Welcome to Eurotimes

Nick Lane, PhD
Verteporfin (Visudyne®, Novartis Ophthalmics/QLT) was launched in April 2000 in the US, and in July 2000 in the EU. Hailed as a breakthrough technology for the 21st century, with good efficacy and few side effects, it won prestigious prizes as one of the ‘top products’ of 2000, and became the biggest ophthalmic launch ever. Approved in some 60 countries, verteporfin has now been used in photodynamic therapy (PDT) in over 250,000 age-related macular degeneration (AMD) patients worldwide.

But this success story is marred by lawsuits, contradictory regulatory decisions, and deep uncertainties over cost and effectiveness in clinical practice. Is PDT with verteporfin up to the hype or should ophthalmologists treat it with suspicion? EuroTimes looked into the turbulent story of verteporfin.

Verteporfin was approved on the basis of the TAP (Treatment of AMD with Photodynamic therapy) Investigation, twin phase III studies of the safety and efficacy of verteporfin PDT versus sham procedure in 609 patients with subfoveal choroidal neovascularisation (CNV) secondary to neovascular AMD. The lesions had to have a classic component.

In the intention-to-treat analysis at month 12, 61% of 402 eyes assigned to PDT with verteporfin had lost fewer than 15 letters (three lines on the ETDRS chart) of visual acuity from baseline, versus 46% of 207 eyes assigned to sham treatment, a relative risk reduction of 32% (P<0.001). A small proportion of patients gained vision: 16% of PDT-treated patients gained more than five letters of visual acuity, versus 7% of sham treated patients, a relative difference of 128%.

However, these bald numbers concealed a heterogeneous response among various different subgroups of patients in a post-hoc analysis. Patients with ‘predominantly classic’ lesions (>50% classic CNV) fared much better with PDT: 67% of eyes treated with PDT had lost less than 15 letters at month 12, versus 39% of sham-treated eyes, a relative risk reduction of 72% (P<0.001).

These positive results were sustained over four years of treatment with PDT. In contrast, patients with minimally classic lesions (<50% classic CNV) gained no benefit (56% of eyes treated with PDT lost fewer than 15 letters, versus 55% of sham-treated eyes).

To confound matters further, a sizable subgroup of patients – nearly 10% of all the patients recruited – were ‘misdiagnosed’ as having a classic CNV component, which the reading centre at the Wilmer Eye Clinic (Johns Hopkins University) reinterpreted as purely occult.

Curiously, this group of patients did better than those with minimally classic lesions: 63% of PDT-treated patients lost fewer than 15 letters visual acuity, versus 30% of sham-treated patients, a relative risk reduction of 110%.
The small, incorrectly enrolled subgroup of patients with purely occult lesions raised more questions than it answered.

"The only reasonable explanation for these results is that the classic and occult forms of AMD are actually quite different forms of neovascular AMD, occult being a more chronic and classic an acute form, with different natural courses and different responses to treatment, and not just different stages of the same disease. Before we can understand the Visudyne results properly, we need to know a lot more about the underlying natural history of AMD," Professor Michael Stur, MD, a retinal specialist at the Allgemeines Krankenhaus, University of Vienna, told EuroTimes.

The inexplicable variations in success rates across the post-hoc analysis of subgroups forced the FDA and EMEA into making unorthodox regulatory decisions.
Verteporfin therefore initially gained approval for use only in patients with predominantly classic AMD lesions.

However, it was ruled inappropriate for patients with minimally classic or occult CNV pending further clinical data. Verteporfin therefore gained approval for a subgroup of patients identified in post-hoc analysis and not, as is usually the case, on the basis of the primary endpoint.

This unorthodox regulatory decision was reasonable, but confusing to both ophthalmologists and Wall Street. Predominantly classic CNV is usually diagnosed by stereoscopic fluorescein angiography, and this typically demands referral to retinal specialists. Many more patients are diagnosed with minimally classic CNV and they gain relatively little from verteporfin PDT.

While it is difficult to justify withholding a treatment that might just work (when no other treatment is available for these patients) ambivalent clinical results have undermined the perception of PDT with verteporfin as an effective treatment, and highlighted its cost – no snippet at about $1750 a treatment, and an average cost per patient of $8--10,000 over several years.

The limited scope of regulatory approval for verteporfin hit trader expectations of the potential market size and was the first of a number of setbacks to QLT, developer and manufacturer of Visudyne.

