Predicting speed of progression of lens opacification after pars plana vitrectomy with silicone oil.

PURPOSE: An increasing number of posterior segment disorders is routinely managed with pars plana vitrectomy (PPV). In older, phakic patients cataract formation is expected within the first two years after surgery. For younger patients its progression is individually fluctuating. This study uses an objective quantitative measurement for lens-status-monitoring after PPV with silicone oil to derive predictions for progression and severity of post-operative lens opacification evaluated in patients with rhegmatogenous retinal detachment (RRD).
METHODS: Data acquisition was performed prospectively between March 2018 and March 2021. PentacamHR® Nucleus Staging mode (PNS) was used to objectively gather data about nuclear cataracts after PPV at different time points. Data was grouped into training and test sets for a mathematical prediction model. Via backward variable selection method a mathematical formula was set up by means of which predictions about lens densitometry (LD) can be calculated.
RESULTS: 20 males [58.8%] and 14 females [41.2%] matched the inclusion criteria (mean age 50.6 years [23-75; ±12.3]). Average follow-up was 8.1 months (3,4-17.4; ±3.4). Mean baseline LD of the treated and fellow eye before surgery was 11.1% (7.7%-17.6%; ±2.0) and 11.2% (7.7%-14.8%; ±1.5), respectively. Predicted LD values by the model for five pre-selected patients closely match the observed data with an average deviation of 1.06%.
CONCLUSIONS: Using an objective parameter like LD delivered by the PentacamHR® PNS mode additionally to the patient's age allows us to make an individual prediction for any time after PPV with silicone oil due to RRD for all ages. The accuracy of the model was stronger influenced by baseline LD as cofactor in the equation than patient's age. The application for the prediction lens opacification [which can be accessed for free under the following link (https://statisticarium.com/apps/sample-apps/LensDensityOil/)] can help vitreoretinal surgeons for patient consultation on the possibility to combine PPV with cataract surgery.

Introduction

An increasing number of posterior segment disorders is successfully managed with pars plana vitrectomy (PPV). As a result vitrectomy related cataractogenesis is a common surgical side effect, which is expected within the first two years after surgery and it is more severe in elderly patients [1, 2]. For patients over 50 years of age it has already been shown several times that significant lens opacification mostly occurs within the first 2 years after PPV [3-5].

However, this is not the case for all patients receiving PPV. The exact biological mechanism that leads to the progression of the lens opacity is still vividly discussed and has not yet been definitely clarified. Postoperative elevated levels of oxygen and free radicals seem to play an important role [6-9].

It has already been shown for C3F8-gas that lens opacity after a PPV can be predicted for every point in time after surgery with sufficient accuracy by using lens densitometry with a mathematical model [10]. PPV nowadays is more and more often combined with cataract surgery in patients over 50 years of age, but it may also be considered in younger patients. However, removing the lens during the vitrectomy procedure also has disadvantages such as an additional risk of surgical complications related to phacoemulsification, increased postoperative inflammation, less predictable postoperative outcome than in sequential surgery and an increased risk of silicone oil spill-over in the anterior chamber. Thus, it may be desirable for advising patients and surgical planning to be able to predict on an individual patient basis how soon (if at all) cataract surgery after vitrectomy with silicone oil will become necessary.

In this study we evaluated patients of various age and lens densitometry values that were treated for rhegmatogenous retinal detachment (RRD) via the aforementioned regimen. To receive a quantitative analysis of lens transparency the PentacamHR® Nucleus Staging (PNS) module of the Scheimpflug tomography system (Oculus, Wetzlar, Germany) was used.

Methods

In our prospective Study patients with RRD treated with 23G PPV using silicone oil as an endotamponade from March 2018 to March 2020 were included. The study followed the tenets of the Declaration of Helsinki and was approved by the Medical Institutional Review Board of Hamburg (PV7250). Patient’s data were gathered fully anonymized. Written consent was waived for this study, but this was obtained verbally.

