A, Late-phase fluorescein angiogramof study participant judged to have a minimally classic lesion at baselinein which the area of classic choroidal neovascularization (CNV) is less than50% the area of the entire lesion (visual acuity, 20/100). B, Subsequent gradingof the lesion composition at the 3-month examination indicates conversionto a predominantly classic lesion in which the area of classic CNV is at least50% the area of the entire lesion (visual acuity, 20/80).
A, Early-phase fluorescein angiogramof study participant judged to have a minimally classic lesion at baselinein which the area of classic choroidal neovascularization (CNV) is less than50% the area of the entire lesion (visual acuity, 20/40). B, Subsequent gradingof the lesion composition at the 3-month examination indicates a minimallyclassic lesion in which the area of classic CNV, while larger than the areaof classic CNV at baseline, remains less than 50% the area of the entire lesion(visual acuity, 20/100).
Left, Number of eyes with a minimallyclassic lesion at baseline that converted to a predominantly classic lesion(n = 98) at follow-up examination when conversion was first noted. Right,Percentage of conversions by the 24-month examination by follow-up examinationwhen conversion was first noted.
Flow diagram for 39 of 98 minimallyclassic lesions (among 207 patients assigned to placebo in the Treatment ofAge-Related Macular Degeneration With Photodynamic Therapy Investigation)that converted to predominantly classic lesions by the 24-month examinationby lesion size and visual acuity when conversion was first noted. MPS indicatesMacular Photocoagulation Study.
Bressler SB, Pieramici DJ, Koester JM, Bressler NM, TAP Study Group. Natural History of Minimally Classic Subfoveal Choroidal NeovascularLesions in the Treatment of Age-Related Macular Degeneration With PhotodynamicTherapy (TAP) InvestigationOutcomes Potentially Relevant to Management—TAP Report No.6. Arch Ophthalmol. 2004;122(3):325-329. doi:10.1001/archopht.122.3.325
To determine if there is a rationale for monitoring patients with age-relatedmacular degeneration who have a minimally classic subfoveal choroidal neovascularlesion and do not receive treatment at initial examination.
Participants assigned to placebo who had a minimally classic lesioncomposition at baseline were identified from the TAP Investigation. Fluoresceinangiograms at baseline and follow-up examinations from these participantswere reviewed by photograph reading center graders to determine if any follow-upangiograms had converted from a minimally classic lesion composition to apredominantly classic lesion composition.
Main Outcome Measures
Proportion of minimally classic lesions at baseline that converted toa predominantly classic lesion composition, time of this conversion, and visualacuity and lesion size at the time of conversion.
Of the 207 patients assigned to placebo in the TAP Investigation, 98were judged to have a minimally classic lesion at baseline in the study eyewhen the fluorescein angiograms were reviewed in 2001. Of these 98 patients,39 (40%) had lesions that converted to a predominantly classic lesion composition,including 21 by the month 3 examination. At the time of conversion, 32 (82%)lesions were no greater than 9 disc areas, including 20 (51%) with visualacuity of 20/200 or better.
These data would suggest that patients with minimally classic lesions,in whom no therapy is recommended initially, should be monitored so that potentialconversion to a predominantly classic lesion can be identified promptly andverteporfin therapy considered.
The Treatment of Age-Related Macular Degeneration With PhotodynamicTherapy (TAP) Investigation showed that photodynamic therapy with verteporfin(Visudyne; Novartis AG, Basel, Switzerland) could reduce the risk of moderate(≥3 lines) and severe (≥6 lines) visual acuity loss compared with asham treatment in selected patients with age-related macular degeneration(AMD). These patients had subfoveal choroidal neovascularization (CNV) judgedby photograph reading center graders to have a predominantly classic lesioncomposition (area of classic CNV ≥50% area of entire lesion).1- 3 Inthe subgroup analysis of minimally classic lesions (area of classic CNV >0%to <50% area of entire lesion) in the TAP Investigation, the therapy didnot reduce the risk of vision loss. This latter finding seemingly was inconsistentwhen the results in predominantly classic lesions from the TAP Investigationwere coupled with results from the Verteporfin in Photodynamic Therapy Trial,in which the therapy reduced the risk of vision loss in selected cases withan occult with no classic composition and presumed recent disease progression.4 Subsequent exploratory analyses made some sense ofthis apparent inconsistency, suggesting that the benefits of verteporfin therapymight be dependent not only on the lesion composition but also on lesion size.4,5 Recent results reported from the Visudynein Minimally Classic CNV Trial also suggest that verteporfin therapy mightreduce the risk of moderate and severe vision loss through at least 1 yearof follow-up in selected eyes with minimally classic lesions.6 Currentexpert opinion suggests that verteporfin therapy could be considered for patientswith minimally classic lesions with smaller lesions (≤4 Macular PhotocoagulationStudy [MPS] disc areas) and lower levels of visual acuity (≤20/50−1).7 However, verteporfin therapyis not widely used for minimally classic CNV lesions because evidence of effectivenessin these lesions is not considered conclusive.
