Unmutated < 3.0 91.8 81.6 0.136
Mutated ≥ 3.0 8.2 18.4

*Cut-off 98% germline homology. Abbreviations: Del, deletion; Tris, trisomy; Rearr, rearrangement. P-values (Pearson Chi-Square) reaching statistical significance are bold.

(A) High expressing patients had a mean overall survival (OS) of 153 months compared to 277 months in low expressing patients (p = 0.001). (B) The mean treatment free survival (TFS) was 75 months in the CNR1 high group vs. 150 months in the CNR1 low group (p<0.0001).

In contrast, no significant differences could be observed between CNR2 high and low expression groups. CNR2 mRNA expression levels were not associated with any of the established prognostic markers, nor did high expression translate into shorter OS, TFS, or PFS (p = 0.763, p = 0.229, p = 0.089, respectively (log rank)). Patient characteristics in relation to CNR2 high and low expression are listed in S3 Table, Kaplan Meier Analyses for OS and TFS are shown in S1 Fig.

Impact of cannabinoids on viability of primary cells

A dose dependent reduction of viability could be observed in CLL primary cells with increasing concentration of drug both in suspension and co-culture experiments ( Fig 2 ).

PBMC from CLL patients were incubated in triplicates both in suspension culture and in co-culture with M2-10B4 mouse fibroblast cells in increasing concentrations of compounds. Viability was determined after 48h, mean values and standard deviations are shown. (A) (R)-(+)-methanandamide (N = 10). (B) (-)-cannabidiol (N = 18). (C) ACEA (N = 16). (D) JWH133 (N = 16). (E) AM251 (N = 16). (F) AM630 (N = 16). For ACEA, JWH133, and AM251, the 50% reduction in viability required for IC50 calculation could not be reached in co-culture. Note different scale on x-axis in A and D.

This reduction was more pronounced in suspension culture compared to co-culture for the selective CB1 antagonist AM251 (p = 0.0431) compared to the selective CB1 agonist ACEA (p = 0.1855) and more pronounced for the selective CB2 agonist JWH133 (p = 0.0527) compared to the selective CB2 antagonist/inverse agonist AM630 (p = 0.1353). No differences between the two culture conditions could be observed for (R)-(+)-methanandamide (RM) (CB1 agonist, activity against vanilloid and other G-protein coupled receptors and ion channels) and (-)-cannabidiol (CNB) (weak CB1 antagonist, CB2 inverse agonist, interacting also with GPR55, vanilloid receptors and opioid receptors) (p = 0.1464 and p = 0.6549, respectively). These data suggest a protective effect of the fibroblasts in incubations with AM251 and JWH133. The loss of this effect, at least for RM, CNB, and AM630, may be explained by the cytotoxicity these compounds exerted on M2-10B4 fibroblast cells alone, IC50 values being in the range of CLL primary cells ( Table 2 ). A dose dependent reduction of viability was also observed for PBMC of healthy individuals in suspension ( Fig 3 ), IC50 values were in the same range as those of primary CLL cells under the same conditions except for RM and AM630 for which IC50 values were approximately twice as high ( Table 2 ).

PBMC from 3 healthy donors were incubated in triplicates in suspension culture in increasing concentrations of compounds. Viability was determined after 48h, mean values and standard deviations are shown. (A) (R)-(+)-methanandamide. (B) (-)-cannabidiol. (C) ACEA. (D) JWH133. (E) AM251. (F) AM630. Note different scale on x-axis in A and D, note different scale on y-axis in D.

Table 2

RM CNB ACEA JWH133 AM251 AM630
CLL suspension 33.19 21.74 31.78 75.68 9.43 12.08
CLL co-culture 29.27 16.78 NR NR NR 27.64
HD suspension 60.13 15.09 39.01 78.12 11.44 28.51
M2-10B4 34.55 13.52 NR NR NR 28.27

IC50 values (μM) are based on the mean values of 10 to 18 CLL PBMC and 3 HD PBMC which were incubated in triplicates in a concentration range of 0 to 100 μM of drugs for 48h. Viability of CLL cells was assessed both in suspension culture and in co-culture with the mouse fibroblast cell line M2-10B4. Healthy donor (HD) cells were incubated in suspension culture only. Abbreviations: RM, (R)-(+)-methanandamide; CNB, (-)-cannabidiol; NR, IC50 not reached.