All patients underwent a complete ophthalmological examination with BCVA, intraocular pressure measurement, slit lamp biomicroscopy, dilated funduscopy, and most importantly PentacamHR® lens densitometry prior to surgery and at every visit during follow-up as it is standard procedure in our clinic. Post-operative follow-up visits were scheduled at about six weeks, three months and six months. If additional visits were necessary due to disease related further clinical controls, these measurements were also used for the analysis. Furthermore, there should not have been any complications during PPV that could accelerate lens opacification.

In addition, only patients were included for whom the surgeon had previously decided not to perform a combined phakovitrectomy.

PentacamHR® measurements were taken after pupil dilation and mean lens densitometry (LD) value was calculated by the PNS module in predefined three-dimensional volumes centered on the apex. The measurement and calculation is performed automatically within less than 3 seconds, provided the PNS module is installed on the device. All LD measurements have been performed by trained medical-laboratory assistants under standard dim-light conditions. The advantage of this system is that the parameters are collected from a three-dimensional portion of the lens nucleus and thus LD is calculated from a volume and not just from a linear section. This way it reflects a better overall approximation of the actual lens opacification. The mean lens densitometry is output as a percentage of total backward scatter of light.

Only PentacamHR® images without reflections or distortions were used for the evaluation. This method has been successfully established by prior various studies [10-16].

Patients with prior cataract surgery on one eye, PPV, or any other intraocular procedure were excluded. Patients with history of ocular trauma, uveitis, topical or systemic corticosteroid therapy, and signs of visually impairing cataract at baseline evaluation were also excluded from the analysis. Additionally, known topical or systemic conditions (diabetes excluded) that could accelerate cataract formation and/or progression after PPV led to exclusion from the study. Patients with ischemic and/or proliferative retinopathy were also excluded [17-19].

The need for cataract surgery during follow-up was marked as the endpoint for this case. LD measurements of fellow eyes were taken to provide reference values for LD progress and to show their natural course. The surgery was performed by 3 vitreoretinal surgeons with at least 3 years of post-fellowship experience in the field of retinal detachment. For details about the surgical technique we refer to the supplemental content.

We selected the following variables to co-evaluate with post-surgical LD changes:

time after surgery (months)

baseline LD (%)

age at the time of surgery (years)

The sample was randomly divided into training and test data sets. Backward variable selection method based on Akaike information criterion was applied to develop a mixed regression model for prediction of LD. The observed LDs from the test data set were compared with the predicted LDs.

Results

Thirty-four eyes of 34 patients matched the inclusion criteria. Demographic and presurgical parameter are summarized in Table 1.

Table 1

Demographics and preoperative parameter.

Patients	Total n = 34
Right eye	n = 17 (50.0%)
Left eye	n = 17 (50.0%)
Female	n = 14 (41.2%)
Male	n = 20 (58.8%)
Age (years)	23–75; 50.6 (±12.3)
BCVA LogMAR [Snellen]	0.0 [20/20]– 2.4 [hand motion]; 1.3 (±0.9)
Lens Densitometry baseline treated eye (%)	7.7–17.6; 11.1 (±2.0)
Lens Densitometry baseline fellow eye (%)	7.7–14.8; 11.2 (±1.5)
Diabetes Type I	n = 0 (0.0%)
Diabetes Type II	n = 3 (8.8%)
Follow-up time (months)	3.4–17.4; 8.1 (±3.4)

Qualitative parameters are presented as percentages, and quantitative ones as min to max and mean (±Standard deviation).

Datasets of 20 males [58.8%] and 14 females [41.2%] were investigated. Seventeen patients each underwent surgery in the right or left eye. Patient’s age ranged from 23 years to 75 years with an average of 50.6 (±12.3) years. Mean baseline LD of the treated and fellow eye were 11.1% (7.7%-17.6%; ±2.0) and 11.2% (7.7%-14.8%; ±1.5), respectively.

Average follow-up was 8.1 months (3.4–17.4; ±3.4). Three patients suffered from diabetes type II, but none of these patients had signs of diabetic retinopathy during the period of investigation.