In this setting, investigators participating in the TAP Investigationquestioned if data from the verteporfin randomized clinical trials could providea rationale for monitoring patients with AMD who have subfoveal minimallyclassic lesions if verteporfin therapy was not applied at initial examination.Specifically, for minimally classic lesions assigned to placebo in the TAPInvestigation, the investigators wanted to determine how often a predominantlyclassic lesion developed. To answer this question, the investigators determinedthe proportion of placebo-treated minimally classic lesions that convertedto a predominantly classic lesion, as well as the lesion size and visual acuityat the time of this conversion.
The methods of the TAP Investigation have been reported previously.1,2 In brief, study participants were enrolledinto 1 of 2 identically designed randomized clinical trials between December1996, and September 1997, at 1 of 22 clinical centers throughout North Americaand Europe after an informed consent process and form were approved by therespective center's institutional review board. Principal eligibility criteriaincluded the following: (1) presence of subfoveal CNV due to AMD in whichthe lesion on fluorescein angiography had evidence of some classic CNV, (2)a greatest linear dimension on the retina no greater than 5400 µm, and(3) a lesion in which at least 50% of the lesion components were CNV (includingany classic and any occult CNV) following specific definitions for these termsas previously described.7- 9 Atstudy entry, patients were assigned randomly to verteporfin therapy or a shamtreatment using a placebo in a double-masked fashion.1 Patientswere to be observed as often as every 3 months for at least 2 years with baselineand follow-up examinations that included best-corrected visual acuity followinga standardized protocol refraction and visual acuity determination, as wellas stereoscopic color fundus photographs and fluorescein angiograms.
At baseline, 2 independent photograph reading center graders categorizedlesions with evidence of classic CNV as predominantly classic (area of classicCNV ≥50% area of entire lesion) or minimally classic (>0% to <50% areaof entire lesion) without knowledge of treatment assignment, followed by openadjudication of any discrepancies between graders and review by a readingcenter ophthalmologist. At each scheduled follow-up examination (every 3 monthsthrough the 12-month examination, then every 6 months through the 24-monthexamination) for initial studies reported,1,2 photographreading center graders identified whether any classic CNV was present andmeasured the size of the entire lesion within specific disc area categories;however, the graders did not determine whether the lesion was predominantlyclassic or minimally classic when classic CNV was identified at follow-up.These baseline and follow-up assessments were completed in the late 1990s.
In 2001, follow-up angiograms were reanalyzed to determine whether lesionswere predominantly classic (Figure 1)or minimally classic (Figure 2)when classic CNV was identified at follow-up. Because follow-up angiogramswere reanalyzed, baseline angiograms also were reanalyzed to minimize temporalvariability in the grading process, described previously.9 Thisreanalysis of baseline angiograms resulted in slight changes in the totalnumber of placebo-treated lesions originally categorized as minimally classic(n = 104) from the TAP Investigation.2 Thissubgroup of 104 patients originally included 98 lesions confirmed as minimallyclassic and 6 lesions graded as questionably having evidence of classic CNVby the reading center graders. These 6 lesions graded as questionably classicoriginally were included in the minimally classic subgroup before outcomeassessment based on the enrolling ophthalmologist's assessment that classicCNV was present to meet the eligibility criteria. Of the 6 lesions originallygraded as questionably classic, 2 were judged to be minimally classic and4 subsequently were judged to have no classic CNV in the 2001 rereview; theselatter 4 lesions were not analyzed further. Of the remaining 98 lesions originallygraded as minimally classic, 2 subsequently were judged to be predominantlyclassic in the 2001 rereview and were not analyzed further. Therefore, 98minimally classic lesions assigned to placebo were analyzed for this article.The 2001 reevaluation of angiograms to determine lesion composition was doneaccording to the same procedures as for initial evaluations (ie, 2 independentgraders, open adjudication, and review by a reading center ophthalmologist).For those lesions that converted to a predominantly classic composition, nofurther follow-up angiograms were evaluated, but the visual acuity and lesionsize at the time of conversion were obtained from the TAP Investigation'sdatabase.
