Improvement in quality of life from photodynamic therapy: a Canadian perspective

Improvement in quality of life from photodynamic therapy: a Canadian perspective

Improvement in quality of life from photodynamic therapy: a Canadian perspective Sanjay Sharma,*tt MD, FRCSC, MSc(Epid), MBA; Hussein Hollands,* MSc(E...

610KB Sizes 0 Downloads 13 Views

Improvement in quality of life from photodynamic therapy: a Canadian perspective Sanjay Sharma,*tt MD, FRCSC, MSc(Epid), MBA; Hussein Hollands,* MSc(Epid); Gary C. Brown,§ MD, MBA; Melissa M. Brown,§ MN, MD, MBA; Gaurav K. Shah,1I MD; Susan M. Sharma,* MD ABSTRACT • RESUME Background: The Treatment of Age-Related Macular Degeneration with Photo-

dynamic Therapy (TAP) Study showed that at I year, photodynamic therapy significantly reduced the chances of severe visual loss (15 letters or greater) compared with placebo treatment in patients with "predominantly classic" subfoveal choroidal neovascularization (CNV). We performed a study to determine the expected gain in quality of life associated with photodynamic therapy for the treatment of subfoveal CNV in a Canadian cohort of patients with age-related macular degeneration. Methods: We created a decision analysis model to determine the incremental gain in quality-adjusted life years (QALYs) associated with photodynamic therapy over placebo over a 2-year period. The analysis was conducted using efficacy data derived from the TAP Study and patient-based utilities collected by means of the time tradeoff technique. We conducted one-way and two-way sensitivity analyses to determine the robustness of our model. A Monte Carlo simulation was used to determine whether the observed gain in QALYs with photodynamic therapy was significant. Results: Photodynamic therapy was associated with a relative increase in QALYs of 11.3% compared with placebo. In one-way and two-way sensitivity analyses, the relative increase in quality of life associated with photodynamic therapy ranged from 4.2% to 25.7%. The Monte Carlo simulation showed that the gain in QALYs conferred from photodynamic therapy was statistically significant (p < 0.00 I). Interpretation: Photodynamic therapy improves the quality of life of Canadians with predominantly classic subfoveal CNV secondary to agerelated macular degeneration.

Contexte : L'etude Treatment of Age-Related Macular Degeneration with

Photodynamic Therapy (TAP) (I'Etude du traitement de la degenerescence macu-

From *the Cost-Effective Ocular Health Policy Unit, Hotel Dieu Hospital, Queen's University, Kingston, Ont., the Departments of tOphthalmology and :l:Community Health and Epidemiology, Queen's University, Kingston Onl., §the Center for E vid ence-Based Healthcare Economics, Flourtown, Pa., andlithe Barnes Retina Institute, Washington University, St. Louis, Mo. Originally received Nov. 9, 2000

332

Photodynamic therapy-Sharma et al

Accepted for publication June 1, 200 1 Reprint requests to: Dr. Sanjay Sharma, Cost-Effective Ocular Health Policy Unit, HotelDieu Hospital, Brock 2, 166 Brock St., Kingston O N K7L 5G2; [email protected] hotmail.com

Can J Ophthalmol 200 I ;36:332-8

Photodynamic therapy-Sharma et al

laire liee it I'age par therapie photodynamique) a demontre qu'it un an, celle-ci reduit de fa~on significative Ie risque de perte grave de la vue (15 lettres ou plus), comparativement au placebo, chez des patients qui avaient une neovascularisation choro'idienne (NYC) sousfoveale ({ it predominance classique ». Notre etude a cherche it etablir les gains prevus en matiere de qualite de vie associes it la therapie photodynamique pour Ie traitement de la NYC sousfoveale chez une cohorte de patients canadiens ayant une degenerescence maculaire liee it I'age. Methodes: Nous avons cree un modele d'analyse de decision pour etablir les gains differentiels en annees-personnes sans invalidite (APSI) associes it la therapie photodynamique, en regard d'un placebo, sur une periode de deux ans. L'analyse a ete faite it partir de donnees d'efficacite derivees de I'etude TAP et d'utilites recueillies it I'aide de la technique de I'echange de temps (time trade-off). Nous avons mene des analyses de sensitivite it facteurs simples et doubles pour etablir la solidite de notre modele. Une simulation Monte-Carlo a perm is d'etablir si les gains en APSI observes it la suite de la therapie photodynamique ont ete importants. Resultats : La therapie photodynamique a ete associee it une hausse relative de APSI de I 1,3 %, comparativement au placebo. Dans les analyses de sensitivite it facteurs simples et doubles, I'amelioration relative de la qualite de vie associee it la therapie photodynamique a varie entre 4,2 % et 25,7 %. La simulation Monte-Carlo a demontre que les gains en APSI obtenus grace it la therapie photodynamique ont ete importants sur Ie plan de la statistique (p < 0,00 I). Interpretation: La therapie photodynamique ameliore la qualite de vie des Canadiens qui ont une NVC sousfoveale a predominance classique derivee d'une degenerescence maculaire liee a I'age.