In the summer of 2000, the Chief Executive Officer and Chief Financial Officer of QLT sold some $20 million shares, and were accused of improper dealing. Investor groups from all over the US sought damages against QLT. At the same time, the Massachusetts Eye and Ear Infirmary filed patent challenges against QLT.
The court-cases, though in favour of QLT, courted bad publicity for extended periods.

The share price crashed from a high of $80 a share in the summer of 2000, to a low of $20 a year later (before 9/11).
In the meantime, ophthalmic PDT products under development by other companies (Miravant and Pharmacyclics) were floundering, and have yet to reach the market. The disappointing clinical data inevitably questioned the efficacy of PDT in general, undermining confidence in verteporfin too.

VIP Science
The VIP (Verteporfin In Photodynamic therapy) Investigations began to clarify the science. Recruiting two subsets of patients, those with occult CNV secondary to AMD (which could include a classic component if visual acuity was better than 20/40), and those with pathologic myopia, the results were especially encouraging for people with pathologic myopia: at month 12, 72% of the PDT with verteporfin group had lost fewer than eight letters visual acuity, versus 44% of sham-treated patients, a relative risk reduction of 64% (P<0.01)

PDT with verteporfin was approved for patients with pathologic myopia in the EU and US in March and April, 2001, respectively.
The findings were less clear-cut for patients with occult subfoveal CNV. At month 12, there were no significant differences between PDT with verteporfin and sham treatment, but by month 24, 45% of the PDT group and 32% of sham-treated patients had lost fewer than 15 letters, a relative risk reduction of 41% (P = 0.032).

Chief Investigator Dr Neil Bressler, MD, of the Wilmer Eye Clinic, Johns Hopkins University, summed up the findings: "Based on these results, Visudyne should be considered for AMD patients with lesions composed of occult without classic CNV, especially if they have signs of recent disease progression."

On the basis of the VIP results, and a formal request from the US Vitreous Society, the CMS (Centers for Medicare and Medicaid Services), announced its intention to expand US national coverage for verteporfin PDT to include AMD patients with occult subfoveal CNV, as well as those predominantly classic CNV.

The FDA, however, insisted on its normal protocol for a second trial of PDT with verteporfin in patients with occult and minimally classic subfoveal CNV. Under pressure to be consistent the CMS reconsidered its position, and in April 2002 reversed its original decision. Ironically, in May 2002, the CPMP recommended approval of PDT with verteporfin in the EU for the treatment of occult CNV, which the EU (EMEA) ratified at the end of August 2002.

Twelve-month data from the phase II VIM (Visudyne in Minimally Classic) trial were presented at ARVO in May 2003. According to Neil Bressler the data showed that "patients with minimally classic lesions treated with verteporfin PDT therapy had a reduced risk of vision loss compared with placebo-treated patients" (P=0.01). A phase III trial begins in autumn 2003.

Who should get Verteporfin?
Cataract and refractive surgeons are used to seeing immediate benefits after surgery, and may be suspicious of benefits measured in slowed progression. But many degenerative diseases must be measured in this manner. It is therefore useful to compare verteporfin PDT with treatments for equivalent degenerative conditions. A useful way of doing this is to compare the average number who must receive the treatment in order for at least one patient to have a clinically and statistically significantly better outcome compared with no treatment (see table).

The numbers-needed-to-treat (NNT) analysis suggest that there are good clinical grounds for the use of PDT with verteporfin to delay visual loss in patients with occult (without classic) and predominantly classic subfoveal CNV secondary to AMD, as well as patients with pathologic myopia.

A re-analysis of clinical data presented by Professor Bressler at the International Congress of Ophthalmology in Sydney, Australia, April 2002, showed that early detection and treatment of small lesions could halve the rate of progression of visual acuity loss, regardless of lesion type.

The difference was most marked in pure occult lesions treated with verteporfin PDT, in which the mean visual loss was nearly four times less than that in the largest lesions. In contrast, and equally importantly, patients who had large occult lesions but good vision lost more lines with PDT than with sham treatment.

This means that the success of treatment can be undermined by slow referral. According to Dr Stur, "We get referrals mainly from general ophthalmologists. This is simply caused by the fact that cataract surgeons refer patients back after surgery to the general ophthalmologist.

"In my opinion, the major problem today is that screening is ineffective, and many patients are referred too late. Thus, the number of PDT-eligible patients is much lower than would be expected. Many general ophthalmologists have three-month waiting lists. Since many AMD patients are not aware of the urgent need for treatment, they wait until seen by the general ophthalmologist, who then finds a disciform scar and does not even refer to a retina specialist."