Of the 104 study eyes of 104 study participants assigned to placebothat were originally included in the minimally classic subgroup from the TAPInvestigation, 98 study eyes in 98 participants were confirmed in 2001 byphotograph reading center graders to have a minimally classic lesion compositionat baseline. Of these 98 cases, 39 (40%) had a lesion composition judged bythe graders to become predominantly classic at a follow-up examination between3 and 24 months after study entry. For the 39 lesions that converted to apredominantly classic lesion, the size of the minimally classic lesion atbaseline was no greater than 4 MPS disc areas in 16 (41%), including 15 eyeswith a visual acuity (approximate Snellen equivalent) no better than 20/50−1. The size of the minimally classic lesion at baseline was greaterthan 4 MPS disc areas in 22 cases (56%). The lesion size at baseline couldnot be determined in 1 case (3%), although the lesion composition could begraded as minimally classic.
Figure 3 shows that most ofthese conversions (21/39 [54%]) were first noted at the 3-month examination,with most conversions (34/39 [87%]) noted by the 9-month examination and only2 (5%) noted at the 24-month examination. For the 16 conversions among lesionsno greater than 4 MPS disc areas at baseline, 10 converted at the 3-monthexamination, 4 at 6 months, 1 at 9 months, and 1 at 18 months. For the 22conversions among lesions greater than 4 MPS disc areas, 11 converted at the3-month examination, 3 at 6 months, 4 at 9 months, 2 at 18 months, and 2 at24 months. One patient with a lesion size that could not be determined atbaseline converted at 9 months.
Figure 4 shows the lesionsize and visual acuity at the time of conversion for those study eyes thatconverted to a predominantly classic lesion. For 7 (18%) of the 39 eyes, thelesion size was greater than 9 MPS disc areas with a visual acuity (approximateSnellen equivalent) ranging from 20/160 to 20/800. For the remaining 32 (82%)of the 39 eyes, the lesion size was no greater than 9 MPS disc areas, including20 (51%) of the 39 eyes with this lesion size that had a visual acuity of20/200 or better at the time of conversion.
Of 98 placebo-treated lesions identified as having a minimally classiccomposition at study entry, 39 (40%) converted to a predominantly classiclesion, including 21 by the 3-month examination and 34 by the 9-month examination,although some conversions were noted for the first time at the 24-month examination.These results were similar whether or not at baseline the minimally classiclesion was small (≤4 MPS disc areas). At the time that the conversion wasfirst noted, about half of the patients had lesion sizes no greater than 9MPS disc areas accompanied by visual acuities of 20/200 or better, such thatverteporfin therapy likely would be considered based on the inclusion criteriaand results from the TAP Investigation.
Several caveats to this analysis should be recognized when consideringusing this information to manage patients with AMD who have subfoveal CNVwith a minimally classic lesion composition. First, while verteporfin therapyhas been shown to be beneficial for lesions with predominantly classic lesioncompositions, one does not know how many of the predominantly classic lesionsat initial examination had converted from a minimally classic lesion. Therefore,one does not know if results from the TAP Investigation regarding predominantlyclassic lesions apply to lesions that had converted from minimally classic,or those that did not convert from minimally classic lesions, or both. Nevertheless,there is no biologic rationale, in our opinion, to suggest that the TAP Investigationresults on predominantly classic lesions would not apply to minimally classiclesions that convert to predominantly classic lesions.
Second, recent information suggests that lesion size is a strong factorin determining if verteporfin therapy is likely to reduce the risk of visionloss. Specifically, smaller minimally classic lesions and smaller occult withno classic lesions (≤4 MPS disc areas) had a greater treatment benefitthan larger lesions within these lesion compositions.3- 5 Basedon this recent information, as well as reports from the Visudyne in MinimallyClassic CNV Trial,6 some ophthalmologists mightconsider recommending verteporfin therapy for minimally classic lesions thatare small, without waiting for conversion to a predominantly classic lesion,perhaps making the findings of this report less relevant for minimally classiclesions that are small at initial examination.