Statistics and prediction model

A backward variable selection method was applied starting with a saturated model with all interaction effects between time, age, baseline lens densitometry. The resulting fixed effect part of the model is:

LogitDensitometry = β0

β1 × Months +

β2 × Age +

β3 × Baseline LD +

β4 × Months × Age +

β5 × Months × Baseline LD +

β6 × Age × Baseline LD +

β7 × Months × Age × Baseline LD

The supplement document of this paper gives further information on statistical calculations, prediction of LD, surgical method and results of the fellow eyes.

The model delivers a set of coefficients that describe the interaction between dependent factors, for example “months” after surgery or “baseline LD”, and the prediction. The factors have been selected according to their significant influence on lens opacification in the statistical analysis.

There is a significant effect of time (months after surgery). But the effect of time enters the model in four ways: as a linear effect, an interaction with age, an interaction with baseline LD and higher order interaction with age and baseline LD. That explains why “months” appears several times and in different ways in our equation further below. Note that quadratic term of time was not significant, because there is no strong tendency to decreasing rate of change in the analysed period.

Furthermore, baseline LD is positively related to postoperative LD, meaning that all being equal the higher baseline LD is, the higher the predicted trajectories will be.

The effect of age is also significant, but not as much as baseline LD. Therefore, baseline LD is a higher influential cofactor than age. In the model we consider age as continuous variable.

All these findings were considered when calculating the coefficients β0 through β7. In this way we received the results summarized in Table 2.

Table 2

Mixed effects regression model parameters β0 through β7.

	Dependent variable:LD1001−LD100
β1 Months	-0.567*** (0.143)
β2 Age	0.018 (0.025)
β3 Baseline lens densitometry	0.164 (0.133)
β4 Months x Age	0.013*** (0.003)
β5 Months x Baseline lens densitometry	0.055*** (0.016)
β6 Age x Baseline lens densitometry	-0.002 (0.003)
β7 Months x Age x Baseline lens densitometry	-0.001*** (0.0003)
β0 Constant	-3.760*** (1.277)
Observations	104
Log Likelihood	52.242
Akaike Inf. Crit.	-84.483
Bayesian Inf. Crit.	-58.039

P-Values are given symbolically with asterisks:

*p<0.1;

**p<0.05;

***p<0.01. In parenthesis are standard errors of the coefficients.

Note that:

β4 is positive → the older the patient the steeper (stronger) is the change.

β5 is positive → the higher the baseline LD the steeper is the change.

We see that time does not significantly enter the model. Both age and baseline LD seem to be significantly related to the LD. Since time is not related to the trajectories the prediction would be a horizontal line that only goes up or down depending on baseline LD and age. The model works by inserting values for each dependent variable into the following equation:

Logit(LD) =  −3.760

−0.567 × Months +

0.018 × Age +

0.164 × Baseline LD +

0.013 × Months × Age +

0.055 × Months × Baseline LD +

−0.002 × Age × Baseline LD +

−0.001× Months × Age × Baseline LD

Where,

To get the final prediction we have to back-transform from the logit scale as follows:

By filling the equation with individual values for “months” after surgery, “baseline LD” and “Age” of the Patient one can calculate and therefor predict an estimated LD for any time after surgery for any patient facing a PPV with silicone due to RRD.

Fig 1 shows the observed post-surgical LD trajectories of just 29 cases, since 5 cases have been excluded from the model creation process to test its power afterwards.

Fig 1

Changes in trajectories of lens densitometry.

Trajectories of lens densitometry (%) during months after surgery for treated eyes.

Our statistical analysis has shown that age of the patient and baseline LD measurement are positively related, meaning that the older a patient is, the higher is his or her baseline LD. When looking at Fig 1 there appears to be two groups of patients. One group does not increase in LD over a period of 10 months. This group has in common that its baseline LD is below 10%. The range of age in this group is 23–44 years. Above 10% baseline LD all cases experience a rapid increase in LD within the investigated period. One of the younger patients with initially low baseline LD developed a steep increase in LD only after 10 months, although there was no notable increase in LD up to 10 months after surgery.

Model performance

Now we used these five patients that have been excluded in advance for validation purpose.

In Fig 2 blue dots show predicted LD values from the model. Red dots depict actual LD values from the patient’s PentacamHR® file, respectively.

