Discussion
RRD is a serious, vision-threatening condition, whose prevalence ranges from 6.3 to 17.9 per 10000 people and can result in blindness unless surgically treated [
10]. At present, SB is one of the most common surgical methods for RRD. The development of postoperative MH following SB for RRD repair is a rare complication [
11,
12]. Additionally, there are no relevant reports on the occurrence this complication in Chinese patients, and current reported cases only involve FTMH [
11,
12]. In this study, we collected clinical data for 6 years from the largest ophthalmic centre in Northwest China. In addition, LMH following SB for RRD repair was first reported.
In our study, the prevalence of MH following SB for treating RRD was 0.83%, which is consistent with available data (0.54–0.86%) [
13,
14]. Although the prevalence is low, the condition still deserves attention owing to its effect on central vision. Nevertheless, the prevalence was higher than those (0.20% [
7] and 0.42% [
15]) in other studies. However, data in those studies were based on patients undergoing RRD repair using treatment protocols other than just SB. The mean time of MH detection following SB was 26 days, with a range between 7 days and 23.5 months as reported in the literature [
12,
13,
16]. In our study, MH occurred in one patient with macula-on RRD and three patients with macula-off RRD, with a prevalence of 0.32% and 1.85% in their respective populations. Additionally, MH was nearly six times more common in patients with macula-off RRD than in those with macula-on RRD but without statistical significance as demonstrated by the chi-squared test. Therefore, more cases should be included. Among the 41 cases reported in eight reports on MH following SB surgery, approximately 85% included patients with macula-off RRD [
7,
11‐
17]. Thus, we should be more vigilant of this complication in patients with RRD involving the macula.
The potential mechanism of MH following SB for RRD repair could include the following: first, SB may increase the axial length, leading to aggravation of VMT; second, SB combined with an encircling band could affect ocular circulation; third, SB does not resolve the inflammation in the vitreous cavity.
In the present study, PVD was observed in Cases 1–3 as well as in other studies; however, VMT is still considered the main cause of MH [
14,
15]. PVD is defined as the separation of the posterior cortical vitreous from the internal limiting membrane [
18]. It usually begins as a shallow, localised separation of the vitreous from the perifoveal retina and progresses slowly until its completion at the time of vitreopapillary separation [
18]. Therefore, Weiss ring lesions have been often used as a criterion for evaluating PVD, but they only represent the detachment between the posterior hyaloid and optic disc and do not indicate the absence of VMT [
12]. If PVD and VMT are to be evaluated objectively, OCT may be the best choice [
19]. However, the status of the detached retina interferes with the reliable evaluation of PVD or VMT in an OCT image (especially macula-off RRD, as shown in Cases 1–3), and there is no published report on the evaluation of PVD using OCT in patients with RRD [
20]. Cases 1–3 had high myopia; in addition, anomalous PVD may have occurred in these patients due to accelerated vitreous liquefaction before adequate weakening of the vitreoretinal adhesion [
21]. Furthermore, preoperative fundus observation (whether with ophthalmoscopy or OCT) would be affected owing to pigment particles and inflammation in the vitreous cavity, which may affect the PVD assessment. As a result, even if Cases 1–3 had PVD, we could not rule out the absence of preoperative VMT, which is a risk factor for postoperative MH. However, another explanation could be that pre-existing PVD damages the inner retina, such as subtle defects or breaks in the internal limiting membrane [
13,
22]. Importantly, the formation of outer LMH observed with OCT in our study confirmed the role of VMT in MH formation following SB, which has not been reported or confirmed using OCT previously [
13]. How to accurately assess preoperative PVD in patients with RRD (especially involving the macula and in those with high myopia) is a challenge for both technologists and surgeons.