A

ge-related macular degeneration (AMD) is the leading cause of visual loss in people over the age of 50 years in North America. 1 Visual loss from this disease is most commonly associated with the development of choroidal neovascularization (CNV).2 Thermal laser photocoagulation was demonstrated to be of benefit for patients with well-defined juxta- and extrafoveal CNV in studies conducted by the Macular Photocoagulation Study Group.3.4 Although the study group also found that laser photocoagulation was of benefit in some patients with subfoveal CNV,5 this treatment has not been readily adopted by most clinicians given the immediate visual loss associated with it. Moisseiev and colleagues6 showed that less than 10% of patients with this problem receive treatment. Photodynamic therapy, consisting of administration of a photosensitizing dye followed by laser treatment, has been suggested as a possible treatment for subfoveal CNV. The Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study, a randomized clinical trial, was designed to evaluate the efficacy and safety of this therapy with Visudyne (verteporfin) (Novartis Ophthalmics, Dorval, Que.) for the treatment of subfoveal CNV.7.8 At 1 year photodynamic therapy significantly reduced the chances of

severe visual loss (defined as a loss of 15 letters or greater) compared with placebo in patients with "predominantly classic" subfoveal CNV.7 We performed a study to determine the improvement in quality of life that may be expected to be gained by a cohort of patients who receive photodynamic therapy for the treatment of predominantly classic subfoveal CNV in Canada. METHODS

We created a decision analysis model for a cohort of hypothetical patients similar to the patients involved in the TAP Study, who had a disciform scar in one eye and predominantly classic CNV in their second and betterseeing eye. Our model followed this cohort for 2 years. The decision analysis model incorporated the following pieces of information: measurement of therapeutic efficacy, utility valuation, and estimation of complications and the disutility associated with those complications.

Therapeutic efficacy Table 1 shows the proportion of patients who

CAN J OPHTHALMOL-VOL. 36, NO.6, 2001

333

Photodynamic therapy-Sharma et al

Table I-Proportion of patients who received either photodynamic therapy or placebo who had a 3-line loss in vision after I and 2 years of followup, as reported by the Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group7,8 Group; % of patients (and 95% confidence interval [CI]*) Follow-up period

Placebo

Photodynamic therapy (n 159 at baseline)

I yr

2 yr

=

(n

0.390 (0.31 ~.470) 0.313 (0.24~.392)

=84 at baseline)

0.670 (0.555-0.766) 0.591 (0.482-0.70 I)

*Exact binomial confidence intervals calculated by us, assuming no loss to follow-up after 2 years.

received either photodynamic therapy or placebo in whom severe visual loss developed at 1 year and at 2 years. Although not provided by the authors, we have included 95% confidence intervals (CIs) (assuming no loss to follow-up over 2 years) in the table. The 1- and 2-year results were obtained from the published TAP Study data. 7,8 Utility assessment

Utility theory has recently been used to quantify the quality of life of patients with various ocular conditions and is known to be highly dependent on visual acuity in the better-seeing eye.9- 12 Between May 2000 and February 2001 we conducted a crosssectional study of patients who presented to one of us (S.S.) for evaluation of AMD. Patients were eligible if they had vision of 6112 or worse in at least one eye and had received a diagnosis of AMD. In addition, the patient had to be competent to answer the questions required. Exclusion criteria were language or other

communication barriers, developmental disability and psychiatric illness. The study was approved by the Human Ethics Review Board of Queen's University, Kingston, Ont. Utilities

We obtained our utilities using the time trade-off technique. This technique has previously been described. 9- 12 Our decision analysis model was based on a dichotomous outcome (development or nondevelopment of a 3-line loss in vision) for patients with CNV secondary to AMD. For our model, to assess the utility associated with the development of a 3-line loss in vision, we used a mean utility for actual patients (n = 40) with visual acuity values ranging from 6/12 to 6/30. The utilities (and 95% CIs) associated with the development and nondevelopment of a 3-line loss in vision over a 2-year period are 0.56 (0.44-0.66) and 0.81 (0.72-0.89) respectively.

