Hematopoietic Stem Cell Gene Therapy for Cystinosis

Abstract

Background

Cystinosis is a multisystemic lysosomal storage disorder caused by mutations in the CTNS gene encoding the transmembrane lysosomal cystine transporter, cystinosin. In cystinosis, cystine accumulates within lysosomes in all organs. Cysteamine delays but does not prevent disease progression.

Methods

This phase 1/2 open-label clinical study is evaluating the primary safety of CTNS-RD-04, consisting of autologous CD34⁺ cells transduced with lentiviral vectors carrying CTNS cDNA. Secondary outcomes were efficacy as white blood cell [wbc] cystine levels and other measures of cystine storage depletion. Oral cysteamine was withdrawn prior to CTNS-RD-04 infusion and cysteamine eye-drops were withdrawn one month after myeloablation.

Results

Six participants (20 to 46 years of age) received CTNS-RD-04 and were followed for 29 to 63 months. CTNS-RD-04 cell doses ranged from 3.63×10⁶ to 9.59×10⁶ CD34⁺ cells/kg, and vector copy numbers (VCNs) ranged from 0.59 to 2.91 copies/diploid genome. All patients had sustained and highly polyclonal hematopoietic reconstitution; VCNs at 24 months ranged from 0.51 to 2.67 copies/diploid genome. In all, 216 adverse events occurred, mostly mild or moderate; these were largely consistent with the procedures and underlying disease. No evidence of monoclonal expansion was noted. The secondary outcome of wbc cystine levels decreased compared to baseline, except in patient 4, who had the lowest VCN .

Conclusion

In this small study of an ex vivo gene therapy approach for cystinosis, showed an acceptable safety profile. Adverse effects were largely consistent with the myoablative regimen and underlying disease profile. Wbc cystine levels decreased following therapy.

Cystinosisis a rare autosomal recessive lysosomal storage disease that creates tremendous burden for patients and caregivers and has poor outcomes, despite availability of cystine-reducing treatment with cysteamine.¹ The disease is caused by mutations or deletions in the ubiquitous CTNS gene (17p13.2), which encodes cystinosin, a lysosomal transmembrane cystine transporter, leading to the accumulation of cystine within lysosomes and cystine crystals within tissues.^2–5 Three allelic forms of cystinosis exist, the most severe and most common of which is the infantile form (OMIM #219800). Affected children develop renal Fanconi syndrome by 6–18 months of age and chronic kidney disease (CKD), which eventually leads to end-stage kidney disease (ESKD).² Non-renal complications of cystinosis include photophobia and corneal erosion,³ cardiovascular complications,⁴ diabetes mellitus,⁵ hypothyroidism,⁵ bone deformities and fragility,⁶ neurological defects,⁷ and distal myopathy that can result in life-threatening respiratory dysfunction, dysphagia, and aspiration pneumonia.^8,9

Intracellular cystine reduction with the cystine-depleting agent ,cysteamine, allows cystine to exit cells, slows the progression of the disease. To reduce corneal cystine crystal accumulation, cysteamine eye drops are needed every waking hour. Premature death has been inevitable, despite these therapies.

Given these issues, delivery of functional cystinosin to tissues throughout the body would be an important therapy. Preclinical studies in a rodent model of cystinosis, Ctns^−/− mice,^10,11 we demonstrated that transplantation of hematopoietic stem and progenitor cells (HSPCs) expressing a functional Ctns gene resulted in tissue integration of bone marrow-derived cells, decreased cystine accumulation, and long-term preservation of kidney, thyroid and eye function.^12–15 A mechanistic study in this model demonstrated cross-correction through tunnelling nanotubes, which appear to act as cellular bridges that support bidirectional lysosomal exchange between the tissue engrafted HSPC-derived macrophages/microglia and the host’s diseased cells.^12,13,16,17

Allogeneic HSPC transplantation carries considerable morbidity and mortality risks; indeed, the death of a cystinosis patient who received an allogeneic HSPC transplant was considered due to graft-versus-host disease.¹⁸ To avoid such an outcome, we developed an autologous transplantation strategy with the use of autologous gene-modified HSPCs using a self-inactivated lentiviral vector (SIN-LV) containing CTNS complementary DNA (cDNA) for cystinosin and had demonstrated efficacy in Ctns^−/− mice.¹⁹

We now report the initial results of an open-label, phase 1/2 clinical study of investigational gene therapy for cystinosis with the use of CD34⁺ enriched HPSCs transduced ex vivo with SIN-LV and containing CTNS cDNA (CTNS-RD-04) encoding functional cystinosin.

