Results of surgery for late sequelae of cicatricial retinopathy of prematurity.

Purpose: To report anatomical and functional results of vitreoretinal surgery in our case series of late cicatricial retinopathy of prematurity (ROP) patients with subtotal retinal detachment.
Methods: This is a retrospective, consecutive case series. Eleven eyes of 10 patients presented with partial tractional retinal detachment secondary to late cicatricial ROP (cicatricial stage 4B) who underwent vitreoretinal surgery were retrospectively reviewed. Anatomical and functional outcomes were evaluated.
Results: The mean gestational age at birth was 28.6 (26-32) weeks. The mean age at surgery was 79 (4-213) months. Patients were followed up for 21.7 (6-40) months. Six eyes (55%) had lens-sparing vitrectomy and five eyes (45%) had lensectomy + vitrectomy. Anatomical success was achieved in 10 eyes (91%). Improvement in visual acuity was noted in nine eyes (82%).
Conclusion: Eye grows but fibrotic tissue does not grow with age, and during this period retinal traction may get worse. Relieving these tractions may lead to good anatomical and visual outcomes in selected late cicatricial ROP cases.

Retinopathy of prematurity (ROP) is a vascular disorder of the retina occurring in newborn preterm infants. With the increase in the survival of preterm babies, ROP has become the leading cause of preventable childhood blindness throughout the world.[1]

The most common outcome of ROP is spontaneous regression by involution and evolution from vasoproliferative phase to fibrotic phase. Involutional sequelae range from innocuous to extremely severe; the more advanced or the more posterior the active disease, the worse the cicatricial sequelae.[2] Such eyes are usually left untreated in this cicatricial stage because the tractional changes are very difficult to relieve in late cases and also because of a probable amblyopia development in later ages. The belief is that if the retina was detached since neonatal period and stayed like that for a long time, it is impossible to get a useful vision even if anatomical success could be achieved. This belief together with risk of losing the present limited vision in a late cicatricial ROP cause surgeons to avoid performing surgery in such eyes. However, a subgroup of cicatricial ROP patients who have some vision may get real benefit from surgery. On the other hand, the risk benefit ratio has to be very well assessed for each case and careful patient selection is very important before offering surgical option.

The purpose of this study is to report anatomical and functional results of vitreoretinal surgery in our case series of late cicatricial ROP (cicatricial stage 4B) patients with subtotal retinal detachment (RD).

Methods

This retrospective case series study was approved by the ethics committee of our institution (name of the institution is masked) and adhered to the tenets of the Declaration of Helsinki.

We retrospectively reviewed the charts of patients with partial tractional retinal detachment (TRD) secondary to late cicatricial ROP (cicatricial stage 4B) who underwent vitreoretinal surgery by one single surgeon between September 2010 and December 2016 after appropriate informed consent. Eyes with cicatricial findings consistent with late sequela of regressed ROP like posterior retinal fold, subtotal TRD, as described by Committee for the Classification of the ROP,[2] without prior treatment for active ROP, were included in this study. Cases who had any kind of previous treatment for ROP and those with total RD (stage 5) were not involved in the study.

Data involving patient gender, birth weight, gestational age, age at surgery, and follow-up duration were collected from records. Any previous treatment for ROP, best-corrected visual acuity (initial and final), detailed ocular findings at presentation, surgery undertaken, complications, and postoperative anatomical and functional results were noted. Anatomical success was defined as complete or partial success. Complete success is total retinal reattachment, and partial success is partial flattening of retinal folds and relocation of the fovea. Functional success was defined as an increase in visual acuity.

All surgeries were performed under general anesthesia, and before each surgery a fundus examination was done with a binocular indirect ophthalmoscope to determine the places for sclerotomies to avoid retinal damage during sclerotomies. The sclerotomies were performed at different distances from the limbus according to the age of the patients as described by Aiello et al. Conjunctival peritomy was performed at the beginning and sclerotomies were sutured at the end of the surgery in all the cases.[3] A three-port vitrectomy was performed in all the cases and sclerotomies were 23 or 25G with or without (valved) cannula. Transiris root entry or limbal entry was preferred if the lens was planned to be removed in those cases with severe peripheral anterior proliferations extending toward the lens. Following central core vitrectomy, an attempt was done to detach the posterior hyaloid from the posterior pole with the help of diluted triamcinolone without being insistent on it. Even if the posterior hyaloid could be detached easily from the posterior pole, a limited detachment no more than just anterior to the arcade vessels was performed. Peripheral hyaloid detachment was not attempted and left in place to avoid retinal break and hemorrhage which may be caused by vascular tractions in the fibrovascular proliferation areas. If the posterior hyaloid was tightly attached, cortical vitreous was trimmed more and the posterior hyaloid was left attached to avoid risk of iatrogenic break formation during posterior vitreous detachment induction. Fibrovascular membranes causing retinal tractions were removed or trimmed as much as possible to relieve the retina. At the end of the surgery, only air was used as a tamponade which aids in watertight suturing of sclerotomies and keeps the retina away from the lens and iris (in lensectomized eyes) preventing readhesion of the retina to the anterior structures which may be the case especially following postoperative hemorrhage. All the patients were examined on postoperative day 1, 2 weeks later, and monthly thereafter.