Cannabinoids reportedly interfere with cell-cell crosstalk which may potentially affect the therapeutic efficacy of drugs. Thus, we combined cannabinoids with fludarabine to explore potential synergies between tested drugs. Pre-incubation with cannabinoids before the addition of fludarabine generally led to a reduction in cell viability compared to cannabinoids alone (N = 5) (S2 Fig). These differences, however, were not statistically significant. Exception was the CB1 agonist ACEA where the concentration dependent cytotoxicity increased reaching statistical significance in the 40 μM incubations (10 μM ACEA vs. 10 μM ACEA+Flu p = 0.117; 20 μM ACEA vs. 20 μM ACEA+Flu p = 0.076; 40 μM ACEA vs. 40 μM ACEA+Flu p = 0.047). Impact of cannabinoids and combinations were not statistically significant either when compared to controls (N = 6) with fludarabine alone, with AMD3100 alone or with AMD3100+Fludarabine (S2 Fig).

Serum factors may interact with cannabinoids resulting in a reduction in biologic drug activity. To investigate such an effect, primary cells were incubated with increasing concentrations of cannabinoids under reduced serum conditions, both in suspension and in co-culture (N = 5). Although not statistically significant within culture conditions, the results suggest that reduced serum concentrations added to the cytotoxic effect of the compounds, particularly at lower drug concentrations (1 and 2.5% vs. 5 and 10% FCS; S3 Fig), while this difference leveled out at higher compound concentrations. A pairwise comparison of corresponding compound concentrations between the two culture conditions ( Table 3 ) showed a significant advantage for CLL cells in co-culture at 1 and 2.5% serum concentrations and in most cases also at 5 and 10% serum concentrations, again underlining the importance of cell-cell interaction for CLL cell survival especially at low serum conditions. Once more CNB appeared to be an exception probably due to its impact on the feeder cells (Tables ​ (Tables2 2 and ​ and3 3 ).

Table 3

P-values of the pairwise comparison of IC50 values between suspension and co-culture in serum reduced experiments.

RM CNB ACEA JWH133 AM251 AM630
1% serum 0.0336 0.125 0.0311 0.0278 0.03 0.0278
2.5% serum 0.7254 0.03 0.0311 0.0412 0.0396 0.0468
5% serum 0.0154 0.3255 0.0279 0.0282 0.0823 0.0993
10% serum 0.05 0.2511 0.0468 0.0386 0.1021 0.0337

Student´s t-test was applied, p-values reaching statistical significance or borderline significance are bold. Abbreviations: RM, (R)-(+)-methanandamide; CNB, (-)-cannabidiol;

Of note, sensitivity to cannabinoids was not associated with mRNA expression of either of the receptors (S4 and S5 Figs) indicating that the cytotoxic affect exerted by the tested cannabinoids was not or only partially mediated by cannabinoid receptors.

Migration assays

In T-cells, cannabinoids were shown to inhibit CXCL12 directed migration. Considering the importance of the CXCL12-CXCR4 axis in CLL, we studied whether and to which degree cannabinoids might interfere with CXCL12 mediated CLL cell migration. As shown in Fig 4 , vehicle did not induce any significant changes in migration towards CXCL12, in contrast AMD3100 inhibited migration significantly (p = 0.0006) compared to controls. Both CB1&2 agonists (ACEA, JWH133) and antagonists (AM251, AM630) did not significantly influence the migratory behavior of purified CLL cells towards CXCL12. Likewise, combination of antagonist plus corresponding agonist (AM251+ACEA for CB1; AM630+JWH133 for CB2) did not lead to significant changes, either, compared to migration in vehicle containing controls. CLL cells incubated without CXCL12 showed a significantly reduced migratory behavior compared to controls (p <0.0001) (Fig 4 ).

B-cell enriched primary cells of 7 CLL patients (97.7% ± 1.04 CD19+CD5+) were incubated in transwell plates for 4h before the number of migrated cells was determined. Control experiments included CXCL12 alone (control), no CXCL12 (control w/o CXCL12), incubation with vehicle (DMSO, ethanol), and incubation with the CXCR4 inhibitor AMD3100. CLL cells were incubated either with agonist (ACEA, JWH133), antagonist (AM251, AM630), or a combination of antagonist plus agonist before migration (CB1: AMS251&ACEA; CB2: AM630&JWH133). Bars represent the mean values of migration indices + standard deviations, hatched lines indicate experimental blocks. *p = 0.0006; **p

See also  Can cbd oil be applied topically for pain

Experiments with unsorted PBMC from CLL patients using the same compound concentrations but pre-incubating the cells before migration led to similar results, with significantly reduced migration in the absence of CXCL12 (p<0.0001), significant inhibition of migration towards CXCL12 in the presence of AMD3100 (p = 0.0016), and no significant block of migration in incubations with cannabinoids (S6 Fig).