METHODS

Study Design and Oversight

We assessed the safety and efficacy of the CTNS-RD-04 drug product in six adult patients with infantile cystinosis. Our ongoing study is being conducted at a single site at the University of California San Diego (UCSD), and more complete descriptions are provided in the Supplementary Appendix, Methods. An additional patient was excluded during screening due to poor pulmonary function; another potential participant was not included due to poor mobilization and unsatisfactory stem cell collection. Oral cysteamine treatment was discontinued two weeks prior and cysteamine eye-drops one month after myeloablation conditioning with busulfan, and CTNS-RD-04 was infused intravenously. In the primary clinical study, the patients were followed for two years and then were offered enrollment in an ongoing 13-year long-term follow-up study.

Cell Product Manufacturing and Transplant

To obtain CD34⁺ HSPCs, each patient underwent leukapheresis, and CD34⁺ HSPCs were selected after mobilization with granulocyte colony-stimulating factor (G-CSF) and plerixafor administered for 4 and 1 days, respectively. Positive selection of CD34⁺ cells was performed prior to lentiviral vector transduction. In the course of this study two different lentiviral vectors were used to generate a CTNS-RD-04 product, pCCL-CTNS (patients 1–3) and pCDY-CTNS vector (patients 4–6), the latter to prepare for transition to a commercially compatible vector (Supplementary Appendix, Methods and Fig. S1). To improve lentiviral transduction efficiency, LentiBOOST^® (Sirion Biotech; 1 mg/mL) was added to the clinical manufacturing process during the transduction step for patients 5 and 6. The minimum target yield for the cell product was 3×10⁶ CD34⁺ cells per kilogram of body weight. Intravenous cell infusion took place through the use of a customized administration set-up after a 24-hour busulfan washout period. All patients were closely monitored in hospital for safety throughout conditioning, infusion, and hematologic reconstitution.

End Points

Patients attended regular follow-up visits at which blood samples were obtained and at which patients were assessed for adverse events and disease progression (see Study Protocol at NEJM.org). Primary end points were safety and tolerability of the drug product, as measured by the incidence and grade of adverse events, and serious adverse events Secondary outcomes were efficacy as manifested by decrease in white blood cell (wbc) cystine levels, an accepted marker of control of cystinosis. Safety end points also included detection of replication-competent lentiviral or clonal expansion by vector-integration site analysis; results were considered abnormal when a single site constituted more than 20% of the total integrations on two separate occasions.

Secondary end points included engraftment of the genetically modified HSPCs, determined by the vector copy number (VCN) and CTNS expression in peripheral blood mononuclear cells. Intracellular cystine content was assayed in mixed leukocytes and isolated granulocytes, and in rectal and skin biopsies (details in Supplementary Appendix, Methods). Cystine crystals were quantified in rectal and skin biopsies by Dr. Robert Newbury and Dr. Katayoon Shayan on histological sections prepared as previously described,²⁰ and in skin through non-invasive in vivo intradermal confocal microscopy.²¹ Corneal crystals were visualized by in vivo corneal confocal microscopy and density scored in de-identified micrographs by Dr. Hong Liang as previously described,²² along with vision assessments. The estimated glomerular filtration rate (eGFR) was calculated using cystatin C and serum creatinine with the use of both the EPI-CKD equation (2021) and CKD-EPI creatinine-cystatin equations (2021). ⁴⁸ Neuromuscular and bone assessments, thyroid function, and respiratory and cardiac functional assessments were performed. Psychometric evaluation was also carried out with the use of the Wechsler Abbreviated Intelligence Scale (WASI) and the Beery Test of Visual Motor Integration (VMI).