Visual acuity was assessed with Snellen equivalent methods. Those too young for this method of acuity testing had fixation patterns assessed and recorded, and then projected to visual acuity measurements using a previously described technique modified from Zipf as indicated in Table 1.[4]

Table 1

Fixation patterns projected to visual acuity measurements

Fixation pattern	Visual acuity
CSM	>20/30
CSUM	20/30-20/100
CUSUM	<20/300
UCUSUM	<5/200

C: Central; S: Steady; M: Maintained; UC: Uncentral; US: Unsteady; UM: Unmaintained

Results

Eleven eyes of 10 patients (3 males and 7 females) underwent surgery; 1 patient had bilateral surgeries (patient 4). Demographics, baseline characteristics, treatment, and outcomes of patients who had vitreoretinal surgery for cicatricial ROP are displayed in Table 2. The mean gestational age at birth was 28.6 ± 1.6 (range 26–32) weeks. The mean birth weight was 1204 ± 308 (range 850–1800) g. All the eyes were stage 4B cases.

Table 2

Demographics, baseline characteristics, treatment, and outcomes of patients who had vitreoretinal surgery for cicatricial ROP

Patient/gender/eye	Gestational age (weeks)/birth weight (g)	Age at surgery (months)	Initial vision	Surgery	Complication	Anatomicaloutcome	Macular ectopia preop/postop	Retinal fold preop/postop	Final vision
1/M/L	28/1060	6	CUSUM <20/300	PPV+PPL	-	No success (no improvement in macular fold)	+/+	+/+	CUSUM <20/300
2/F/L	30/1370	10	LP	PPV	-	Partial success	+/−	−/−	CUSUM <20/300
3/F/L	28/1500	11	LP	PPV+PPL	-	Partial success	+/+	+/−	20/400
4/M/R	28/850	79	20/100	PPV	-	Complete success	+/+	+/−	20/30
4/M/L	28/850	81	20/100	PPV	-	Complete success	+/+	+/−	20/60
5/F/R	28/1800	4	LP	PPV	-	Complete success	+/+	+/−	20/400
6/F/R	30/1290	104	CF at 2 m	PPV	-	Partial success	+/+	+/−	20/400
7/F/R	28/900	213	CF at 1 m	PPV+PPL	-	Complete success	+/+	+/−	20/400
8/M/R	29/950	54	LP	PPV+PPL	-	Complete success	+/−	+/−	CUSUM <20/300
9/F/L	26/1270	156	20/200	PPV	-	Complete success	+/−	−/−	20/200
10/F/R	32/1400	151	LP	PPV+PPL	-	Complete success	+/+	+/+	CF at 2 m

LP: Light perception; CF: Counting fingers; RD: Retinal detachment; PPV: Pars plana vitrectomy; PPL: Pars plana lensectomy

The mean age at surgery was 79 ± 71 (range 4–213) months. Patients were followed up for 21.7 ± 11 (range 6–40) months.

On initial examination, 10 eyes (91%) had partial TRD involving macula (stage 4b), and 1 eye had total leukocoria simulating stage 5 ROP but was found to have partial RD with minimal attached retina after dissection of fibrotic membranes [Fig. 1]. All of them had macular detachment and ectopia, and nine eyes (82%) had retinal fold.

Figure 1

Case 8: anterior segment examination of the RE showed leukocoria (a). B-scan US revealed vitreous membranes causing anteroposterior tractional RD (b). Peroperatory pictures after partial dissection of the membranes show that there was TRD associated with retinal folds and some limited areas of attached retina (stage 4b) (c and d). Postoperative anterior segment and fundus pictures. Note that leukocoria has disappeared (e) and retinal folds decrease in size with larger areas of attached retina (f)

Visual acuities at initial and final examination are shown in Table 2. The initial visual acuity in the affected eyes was light perception in five eyes (46%), counting fingers in two eyes (18%), and between 20/400 and 20/100 in four eyes (36%).

Six eyes (55%) had lens-sparing vitrectomy and five eyes (45%) had a combined lensectomy and vitrectomy. There was no intraoperative complication. None of the eyes ended up with phthisis or enucleation.

Anatomically, 10 eyes (91%) ended up with successful results: 7 eyes (64%) with complete retinal reattachment and 3 eyes (27%) with partial improvement. In one eye (9%), surgery did not lead to any improvement in preexisting RD, retinal fold, and macular ectopia after surgery.

Visual acuities were noted to improve in nine eyes (82%) and remained the same in two eyes (18%) at the final visit. The median postop visual acuity was 20/400 at the final follow-up. Final visual acuities were between 20/400 and 20/200 in eight (73%) eyes. Patient no. 4 was exceptional because he achieved 20/30 in the right eye and 20/60 in the left eye following surgery, which is described in detail in Fig. 2.

