The CXCR4 antagonist plerixafor corrects panleukopenia in patients with WHIM syndrome

WHIM syndrome is a rare primary immunodeficiency disorder characterized by warts, hypogammaglobulinemia, infections, and myelokathexis.^1-3  Myelokathexis refers to the retention of mature neutrophils in BM, resulting in neutropenia. Most patients with WHIM syndrome are panleukopenic, which partially corrects if they undergo severe infection.⁴  The cause of almost all cases of WHIM syndrome is autosomal-dominant inheritance of carboxy-terminal truncation mutations that remove 10-19 amino acids from the chemokine receptor CXCR4.⁵  CXCR4^R343X, shortened by 19 amino acids, is the most common and best-studied variant in WHIM syndrome. When expressed in cell lines lacking endogenous CXCR4, CXCR4^R343X exhibits ∼ 2-fold enhanced signaling on exposure to its ligand CXCL12 (also known as stromal cell–derived factor-1).⁶  The mechanism involves loss of negative regulatory elements in the missing C-terminus and defective ligand-induced down-regulation of the receptor from the cell surface.^7,8  Because CXCR4 normally regulates leukocyte trafficking and in particular is important for neutrophil adhesion in BM, increasing its function may cause myelokathexis and neutropenia. Consistent with this, gene transfer of CXCR4^R343X phenocopies myelokathexis in both zebrafish and Nod-scid (ie, nonobese diabetes/severe combined immunodeficiency) mice.^9,10 

Current treatments for WHIM syndrome include G-CSF (filgrastim [Neupogen; Amgen Inc]) and intravenous immunoglobulin but are nonspecific, expensive, difficult to administer, and only partially effective.⁴  Recently, a small molecule CXCR4 antagonist named plerixafor (trade name Mozobil [Genyzme Corporation] and formerly known as AMD3100) was approved by the Food and Drug Administration (FDA) for mobilizing HSCs from BM to blood with G-CSF for the purpose of autologous transplantation after cytoreductive therapy in patients with non-Hodgkin lymphoma or multiple myeloma.^11,12  In this role it is usually given at an FDA-approved dose of 0.24 mg/kg before harvesting HSC.

We have previously established that AMD3100 is equipotent and equieffective at blocking both endogenous and recombinant forms of wild-type CXCR4 and CXCR4^R343X.⁶  Because mice lacking Cxcr4 are not viable and have defective myelopoiesis and B-cell lymphopoiesis as well as defects in cerebellar and vascular development,^13-15  there are important safety concerns about chronically blocking CXCR4 with drugs in any disease. However, we hypothesized that partial blockade of CXCR4 with plerixafor would be sufficient to correct panleukopenia and possibly other phenotypes in patients with WHIM syndrome and would be safe. Here we test this hypothesis in a phase 1 clinical trial in adults.

All patients signed informed consent consistent with the Declaration of Helsinki under clinical protocols approved by the National Institute of Allergy and Infectious Diseases before taking part in research at the National Institutes of Health Clinical Center. Three unrelated white adults with WHIM syndrome were recruited: a 44-year-old woman (patient 1), a 30-year-old woman (patient 2), and a 51-year-old man (patient 3). Patient 1 was previously identified as P1 in Wetzler et al,³  whereas patients 2 and 3 are the only identified adult survivors from 2 previously unreported families. All 3 patients were found to have the 4 defining clinical features of WHIM syndrome and were heterozygous for the CXCR4^R334X mutation.

Plasma plerixafor concentrations were determined by Tandem Labs by the use of a previously validated proprietary assay. Prism v5 (GraphPad Software) was used to plot the concentration versus time data, and WinNonlin v5 (Pharsight Corporation) was used to generate standard pharmacokinetic parameter values via the use of noncompartmental methods. See supplemental Methods for additional details (available on the Blood Web site; see the Supplemental Materials link at the top of the online article).

Flow cytometry was performed on whole blood and purified cells after staining for 30 minutes at 4°C with directly conjugated antibodies (BD Biosciences/Pharmingen) with the use of standard techniques. See supplemental Methods for additional details.

Reactive oxygen species (ROS) production was tested by flow cytometry by use of the ROS-sensitive fluorescent dye dihydrorhodamine (DHR). Leukocytes isolated by ammonium chloride lysis were loaded with DHR and then stimulated with phorbol myristate acetate (PMA; Sigma-Aldrich), as previously described.^16,17  Neutrophils were purified from diluted whole blood by centrifugation over lymphocyte separation media followed by dextran sedimentation and hypotonic saline lysis, as previously described,¹⁷  and their ability to kill bacteria was assessed by incubation with Staphylococcus aureus for 90 minutes at 37°C.