Discussion

Known for their psychoactive effects, cannabinoids have been used for centuries both for recreational and medical purposes. Upon the discovery of the receptors involved, drugs were developed exploiting the receptors in order to treat accompanying symptoms in various diseases [4, 64, 65]. In addition, cannabinoids displayed a number of anti-tumor effects in solid tumors [4, 20, 22, 24, 26, 27, 29] and in hematologic malignancies [23, 66] which led us to determine the expression of cannabinoid receptors and to evaluate the therapeutic potential of cannabinoids in CLL.

We found that both cannabinoid receptors were overexpressed in CLL cells compared to healthy B-cells. On the mRNA level, CNR2 had a higher median expression than CNR1, which, on the other hand, had a wider range of expression compared to CNR2. When split into high and low expressing groups based on median receptor expression, only CNR1 could be shown to have prognostic value, not CNR2. Although previous publications reported an overexpression of one or both cannabinoid receptors in hematologic malignancies [22, 23, 34, 49, 67], reports regarding a potential role of this overexpression in the clinics are few [49–51]. More information is available for solid tumors, where in most studies either CB1 or CB2 expression was linked to poorer patient outcome [32, 43–48, 68]. Here, we provide for the first time evidence pertaining prognostic relevance of cannabinoid receptors in lymphoid malignancies.

On the protein level, however, we could not detect any differences in expression in high and low expressing patient groups. Although some studies reported the detection of receptor protein levels in relation to mRNA expression [22, 23], we were not able to reproduce this observation. This was likely due to a lack of specificity of the antibodies used in this study. Our data are corroborated by two studies in which several commercially available antibodies were tested for CB1&2, respectively. Both groups reported great differences between the claimed specificity and the observed high degree of unspecific detection of protein by these antibodies [69, 70]. Therefore, cannabinoid receptor protein expression was excluded from further analysis.

We continued by evaluating the two receptors as potential therapeutic targets in CLL. For this, a specific agonist and antagonist pair was chosen for each receptor (CB1: agonist ACEA, antagonist AM251; CB2: agonist JWH133, antagonist/inverse agonist AM630). While these pairs are highly selective for either CB1 or CB2, respectively, a certain degree of promiscuous activity appears to be a common feature to them all. AM251 and AM630 are also ion channel activators [71], AM251 additionally acts as agonist at GPR55 and antagonist at μ-opioid receptors [40, 72]. JHW133 not only acts on CB2 but also the TRPV1 vanilloid receptor plus showed off-target effects in chemotaxis experiments in macrophages [73, 74]. Besides these two ligand pairs, two compounds were included which are being widely used in cytotoxic experiments and which have long been known for broader activity: R-(+)-methanandamide (RM) is CB1 agonist, shows activity at GPR and ion channels [1, 25], and (-)-cannabidiol (CNB) is a weak CB1 antagonist/CB2 inverse agonist interacting also with GPR55, TRPVR1 vanilloid receptors and μ-opioid receptors [1, 39, 75, 76].

The dose dependent reduction in viability was more pronounced in suspension compared to co-culture. Although not statistically significant, this was reflected by different IC50 values under the two culture conditions and the fact that in some co-culture incubations IC50 values could not be calculated, in particular where the viability of M2-10B4 cells was not much impacted by the compounds. The protective effect of the microenvironment and of supporting cells in culture, particularly under treatment conditions, is known for CLL and for CLL cells [77–79].

The exceptions observed—RM, CBD, and AM630—can be explained by the toxicity these drugs exerted on the feeder cells. Likewise, healthy PBMC showed IC50 values similar to primary leukemic cells alone except for RM which appeared to have much less impact on healthy PBMC compared not only to the mouse fibroblasts but also to neoplastic cells, an observation that has been reported before [33]. In this line, Almestrand reported that normal PBMC subsets did not show significant changes in persons treated with rimonabant for obesity [80] while this drug induced cell death in CB1 expressing primary MCL cells in cell culture [22]. Likewise in vitro, rimonabant, a CB1 antagonist/inverse agonist and μ-opioid antagonist, appeared to influence healthy PBMC to a much lesser degree compared to Jurkat and U937 cells [27]. In the presented study, also AM630, the CB2 antagonist/inverse agonist, had less cytotoxic impact on healthy PBMC compared to primary CLL cells. Together this information suggests that differential or equal cytotoxic impact of cannabinoids on normal and malignant cells may be highly compound specific.

It is noteworthy, that the cytotoxic effect was not associated with either CNR1 and/or CNR2 mRNA levels. Our data differ in this respect from other studies in which the cytotoxic effect of cannabinoids observed in MCL, CLL, and HL cell lines were attributed to the overexpression of cannabinoid receptors in these cells [22, 23, 34] while other, low expressing cell lines were not inhibited [22, 23]. They also suggest that the cytotoxic effects of cannabinoids very likely are mediated by more than one receptor and one mechanism even in cases where molecules are reportedly highly specific. As mentioned above, a certain degree of unspecific action seems to be inherent to the nature of cannabinoids, an aspect that is increasingly reflected by the literature reporting a broad range of cannabinoid activity on a variety of G-protein coupled and other receptors [1, 39–41, 66, 71–74, 81–83].