Fig 2

Predicted and observed lens densitometries over time.

Predicted (blue dots) and observed (red dots) postoperative lens densitometries (%) over time (months after surgery).

Apart from ID = 5, blue and red dots are very close together. The prediction ends when there were no more recordings in the patient’s file, but predictions can be done for any time after surgery for each individual. To prove that the prediction is working even for a period after 10 months one must look at ID = 5 again. His or her last LD measurement was made after 12 months and matches the predicted LD perfectly. This patient is 27 years of age and had a baseline LD of 8,3%.

Web application for predicting postoperative lens densitometries

As we did for PPV with C3F8-gas previously [10], we provide a web application that calculates individual postoperative trajectories for LD based on the model shown above. Only “age" in years and “baseline LD” must be provided by the user and the application displays the results as a line plot. The application can be accessed free under the following link: https://statisticarium.com/apps/sample-apps/LensDensityOil/.

Discussion

Cataract formation is the most frequent complication after PPV in phakic eyes. But precise estimations when and to what degree it will occur are only vaguely possible so far. The cause for lens opacification may relate to increased oxygen levels and the ensuing oxidative stress, as well as the altered biochemistry in the vitreous cavity following vitreous removal [6, 7, 20].

Previous studies underlined that oxidative stress to the lens, for example by hyperbaric oxygen therapy or removal of the vitreous body, increases the exposure to molecular oxygen and can cause lens opacification [21-24].

Among other factors, age plays a key-role in causing not only age-related cataract, but also postsurgical lens opacification. For patients >50 years of age combined PPV is performed more frequently, but it is sometimes needed in younger patients, too. Kataria et al. and Thompson et al. suggest that NSC progresses at a rate 6- to 9-fold higher in patients >50 years of age compared with patients <50 years of age. This finding suggests that cataract surgery should not be performed routinely combined with PPV in patients less than 50 years of age because cataract surgery is usually not needed for many years [19, 23, 25–28].

Whether combined cataract and PPV surgery should be generally performed for patients over 50 years of age has not yet been clearly recommended.

The rate of lens opacification is simply too difficult to estimate for individual cases without objective parameters. Especially for the age group between 41–55 years there is no reliable evidence about how long it takes until a significant lens opacification occurs after PPV. Individual fluctuations are too wide in this regard. Using silicone oil as an endotamponade may lead to significantly earlier cataract formation after PPV than air or gas. This finding is supported when looking at the data of the average LD 6 months after surgery from our previous study on the progression of lens opacity after PPV with C3F8 gas. With silicone oil mean LD of the cohort increases by approximately 5.5%, whereas under C3F8 gas {(n = 34; age (years) 32–77; 58.5 (±4.3); baseline LD treated eye (%) 8.7–14.8; 10.9 (±0.8)} there was only an increase of about 2.7% within 6 months after PPV [10].

Moreover, posterior subcapsular cataract (PSC) is more likely to develop in silicone oil filled eyes [29].

But based on the endotamponade used alone, it cannot be concluded when exactly vision impairing cataract will occur. Other factors, such as the surgeon’s experience, sometimes play a role as well. For instance, Xu et al. were able to show that the time between PPV and cataract surgery was significantly longer for surgeons with more experience than for inexperienced surgeons [30].

As has already been shown in other studies, lens opacification progressed relatively fast after surgery in most of our cases. But LD trajectories in younger patients were flat and lens opacification progressed slower compared to older patients (see Figs 1 and 3) for at least 10–12 months following PPV. The fact that baseline LD of older people is naturally higher was considered when developing our model. However, the increase in LD with age did not always follow a linear fashion. Patients with higher age may have a relatively low baseline LD and younger patients may have an increased LD. Thus, baseline LD might be a better indicator than age for the decision whether a combined phakovitrectomy should be performed or not.

Fig 3

Observed trajectories for patients under 51 years of age (yoa).

Changes in lens densitometry (%) after pars plana vitrectomy over time (months) for individual cases with younger age.