For safety and to find the lowest dose that might be used for chronic treatment, we designed a phase 1 dose-escalation study by using single daily injections at doses that had previously been tested in normal volunteers, up to a maximum of 0.24 mg/kg, the FDA-approved dose for HSC mobilization in patients with normal renal function (Figure 1). The primary end point of the protocol was safety, and the secondary end point was change in numbers of leukocytes in the blood with a doubling of average baseline value considered significant. Because there were no previous peer-reviewed published data on effects of plerixafor in children, we restricted the study to adults, defined as age 18 or older. G-CSF was 1 week before we initiated plerixafor. Patient 1 was receiving G-CSF on entry into the study, whereas patients 2 and 3 had not received G-CSF for several years prior to enrollment.

To monitor leukocyte mobilization dynamics, complete blood count with differential was performed every 3 hours 4-5 times per day after the patients received plerixafor. Dose escalation was designed to stop if neutrophils reached 4000/μL. At 2 days after patients received the maximal dose, we determined the pharmacokinetics of plerixafor by readministering the maximal dose followed by 9 blood draws over the course of 24 hours. Erythrocyte sedimentation rate and serum chemistries were performed daily. Participants had to be free of infection, not pregnant (if female), willing to use 2 methods of contraception for 6 weeks before and after drug administration, and willing to temporarily discontinue G-CSF. Other exclusion criteria were the presence of hematologic malignancy, serious heart disease or conduction abnormalities, and poor renal function (creatinine clearance < 15 mL/min or requiring dialysis).

The pharmacokinetic data after a single subcutaneous dose of 0.24 mg/kg revealed a pattern of absorption, distribution, and elimination in WHIM patients 1 and 2 similar to that previously reported for normal volunteers (Figure 2).^18-20  The average peak plasma concentration was 616 ng/mL at ∼ 30 minutes after injection. The mean half-life was 5.2 hours with an apparent clearance and volume of distribution of 5.2 L/h and 39.2 L, respectively. A 2-compartment model with first-order elimination was found to closely predict the measured values, consistent with previous studies in normal volunteers.^18-20 

Off drug, all 3 subjects were severely panleukopenic at levels that were stable during the 2-day prelude (Figure 3). The 0.02 mg/kg dose of plerixafor induced a rapid increase in white blood cells (WBCs; peak 3-6 hours after the dose) in all subjects. Unexpectedly, lymphocytes accounted for most of the increase in the total WBC and by the 0.04 mg/kg dose exceeded the upper limit of normal, unlike in normal volunteers, in whom the majority of the increase in WBC is because of neutrophils.^18-20  Monocytes were also highly responsive to plerixafor (peak 3-6 hours) in all 3 subjects and exceeded the upper limit of normal by the 0.04-0.08 mg/kg doses. The neutrophil increase was slower (peak 9-12 hours) and more sustained but never exceeded the upper limit of normal even at the highest dose.

Nevertheless, even at the 0.02-mg/kg dose, the absolute neutrophil count (ANC) exceeded the safe level of > 500 neutrophils/μL of blood in all 3 subjects. Comparison of leukocyte dynamics after the 0.16-mg/kg and 0.24-mg/kg doses of plerixafor to leukocyte levels maintained by chronic daily G-CSF administration (Figure 4) revealed that peak ANC after plerixafor was similar to steady-state ANC on G-CSF; however, the peak absolute lymphocyte and monocyte counts (ALC and AMCs, respectively), and thus total peak WBC, were much greater after the administration of plerixafor than with G-CSF. All 3 major subsets of lymphocytes (CD19⁺ B cells and CD4⁺ and CD8⁺ T cells) were strongly mobilized by plerixafor but not by G-CSF (Figure 4, Table 1). T cells were primarily of the memory phenotype (CD45RO⁺), specifically the effector memory subtype (CD45RO⁺CD62L⁻CCR7⁻). CD4⁺CD25⁺ and CD4⁺CCR6⁺ T cells, which contain T_reg cells and TH₁₇ cells, respectively, were also strongly mobilized (data not shown). Mobilized B cells appeared to be primarily an immature phenotype (CD19⁺CD27⁻) and were largely composed of (CD27⁻IgD⁺IgM⁺) naive (transitional) type cells, although CD27⁻IgD⁻IgM⁺ immature B cells were also present. Mobilized monocytes were predominantly classic monocytes (CD14⁺CD16⁻), but inflammatory monocytes (CD14⁺CD16⁺) were also mobilized.