Cannabinoids supposedly interact with factors in serum which may lead to an inhibition of cellular metabolism, cannabinoid functionality, and reduction in cytotoxicity [66, 84–86]. As described previously [20, 22, 33], we found that particularly low serum concentrations enhanced the reduction in viability under both culture conditions, although less so in co-culture. These serum dependent differences, however, diminished at the highest drug concentrations where the cytotoxic effects of cannabinoids gained impact. While we do not exclude an interaction of cannabinoids with factors in serum causing an attenuation of cannabinoid cytotoxicity, we think that the impact of such effects in the course of CLL therapy is limited. In peripheral blood, under normal serum conditions, higher concentrations of cannabinoids would be required that would also lead to a high degree of toxicity in healthy cells. Under serum reduced conditions, in lymphoid organs, again higher concentrations of cannabinoids would be required since the microenvironment attenuates at least part of the cytotoxic effect of cannabinoids, and, again, healthy cells would also be influenced severely. The net effect under both conditions would be the same: a necessity of higher drug concentrations, a higher portion of healthy cells also being effected, but no real gain in killing CLL cells.

Another feature of cannabinoids should be noted. The ambivalent and variable mechanism of cannabinoid action may cause block or activation of a receptor at low concentrations while high concentrations will lead to cell death [66, 83]. Such bimodal behavior may explain the increases of viability at low concentrations in some experiments.

Interaction with the microenvironment is extremely important for CLL cells and relies to a large part on the CXCR4-CXCL12 axis, which also serves as target for therapies [79, 87]. Based on the blockage of CXCL12 induced chemotaxis and inhibition of similar receptor-ligand pairs in various lineages of peripheral blood cells [17, 88–91], we studied potential synergistic and inhibitory effects of cannabinoids with regard to the standard therapeutic agent fludarabine. Although enhanced reduction in viability could be observed in all experiments, pre-incubation with cannabinoids before addition of fludarabine did not add significantly to the reduction in viability except for the CB1 specific agonist ACEA at the highest concentration. On the other hand, our controls using the CXCR4-specific inhibitor AMD3100 did not lead to a significant synergistic effect, either. This is in contrast to a previous study where pre-treatment with AMD3100 led to a statistically significant reduction of viable CLL cells in co-culture after incubations with different drugs [79]. Although both studies used less than 10 samples for this assay (N = 8 in the Stamatopoulos study, N = 5 in this study), the difference might be attributed to the high biological variability of CLL samples and/or the different type of feeder cells used (mesenchymal stromal cells in the Stamatopoulos study, M2-10B4 mouse fibroblasts in this study) [79].

Although the impact of AMD3100 in co-incubatory experiments was limited, we found this drug to significantly block migration of CLL cells towards CXCL12. In contrast, none of our incubations with cannabinoids led to a significant inhibition of the migratory behavior of primary cells. This indicates that AMD3100 successfully interfered with the CXCR4 receptor in our experiments, but also suggests that the previously reported inhibition of the CXCR4-CXCL12 axis using cannabinoids in different blood cells [17, 89] is not valid in the CLL setting and will most likely not beneficially contribute to therapeutic regimens.

It is still unclear what role the endocannabinoid system my play in cancer and there still is much information to be gathered on how to exploit this system for anti-cancer therapy. At least for MCL some knowledge has accumulated on how cannabinoids effect malignant cells, in addition to the possibility that the endocannabinoid system even may be involved in leukemic development [22, 24, 51, 66, 92]. Also very prominent are the studies evaluating the interference of cannabinoids with cell-cell cross-talk by different chemotaxis pairs [17, 89], which also would constitute an important aspect of targeting therapies.

We, however, could not corroborate these previous findings regarding a substantial inhibition of chemotaxis and a block of the CXCR4-CXCL12 axis by cannabinoids for CLL cells. While we did find that cannabinoids reduced viability of CLL primary cells considerably independent of CNR mRNA expression, we found healthy cells to be affected to the same degree. Thus—and in contrast to other malignancies—our data suggest cannabinoids to be of poor therapeutic potential for treatment of CLL although CNR1 mRNA expression could be determined as novel prognostic marker. Their role in CLL notwithstanding, cannabinoids may still proof useful for anti-tumor therapy in other, selected hematologic malignancies and solid tumors in which the potential of cannabinoids will have to be studied accordingly.