Figure 2

Case 4: fundus examination revealed TRD caused by fibrotic membranes extending from optic nerve head to the back of the nasal part of the lens anteriorly and nasal periphery resulting in inferonasal foveal ectopia and detachment (a and b). PPV was performed on both eyes and macula was attached (although displaced inferonasaly) (c and d) with a very well-formed foveal contour in OCT (e and f)

Discussion

The incidence of ROP has increased in the developing world.[1] ROP spontaneously regress with no visual impairment in mild cases, but it can cause severe visual impairment in advanced cases.[5] After acute phase of the disease, treated or untreated, a majority of the cases go into regressed stages. Committee for the Classification of the ROP described the features of regressed ROP which are vascular changes, pigmentary changes in the retina, peripheral folds, vitreous membranes, lattice-like degeneration, retinal breaks, and tractional/rhegmatogenous RD (RRD).[6]

Machemer reported that the retinal and vitreous changes seen in ROP can be explained by proliferation and contraction of tissue originating in the shunt area, and the detachments of ROP are, typically, TRDs.[7] He reported five cases of TRD occurring later in life as a sequela of cicatricial ROP or showing the clinical picture of ROP.[8] They presented with taut membranes in vitreous cavity, causing TRD, and often showed preretinal membranes that were collagen-rich and contained cells with glial characteristics.[8]

Park et al. have reported 16 eyes of 15 patients that required surgical treatments due to RD of regressed ROP; after treatment of acute phase of ROP, 8 eyes showed grade II of cicatricial changes and 8 eyes showed grade III. All eight eyes with grade III of cicatricial ROP showed TRD and five of eight eyes (62.5%) with grade II showed RRD.[9] Their series has some similarities with our series in that they are all late cicatricial cases occurring after the regression of initial active disease, 10 of them being TRD and 5 of them RRD. Their series included both previously treated and untreated cases. The anatomical and functional outcomes of surgery were relatively inferior to those of an adult-onset RD as a late complication of ROP probably because of delayed detection of the late RD in preverbal children.

In our study, eyes with any previous treatment for ROP were excluded and only the eyes where the active disease had cooled down itself leaving fibrotic tissues behind causing retinal tractions were included. So, they were all late stage 4B cases. Anatomical success rate was 91% (seven eyes with complete success and three eyes with partial success) in the presented series. It is well known that fibrotic tissue is not elastic and becomes even tighter with time, and during the growth of the eye, that fibrotic tissue usually causes gradual increase in traction causing RD within months to years. Posterior hyaloid contraction is an additional problem within time. Which means that macula may be “on” during the early years of life and may be involved in detachment later in a cicatricial ROP case. Case 4 is a very striking example for this. Although the case was 6 years old during surgery, vision has improved to a surprisingly high level after surgery. This is contrary to the belief that late surgery for cicatricial ROP may be useless especially because of amblyopia problem. However, these eyes may not be amblyopic since the vision was probably better during early years of life in these babies.

In Parks et al.'s series of regressed ROP-related TRD, surgical treatment ended up with an anatomical success rate of 54.5%. Vision was reported to improve in three eyes, did not change in six eyes, and decreased in two eyes in this series postoperatively, and only 12% of the eyes with late RD achieved a visual acuity of 20/200 or better.[9] In our study, the rate of eyes with improvement in visual acuity (82%) and the rate of eyes with a final visual acuity of 20/200 or better (27%) were higher in the presented series. This may be explained by patient selection and the improvement in the vitreoretinal surgical equipment facilities within more than 10 years.

The most important complications of vitrectomy for ROP are intraoperative bleeding, iatrogenic retinal breaks, retinal redetachment, proliferative vitreoretinopathy, postoperative hypotony, endophthalmitis, formation of cataract, and phthisis.[10] Jandeck et al. reported the indications and results of pars plana vitrectomy for late RD in patients born prematurely.[11] They have reported retinal redetachment in 3 of 11 eyes, 2 of which were secondary to iatrogenic breaks and proliferative vitreoretinopathy and 1 of them developed phthisis. In our study, none of the eyes developed such complications. We believe that patient selection is again the main determinant to avoid such complications in these cases. The eyes with membranes causing a clearly visible anteroposterior traction which look relievable without taking too much risk of retinal break are the best candidates for this surgery. These are the eyes with increased anteroposterior traction during growth of the eye with the increase in anteroposterior length of the globe within years. However, those membranes causing tangential traction over the retina which looks almost totally adherent to the retina along its all length are not good candidates for this surgery. We think that even a late cicatricial ROP case with leukocoric RD which has some vision (LP or hand motions) in the better eye could be operated to get useful vision (Case 8). On the other hand, this cannot be generalized to all cicatricial ROP cases and every case should be assessed individually.

This study is limited by small sample size, and hence our results cannot be generalized. We want to emphasize that to get the best results, patient selection is very important.

Conclusion

In conclusion, eye grows but fibrotic tissue does not grow with age, and during this period retinal traction may worsen. Although vitreoretinal surgery for cicatricial ROP is controversial, by relieving these tractions, it may be possible to get satisfactory anatomical and functional outcomes in selected cicatricial ROP cases.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.