Virtually all blood neutrophils harvested at the peak of the plerixafor response were capable of producing normal amounts of ROS (Figure 5A). We also performed a classic neutrophil bacterial killing assay to measure staphylocidal activity at 2 target:effector ratios (2 and 8 bacteria/neutrophil). Neutrophils collected at the peak of the plerixafor response killed bacteria normally (Figure 5B).

In terms of safety parameters, there were no arrhythmias, palpitations, electrocardiogram changes, thrombocytopenia or changes in electrolytes, kidney function, or liver enzymes (supplemental Figure 1). Patients 1 and 3 had asymptomatic, postprandial elevations of glucose (range, 116-231 mg/dL) on the day after the greatest plerixafor dose; however, patient 1 had been previously diagnosed as prediabetic. Mild injection-site reactions occurred in all 3 patients, consisting of pruritic erythematous induration that resolved in 30-90 minutes without treatment (supplemental Figure 2). This was not dose-dependent but was more common in stomach than arm injection sites and was associated with shallower injection technique. Patients 1 and 3 exhibited insomnia and headache on days 5-7 that were alleviated with narcoleptics and analgesics. Patient 2 had lumbar and femoral pain with some associated nausea and emesis similar to previous treatment with G-CSF that was relieved by analgesics. There was no grade 3 or 4 toxicity.

We have demonstrated, in a phase 1 dose-escalation drug treatment trial in WHIM syndrome, that plerixafor, which blocks the precise molecular target in the disease, appears to be safe and effective for 1 week in correcting panleukopenia, the major cellular mechanism of immunodeficiency in the disease.

With regard to scientific significance, the data provide the first in vivo pharmacologic evidence in support of the hypothesis that panleukopenia in WHIM syndrome is caused by increased CXCR4 signaling. This hypothesis draws its strength from multiple independent lines of evidence, including: (1) CXCL12-CXCR4 signaling is known to normally mediate neutrophil retention in the BM of mice^9,21,22 ; (2) CXCR4 is the disease gene in WHIM syndrome⁵ ; (3) the CXCR4^R334X variant found in our WHIM patients exhibits increased agonist-dependent signaling in multiple independent assays and mediates myelokathexis when expressed in mice and zebrafish^9,10 ; (4) the administration of plerixafor mobilizes many leukocyte subsets, including HSCs, lymphocytes, monocytes, neutrophils, and eosinophils, to the blood in healthy subjects^18-20 ; (5) plerixafor is a specific and equipotent antagonist of both wild-type CXCR4 and CXCR4^R334X6 ; and (6) heterozygous Cxcr4^+/− mice appear phenotypically normal.^13-15 

It is important to note that the rank order of responsiveness of leukocyte subsets to plerixafor (ALC > ANC > AMC) was consistent for all 3 of our WHIM patients but different from what has been reported for healthy subjects (ANC > ALC > AMC), which possibly reflects differential expression and activity of the mutant receptor in these lineages. The difference is not as great, however, when the rank order is determined not from peak levels after drug but as -fold increase from baseline cell counts in which the WHIM patients increased their baseline WBC by 8.8-fold, ANC by 13.9-fold, ALC by 10.3-fold, and AMC by 12.6-fold. In healthy volunteers and patients with HIV, the increase peaked at less than 4-fold for all of the aforementioned leukocyte subtypes even with prolonged administration.^18-20,23  Remarkably, in our patients CD19⁺ B cells were almost completely absent from the blood before drug and increased 54-fold to above the upper limit of normal after peak drug effect. The sensitivity of B cells to plerixafor has not been previously reported, but our data reveal that CD19⁺ B cells express a high level of CXCR4 in both healthy donors and patients with WHIM (not shown), which may explain their responsiveness. The slower dynamics of the neutrophil compared with lymphocyte response may suggest that their movement is not as dependent on drug concentration and permit less frequent dosing to achieve the safe target level of ANC > 500/μL.

With regard to medical significance, our data support continued investigation of plerixafor as a safe and effective treatment for all phenotypes that define WHIM syndrome. Although our study was not designed to address clinical efficacy and was restricted to adults, plerixafor increased all major leukocyte lineages from below the lower limit of normal to above the upper limit of normal (ALC and AMC) or to the normal range (ANC). Importantly, the panleukopenia in all study subjects was rapidly corrected with very low doses of drug, 1/12th that of the FDA-approved dose for HSC mobilization. At the lowest dose, we estimate that at peak the drug achieves at most ∼ 30% receptor coverage, which is important for safety.