Supporting Information

S1 Fig

(A) Mean overall survival (OS) for high expressing patients was 196 months vs. 230 months for low expressing patients (N = 107; p = 0.763). (B) Mean treatment free survival (TFS) in CNR2 high and low mRNA expressers was 100 months vs. 135 months in high and low expression groups, respectively (N = 107; p = 0.2290). One hundred and seven patients were included in the analysis, median mRNA expression of CNR2 (3.77) was used as cut-off.

S2 Fig

CLL primary cells (N = 5) were incubated in triplicates in co-culture with M2-10B4 mouse fibroblasts and incubated for 30 minutes with increasing concentrations of cannabinoids before fludarabine (5 μM) was added. Viability was determined after 48h. Incubations with vehicle served as control. For comparison, cells were incubated with fludarabine alone, with AMD3100 alone, and with AMD3100 in combination with fludarabine (N = 6). Mean values and standard deviations are shown. Hatched lines mark experimental blocks. The synergistic effect of the combination 40 μM ACEA with 5μM fludarabine was significantly different from the effect of 40μM ACEA alone. *p = 0.047. Abbreviation: CNB, (-)-cannabidiol.

S3 Fig

PBMC of 5 CLL patients were incubated in triplicates for 48h in increasing compound concentrations at 1%, 2.5%, 5%, and 10% serum containing medium in suspension and in co-culture with M2-10B4 mouse fibroblasts before viability was measured. Mean values and standard deviations are shown. (A) (R)-(+)-methanandamide in suspension and (B) in co-culture. (C) (-)-cannabidiol in suspension and (D) in co-culture. (E) ACEA in suspension and (F) in co-culture. (G) JWH133 in suspension and (H) in co-culture. (I) AM251 in suspension and (J) in co-culture. (K) AM630 in suspension and (L) in co-culture. Note different scales on x- and y-axes.

S4 Fig

PBMC from CLL patients were incubated in triplicates in increasing compound concentrations in suspension and co-culture with M2-10B4 mouse fibroblast cells for 48h before viability was measured. (A) (R)-(+)-methanandamide (N = 10). (B) (-)-cannabidiol (N = 18). (C) ACEA (N = 16). (D) JWH133 (N = 16). (E) AM251 (N = 16). (F) AM630 (N = 16). The x-axis shows the measured mRNA expression for each CLL sample tested (healthy CD19 sorted cells set as 1) from highest (left) to lowest (right) expression. Absent values may indicate that i) sample was not tested, or ii) IC50 could not be calculated, or iii) 50% viability reduction could not be achieved. Note different scales on Y-axis for A and D.

S5 Fig

PBMC from CLL patients were incubated in triplicates in increasing compound concentrations in suspension and co-culture with M2-10B4 mouse fibroblast cells for 48h before viability was measured. (A) (R)-(+)-methanandamide (N = 10). (B) (-)-cannabidiol (N = 18). (C) ACEA (N = 16). (D) JWH133 (N = 16). (E) AM251 (N = 16). (F) AM630 (N = 16). The x-axis shows the measured mRNA expression for each CLL sample tested (healthy CD19 sorted cells set as 1) from highest (left) to lowest (right) expression. Absent values may indicate that i) sample was not tested, or ii) IC50 could not be calculated, or iii) 50% viability reduction could not be achieved. Note different scales on Y-axis for A and D.

S6 Fig

Primary cells of 5 CLL patients were pre-incubated with cannabinoids before being transferred to transwell plates and incubated for 4h for migration. Control experiments included CXCL12 alone (control), no CXCL12 (control w/o CXCL12), incubation with vehicle (DMSO, ethanol), and incubation with the CXCR4 inhibitor AMD3100. CLL cells were incubated either with agonist (ACEA, JWH133) or antagonist (AM251, AM630) before migration. In addition, cells were treated with antagonist before agonist incubation before migration was allowed (CB1: AMS251&ACEA; CB2: AM630&JWH133). Bars represent mean values of migration indices + standard deviations, hatched lines indicate experimental blocks.* p = 0.0016; ** p

Cannabis Extract Treatment for Terminal Acute Lymphoblastic Leukemia with a Philadelphia Chromosome Mutation

This is an Open Access article licensed under the terms of the Creative Commons Attribution-NonCommercial 3.0 Unported license (CC BY-NC) (www.karger.com/OA-license), applicable to the online version of the article only. Users may download, print and share this work on the Internet for noncommercial purposes only, provided the original work is properly cited, and a link to the original work on http://www.karger.com and the terms of this license are included in any shared versions.

