Lecture: Microbial Keratitis Decoded: Unusual Presentations

Microbial keratitis is a leading cause of corneal opacities and blindness. The prevalence of corneal infections is disproportionately distributed to low and middle income countries. These are also the regions which have limited access to a cornea service. Patients often self medicate or opt for alternative traditional medicines. In this scenario patients present late, often lost to follow up or present with an altered clinical picture. Infection with atypical microorganisms, mixed infection, and unusual clinical presentations are not uncommon. We discuss the clinical approach and management strategies to unusual microbial keratitis.

Lecturer: Dr. Aravind Roy, L V Prasad Eye Institute, India


DR ROY: Hello, everybody. Today we shall be discussing another unusual presentation of microbial keratitis. In our series of lectures on different aspects of microbial keratitis that we had presented earlier. I am Aravind Roy. I’m your host for this session. And I am a cornea and anterior segment faculty at LV Prasad Eye Institute in India. Before we begin, could you please cite your position? Thank you. Okay. So we have a fair mix of all eyecare workers, as well as surgeons in training, in the audience. Thank you very much. Microbial keratitis is a leading cause of corneal blindness and a significant cause of ocular morbidity, especially in Africa and Southeast Asia. If we look at the incidence of reports that have been published, one can very easily note out that the incidence of new cases of microbial keratitis is almost ten times more in developing nations, as compared to developed nations. In a report that was published from southern India, the projections were that in India alone, there were 84,000 new corneal ulcers that occurred. Based on this, we need to understand that microbial keratitis continues to be one of the leading ocular emergencies. In a recent paper that was published from our institute during the ocular responses and emergencies that were encountered during the COVID-19 pandemic, microbial keratitis and trauma were the leading two causes of ocular emergencies that needed immediate attention of the ophthalmologist. Microbial keratitis needs a lot of support from the diagnostics, including microbiology, access to a good microbiology laboratory service is essential. Culture has and continues to remain the gold standard for diagnosis of microbial keratitis. How do we consider growth to be significant? Growth is significant if the same organism grows in two solid media. One solid media, and turbidity in liquid media, confluent growth in one solid media that is consistent with microscopy, or there’s growth of the organism on repeat scraping, the same organisms. A host of media are required. And a typical workflow of scraping material goes through three smears, which are the 10% KOH wet mount, Gram stain, and Giemsa stain. So not only that. But there are several other solid and liquid media that one needs to take into consideration for identifying the material. This includes typically the sheep blood agar, 5% sheep blood agar, one plate into aerobic and the other into anaerobic incubation, chocolate agar with 5% carbon dioxide, slides for fungal growth including dextrose agar, potato dextrose agar, and heart and lung infusion broth. This analysis of 563 corneal scrapings — the smears were compared to the gold standard, which is the culture positivity. And Gram, KOH, and Giemsa were compared for sensitivity and specificity to bacteria, fungi, and parasites. The sensitivity of bacteria was pretty low. The specificity was very high. What does it mean? It means if one gets the organism in the smear, then one can be sure that you are dealing with a particular type of bacteria. It could be rods or cocci likewise. But if you do not get it, that means the sensitivity is low, then one cannot be sure whether we are dealing with a bacteria or not. But that is not the case for fungi and parasites. The sensitivity and specificity is pretty much higher, and even higher if we compare the sensitivity and specificity for the KOH-CFW mount. You see the Calcofluor-white shows 90% sensitivity and 94% specificity. If we do get fungus, we can be pretty sure we’re dealing with a case of fungal keratitis. And if we get acanthamoeba cysts, we are sure we are dealing with acanthamoeba, and we can start the antiparasite drugs. Why is it important? Because fungi and parasites can have chronic infection, and often they need specific treatment for longer durations of time. Therefore one should not blindly start antifungal or antiparasitic without a definitive evidence. And the most simple and definitive evidence is a 10% KOH wet mount, which can be put up in any simple place, and it can be a very good accessory to a cornea service. So based on this, the treatment paradigms are that because most incidence studies on microbial keratitis have found that bacteria are the commonest cause of keratitis, once you start with the empirical broad spectrum treatment. When I say empirical broad spectrum treatment, we mean that we should start with the antibiotic which has a wide range of coverage against Gram positive and Gram negative organisms. This includes fortified antibiotics, which could be ciprofloxacin, aminoglycosides, or a combination of these. The most common we use is fortified cefazolin with moxifloxacin. And then we start the treatment with 5% natamycin