ARTICLE IN BRIEF
In a small open-label trial, intrathecal 2-hydroxypropyl-β-cyclodextrin was well tolerated and appeared to slow progression of Niemann-Pick, type C1.
Intrathecal 2-hydroxypropyl-β-cyclodextrin (HPβCD) has an acceptable safety profile and appears to slow disease progression in patients with Niemann-Pick disease, type C1 (NPC), a lysosomal storage disorder characterized by progressive neurodegeneration, according to findings from a non-randomized, open-label, phase 1/2 trial published in the August 10 online edition of TheLancet.
NPC, which has an overall prevalence of about one in 100,000, is usually diagnosed in childhood. Its initial presenting symptom can be enlargement of the spleen or liver, but the disease often goes undiagnosed until neurologic manifestations begin. Cerebellar ataxia and cognitive impairment, among other neurologic symptoms, progress relentlessly, and most people with the disease die within 10 to 15 years after the onset of symptoms. There are currently no approved treatment options for NPC in the United States, other than supportive care.
Most cases of NPC are caused by mutations in the NPC1 gene, which produces a dysfunctional cholesterol trafficking protein; as a result, the body is unable to metabolize cholesterol and other lipids, and these build up in the lysosomes, ultimately leading to cell death and organ damage. Cyclodextrins readily bind cholesterol and have commonly been used to modulate cellular cholesterol content in cell culture systems.
Early preclinical work with HPβCD, in mouse and cat models, has been extremely promising, with significantly delayed loss of cerebellar Purkinje cells, slowed progression of neurological symptoms, and increased lifespan.
The phase 1/2 trial was conducted at the National Institutes of Health (NIH), with additional patients treated through a similar protocol at Rush University Medical Center in Chicago. It is the first trial of HPβCD in human patients.
The NIH site enrolled a total of 14 patients to receive monthly HPβCD, non-randomly and sequentially assigning them in cohorts of three to receive starting doses of 50 mg, 200 mg, 300 mg, 400 mg, or 900 mg intrathecal HPβCD. The dose was advanced based on tolerance and safety data from higher-dose cohorts. The Rush site enrolled a total of three patients to receive HPβCD every two weeks; two were assigned a 200-mg starting dose, while the third was assigned a 400-mg starting dose. The dose was also escalated as tolerated to as much as 900 mg in the Rush patients.
After 18 months, results were assessed in 14 patients; three of the NIH patients were excluded from the analysis, as one had been removed from the trial after a diagnosis of hepatocellular carcinoma, while two patients had not reached 18 months on therapy at the time of the analysis.
Both biomarker and clinical evidence suggested efficacy for the drug, and other than transient unsteadiness, fatigue, and post-lumbar puncture headache, the only significant adverse events involved ototoxicity, an anticipated side effect of the drug. Hearing loss is common in NPC itself, and was present at baseline in all study participants, but after the trial, all participants had additional mid-frequency to high-frequency hearing loss.
Clinical efficacy was measured using the NPC Neurological Severity Scores (NNSS), comparing progression of the disease to that observed in an NIH historical cohort of 21 patients who were of similar ages and were similarly affected by the disease. The total NNSS for the 14 participants treated monthly increased at a slower rate of 1.22 points per year, compared with 2.92 points per year (p=0.0002) in the comparison cohort. That difference was even more significant when the components related to hearing were removed from the NNSS; in that analysis, HPβCD-treated participants had a progression rate of 0.69 points per year compared with 2.67 points per year (p<0.0001).
“By comparison with the controls, the HPβCD cohort had significantly decreased progression in cognition (p=0.0040) and speech domains (p=0.0423),” the authors noted. “The ambulation domain was also decreased (p=0.0622).”
Elizabeth Berry-Kravis, MD, PhD, professor of pediatrics and neurological sciences at Rush University Medical Center in Chicago, contributed the Rush data to the trial analysis and is a co-investigator on a follow-up phase 2/3 study that is now underway and expected to conclude in March 2018.
She had spent the previous 12 years of her career conducting clinical research into targeted therapies for Fragile X Syndrome, when Harry Koujaian came to see her about his daughter, Hayley. In 2009, at age 7, the previously healthy girl experienced cognitive decline; at 10 years old began having seizures; and two years later, she was diagnosed with NPC.
Dr. Berry-Kravis agreed to begin working with the family, and when Hayley didn't qualify for the primary HPβCD site at the NIH because of her vagus nerve stimulator as well as the number of seizure medications she was taking, she requested permission for expanded access to the drug from the US Food and Drug Administration (FDA).
“I never say no to things that might help my patients, even if they are a huge amount of work,” she quipped. “And I didn't think they'd have any other options in the Chicago area.” With her large clinical trial staff, she could take on the work that the FDA required; she also enrolled two other patients with NPC on the compassionate-use protocol, and that cohort was combined with the NIH cohort for the Lancet paper.