Statistical Analysis

This study was designed as a preliminary assessment of safety and wbc cystine levels (coprimary outcome) and other aspects of efficacy (secondary outcomes) of gene therapy for cystinosis, with all results presented descriptively, given the small patient cohort. No separate statistical analysis plan was developed.

RESULTS

Patients and Treatment

Six patients, five men and one woman, between 20 and 46 years of age, were enrolled between July 2019 and May 2022 and received CTNS-RD-04, Demographics and vital signs are summarized in Table 1, Tables S1, S2 and S3. Patient representativeness compared with the broader cystinosis population is detailed in the Supplementary Appendix, Table S1. While our small cohort limited generalizability, participant characteristics aligned with adult cystinosis patients eligible for advanced therapies. Two of the six patients had stage 3 chronic kidney disease (CKD) and had not yet required kidney transplantation, one had received a kidney transplant, and three had had two kidney transplants prior to stem cell infusion. All six patients were receiving oral cysteamine, with a mean pill burden of 35.8 pills per day (Table S4). The present analysis from July 2019 to January 2025, marked the end of the parent study (at least 24 months follow up for all patients) and a five-year follow-up for patient 1, with median patient follow-up 36 months (range, 29 to 63 months). Five of the six patients enrolled in the long-term follow up study; patient 2 declined.

Table 1.

Demographics

	Patient 1	Patient 2	Patient 3	Patient 4	Patient 5	Patient 6
Age of symptom onset/diagnosis	0 year / 8 months	0 years 6 months	4 years	6 years	8 months	2 years
Age at the time of treatment with CTNS-RD-04	20 years, dosed in October 2019	46 years, dosed in June 2020	22 years, dosed in November 2020	33 years, dosed in November 2021	31 years, dosed in March 2022	30 years, dosed in October 2022
Gender	Male	Male	Male	Male	Female	Male
Mutation	57-kb deletion nt1035(ins C), p.Val233Argfs*63	57-kb deletion c.473t>C, p.Leu158Pro	c.18_21del, p.Thr7Phefs*7 c.295_298del, p.Val99Ilefs*18	57-kb deletion c.473T>C, p.Leu158Pro	57-kb deletion c.414G>A, p.Trp138*	Homozygous 57 kb deletion
Kidney transplant status	Stage 3 (moderate CKD) renal failure at enrollment and 1 kidney transplant 34 months Post Drug Infusion (2022)	2 renal transplants (1987 and 1999)	1 renal transplant (2010)	2 renal transplants (2008 and 2017)	No renal transplant; stage 3 (moderate CKD) renal failure	2 renal transplants (2010 and 2019)

Patients received pre-infusion conditioning with busulfan at a target area under the curve (AUC) of 85–90 mg x h/L, a dose selected because it provides a acceptably-tolerated yet effective level of myeloablation.²³ The AUC range was from 81.8 mg x h/L to 90 mg x h/L (Table S2). One day later, patients were infused with the CTNS-RD-04 product, which contained 3.63 to 9.59 million CD34⁺ cells per kilogram of patient body weight, with a vector copy number of 0.59 to 2.91 copies per diploid genome (Table S2). LentiBOOST^® was introduced to the clinical manufacturing during the transduction step for patient 5 due to transduction efficiency falling below the release criteria of the drug product (VCN=0.4). Small-scale studies with patient 5’s own CD34⁺ cells confirmed that transduction efficiency was low, irrespective of the cell density, whereas adding Poloxamer 338 (research grade LentiBOOST^®) was associated with higher VCN (Table S5). LentiBOOST^® was subsequently used to manufacture the drug product for patient 6 to proactively mitigate the risk of low VCN yield product. Following mobilization, the CD34⁺ cell yield was low for patients 4 and 6, requiring repeated leukapheresis; patient 6 had two drug products manufactured that were infused sequentially the same day (Table S2). After infusion, all patients reached absolute neutrophil and platelet counts considered to constitute engraftment at a median of 13 days and 19.5 days, respectively (Fig. S2).