Table 2-Utilities and probabilities associated with various complications reported in the TAP Study7,8 Group; incidence of complications, % Complication Visual disturbance Vitreous hemorrhage Retinal capillary non perfusion Injection site adverse event Infusion-related back pain Allergic reaction Photosensitivity reaction

334

CAN J OPHTHALMOL-VOL. 36, NO.6, 2001

Utility

Photodynamic therapy

Placebo

1.0 0.8 0.95 0.99 0.99 0.95 0.98

71 (17.7) 4 ( 1.0) I (0.2) 54 (13.4) 9 (2.2) 5 ( 1.2) 12 (3.0)

24 (11.6) I (0.5) 2 ( 1.0) 7 (3.4) 0 (0.0) 7 (3.4) 0 (0.0)

Photodynamic therapy-Sharma et al

Visual outcomes (Markov states) • Development of visual loss ~ 3 lines • Nondevelopment of visual loss ;;::: 3 lines

Overall utility of PDT or placebo treatment PDT or placebo

Potential complications

Fig. I-Influence diagram of decision tree. Rectangle represents decision nodes, ovals represent chance nodes, and diamond represents outcome nodes. Arrows represent influence from one part of the tree to another. PDT = photodynamic therapy.

Complications and associated dis utility The various complications listed in Table 2 were reported in the TAP Study,?8 The disutilities associated with these potential complications were determined after lengthy discussions by a Delphi panel, consisting of a group of ophthalmologists (S.S., G.C.B., M.M.B. and G.K.S.). The probabilities and utilities associated with potential complications were incorporated into the model multiplicatively.

Decision analysis model To determine the expected utility gained from photodynamic therapy, we entered the utilities and probabilities into the medical decision analysis program TreeAge. 13 The model is depicted graphically as an influence diagram in Fig. 1. Our decision tree was designed to decide whether patients with AMD in whom predominantly classic subfoveal CNV develops in their second and better-seeing eye should receive photodynamic therapy. The decision tree was "rolled back" to obtain the preferred method of action, defined as the option that maximizes overall expected utility. In addition, we were able to calculate the expected quality-adjusted life years (QALYs) associated with both treatments.

Sensitivity analyses We performed sensitivity analyses on the decision

tree to determine our confidence in our models given the inherent variability of the estimates of efficacy and utilities. A one-way sensitivity analysis was performed on each of the efficacy probabilities, with the probability variables being varied over the entire range of their respective 95% CIs. When one of the probabilities was varied, the "remaining" probabilities were proportionally affected, such that the sum of probabilities at the node always equalled 1. Sensitivity analysis was also performed on each of the utilities and was varied within the entire range of their 95% CIs. In addition, we performed a two-way sensitivity analysis by varying both utilities and efficacy probabilities simultaneously.

Monte Carlo simulation In medical decision analysis, a Monte Carlo simulation allows for the opportunity to actually put a hypothetical cohort of patients through the decision model and to look at an observed outcome, as opposed to an expected outcome. If a very large cohort of patients is put through the Monte Carlo simulation, the observed outcome should approximate the expected value from a classic decision analysis output. A Monte Carlo simulation allows an investigator to measure the variability in a decision analysis; consequently, statistical tests can be performed to determine the significance of the observed result.

CAN J OPHTHALMOL-VOL. 36, NO.6, 2001

335

Photodynamic therapy-Sharma et al

Table 3-0ne-way sensitivity analysis varying efficacy probabilities within the 95% Cis

Treatment; follow-up period Photodynamic therapy I yr 2 yr Placebo I yr 2 yr

Probability of 3-line loss in vision* (and range)

Range of QALYst gained from photodynamic therapy

Range of % increase in quality of life from photodynamic therapy

0.390 (0.314-0.470) 0.313 (0.244-0.392)

0.106-0.172 0.128-0.151

8.6-13.9 10.4-12.2

0.670 (0.555-0.766) 0.591 (0.482-0.70 I)

0.099-0. 173 0.131-0.149

7.8--14.5 10.5-12.1

*Taken from the TAP Study.1.8 Efficacy probabilities are varied over the width of 95% Cis calculated by us. tQALY = quality-adjusted life year.