The goal in treatment in WHIM syndrome is to target the molecular mechanism, which should require only partial receptor blockade to reduce CXCR4 signaling toward normal. This may avert potential toxicities associated with complete loss of normal CXCR4 function observed in previous murine studies.^13-15  The relatively high doses of plerixafor used in the dose-escalation phase of our study were accompanied by very few side effects compared with what has been reported in clinical trials in patients with cancer and HIV.^11,12,23  Minor skin hyperemia occurring within minutes after injection at the site of injection was the only definitely associated side effect. This finding confirmed a previous report and resolved spontaneously within 90 minutes.²⁰ 

Unlike in studies of healthy volunteers and HIV and cancer patients, there was no association of plerixafor with heart arrhythmia, thrombocytopenia, or gastrointestinal symptoms in our patients, which may indicate that these side effects may be because of the underlying illness or other drugs or that WHIM patients are more resistant to these side effects; additional WHIM patients will need to be treated with plerixafor to provide a definitive answer. Importantly, our patients did not develop any infections while they were taking the drug. Consistent with this finding, patient neutrophils mobilized to the blood by plerixafor were able to produce ROS ex vivo at normal levels in response to PMA and to kill S aureus ex vivo. The corollary of this is that in WHIM syndrome susceptibility to infections is probably because of defects in the number of neutrophils in the blood, not their function. Whether this is also the case for monocytes remains to be determined. For lymphocytes, there is evidence of dysfunction in WHIM syndrome because patients are usually hypogammaglobulinemic and may respond poorly to vaccines. Future studies will be needed to determine whether these problems can be resolved by simply mobilizing cells with plerixafor.

BM is probably the major source of neutrophil sequestration as suggested by direct BM examination of WHIM patients and, as mentioned previously, by direct data establishing a role for CXCR4 in neutrophil retention in BM in the mouse.^9,21  The sequestration site for other leukocyte subsets mobilized by plerixafor in WHIM patients is not directly addressed by our study, nor has it been defined in healthy subjects receiving the drug. Monocytopenia in WHIM syndrome has only recently been recognized,²⁴  and our results suggest that both classic and inflammatory monocyte numbers in the circulation are controlled by CXCR4. It is unknown whether monocytopenia causes any of the phenotypes in the syndrome, although monocytes are known to play a significant role in controlling infection, particularly by intracellular pathogens. The deficiency of naive (transitional B cells) in patient blood may explain the poor vaccine responses and hypogammaglobulinemia seen in WHIM syndrome, although the full mechanism of action may be more complicated because CXCR4 is also important in precisely locating B cells with T cells in the germinal centers where they are stimulated by antigen and undergo class switching.^25,26  CD4⁺ and CD8⁺ effector memory T cells were both released in large numbers by plerixafor. Vaccination studies have identified these cells in mouse BM, which may serve as a reservoir.^27,28  A recent study suggests that this may also occur in humans.²⁹  Thus, BM is potentially a major source of these cells mobilized to the blood by plerixafor.

In conclusion, our study provides initial human in vivo evidence supporting CXCR4 signaling as the critical molecular mechanism in patients with WHIM syndrome and justifies continued study of plerixafor as a potentially safe and effective mechanism-based treatment for adult patients with this disease. A longer, therapeutic trial at the National Institutes of Health is currently underway. Future work will need to address optimal dosing, and long-term safety and efficacy in more patients, including children.

The authors thank the patients with WHIM syndrome and their families who have volunteered to make this study possible. They also thank Dr Kathryn Zoom for support.

This work was supported by the Division of Intramural Research of the National Institute of Allergy and Infectious Diseases and the Office of Rare Diseases Research, National Institutes of Health, and was funded in part with federal funds from the National Cancer Institute, National Institutes of Health under Contract number HHSN261200800001E. Plerixafor was provided at no cost for the study by Genzyme Corporation.

The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government.

National Institutes of Health

Contribution: D.H.M., N.K., S.R.P., H.L.M., and P.M.M. designed the protocol; D.H.M., J.U., S.A., C.K., M.G., P.L., M.M.M., D.H., R.D., H.L.M., and P.M.M. provided patient recruitment and care; Q.L., J.O.F., S.R.P., and D.A.L.P. generated and analyzed experimental data; and D.H.M., T.A.F., D.B.K., H.L.M., and P.M.M. supervised the experiments and analyzed data. All authors participated, but D.H.M. and P.M.M. were principally responsible for the writing of the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: David H. McDermott, MD, 9000 Rockville Pike, Bldg 10, Rm 11N107, Bethesda, MD 20892; e-mail: dmcdermott@nih.gov.