with acanthamoeba, in case of acanthamoeba. In addition to microbiology, confocal microscopy is a big aid to the corneal diagnostics. The confocal is based on the principle of a common focal point of the observer and the illumination. Most of the illumination or the magnification that can be noted in ophthalmic diagnostic systems is limited to about 40x. Meaning that if we enlarge the images, then we’ll just end up with big and blurry images. Whereas the confocal way, adjusting the focal lens, can increase the axial illumination and magnification in such a manner that the resolution can be up to 1 to 2 microns. The normal slit lamp gives you a resolution of a maximum of 20 microns, whereas the confocal can give up to 1 to 2 microns, axially, and laterally, it gives resolution of 5 to 10 microns. That gives us the advantage of detecting in vivo, in the cornea, fungi, parasites, and other pathologies. But before we understand the pathology, it’s very important to understand the normal corneal physiology. The anterior stroma typically has elongated nuclei, activated keratocytes. Nerve fibers. The posterior stroma is comparatively less cellular. And the endothelium presents as a normal layer of hexagonal or polygonal cells. The margins are well defined, and the nuclei are usually not seen. Some epithelial plexuses can be beneath the Bowman’s or above the Bowman’s, and they are typically consisting of wavy, beaded branches nerve fibers, in the diameter of three to seven microns. The stromal nerves appear to pierce through the stroma. They are linear. They have v or y shaped branching. They are usually much thicker, in the range of 10 to 15 microns. Fungi can present as irregular, branched, reflective, double walled filaments. They have irregular branching, and they are typically in bunches. So this is the characteristic picture of a fungal filament in the corneal stroma. The bacterial keratitis is characterized by plenty of activated keratocytes and infiltration with leukocytes and Langerhans cells. Bacteria typically are not visualized. Fungal keratitis has filaments, and viral keratitis is characterized by ovoid dendritic cells at the level of the subepithelial layer. These are typically not visualized in the slit lamp. Acanthamoeba presents as highly reflective, double walled round particles. They have a diameter of 10 to 20 microns. The inner wall typically has a hexagonal configuration. These are pointers that perhaps we are dealing with acanthamoeba cysts, and not just something which is an artifact. So with this, let’s come to the question that: What would be true for confocal, except? Okay. So as we discussed, the tandem and slit scanning techniques may be used to image. So that’s true for confocal. They’re usually useful for atypical keratides. In the case of organisms such as fungi or acanthamoeba, often it’s not possible to get a biopsy. In these cases, in realtime, the confocal helps in imaging, and if the confocal is positive, one can be reasonably sure that we are dealing with these atypical or unusual organisms. Bacteria could not be visualized. In addition to all of this, the confocal is usually helpful to get the pachymetry, as well as to measure the endothelial cell density. Therefore the confocal can certainly help you get a measure of the pachy and the cell density. In a publication from our group, we found that when we compared confocal microscopy in the setting of fungal keratitis and acanthamoeba keratitis, and compared it with the gold standard that is culture positivity, we found that the sensitivity and the specificity of this method of diagnosing keratitis was 88.3 and 91.1%. And the internal observer agreement was very good, with a kappa of 0.6. So when we looked at the way microbial keratitis is treated everywhere, we found that in spite of very definitive signs of suppurative keratitis, the microbiology would be negative, and this is true in 50% of cases in Ghana and 30% in India, in a comparative study between these two places. Why is that? It was because that perhaps in such scenarios, the amount of scraping that can be taking can be limited. The viability of the organisms may be questionable. The organism may get inactivated when they are transported to the laboratory. So it is a reality for the cornea surgeon that in almost half of the microbial keratitis, even if there are very, very clearcut signs of an active infectious process, the microbiology is going to be negative. What should one do in such a consideration? How are we going to approach the patient? How are we going to treat in such a scenario? To aid our treatment, to help in prognostication, a scoring system was also proposed by these authors. This was true for fungal keratitis. And they also issued a caveat. That perhaps one should use the scoring system in those areas or those regions of the world where fungal keratitis is common. So they found that the odds ratio for typical features of fungal keratitis, such as serrated, ill defined margins, raised slough, dry texture of the slough, or a color which is not typically yellow, but a typical color of the infiltrate, can have a higher odds of the keratitis presumably being of fungal etiology. And this — they also specifically mentioned in the paper — would be true for those regions of the world where fungal keratitis is common. And that would be Southeast Asia, India, Africa, South Florida. But it may not be true in a temperate climate, such as