Safety Outcomes

As of data cut-off on January 30, 2025, 216 adverse events (AEs) had occurred, mostly mild (n=179) or moderate (n=33), consistent with known risks associated with pre-infusion conditioning, therapy post-conditioning and the participants’ underlying cystinosis and-or pre-existing conditions (Tables 2, S6 and S7A–C). No AEs appeared to be related to CTNS-RD-04. Four severe AEs were identified-- appendicitis (Patient 1), two episodes of worsening chronic kidney disease (Patient 1), and pre-existing coronary artery disease (CAD) (Patient 6), which was also classified as a serious AE because it required hospitalization (Table 2). Patients 1, 4, and 6 were thrombocytopenic from baseline (Fig. S2); the etiology was unclear but presumably attributable to underlying cystinosis.²⁴ Two patients had SARS-CoV-2 infection shortly after infusion of CTNS-RD-04, and experienced delayed wbc reconstitution. Patients 4 and 5 became infected at 10- and 13-weeks post infusion, respectively, (tested positive by PCR and antigen tests for 14 and 23 days, respectively), and reached normal range wbc counts at 9- and 30-months post infusion, respectively (Fig. S2).

Table 2 –

Major Adverse Events

Category	Timepoints1	Attribution	N of Events	N of Patients
Gastrointestinal			26	6
Vomiting	S(1), M(1),P(4),F(2)	Gastric erosions, Plerixafor, post-transplant follow-up, immunization reaction, appendicitis	8	4
Diarrhea	M(5), C(1), P(2), F(1)	Plerixafor, Busulfan, Post-transplant follow-up, Erythromycin	9	5
Constipation	P(3)	Opioid	3	2
Mucositis, oral	C(6)	Busulfan	6	6
Hematologic			20	5
Thrombocytopenia	C(7)	Busulfan	7	5
Leukopenia	C(4)	Busulfan	4	4
Neutropenic fever	C(5)	Busulfan	5	4
Epistaxis	C(3), F(1)	Busulfan, Mild trauma	4	3
Renal/Electrolyte			18	4
↑Creatinine, BUN, Chloride,  ↓ Bicarb, K, Phosphorus	S(5), M(4), C(4), P(1)	CKD (4), Renal transplant (1), acyclovir and Septra (1), Busulfan(2), Plerixafor(1), Fanconi syndrome(1)	14	4
Hypomagnesemia	M(3), P(1)	Apheresis(3), Tacrolimus(1)	4	3
Endocrine			15	6
Azoospermia	S(5)	Cystinosis	5	5
Gonadal Failure	C(6)	Busulfan	6	6
Hypothyroidism	P(1), F(3)	Cystinosis	4	4
Constitutional			14	5
Anorexia	S(1), C(1), P(2), F(1)	Eosinophilic esophagitis, Busulfan, Appendicitis, Constipation, Immunization reaction	5	2
Fatigue/ Lethargy	M(2), C(4)	Apheresis, Stress and anxiety, Busulfan	6	5
Unintentional Weight Loss	S(1), F(2)	Eosinophilic esophagitis, Appendicitis, Coronavirus	3	2
Dermatologic			13	6
Alopecia	C(6)	Busulfan	6	6
Rash	M(1),P(6)	Tape sensitivity, PICC dressing, Folliculitis, Bactrim, Vancomycin	7	6
Swelling			5	2
Edema	P(2)	Renal insufficiency	2	2
Swelling of leg, arm, knee	P(1), F(2)	Popliteal cyst, IV infiltration, Mild trauma	3	2
Infection			3	3
Covid	P(1), F(2)	Coronavirus	3	3
Cardiac			1	1
Mitral valve surgery2	F(1)	Pre-existent valvular calcification	1	1
N of Mild AEs: 178	N of Moderate AEs: 35	N of Severe AEs: 4	N of SAEs: 1	Total Number of Unique AEs: 217

There was a total of 217 individual events recorded, of which the table enumerates the 113 most significant, categorized by system, with the timepoints at which the events arose.