Furthermore, whether PVD affects the choice of surgical method in patients with RRD is worthy of discussion. The development of RRD typically involves three factors: one or more full-thickness breaks in the retina, PVD, and passage of fluid from the vitreous cavity through the retinal breaks into the potential subretinal space [
23]. However, RRD without PVD can also develop [
24]. Research has shown that simple SB is suitable for RRD with less liquefied vitreous humour and without PVD, as the formed vitreous can act as a ‘bio-tamponade’ to block the passage of fluid, and traction to the retinal break associated with PVD and vitreous liquefaction may prevent break closure [
20]. However, SB can also be used in patients with RRD and PVD and may need to be combined with other external procedures, such as drainage of the SRF or injection of gas [
20]. There are no unified formal guidelines for selecting the optimal surgical procedure for the repair of RRD [
23]. The surgical treatment of RRD remains a highly individual matter that is influenced by the preoperative findings, patient characteristics, available tools for surgery, and experience and ability of the operating surgeon [
25]. A multicentre, prospective clinical trial showed a benefit of SB in phakic eyes with respect to BCVA improvement compared with PPV [
25]. Therefore, in clinical practice, it is generally believed that SB is an ideal procedure for RRD in young and phakic eyes [
26], such as the patients in this study. It is also important to note that the risk of postoperative MH is a challenge if patients without PVD are selected for SB.
Considering the ocular circulation, retinal ischaemia in the central fovea owing to retinal separation from the choroidal vascular supply plays an important role [
14]. In the present study, the range of retinal detachment in a patient with outer LMH postoperatively did not involve the macula, albeit it was close to the macula, which could still affect macular circulation. Moreover, the pressure from SB and the encircling band could affect the blood supply of the eyeball. Lastly, SB did not disturb the intraocular microenvironment, but it did not resolve the intraocular inflammation, which could exacerbate the macular disorder.
There are other viewpoints that explain the pathogenesis of MH following SB, such as epiretinal membrane (ERM) and cystoid macular oedema (CME) [
12]. In our study, we did not observe ERM or CME in any of the cases. The prevalence of ERM is positively correlated with age [
27]. Half of the participants in the study by Garcia et al. [
12] had ERM, and the average age of the patients was 54 years, whereas that of patients in the present study was 43.5 years. Considering CME, we reviewed the original literature demonstrating that CME usually appeared after PPV or in patients with diabetic retinopathy. In the present study, except for Cases 1–3 with high myopia, there were no CME-related ocular factors as mentioned above. Furthermore, discussions on the relationship between high myopia and postoperative MH are scant. In the previous eight studies, only 10% of the patients had high myopia, and some studies did not mention the refractive status of the patients [
7,
11‐
17]. However, 75% of the patients in our study had high myopia and postoperative FTMH. Nearly all previous reports were from Europe and America; however, myopia is a common disease in East and Southeast Asia [
28]. An unhealthy vitreomacular interface, thin retina and choroid, and poor tolerance to decreased circulation make patients with high myopia more susceptible to MH. Thus, we should pay more attention to these patients.
All three patients with RD in this study had high myopia. The most commonly used measurement method, such as optical biometry or A-scan ultrasound, may not provide accurate axial length measurements in patients with retinal detachment as the patients’ fixation is hampered by the detached retina or macula [
29]. In this study, the B-scan mode of Quantel was used to measure the axial length. Ultrasonic measurement of axial length in patients with retinal detachment may be fallacious; thus, A-scan was introduced to the module to combine A- and B-scan, which was similar to that employed in Mohsen’s study. Moreover, this study showed that A- combined with B-scan is a good choice for patients with retinal detachment [
29]. Nevertheless, accurate measurement of the axial length in these patients is challenging.
In our study, all patients with FTMH underwent reoperation owing to MHRD, and the FTMH closed after the operation, thus enabling retinal reattachment. Cases of FTMH that developed after SB and closed spontaneously without surgical treatment have also been reported [
11]. However, in previous studies, cases of FTMH closure without reoperation accounted for a minority of the total observed cases. Additionally, the closure rate of MH also varied among studies [
7,
15]. In the present study, postoperative LMH following SB was reported for the first time, and we were able to conduct a close follow up without surgical treatment owing to the patient’s compliance. Therefore, based on different situations, reasonable treatment judgments should be made.
Although our study summarised clinical data of 6 years from the largest ophthalmic centre in Northwest China, the present study still has certain limitations. First, it was a retrospective study with a small sample size. Second, preoperative OCT was performed in the patients; however, complex and severe RD may affect the OCT results, which is a limitation of this study. Therefore, adequate fundus examination should be performed before surgery to minimise the omission of MH. Moreover, the prevalence of MH following SB could have been underestimated since not all 483 patients underwent regular OCT after surgery in our study. Additionally, some patients were not followed up regularly, possibly owing to the impact of the coronavirus disease pandemic. Therefore, larger studies, preferably from multiple centres, with more data concerning MH following SB for the treatment of RRD, as well as deeper research and greater effort, are warranted.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.