the United Kingdom. So with this background, there can be a lot of variability between the patients who have a clear cut feature of microbial keratitis, vis-a-vis a diagnosis which is elusive. The microbiology being consistently negative. We can often end up with a number of co-infections, superinfections, unusual presentations, and in the rest of the talk, we shall be highlighting these with second examples and second learning objectives that we gain from each of these scenarios. We’ll take a few questions at the end of the session. But I will just take you through the patterns in which microbial keratitis can present in very confusing manners. So we’ll start with our first patient, who was a 51-year-old female, who presented with a history of graft infiltrate. The patient had undergone therapeutic keratoplasty for a fungal keratitis, which was performed two years ago. The graft had eventually failed. The scrapings showed pseudomonas aeruginosa. And the graft resolved. The graft infiltrate — I’m sorry — resolved with topical antibiotics. She was administered ciprofloxacin hourly for two weeks. However, the graft surface had a scar. Two quadrants of vascularization. And had a generally poor surface. This was the background for the subsequent infection, which was very interesting. She presented after one year. She was maintained on topical lubricants and low potency topical corticosteroid. She presented with five days history of irritation in her right eye. There was a small area of opacity. So what are the features that alerted us that this could perhaps be infection? Number one, there was some mild irritation. Non-specific history. And a coarse surface. It stained with fluorescein, so that’s why the color is slightly greenish. Overall, there looked to be some areas of serrated margins. Are we dealing with fungus here? We’re not sure. There are some satellites. And one of the corneal sutures had some amount of infiltrates around it. Which was indicative of possibly an inflammatory component in the anterior segment. Nevertheless, we started her off on broad spectrum antibiotics, after taking the samples from the cornea. And we also stopped the topical corticosteroids. The scraping yielded yeast. The culture was candida albicans. So she was shifted to topical fluconazole, 8 times a day, for three weeks. So there are some questions from the participants, but let us take the questions once we finish the cases. So then this was the clinical course of the patient. She gradually increased — the infiltrates increased in size, intensity, and dimensions, and over a period of three weeks with topical fluconazole alone, there was gradual improvement, ultimately leading to a scar in the graft. Our understanding of candida keratitis comes from this paper by Sun et al.. Candida species are opportunists that occasionally complicate chronic keratopathy and corneal grafts. Usually there is a context of immunosuppression and diabetes that may predispose to candidiasis. They can be misdiagnosed and if they do not respond to medical treatment, they may require a penetrating keratoplasty. The demographics of the study in this paper found that most of the patients had an ocular surface disorder, a chronic keratopathy, or a toxicity. Usually there was a background of pretreatment with corticosteroids or antibacterial agents. They were often misdiagnosed as bacterial keratitis. And candida albicans was the most commonly isolated organism. This is another very interesting scenario. I’m sure all of us would have seen this in the cornea clinic. Peripheral ulcerative keratitis is another corneal emergency. There’s a progressive ulceration of the peripheral cornea. It could be due to several causes, which can be broadly categorized into non-infectious and infectious. The infectious causes could be viral, bacterial, fungal. In fact, there are myriad causes that might cause a peripheral presenting lesion. And when there is an immune mediated peripheral ulcerative keratitis, even those can be secondarily contaminated by microorganisms, including the patient’s own microflora. The typical features are infiltrate with signs of inflammation, as can be very obvious in this picture. Sometimes it’s not so obvious. In a 66-year-old male, who was a laborer, presented with 10 day history of trauma, following which was pain and redness. The clinical picture was a peripheral ulcerative keratitis. There were some areas of cellularity and some of thinning of the cornea. Because there was a clear history of injury, we withheld the steroids and gave the patient empirical broad spectrum antibiotic regimen. The patient did not improve. But he did not worsen either. And the infiltrates continued to remain as such. They were very minimal and not really amenable to scraping. In view of the progressive thinning of the peripheral cornea, he was advised resection with tissue analysis and bandage contact lens. Usually in these cases, we prescribe a topical corticosteroid, which is usually 1% prednisolone acetate, and when the patient presented after one month, this was the clinical picture. There was obviously a superinfection here, and there was a florid infection, with a lot of corneal melt, involving the entire inferior cornea, which was dense hypopyon, and this kind of a melt is usually typical of a very aggressive organism. And in these cases, usually we suspect that it could be either pseudomonas or any other Gram negative bacteria. And