See also  Cbd oil for sale red deer

Abstract

Acute lymphoblastic leukemia (ALL) is a cancer of the white blood cells and is typically well treated with combination chemotherapy, with a remission state after 5 years of 94% in children and 30–40% in adults. To establish how aggressive the disease is, further chromosome testing is required to determine whether the cancer is myeloblastic and involves neutrophils, eosinophils or basophils, or lymphoblastic involving B or T lymphocytes. This case study is on a 14-year-old patient diagnosed with a very aggressive form of ALL (positive for the Philadelphia chromosome mutation). A standard bone marrow transplant, aggressive chemotherapy and radiation therapy were revoked, with treatment being deemed a failure after 34 months. Without any other solutions provided by conventional approaches aside from palliation, the family administered cannabinoid extracts orally to the patient. Cannabinoid resin extract is used as an effective treatment for ALL with a positive Philadelphia chromosome mutation and indications of dose-dependent disease control. The clinical observation in this study revealed a rapid dose-dependent correlation.

Presentation of the Case

A 14-year-old female, P.K., presented with symptoms of weakness, shortness of breath and bruising when she was taken to the Hospital for Sick Children, Toronto, Canada, on the 10th March 2006. She was diagnosed with acute lymphoblastic leukemia (ALL), with >300,000 blast cells present. Acute chemotherapy followed by a standard chemotherapy regimen went on for 6 months after the diagnosis. Upon further analysis, she was found to be positive for the Philadelphia chromosome mutation. A mutation in the Philadelphia chromosome is a much more aggressive form of ALL. When standard treatment options were unsuccessful, a bone marrow transplant was pursued. She successfully received the transplant in August 2006 and was able to be released from isolation 45 days later. She was observed posttransplant by following the presence of blast cells, noted 6 months after treatment. Consequently, in February 2007, aggressive chemotherapy procedures (AALL0031) were administered along with a tyrosine kinase inhibitor, imatinib mesylate (Gleevac), 500 mg orally twice a day. In November 2007, 9 months after the transplant, the presence of premature blast cells was observed and it was determined that another bone marrow transplant would not be effective. In February 2008, in an effort to sustain the patient, another tyrosine kinase inhibitor, disatinib (Sprycel), was administered at 78 mg twice a day with no additional rounds of chemotherapy. The patient experienced increased migraine-like headaches in June 2008. After conducting a CT scan of the head in July 2008, cerebellitis was noted. It was assumed by the primary oncologist that the blast cells could have infiltrated the CNS and be present in the brain, although none were noted in the blood. By October 2008, ten treatments of radiation therapy had been administered to the brain.

On the 4th February 2009, blood was noted in the patient’s stools and a blood cell count revealed the presence of blast cells. As a result, all treatment including the disatinib was suspended and the patient’s medical staff acknowledged failure in treating her cancer. It was charted by the patient’s hematologist/oncologist that the patient ‘suffers from terminal malignant disease. She has been treated to the limits of available therapy… no further active intervention will be undertaken’. She was placed in palliative home care and told to prepare for her disease to overwhelm her body and from which she would suffer a stroke within the next 2 months.

Cannabinoid Treatment

After this, disease progression was observed with rising counts of blast cells. The patient was receiving frequent blood transfusions and platelets during this period. Through research conducted by the patient’s family, it was observed, in a particular paper by Guzman [1] published in Nature Reviews Cancer, that cannabinoids have been shown to inhibit the growth of tumor cells in culture and in animal models by modulating key cell-signaling pathways. Cannabinoids are usually well tolerated and do not produce the generalized toxic effects of conventional chemotherapies. The family found promise in an organization known as Phoenix Tears, led by Rick Simpson who had treated several cancers with hemp oil, an extract from the cannabis plant. Rick worked with the family to help them prepare the extract.

From the 4th to the 20th of February, the patient’s blast cell count had risen from 51,490 to 194,000. The first dose of cannabinoid resin, also referred to as ‘hemp oil’, was administered orally (1 drop about the size of half a grain of rice) at 6:30 a.m. on the 21st February 2009 (day 0 in fig. ​ fig.1). 1 ). A 2-ounce Cannabis indica strain (known as ‘Chronic Strain’) was used to extract 7.5 ml of hemp oil using 1.2 liters of 99%-isopropyl-alcohol solution, which was boiled off in a rice cooker. Immediately after the dosing, the patient attempted to vomit; nausea had been observed previously and is common with this condition. To deal with the bitter taste and viscous nature of the hemp oil, it was suspended in honey, a known natural digestive aid, and then administered to the patient in daily doses. The objective was to quickly increase the frequency and amount of the dose and to hopefully build up the patient’s tolerance to cannabinoid resin (refer to fig. ​ fig.1). 1 ). The patient was observed to have periods of panic early on during administration of the hemp oil, along with increased appetite and fatigue.