¹The abbreviations for timepoints indicate the experimental phase: S = screening, M = mobilization, C = conditioning, P = post-transplant (<=90 days), F = follow-up (>90 days).

²The single Serious Adverse Event is indicated. The majority of events (71%) occurred during the mobilization, conditioning, and post-transplant periods.

Biomarkers

Vector Integration

Vector integration site distributions in patients’ peripheral blood mononuclear cell DNA at different timepoints exhibited highly polyclonal profiles, without evident events of monoclonal expansion or leukoproliferative complications. No unusual frequency of integration in or near cancer-associated genes was noted. The total number of unique vector integration sites across all patients was between 396 and 85,868 (mean of 12,640 unique sites; Fig. S3). We infer that the difference in VCN ranging from 0.4 to 2.5 includes a range in terms of the integration number. The VCNs showed long-term sustained stem cell engraftment in peripheral blood ranging from 0.46 to 2.89 at 6 months and from 0.51 to 2.67 at 24 months (Fig. 1A). CTNS expression was increased in the patients’ peripheral blood compared with baseline, ranging from 11 to 49-fold at 24 months (Fig. 1B). In colon biopsies, gene marking as exploratory endpoint was used to indicate tissue engraftment of HSPC-derived cells (Fig. S4A and S4B).

Figure 1. Vector Copy Number and CTNS Expression in Peripheral Blood Over Time Post-Infusion.

Panel A shows the vector copy number in CTNS-RD-04 drug product at the time of infusion and in peripheral blood at various time points post-infusion for each of the six patients. *Patient 6 received two separate drug product infusions, and the reported VCN represents the average of both drug products.

Panel B shows CTNS expression in the peripheral blood for each of the 6 patients at various time points after infusion of CTNS-RD-04. For comparison, the mean CTNS expression in peripheral blood from healthy donors was measured at a fold change of 2.97 (n = 3, SD = 0.65).

Cystine and Cystine Crystals

At 24 months post-infusion, leukocyte cystine levels had decreased by 25%−86% in all patients; granulocyte cystine content in five out of six, had decreased by 22%−81%, compared with pre-transplant levels (Fig. 2A, 2B, Fig. S5A). Patients 4 and 5, who had the lowest VCN, had the least reduction in cystine levels, and were restarted on low-dose cysteamine at 14 and 36 months, respectively (<25% of their original dose; Table S2). From baseline to the last rectal biopsy, three patients had reductions in crystal number and cystine content (Fig. S4C–E). We used the non-invasive intradermal confocal microscopy technique as an exploratory outcome to quantify cystine crystals in the skin to obtain a normalized confocal crystal volume (nCCV).²¹ Four patients had decreased nCCV from baseline with patient 3 reaching close to the levels of persons without cystinosis (Fig. S5B). Similarly, corneal confocal microscope examinations showed decreased or stable cystine crystal density in the 3 patients in whom it was performed (Fig. S6).

Figure 2. Leukocyte Cystine, Granulocyte Cystine and Normalized Confocal Crystal Volume (nCCV).

Panel A shows the leukocyte cystine level at baseline and at various time points after infusion. *Patient 1 leukocyte cystine was not measured at baseline. The therapeutic upper limit is indicated by the dotted line. Panel B shows the granulocyte cystine level at baseline and at various time points after infusion. # Patient 4 re-started oral cysteamine (150 mg/day) at 36 months post-infusion. ^Patient 5 restarted oral cysteamine (450 mg/day) at 14 months post-infusion.

Clinical Status and Outcomes

Patient 1 had CKD stage 3 and stable eGFR until 12 months post-infusion; his eGFR then declined and progressed to ESKD at 30 months post-infusion (Table S8). Patient 5 had CKD stage 3 and had a 47% decrease in eGFR, from 54 at baseline to 25 mL/min/1.73m² at 24 months post infusion. Patient 5 had had SARS-CoV-2 infection 13 weeks after product infusion and slow wbc recovery with low monocytes, possibly related to the steep decline in GFR. Patients 2, 3 and 6 had CKD-T stage 2 and patient 4 had CKD-T stage 3, and their eGFRs remained stable overall (Table S8).