This assumption becomes clear when looking at the trajectories of patients who are under 51 years of age (see Fig 3). In the case of the 44-year-old patient with a baseline LD of 10.1%, LD does not increase significantly over time. In contrast, the 39-year-old patient with a baseline LD of 11,0% shows a significant increase in LD after just a few months. These two examples show that age alone is not a reliable parameter for estimating possible lens opacification after PPV. So, taking baseline LD into account and giving it a higher weight in the model than age, makes the difference when predicting the speed of progression of lens opacification after PPV.

When looking at Fig 3 there appear to be 2 groups in this study. One group which shows little or no increase in LD over a longer period of time and a second group which shows a significant increase shortly after PPV. Most of the patients in the group with a rapid increase in LD are over 51 years of age, but not all of them. On the other hand, one can see that the group with a slow increase in LD also includes patients who are at least over 40 years old. Individual fluctuations that cannot be attributed solely to the patient’s age also seem to be present in this study. It remains unclear why some lenses opacify faster than others. However, younger lenses seem to be more stress-resistant to external influences such as intraocular operations or increased oxygen levels.

Various lens opacity grading systems exist for grading cataracts. The Lens Opacity Classification Systems II and III (LOCS II and III) use color lens photograph standards or slit-lamp photographs of the lens to quantify its opacity [31]. However, the LOCS grading systems are rather subjective and depend on human raters.

Using a rotating Scheimpflug camera PentacamHR® provides precise, three-dimensional images of the lens. It delivers reproducible data related to quantify LD and is an objective method to assess the lens transparency [12–14, 32–35].

Limitations are that the PentacamHR® PNS module only depicts nuclear and no cortical or PSC changes. PSC occurrence is described frequently under silicone oil and leads to considerable visual impairment quickly [36].

However, since PSC is transient in most cases and is usually followed by NSC after PPV with silicone oil we do not consider this weakness as a negative influence on the results of our study [37].

The cohort is also very homogeneous, as patients with previous intraocular surgeries or lesions were excluded, which limits the transferability of the results to a larger population.

Using age and baseline LD as individual factors to predict LD for any time after surgery showed precise results (Fig 2). Presumably, the accuracy of the prediction model will improve when more data is used to train the model.

The number of cases in our study is still relatively small. Therefore, the decision for or against a combined phakovitrectomy cannot be completely based on the mathematical predictions.

In general, it is always the decision of the surgeon whether and when a cataract operation is performed. This decision is made on the basis of various clinical and epidemiological criteria. However, many of these criteria are subject to subjective perception. Additional objective parameters such as LD measurements might probably simplify and complement this decision in some cases. On the other hand, this mathematical model can be used to make predictions that also help the patient to understand why a combined phakovitrectomy would make sense or not.

Conclusions

PPV will sooner or later lead to vision impairing cataract formation. Individual estimates of when and to what degree it will occur are difficult. Using objective parameters like LD values and patient’s age allowed us to make precise predictions about the speed of lens opacification in silicone oil filled eyes after PPV for RRD. The accuracy of the model was especially obtained by giving baseline LD a higher weight in the equation than patient’s age.

We developed a statistical model that can be accessed for free under the following link (https://statisticarium.com/apps/sample-apps/LensDensityOil/) to help vitreoretinal surgeons for consultation on the possibility to combine PPV with cataract surgery.

(DOCX)

Click here for additional data file.

19 Jan 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Xingjun Fan, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

Additional Editor Comments:

Both reviewers showed some favor of this study but at the same time raised significant concerns that require the authors to address accordingly. Notably, the sample size and exclusion criteria need to be carefully addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This is a good paper that adds clinical tools to the body of knowledge. Things that I noticed that are missing include how many different surgeons with listed experience level conducted these surgeries (you mention this to be an influential factor in the discussion) and intraoperative details (laser and amount used, time of surgery, fluid used, etc). A nice next step would be nice to apply this to other external data sets to see if the prediction model holds up.

Reviewer #2: The authors present a prospective study describing the predicting abilities of a regression model using Lens densitometry (LD) measurements of the PentacamHR® Nucleus Staging mode (PNS) and other parameters to estimate postoperative cataract formation after Pars Plana Vitrectomy (PPV) performed due to Rhegmatogenous Retinal Detachment (RRD).