We performed a Monte Carlo simulation with a hypothetical cohort of patients equal to the number of patients with classic CNV in the TAP Study (159 patients in the photodynamic therapy arm and 84 patients in the placebo arm). We then performed an F-test using data from a one-way analysis of variance to determine whether there was a significant difference in observed gain in QALY s between the two treatment arms. RESULTS

The results of our decision analysis model showed that patients with predominantly classic CNV who received photodynamic therapy gained 1.373 QALYs. Patients who did not receive photodynamic therapy gained 1.234 QALY s. Therefore, photodynamic therapy confers an absolute improvement of 0.139 QALY s, equivalent to a relative increase in quality of life of 11.3%. When sensitivity analyses were performed varying both I-year and 2-year efficacy probabilities and util-

ity values, photodynamic therapy always remained the preferred outcome. One-way sensitivity analyses varying the efficacy probabilities and utility values are shown in Tables 3 and 4 respectively. When the 1year and 2-year treatment efficacy probabilities were varied over their entire 95% CIs, the percent gain in quality of life associated with photodynamic therapy ranged from 7.8% to 14.5% (Table 3). When the utility values associated with the two visual outcomes were varied within their 95% CIs, the relative increase in quality of life associated with photodynamic therapy ranged from 6.0% to 19.7% (Table 4). When both utilities and probabilities were varied simultaneously within their respective 95% CIs, the relative increase in quality of life associated with photodynamic therapy ranged from 4.2% to 25.7% (Table 5). When we performed the Monte Carlo simulation on our model, we observed a gain of 1.358 (standard deviation 0.228) QALY sfrom photodynamic therapy, compared with 1.228 (standard devi':ltion 0.164) QALYs from placebo treatment. This corresponded to a relative increase in quality of life of 10.6%, which was smaller

Table 4-0ne-way sensitivity analysis varying utilities within the 95% Cis

Visual outcome

Point estimate of utility (and range)

Range of QALYs gained from photodynamic therapy

Development of 3-line loss in vision Nondevelopment of 3-line loss in vision

0.56 (0.44-0.66) 0.81 (0.72-0.89)

0.089-0.188 0.083-0.206

336

CAN J OPHTHALMOL-YOL. 36, NO. 6, 2001

Range of % increase in quality of life from photodynamic therapy 7.5-14.8 6.0-19.7

Photodynamic therapy-Sharma et al

Table 5-Two-way sensitivity analysis varying efficacy probabilities and associated utilities within the 95% Cis Visual outcome Development of 3-line loss in vision (utility 0.56 [range 0.44-0.66]) Range of QALYs gained from photodynamic therapy

Range of % increase in quality of life from photodynamic therapy

Range of QALYs gained from photodynamic therapy

Range of % increase in quality of life from photodynamic therapy

in vision at I yr* in vision at 2 yr*

0.063-0.254 0.076-0.223

4.6-25.2 5.5-21 .2

0.067-0.227 0.081-0.198

5.7-17.8 6.8-15.6

in vision at I yr* in vision at 2 yr*

0.059-0.257 0.078-0.221

4.2-25.7 5.6-21.3

0.063-0.229 0.095-0.173

5.2-18.7 8.0-15.6

Treatment Photodynamic therapy Probability of 3-line loss Probability of 3-line loss Placebo Probability of 3-line loss Probability of 3-line loss

Nondevelopment of 3-line loss in vision (utility 0.81 [range 0.72-0.89])

*Taken from the TAP Study.7.8 Efficacy probabilities are varied over the width of 95% Cis calculated by us.

and therefore more conservative than the gain of 11.3% that was expected from our model. This "conservative" gain in QALYs conferred from photodynamic therapy was statistically significant (p < 0.001). INTERPRETATION

In our analysis, based on quality-of-life data obtained from a series of Canadian patients with predominantly classic CNV secondary to AMD, photodynamic therapy conferred a relative increase in quality of life of 11.3% compared .to placebo treatment. Our sensitivity analyses showed that our model was particularly robust, as photodynamic therapy treatment was always preferred over placebo. In addition, a Monte Carlo simulation demonstrated that these results were highly statistically significant. The validity of decision analysis is dependent on the quality of the data on which the model is based. The TAP Study7,8 is a well-designed clinical trial that demonstrated positive treatment results. As such, it can be considered to be level 1 evidence. 14 It should be noted that our models included only a dichotomous state (development or nondevelopment of a IS-letter loss in vision) because no other data were available regarding movement from specific visual acuity groups in the predominantly classic CNV subgroup at the time of manuscript preparation.