“This study is very exciting,” she said. “It shows that we are engaging the target and causing cholesterol to be mobilized, and that markers of central nervous system damage are improving. Comparison with a natural history group is never perfect, but we matched them as well as possible, and it does appear that the drug is causing a difference from the natural history that is in some cases substantial.”
Hayley Koujaian, for example, was functioning at the cognitive level of an 18-month-old prior to starting treatment in late 2013. Now aged 17, Dr. Berry-Kravis said that her cognitive and language abilities have improved to a 2- or 3-year-old level. She talks much more, sings, can chew and swallow her food more easily, and can read simple sentences—something she was having trouble doing prior to treatment.
“The response is variable,” Dr. Berry-Kravis noted. “Some patients did better than others; some patients, like Hayley, actually improved, while others simply didn't have any further disease progression. But even to have a treatment that stops regression in a subset of patients is an amazing thing in a fatal neurodegenerative disease.”
Hayley Koujaian and another of Dr. Berry-Kravis' patients, Samantha Berns, have now been taking HPβCD for more than three years, and have either improved or not regressed on multiple functional measures, including gait analysis, language abilities, fine motor skills, and IQ testing, during that time. “This has been life-changing for these families,” said Dr. Berry-Kravis.
Marc C. Patterson, MD, FAAN, FRACP, a professor of neurology, pediatrics and medical genetics, and chair of the division of child and adolescent neurology at the Mayo Clinic in Rochester, MN, has led studies of potential therapies for NPC since the early 1990s, when standard cholesterol-lowering agents, singly and in combination, were studied and found not to be effective against the disease.
“This is an important breakthrough,” he said. “The data are the most impressive I've seen so far in terms of influencing the course of the disease. Not all patients are responders, so even if it does continue to perform, it is not the whole answer. But it may be that this drug, for a significant number of patients, converts what has been a relentlessly progressive disease into a chronic one, much like AIDS drugs have done. That's a tremendous accomplishment.”
There are some key limitations, he noted. “Of course, the authors very appropriately make the caveat that it's uncontrolled. That's a big issue for this and any other neurodegenerative condition; there's so much variation, comparing with historical data is always difficult,” he said. “I co-authored the NIH paper published in 2009 showing that statistically, in the historical cohort, patients with NPC progressed at the same rate. But a lot of the literature disagrees with that, particularly when you're looking at a broader range of patients. So we do need to look at these results with some caution.”
The only other approved therapy for NPC is miglustat (OGT 918, N-butyl-deoxynojirimycin), marketed under the name Zavesca and primarily used to treat Gaucher disease type 1. Dr. Patterson led several studies of miglustat beginning in the late 1990s, and the drug was approved for NPC in Europe, Canada, and Japan, but the US FDA declined to approve it for several reasons, including a lack of strong natural history data.
“Miglustat has shown favorable markers in some studies, and some indicate stabilization of disease, but the big problem is that of lack of controls,” Dr. Patterson said. “And since it's approved in most other countries, most patients who can take it have been taking it.” He noted that a European registry with some 400 patients taking miglustat shows little evidence of harm other than transient diarrhea, and indicates that patients on the drug are progressing slowly. “But is it more slowly than they would without the drug? That's very difficult to answer.”
In the HPβCD studies, the NIH group insisted on natural history data. “That's the beauty of doing a study primarily at the NIH — they can do that. It's a very different environment from getting competitive funding externally,” Dr. Patterson said. “They have made a big investment in supporting this, and it's paid off. I would be very hopeful that the next phase study is going to show similar data with evidence of efficacy compared to controls.”
Assuming that is the case and the drug is approved, Dr. Patterson raised the question of cost. “This is an old drug that's been around forever. Johnson & Johnson made it available gratis for the study, the NIH has done the hard slog and the years of preparation, and the company that was formed to take it to market has now been taken over by another company. It will be very interesting to see what happens in terms of price if it's approved. That's something the community will be watching closely.”
As with many neurodegenerative diseases, earlier treatment with HPβCD appears to produce better results in NPC, at least in the animal models. “In the cat model of naturally-occurring NPC, the animal usually dies at about six months of age, and the drug extended the lifespan by at least six to eight times. The results in cats weren't as good if you started treating them later,” Dr. Berry-Kravis said.
“This suggests that if we can stabilize our patients when they are barely symptomatic, the outcomes will be better. While younger patients with NPC who have an enlarged liver or spleen are taken to geneticists, older children, teens and young adults almost always are seen by neurologists, meaning that it's critical for neurologists to be aware of this disease and able to make that diagnosis as early as possible.”