The authors reported accurate predictions, with accuracy mainly attributed to the inclusion of LD baseline measurements, and to a lesser extent, the patient's age. The authors also provide a link in which doctors can use the suggested model for clinical application, which is appropriate and adds contribution. Data were properly made available.

The article is overall decently written. The statistical analysis performed was sound and sufficiently detailed.

However, unfortunately there are some flaws in this study that should to be addressed:

1) The authors have recently published a similar study, regarding C3F8 as opposed to silicone oil, which was indeed referenced in this study. It is hard to understand why the results of both studies were published separately. This causes a significant extent of this study to be replicated from the previous one. The authors should elaborate on this issue, as methods, results, and conclusions are also very similar.

2) The clinical correlation of the Lens densitometry (LD) is hardly addressed in this study. The decision to operate (if not performed simultaneously during PPV) is ideally made clinically, with respect to the patient's wishes. Do the LD measurements necessarily indicate the need for surgery, rather than the clinical presentation? The authors should elaborate on this premise. How this affects the surgeon's decision, to the reader, is left unclear.

3) The sample size may be too small to draw the aforementioned conclusions. Authors are advised to expand their sample size or address this limitation with more modest phrasing.

4) Can the authors provide information regarding the Variance inflation factors (VIFs) in the model?

5) It would be best to add a citation to the sentence in lines 257-259.

6) The authors should provide a reasonable explanation as to the major difference between the outcomes of different Lens densitometry (LD) baselines (10% vs 11%)? Could this be affected by other Confounding variables? Sample size?

7) The authors mention exclusion criteria by stating:

"Patients with prior cataract surgery on one eye, PPV, or any other intraocular procedure were excluded. Patients with history of ocular trauma, uveitis, topical or systemic corticosteroid therapy, and signs of visually impairing cataract at baseline evaluation were also excluded from the analysis. Additionally, known topical or systemic conditions (diabetes excluded) that could accelerate cataract formation and/or progression after PPV led to exclusion from the study. Patients with ischemic and/or proliferative retinopathy were also excluded" How many patients did this entail? Was this a fair amount that represents the cohort? What was the reasoning behind their exclusion criteria? The authors should also provide an explanation as to why Diabetes Mellitus was not excluded, as opposed to other conditions.

8) Do the authors have information on how long it took to perform the Lens densitometry (LD) measurements? Many readers might be interested in this, particularly those who have not clinically used the device.

9) Who performed the Lens densitometry (LD) measurements? Please provide details about the qualifications and background of this/these individual(s).

10) When giving the BCVA, could the authors also add the Snellen 20/X equivalents of the logMAR visual acuities listed?

11) The authors state: "PentacamHR® PNS module only depicts nuclear and no cortical or posterior subcapsular changes. PSC occurrence is described frequently under silicone oil and leads to considerable visual impairment quickly [36]. However, since PSC rarely appears isolated after silicone oil, but mostly in combination with NSC, we do not consider this weakness as a negative influence on the results of our study." Given the broad nature of readership of PLoS One, I think a supportive citation should be added.

12) The authors should add a proper "limitations" paragraph to their discussion section and address the various limitations. Are there no other significant limitations to the study besides the sole depiction of nuclear cataracts?

13) More clinical details on the patients would make this a stronger study, such as retinal and non-retinal co-morbidities. Given the small sample size, it is very important to understand the sample to know how applicable this to other patient populations. This should also be addressed in the limitations section.

14) Are there any data about the predictive value of other measures in the literature? It would be valuable to mention that in the discussion and provide relevant information, to get a sense of how well it does overall in comparison to other, perhaps more widely available, methods/instruments.

15) Due to the aforementioned, I think the conclusions drawn, especially in the way they were phrased – are too conclusive. It would best to use more cautious terminology.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

11 Mar 2022

Dear Reviewers,

First of all we want to thank you very much for your constructive criticism. We appreciate your comments and hope that we have answered your questions sufficiently clearly and satis-factorily.

Following you will find a list of your remarks and our corresponding answers.