Thyroid function was assessed in all participants by measuring thyroxine (T4) and thyroid stimulating hormone (TSH) levels. T4 levels were normal at baseline and remained stable in all patients (Fig. S7A). TSH levels increased in two patients over the study period, patient 1 and patient 6 (Fig. S7B). Overall, ocular examinations in the patients remained unchanged compared with baseline, and photophobia stabilized or improved in all patients except for patient 5, who went from grade 4 at baseline to grade 5 at 30 months post-transplant (Table S9).

All six participants had a normal neurological examination at follow up, with no evidence of weakness or oromotor dysfunction. Finger tapping, a measure of motor speed remained stable in our patients. Visual motor integration and visual perception remained stable; motor coordination, which was below average at baseline improved modestly to a score within the normal range (Fig. S8A–C). Measures of strength (motor examination, grip strength, and timed walking) and cognitive function were stable over time (Fig. S8D and Fig. S9).

Bone density remained stable after stem cell transplant in all patients (Table S10). Pulmonary function and cardiology assessments were stable, except for patients 3 and 5, who had decreased pulmonary diffusion capacity (DLCO) following transplant (Tables S11 and S12).

DISCUSSION

The present study of gene-modified autologous CD34⁺ cell transplantation was associated with interim safety, and tolerability results supporting an acceptable risk-benefit profile and continued clinical development of CTNS-RD-04 for treatment of cystinosis. Evidence of sustained donor cell engraftment was observed during follow-up (over five years in one patient) with stable gene marking and CTNS gene expression in peripheral blood and tissues post infusion of CTNS-RD-04. No adverse event appeared to be related to CTNS-RD-04; there were no reported fatal or life-threatening adverse events, clonal expansions, or leukoproliferative complications. We interpret the results of our study to suggest that this approach may provide a therapeutic advance for patients with nephropathic cystinosis and suggest a need for additional study.

Peripheral white blood cell cystine content is an established surrogate marker for the diagnosis and therapeutic monitoring of cystinosis. Decrease in cystine was observed both by the classical mixed leukocyte cystine assay and the purified granulocyte cystine determination,²⁵ and were associated with the VCN in the hematopoietic cell progeny. These results along with longer follow up periods will be necessary to assess the therapeutic threshold of CTNS-RD-04 more completely.

Cystine crystals in the intestinal mucosa biopsies, estimated using the method previously described by Dohil and colleagues,²⁰ were decreased in three patients. Decrease in cystine crystals in the skin using our intradermal confocal microscopy method²¹ compared with baseline occurred in four patients.

Assessment of the effects of CTNS-RD-04 on clinical outcomes in these six treated patients poses ongoing challenges due to the diverse disease presentations in cystinosis. A report of 86 cystinosis patients reported that 80% of patients who started cysteamine before 5 years of age still developed end-stage renal disease at a mean age of 13.4 ± 4.8 years of age.²⁶ Kidney function during the present study’s follow-up period were as anticipated--patients with prior renal transplants had stable kidney function, and those with pre-existing advanced kidney disease had expected decline.

Neurological symptoms in cystinosis patients include reduced gross and fine motor coordination, hypotonia, oromotor dysfunction, muscle weakness, and visual spatial and visual motor deficits, which do not improve over time.^7,27,28 In this study the neuro-motor and cognitive functioning of the participants remained stable over the entire period of observation, improving in some patients. Post-transplant, HSPCs have been shown to engraft in the central nervous system and differentiate into microglia-like cells in mice and humans.^29,30 We speculate that gene modified HSPCs may lead to decreased cystine storage in the brain, resulting in improved motor coordination. With respect to peripheral nervous system involvement, cystinosis patients present with distal vacuolar myopathy leading to progressive distal muscle wasting and weakness.³¹ In the present study, grip strength remained generally stable or improved.

In conclusion, the ex vivo gene therapy approach reported here for cystinosis, a multi-system condition due to a transmembrane lysosomal protein variant, demonstrated an acceptable safety profile and reduction in wbc cystine levels.