Third, this analysis may underestimate the number of lesions that mayconvert to a lesion composition for which therapy might be beneficial. Specifically,the Verteporfin in Photodynamic Therapy Trial4 resultsin patients with AMD showed that verteporfin therapy could reduce the riskof moderate and severe visual acuity loss in selected patients with presumedrecent disease progression who have subfoveal CNV with a lesion compositionthat has occult CNV with no classic CNV. The methods for this study did notallow one to determine how many minimally classic lesions converted to a lesioncomposition that had occult CNV with no classic CNV. If conversion to a lesioncomposed of occult with no classic CNV occurred in the setting of presumedrecent disease progression, verteporfin therapy also might reduce the riskof vision loss, especially if the lesion was smaller (as noted when usinga cutoff of ≤4 MPS disc areas) or associated with lower levels of visualacuity (as noted when using a cutoff of ≤20/50−1). Underestimationalso is possible if one considers that more cases might have been identifiedif follow-up fluorescein angiograms were done more frequently.
Fourth, this study does not differentiate among cases that developeda small amount of increased classic CNV (eg, from 45%-55% of the lesion) fromthose developing a large amount (eg, from 10%-90%). Both examples would belabeled as progressing to a predominantly classic lesion. The reliabilityof grading areas of classic CNV within categories smaller than predominantlyor minimally classic has not been evaluated, and such information was notcollected in this trial, as it was judged to be difficult to make finer differentiationsthan predominantly and minimally classic. Nevertheless, this study suggeststhat some minimally classic cases will be considered predominantly classicat follow-up with a visual acuity and lesion size for which verteporfin therapylikely would be considered to reduce the risk of additional moderate and severevisual acuity loss.
Fifth, as only 20 eyes (20% of the cases available for this study) convertedto a predominantly classic lesion composition with a lesion size no greaterthan 9 MPS disc areas and a visual acuity of 20/200 or better, this numberis too small to identify factors associated with this conversion. Even ifone were to treat minimally classic lesions with a small size, this studysuggests a rationale to observe those with a larger size. Some of these largerminimally classic lesions will convert to a predominantly classic lesion witha size no greater than 9 MPS disc areas and a visual acuity of 20/200 or better,a situation in which many experts would consider recommending verteporfintherapy.7 Unfortunately, the number of casesevaluated in this study is too small to determine which minimally classiclesions will convert if monitored, necessitating that one consider observingall minimally classic lesions that do not receive treatment at initial examination.
In conclusion, this analysis suggests that ophthalmologists who identifypatients with subfoveal CNV with a minimally classic composition for whomno verteporfin therapy is recommended should consider continued follow-uprather than discharging the patient from continued monitoring. Prudent follow-upwould provide an opportunity to identify cases that may convert to a predominantlyclassic composition for which verteporfin therapy likely would reduce therisk of moderate and severe visual acuity loss compared with no therapy. Asmany of these conversions may occur within the first 3 months after a patientis initially seen with a minimally classic lesion, frequent monitoring ofsuch patients within the first 3 months of examination seems warranted totry to identify these conversions while the lesion is still small enough andthe visual acuity is still good enough to potentially benefit from verteporfintherapy. Because conversions may occur at least up to 24 months after initialexamination, monitoring may be indicated as long as the lesion has not becometoo large or the visual acuity too poor before the conversion to suggest thatreducing the risk of moderate and severe visual acuity loss with verteporfintherapy no longer would be helpful to the patient. Because few conversionsoccurred beyond the 9-month examination, the frequency of monitoring beyond9 to 12 months probably can be based on whether the situation (eg, consideringthe visual acuity, lesion size, and lesion composition) appears stable fromone examination to the next. These recommendations may change as new datarelevant to the management of minimally classic lesions in patients with AMDare obtained.
Corresponding author: Neil M. Bressler, MD, The Retinal VascularCenter, The Wilmer Ophthalmological Institute and Department of Ophthalmology,The Johns Hopkins University School of Medicine and The Johns Hopkins Hospital,550 N Broadway, Suite 115, Baltimore, MD 21205-2002 (e-mail: email@example.com).
?Reprints: Novartis Ophthalmics Medical Affairs, One HealthPlaza, Building 118, East Hanover, NJ 07936.
Submitted for publication May 28, 2003; final revision received October22, 2003; accepted November 6, 2003.
This study was sponsored by Novartis AG and QLT Inc.
The authors had complete access to the raw data needed for this article.