Epidemiological and Clinical Baseline Characteristics as Predictive Biomarkers of Response to Anti-VEGF Treatment in Patients with Neovascular AMD

Table 1

Summary of studies reviewed.

Article	Duration	Tx	Indication	Patient/eye number	Study findings
Phase 3 prospective studies
Altaweel et al., 2015 [31]	CATT: post hoc analysis (24 months)	RBZ/BVZ	nAMD	1185	Percentage blood composition of lesion did not affect VA gains at 12 and 24 months.
Bhisitkul et al., 2015 [18]	SEVEN-UP: post hoc analysis (7-8 years)	RBZ	nAMD	65	Long-term vision outcomes related to patient age but not patient gender or ethnicity. Macular atrophy lesion size was associated with VA.
Boyer et al., 2007 [16]	MARINA: post hoc analysis (24 months)	RBZ	nAMD	716	Increasing age, larger CNV lesion size at baseline, and a higher baseline VA score were all associated with less gain of VA. Most important predictors of VA outcomes were BCVA, CNV lesion size, and age.
Hariprasad et al., 2012 [29]	MARINA, ANCHOR, PIER, SAILOR: post hoc analysis (12 months)	RBZ	nAMD	1824	Lower baseline VA was associated with an early response to treatment.
Kaiser et al., 2007 [17]	ANCHOR: post hoc analysis (12 months)	RBZ	nAMD	423	Lower baseline VA, smaller baseline CNV lesion size, and younger baseline age were associated with greater gain of letters at study endpoint.
Lee et al., 2014 [30]	CATT: post hoc analysis (24 months)	RBZ/BVZ	nAMD	368	Poor VA, GA, and greater lesion size at baseline were associated independently with greater risk of ORT at 104 weeks.
Mayr-Sponer et al., 2013 [32]	EXCITE: post hoc analysis (12 months)	RBZ	nAMD	252	Patients with PVD or VMA at baseline showed no significant differences in efficacy at month 12.
Schmidt-Erfurth et al., 2015 [33]	VIEW: post hoc analysis (96 weeks)	RBZ/AFL	nAMD	1202	IRC at baseline had a negative impact on BCVA. BCVA had a robust influence on VA outcome; visual gains were higher with lower BCVA.
Simader et al., 2014 [34]	EXCITE: post hoc analysis (12 months)	RBZ	nAMD	353	IRC at baseline associated with significantly less gain in BCVA during follow-up. There was no significant difference in outcome between patients with and without SRF at baseline.
Waldstein et al., 2014 [35]	MONT BLANC: post hoc analysis (12 months)	RBZ/vPDT	nAMD	237	Patients with PVD and VMA at baseline receiving ranibizumab achieved similar VA gains and reductions in CRT at month 12.
Ying et al., 2013 [20]	CATT: post hoc analysis (12 months)	RBZ/BVZ	nAMD	1105	Older age, better baseline VA, large CNV area, predominantly or minimally classic lesion, absence of RAP lesion, presence of GA, greater foveal thickness, and RPE elevation were associated with less improvement in VA at 1 year.
Ying et al., 2014 [36]	CATT: post hoc analysis (24 months)	RBZ/BVZ	nAMD	1030	Nonfoveal GA and larger area of CNV at baseline were independently associated with a higher incidence of sustained visual loss.
Prospective studies (not phase 3)
Hoerster et al., 2014 [37]	24 months	RBZ	nAMD	75	Volume of fibrovascular PED at baseline correlated most with impaired BCVA after 24 months.
Shin and Yu, 2014 [19]	24 months	RBZ	RAP	31	Older age, larger CNV size, and poor initial BCVA were associated with poor VA outcome. Among factors associated with poor VA outcome, only the stage of RAP remained statistically significant on multiple linear regression analysis.
Weingessel et al., 2015 [38]	12 months	RBZ	nAMD	34	Better baseline BCVA was the most important predictive factor for final BCVA.
Wickremasinghe et al., 2012 [39]	12 months	RBZ	nAMD	88	Larger baseline retinal venular caliber was significantly associated with a poorer response to treatment.
Retrospective studies
Ahlers et al., 2009 [4]	3 months	RBZ	nAMD	30	ODR from subretinal fluid correlated with BCVA at weeks 4 and 12. Strong association between baseline BCVA and visual function at subsequent visits.
Alkin et al., 2013 [40]	24 months	BVZ	nAMD	63	Patients with idiopathic epiretinal membranes at baseline showed no significant differences in efficacy at month 12 and month 24 compared to those without.
Bloch et al., 2013 [41]	12 months	RBZ	nAMD	279	Great BCVA (≥70 letters) and smaller lesion size (<4 DA) associated with better efficacy outcomes.
Byun et al., 2010 [25]	12 months	BVZ	nAMD	113	SRT thickness and CME were associated with efficacy outcomes.
Chhablani et al., 2012 [42]	6 months	BVZ	nAMD	85	Baseline BCVA was a predictive factor for the visual outcome.
Chhablani et al., 2013 [43]	11.2 months	BVZ	nAMD	50	BCVA, IS/OS junction, and ELM damage were significant predictors for treatment effect and visual improvement.
Clemens et al., 2014 [44]	No time frame given	RBZ/BVZ	nAMD	103	Serous vascularized PED at baseline was associated with RPE tears, whereas fibrovascular PED was not.
Coco et al., 2014 [45]	1.9–5.4 years	RBZ	nAMD/PCV	299	Patients with PCV had a worse final outcome. Worse initial VA associated with atrophy at the final visit.