Reviewer #1:

Answer: Our plan is to spread the statistical model among other surgeons with access to Pen-tacamHR PNS measurements. When they plan to perform PPV without combined cataract surgery, we ask them to perform a PNS scan before surgery. Predictions by our model can be performed for any timepoint, for example one year after surgery. This can be made online on a freely accessible platform. After one year PNS scans should be repeated and compared to the calculated predic-tions.

Reviewer #2 remark 1): The authors have recently published a similar study, regarding C3F8 as op-posed to silicone oil, which was indeed referenced in this study. It is hard to understand why the results of both studies were published separately. This causes a significant extent of this study to be replicated from the previous one. The authors should elaborate on this issue, as methods, re-sults, and conclusions are also very similar.

Answer: Yes, statistical methods are the same. Dividing sample in training and test data sets -> mixed regression modelling with backward variable selection -> applying the resulted model on the test data set.

What essentially differs is the data set: In previous study the model is derived based on patients treated with gas and in the current study are patients treated with oil.

Applying same methodology on different data may produce different result. As you can see the derived models are slightly different. In model for patients treated with C3F8-gas the polynomial term for month is retained. And in the model for silicone oil there is no polynomial for month but there is significant higher order interaction: Months X Age X Baseline_LD

Literature and clinical observations showed that silicone oil is thought to have different impact on lens opacification than C3F8 gas. It was our goal to get reliable data for patients treated with oil vs. patients treated with C3F8 gas. Developing one mathematical model from a dataset containing both groups, oil and gas, would impede the differentiation of their effect. The aim of both studies is to describe progression of lens opacification with objective parameter as precise as possible and from that data develop a statistical tool that adds clinical relevance for the understanding of cataract pro-gression and patient consultation for both types of endotamponade.

Reviewer #2 remark 2) and 3):

Answer: Please see lines 340-352 in the revised manuscript with track changes

Reviewer #2 remark 4): Can the authors provide information regarding the Variance inflation fac-tors (VIFs) in the model?

Answer: It is not clear what you are implying. But, as expected, the VIFs for the parameters in the final model are very large. We should emphasize AS EXPECTED because we use interaction terms in our modelling. In models with interactions the large VIFs are not a problem and are expected. Fur-thermore the goal of the study is to make a prediction about densitometry trajectories and there-fore we are interested much more in the parameter estimates than in their standard errors. Recall that no matter how great the multicollinearity among a set of variables is, it in no way compromises the estimates associated with the other variables in the regression. Nevertheless we are willing to give you some idea about the correlation among the three main predictors in the model: age, month and baseline densitometry. If we estimate the model with only these free parameters without interaction among them we derive following Variance Inflation Factors: VIFmonth=1.005, VIFage=1.62 and VIFbaseline_density=1.61. These VIFs are fairly low to cause any

concern.

Reviewer #2 remark 5): It would be best to add a citation to the sentence in lines 257-259.

Answer: We are not sure which sentences of the manuscript you are referring to, since line 257-259 already have several citations [21]–[24] and [19], [23], [25]–[28]. For a better aspect we added par-agraphs.

Reviewer #2 remark 6): The authors should provide a reasonable explanation as to the major dif-ference between the outcomes of different Lens densitometry (LD) baselines (10% vs 11%)? Could this be affected by other Confounding variables? Sample size?

Answer: Please see lines 298-314 in the revised manuscript with track changes. We rephrased this parts for a better clarification.

Reviewer #2: The authors mention exclusion criteria …How many patients did this entail? Was this a fair amount that represents the cohort? What was the reasoning behind their exclusion criteria? The authors should also provide an explanation as to why Diabetes Mellitus was not excluded, as opposed to other conditions.