We used the time trade-off technique to assess quality of life. This method has been used extensively among patients with retinal disorders. 9- i2 In addition, this method of utility assessment has been proven to be reliable when applied to patients with ophthalmic disease, as its test-retest reliability is very high. is One limitation of our study is that the Canadian utilities were based on only 40 patients, which could have given an unstable result. However, the relative incremental gain in QALY s associated with photodynamic therapy was very similar to that estimated by a decision model in which utilities from a series of 80 US patients with AMD were used.1O, 16 Another limitation is the fact that the utility valuation for potential complications was based on a physician panel. It may be argued that physicians' valuations of the dis utility of complications may be inaccurate. We recently found that physicians tend to underestimate patient utilities.17 Even if this is so, our sensitivity analyses had little effect on our overall result given the small probability of these events. A final limitation is that we used utilities that were derived from a static situation (i.e., a cross-sectional study design) and applied them to a dynamic situation (two time periods). The only way to obtain dynamic utilities is to obtain baseline utilities from a group of patients with AMD and follow them prospectively to determine whether utilities change over time as the

CAN J OPHTHALMOL-VOL. 36, NO. 6, 2001

337

Photodynamic therapy-Sharma et al

patients move from one visual health state to another. We are currently performing such a study. In summary, in this decision analysis using patientbased utilities for a series of Canadian patients, photodynamic therapy was associated with an expected improvement in quality of life of 11.3% compared with placebo. ~ This work was funded by the Canadian Foundation for Innovation, Toronto, the Jeanne Mance Foundation, Hotel Dieu Hospital, Kingston, Ont., the E.A. Baker Foundation for the Prevention of Blindness, Canadian National Institute for the Blind, Toronto, the JP Bickell Foundation, Toronto, and the Retina Research Foundation, Wills Eye Hospital, Philadelphia, Pa. REFERENCES

I. Klein R, Wang Q, Klein BE, Moss SE, Meuer SM. The re-

2.

3.

4.

5.

6.

7.

lationship of age-related maculopathy, cataract, and glaucoma to visual acuity. Invest Ophthalmol Vis Sci 1995; 36:182-91. Ferris FL III, Fine SL, Hyman L. Age-related macular degeneration and blindness due to neovascular macu1opathy. Arch OphthalmoI1984;102:1640-2. Macular Photocoagulation Study Group. Laser photocoagulation of juxtafoveal choroidal neovascularization: fiveyear results of randomized clinical trials. Arch Ophthalmol 1994;112:500-9. Macular Photocoagulation Study Group. Argon laser photocoagulation for neovascular maculopathy. Five-year results from randomized clinical trials. Arch Ophthalmol 1991;109: 1109-14. Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions of age-related macular degeneration. Updated findings from two clinical trials. Arch OphthalmoI1993;111:1200-9. Moisseiev J, Alhalel A, Masuri R, Treister G. The impact of the Macular Photocoagulation Study results on the treatment of exudative age-related macular degeneration. Arch Ophthalmol 1995; 113: 185-9. Treatment of Age-Related Macular Degeneration with Pho-

338

CAN

J OPHTHALMOL-YOL. 36, NO.6, 2001

8.

9.

10.

11.

12.

13. 14. 15.

16.

17.

todynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: one-year results of 2 randomized clinical trials - TAP Report 1. Arch OphthalmoI1999;117:1329-45. Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: two-year results of 2 randomized clinical trials - TAP Report 2. Arch Ophthalmol 2001; 119: 198-207. Sharma S, Brown GC, Brown MM, Shah GK, Snow K, Brown H, et al. Converting visual acuities to utilities. Can J OphthalmoI2000;35:267-72. Brown MM, Brown GC, Sharma S, Garrett S. Evidencebased medicine, utilities, and quality of life. Curr Opin Ophthalmol 1999; 10:221-6. Brown GC, Brown MM, Sharma S. Utility values and agerelated macular degeneration. Arch Ophthalmol 2000; 118: 47-51. Brown MM, Brown GC, Sharma S, Shah G. Utility values and diabetic retinopathy. Am J Ophthalmol 1999;128: 324-30. Data™ 3.5 healthcare user's manual. Williamstown (MA): TreeAge Software; 1999. Sharma S. Levels of evidence and interventional ophthalmology. Can J OphthalmoI1997;32:359-62. Hollands H, Lam M, Pater J, Albiani D, Brown GC, Brown M, et al. Reliability of the time trade-off technique of utility assessment in patients with retinal disease. Can J OphthalmoI2001;36:202-9. Sharma S, Brown GC, Brown MM, Shah GK. The costeffectiveness of photodynamic therapy for fellow eyes with subfoveal choroidal neovascularization secondary to age-related macular degeneration. Ophthalmology (in press). Brown GC, Brown MM, Sharma S. Difference between ophthalmologists' and patients' perceptions of quality of life associated with age-related macular degeneration. Can J OphthalmoI2000;35:127-33.

Key words: photodynamic therapy, verteporfin, Visudyne, health policy, cost-effectiveness, quality of life, age-related macular degeneration