Fang et al., 2013 [23]	6 months	BVZ	nAMD	144	Younger age, lower baseline VA, and shorter duration of disease were significantly associated with greater VA score improvements.
Fong et al., 2008 [46]	9.4 months	BVZ	nAMD	109	Large ICR (gross CME) before treatment had increased risk of worse vision.
Framme et al., 2010 [47]	3 months	RBZ	nAMD	61	Baseline amount of SDPs correlated positively with the increase in BCVA; larger number associated with better outcome with ranibizumab therapy.
Kang et al., 2014 [48]	6 months	RBZ	nAMD	40	Baseline BCVA, baseline CNV size, and subfoveal choroidal thickness were significant prognostic factors for visual outcome.
Kang and Roh, 2009 [49]	12 months	RBZ	nAMD	64	Baseline VA and CNV size influenced VA outcomes.
Kim et al., 2014 [27]	6 months	RBZ	nAMD	91	Longer duration of symptoms, greater extent of hemorrhage, and greater CFT at baseline were correlated with poor BCVA at month 6.
Kolb et al., 2012 [50]	12 months	RBZ/BVZ	nAMD	75	CRT, IS/OS integrity, and retinal fluid did not have a predictive value regarding VA outcome.
Kwon et al., 2014 [51]	3 months	RBZ	nAMD	59	Better initial VA and greater ELM length at baseline were associated with less change in VA. Initial IS/OS-D, ELM length, and particularly ELM-D can be useful predictors of the visual outcome.
Leitritz et al., 2008 [52]	No time frame given	BVZ	nAMD	393	Risk of an RPE tear correlates with the height of the PED on OCT.
Levy et al., 2009 [24]	6 months	BVZ	nAMD	65	Eyes with better VA at baseline and without previous PDT treatment achieved better final VA. Classic membrane type and lower age somewhat associated with better posttreatment VA.
Mathew et al., 2013 [53]	12 months	RBZ	nAMD	100	Intact EZ and the ELM in the subfoveal area at BL indicated final VA at month 12. Patients with ELM have VA nearly 20 letters higher than those without.
Matsumiya et al., 2015 [22]	24 months	RBZ	nAMD/PCV	59	Typical nAMD associated with greater BCVA improvement compared with PCV at BL. Age was associated with response in PCV, but not nAMD. Greater height of PED associated with VA outcome in PCV group.
Menghini et al., 2010 [54]	12 months	RBZ	nAMD	60	Baseline VA was statistically significantly lower in good responders than in bad responders.
Nomura et al., 2014 [55]	12 months	RBZ	nAMD	123	VMA at baseline associated with poor treatment outcomes. Better baseline BCVA was associated with poor visual response.
Oishi et al., 2013 [56]	7.7 months	RBZ	nAMD	76	Baseline BCVA was the most powerful predictor for VA prognosis. ELM length, IS/OS length, and foveal thickness showed weaker correlation.
Shona et al., 2011 [57]	12 months	RBZ	nAMD	77	Poor baseline VA was a predictor of maximum gain in VA. Eyes with better baseline VA had a better final VA.
Singh et al., 2009 [28]	6 months	BVZ	nAMD	73	Worse BL BCVA associated with better final VA. Thicker CRT associated with a greater reduction in CRT. Treatment-naïve patients had a greater mean CRT reduction than  those who had previously received treatment.
Suzuki et al., 2014 [58]	12 months	RBZ	nAMD	141	Initial fibrovascular PED and serous PED were associated with nonresponse as judged by BCVA. Initial fibrovascular PED and type 1 CNV were associated with nonresponse, as judged by fundus findings.
Toth et al., 2015 [59]	36 months	RBZ	nAMD	420	Regression analysis identified atrophy and fibrosis as predictors of best BCVA.
Tran et al., 2011 [60]	12 months	RBZ	nAMD	59	Baseline VA is a predictor of visual gain.
Üney et al., 2014 [61]	22.3 months	RBZ/BVZ	nAMD	61	Patients with PVD at baseline were associated with a greater rate of improved or stable BCVA, compared with patients with VMA.
van Asten et al., 2014 [26]	3 months	RBZ	nAMD	391	Independent predictors for nonresponse were age and baseline VA.
Yamashiro et al., 2012 [21]	12 months	RBZ	nAMD/PCV	105	Age, VA, and size of GLD (lesion size) were significantly associated with visual prognosis in nAMD, but not PCV.

AFL = aflibercept; BCVA = best-corrected visual acuity; BL = baseline; BVZ = bevacizumab; CFT = central foveal thickness; CME = cystoid macular edema; CNV = choroidal neovascularization; CRT = central retinal thickness; DA = disc areas; ELM = external limiting membrane; EZ = ellipsoid zone; GA = geographic atrophy; GLD = greatest linear dimension; IRC = intraretinal cysts; IS = inner segment; nAMD = neovascular age-related degeneration; ODR = optical density ratio; ORT = outer retinal tubulations; OS = outer segment; PCV = polypoidal choroidal vasculopathy; PED = pigment epithelial detachment; PVD = posterior vitreous detachment; RAP = retinal angiomatous proliferation; RBZ = ranibizumab; REP = retinal pigment epithelium; SDP = small dense particle; SRF = subretinal fluid; VA = visual acuity; VMA = vitreomacular adhesion; vPDT = verteporfin photodynamic therapy.