Answer: In our clinic we approximately perform 1000 vitrectomies every year. The number of pa-tients presenting to our clinic with primary rhegmatogenous retinal detachment varies between 300-400 cases per year. These patients usually come as emergencies. Among those patients the majority is already pseudophakic. Remaining phakic cases (on both eyes) receive Pentacam PNS and IOL-Master measurements within our routine diagnostic procedure independent of whether PPV is combined with cataract surgery or not. This ensures that surgery will be able to start as soon as possible without any further delay due to missing diagnostic examinations. When the surgeon has made his or her decision which procedure will be performed, we screen those cases for inclu-sion. So, the clinical decision for or against phakocvitrectomy was made completely independent of this study. About 10 to 15% of the cases that would have met the criteria were lost because these cases arrived at the weekend and no PNS diagnostics were available. Included patients received verbal information about the upcoming measurements including Pentacam PNS during their rou-tine post-operative visits. An exact number of excluded patients cannot be given.

Diabetes mellitus (DM) was not excluded because of two reasons: First, DM is a frequently occur-ring condition in Europe and excluding those cases without any sings of diabetic retinopathy would lead to an even smaller cohort. Number two, diabetes is a known risk factor for cataract formation indeed, but if visually impairing cataract was present at the time of the first presentation of the patient, he or she would most likely have had combined phakovitrectomy. In addition, a recent meta-analysis showed that DM might not lead to nuclear sclerotic cataract (Li L, Wan XH, Zhao GH. Meta-analysis of the risk of cataract in type 2 diabetes. BMC Ophthalmol. 2014 Jul 24;14:94. doi: 10.1186/1471-2415-14-94. PMID: 25060855; PMCID: PMC4113025.) And some recent studies showed that DM could probably have some protective effect against cataract formation in vitrectomized eyes.

Reviewer #2 remark 8) and 9)

Answer: Please see lines 89-92 in the revised manuscript with track changes

Reviewer #2 remark 10): When giving the BCVA, could the authors also add the Snellen 20/X equivalents of the logMAR visual acuities listed?

Answer: units have been added in Table 1

Reviewer #2 remark 11) : Given the broad nature of readership of PLoS One, I think a supportive citation should be added.

Answer: We added a further citation for lines 340-343

Reviewer #2 remark 12) and 13):

Answer: Please see lines 344-346 and 350-359 in the revised manuscript with track changes

Reviewer #2 remark 14): Are there any data about the predictive value of other measures in the literature? It would be valuable to mention that in the discussion and provide relevant information, to get a sense of how well it does overall in comparison to other, perhaps more widely available, methods/instruments.

Answer: To our knowledge this and the predecessor study with C3F8-gas are the first studies that try to depict objective parameters for changes in lens status after PPV and, additionally, derive a mathematical model from that data to make future predictions for individual cases. Other models exist using artificial intelligence and deep learning to predict a risk for cataract from epidemiological data or to calculate the acuity of refractive power after cataract-surgery from preoperative measures.

Reviewer #2 remark 15) Due to the aforementioned, I think the conclusions drawn, especially in the way they were phrased – are too conclusive. It would best to use more cautious terminology.

Answer: We rephrased our conclusions according to the reviewer’s recommendations.

We want to thank you for your efforts.

Yours sincerely,

Dr. med. Phillip Schindler

Specialist for Ophthalmology

University Medical Center Hamburg-Eppendorf

Submitted filename: Response to reviewers.docx

Click here for additional data file.

29 Apr 2022

Predicting speed of progression of lens opacification after pars plana vitrectomy with silicone oil

PONE-D-21-38031R1

Dear Dr. Philipp Schindler,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Xingjun Fan, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The authors are required to address the exclusion criteria.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: A lot of the comments were addressed. Unfortunately, some remarks were brushed off, and major revisions were treated as minor ones.

1) The reply for the first inquiry was dismissive and non satisfactory. Of course, different data sets can produce different results using the same methodology, but this was not the issue raised.

2) Exclusion criteria was not propely addressed. In the case that no further details can be provided, this should be explained to the readers.

3) I believe that a bigger sample size and a broader inclusion criteria, along with co-morbidities and retinal co-morbidities (which are common with RD) would have made this a better study.

Other than the aforementioned, the authors have addressed the rest of the suggestions in a satisfactory fashion, and I believe their manuscript is better in its current form.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

12 May 2022

PONE-D-21-38031R1

Predicting speed of progression of lens opacification after pars plana vitrectomy with silicone oil

Dear Dr. Schindler:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Xingjun Fan

Academic Editor

PLOS ONE