Blast cell counts, days 0–15: Chronic Strain.

The blast cell count reached a peak of 374,000 on the 25th February 2009 (day 5), followed by a decrease, which correlated with the increasing dose. The daily dosing is the amount administered per dose; the doses were initially given once per day up to a total of 3 times per day by day 15, and were continued with the same average frequency throughout the treatment. A decreased use of morphine for pain, an increase in euphoria symptoms, a disoriented memory and an increase in alertness were observed; these are typical with cannabinoid use.

After day 15, the original Chronic Strain had been consumed and administration of a new strain (referred to as Hemp Oil #2) was started. This was obtained by the family from an outside source. It was noted that administering the same dose yielded a decreased response in terms of the side effects of euphoria and appetite, and the patient suffered more nausea with this hemp oil. The blast cells began to increase, demonstrated in figure ​ figure2 2 .

Blast cell counts, days 18–39: Hemp Oil #2.

There is a wide amount of variance in cannabinoid concentration amongst different strains and even in the same strain with changes in growing conditions. The amount of each dose was increased to match the response of the blast cells that had been declining previously (fig. ​ (fig.1). 1 ). After day 27, there was another peak blast cell count of 66,000 followed by a rapid decrease. There were elevated levels of urate present in the blood with corresponding joint pain; it was established that this was caused by tumor lysis syndrome of the blast cells. Allopurinol was administered.

On the 1st April 2009 (day 41), an infected central line with tunnel infection was noted on a blood test and the patient was admitted with a heavy antibiotic regimen of intravenous tazocin, gentamicin and vancomycin. On day 43, a new batch of hemp oil from an Afghan/Thai strain (referred to as Hemp Oil #3) prepared by the family was administered. A stronger psychosomatic response and increased fatigue were observed, so dosing was adjusted to 0.5 ml, shown in figure ​ figure3. 3 . Due to hospital restrictions, dosing was limited to twice a day.

Blast cell counts, days 44–49: Hemp Oil #3.

A new batch of hemp oil was obtained by the family from an outside source and the dosing regimen continued twice a day, shown in figure ​ figure4 4 .

Blast cell counts, days 50–67: Hemp Oil #4.

After returning home from the hospital on the 11th April (day 51), the patient suffered from intense nausea, an inability to eat and overall weakness. On the 13th April, the patient was readmitted to the SickKids Hospital and was treated for refeeding syndrome. This was the outcome of stopping total parenteral nutrition too quickly and causing shock to the patient’s body while she was being treated for the infection. The dosing regimen was intermittent until day 59, remaining at 1–2 doses per day of 0.5 ml. As the blast cells began to increase and the patient’s appetite increased, the dosing frequency was again increased to 3 times per day starting on day 62, and the amount administered was increased from day 65.

On day 68, a new batch of medicine was obtained by the family from an outside source (referred to as Hemp Oil #5), shown in figure ​ figure5 5 .

Blast cell counts, days 69–78: Hemp Oil #5.

Dosing was maintained 3 times a day at 1.0 ml. On day 78, the patient had stomach pain in the morning and was admitted to hospital. Upon X-ray, it was noted that gastrointestinal bleeding had occurred. The patient was under a DNR order and ultimately passed away due to the bowel perforation. A prior history of pancolitis documented by CT scan in March 2009 pointed to neutropenic colitis with perforation as the cause of death. Furthermore, prior to starting on the hemp oil treatment, the patient had been extremely ill, severely underweight and had endured numerous sessions of chemotherapy and radiation therapy in the course of 34 months.

As reported by Hematology/Oncology at SickKids: ‘At admission her total WBC was 1.4, hemoglobin was 82, platelet count 8,000. She was profoundly neutropenic… a prior history of pancolitis documented by CT scan in March 2009 was neutropenic colitis with perforation… her abdomen was distended and obviously had some signs of diffuse peritonitis. The abdomen X-ray was in favour a perforation…she passed away at 10:05 in the present (sic) of family…’.

Discussion

Figure ​ Figure6 6 is a summary of dose response to all the batches of hemp oil administered over a total of 78 days.

Response to hemp oil treatment over 78 days.

The results shown here cannot be attributed to the phenomenon of ‘spontaneous remission’ because a dose response curve was achieved. Three factors, namely frequency of dosing, amount given (therapeutic dosing) and the potency of the cannabis strains, were critical in determining response and disease control. In the figure, it can be seen that introducing strains that were less potent, dosing at intervals >8 h and suboptimal therapeutic dosing consistently showed increases in the leukemic blast cell count. It could not be determined which cannabinoid profiles constituted a ‘potent’ cannabis strain because the resin was not analyzed. Research is needed to determine the profile and ratios of cannabinoids within the strains that exhibit antileukemic properties.

These results cannot be explained by any other therapies, as the child was under palliative care and was solely on cannabinoid treatment when the response was documented by the SickKids Hospital. The toxicology reports ruled out chemotherapeutic agents, and only showed her to be positive for THC (tetrahydrocannabinol) when she had ‘a recent massive decrease of WBC from 350,000 to 0.3’ inducing tumor lysis syndrome, as reported by the primary hematologist/oncologist at the SickKids Hospital.

This therapy has to be viewed as polytherapy, as many cannabinoids within the resinous extract have demonstrated targeted, antiproliferative, proapoptotic and antiangiogenic properties. This also needs to be explored further, as there is potential that cannabinoids might show selectivity when attacking cancer cells, thereby reducing the widespread cytotoxic effects of conventional chemotherapeutic agents. It must be noted that where our most advanced chemotherapeutic agents had failed to control the blast counts and had devastating side effects that ultimately resulted in the death of the patient, the cannabinoid therapy had no toxic side effects and only psychosomatic properties, with an increase in the patient’s vitality.

The nontoxic side effects associated with cannabis may be minimized by slowly titrating the dosing regimen upwards, building up the patient’s tolerance. The possibility of bypassing the psychoactive properties also exists, by administering nonpsychoactive cannabinoids such as cannabidiol that have demonstrated antiproliferative properties. Furthermore, future therapies could examine the possibility of upregulating a patient’s endogenous cannabinoids to help combat leukemic cells. It goes without saying that much more research and, even more importantly, phase clinical trials need to be implemented to determine the benefits of such therapies. Laboratory analysis is critical to figure out the constituents/profiles/ratios of the vast cannabis strains that show the most favored properties for exerting possible anticancer effects. Despite the nonstandardization of the medicines, the dose was readily titrated according to the biological response of the patient and produced a potentially life-saving response, namely, the drop in the leukemic blast cell count.

There has been an abundance of research exhibiting the cytotoxic effects of cannabinoids on leukemic cell lines in the form of in vitro and in vivo studies [1, 2, 3, 4]. An oncology and hematology journal, Blood, has published numerous papers [2] over the years constructing the biochemical pathway to be elicited by the anticancer properties of cannabinoids. Our goal, upon examination of this significant case study which demonstrated complete disease control and a dose response curve, is to invest effort in and to focus on research and development to advance this therapy. An emphasis needs to be placed on determining the correct cannabinoid ratios for different types of cancer, the best method of administration, quality control and standardization of the cannabis strains and their growing conditions as well as therapeutic dosing ranges for various cancers contingent on staging and ages. Toxicity profiles favor therapies deriving from cannabis because toxicity within the body is greatly reduced and the devastating side effects of chemoradiation (i.e. secondary cancers or death) can be eliminated. It is unfortunate that this therapy does come with some unwanted psychosomatic properties; however, these might be eliminated by target therapies of nonpsychoactive cannabinoids such as cannabidiol which has garnered much attention as being a potent anti-inflammatory and possible antileukemic and anticancer agent. It is acknowledged that significant research needs to be conducted to reproduce these results and that in vitro studies cannot always be reproduced in clinical trials and the human physiological microenvironment. However, the numerous research studies and this particular clinical case are powerful enough to warrant implementing clinical trials to determine dose ranges, cannabinoid profiles and ratios, the methods of administration that produce the most efficacious therapeutic responses and the reproducibility of the results. It is tempting to speculate that, with integration of this care in a setting of full medical and laboratory support, a better outcome may indeed be achieved in the future.

References

1. Guzman M. Cannabinoids: potential anti-cancer agent. Nat Rev Cancer. 2003; 3 :745–755. [PubMed] [Google Scholar]

2. Powles T, Poele RT, Shamash J, Chaplin T, Propper D, Joel S, Oliver T, Liu WM. Cannabis-induced cytotoxicity in leukemic cell lines: the role of the cannabinoid receptors and the MAPK pathway. Blood. 2005; 105 :1214–1221. [PubMed] [Google Scholar]

3. McKallip RJ, Jia W, Schlomer J, Warren JW, Nagarkatti PS, Nagarkatti M. Cannabidiol-induced apoptosis in human leukemia cells: a novel role of cannabidiol in the regulation of p22 phox and Nox4 expression. Mol Pharmacol. 2006; 70 :897–908. [PubMed] [Google Scholar]

4. Murison G, Chubb C, Maeda S, Gemmell MA, Huberman E. Cannabinoids induce incomplete maturation of cultured human leukemia cells. Proc Natl Acad Sci. 1987; 84 :5414–5418. [PMC free article] [PubMed] [Google Scholar]