true to our suspicions, the scrapings indeed yielded GNB, and the culture was pseudomonas. With significant growth. This was fortunately not a multidrug resistant pseudomonas, but pseudomonas that was very sensitive to almost all the antibiotics. So the patient was continued with ciprofloxacin, steroids were stopped, and the patient eventually resolved with scarring. Often peripheral ulcerative keratitis can have a scleral component as well. And it’s important to make the distinction between infectious and non-infectious etiology. The pointers towards an infectious etiology is pus pointing. Involvement of adjacent areas. Sometimes that might be satellite nodular-like lesions, cellularity, infiltrates, and anterior chamber reaction, including hypopyon. All of these point that we are perhaps dealing with infection, and not purely an immune mediated lesion of the cornea anterior segment or the adjacent sclera. With this, we go to our next case scenario, which is… A 44-year-old agricultural worker, who presented with a history of a fall of foreign body, with watering and pain since the last ten days. He was prescribed topical moxifloxacin and ciprofloxacin ointment at bedtime. When we examined the lesion, it appeared as a raised dry appearing slough with ill defined margins. There were some areas of pigmentation. Over the corneal ulcer. The peripheral cornea had a lot of corneal edema, and there was a 1 to 1.5 millimeter of hypopyon in the anterior chamber. The scraping yielded septate hyaline filamentous fungus, but the culture yielded presence of two fungi, one of which was dematiaceous fungus. So you can see the pigment there. Pigment in the setting of a rough dry appearing infiltrate would suggest a dematiaceous fungus, which causes pigmentation of the ulcers, and usually they need to be treated aggressively. The first line of treatment is natamycin. The patient was eventually lost to follow-up, but when he presented back almost after a year during the lockdown last year, he had healed well, and the cornea was scarred. So what is true in the management of dematiaceous fungus? So… Usually voriconazole is not the first line of treatment. The first line of treatment is natamycin. But it’s true that while they respond well to management, they may need surgical management, and there is a poor visual prognosis. In case they do not respond. As you can see in this picture, this is the clinical picture of a dematiaceous fungus, where there is almost a fluoride microscopic pigmentation. This doesn’t belong to the previous patient. This belongs to another patient, and I have put it for representative purposes. This is how the ulcers appear. It appears as if there is a perforation in the cornea with entire uveal prolapse. There’s a lot of uveal pigments on the surface. But actually this is just a plaque and it is heavily pigmented, because of the dematiaceous fungi. So our experience has been published both from South and North India, and our paper on keratomycosis was published earlier in the Journal of Ophthalmology. Sengupta have also published their experience from South India in a recent paper in 2011, and they have compared our study with two studies from Western literature. All in all, in summary, these papers found that Curvularia is the most common dematiaceous fungus. The study from Sengupta et al. showed they were dealing with more severe ulcers, the average size of which was 10.5 millimeters. The presenting visual acuity was less than 20/400 in 80% of patients. Macroscopic pigmentation was found in 14% of patients. In our study, we found it in 27% of patients. The final visual acuity was not studied in our series, but it was better than 20/40 in all the Western studies, that ranged from 44 to 78%. However, it was lower in the Sengupta et al. study, probably because they were dealing with more severe cases. 48% of their ulcers healed, as compared to 81 and 72% in the other studies. Penetrating keratoplasty was required, 15% in our series, and 8% in the Sengupta study. Evisceration was required in 9 to 11%. So all in all, natamycin is the first line of treatment. And Curvularia is the commonest species. It does present with macroscopic pigmentation. It responds well. But in 10% of cases, it may require a penetrating keratoplasty, superficial keratectomy, or even evisceration. Our next case is that of a 66-year-old homemaker. Who presented with a five day history of watering, pain, and redness in her left eye. The clinical picture was typical of a fungal keratitis. There were ill defined margins. There were satellite lesions. Dry texture of the slough. And she was responding well with 5% natamycin, which is the first line of treatment. There was peripheral scarring and the area of infiltrates were constantly reducing in size and intensity. After six weeks of treatment, she presented with this clinical picture. She was very symptomatic to start with. There was increase in the infiltrates. There was melting. And sudden increase in hypopyon. When this happens in the setting of fungal keratitis, always think of a possible superinfection. Likewise, she was scraped again, and we found that in addition to the septate hyaline filamentous fungi, we also had Gram positive cocci in pairs. The culture yielded aspergillus flavus and streptococcus pyogenes. This was the susceptibility profile of the organisms. She was placed on topical ciprofloxacin, and after two months of treatment, with both antifungals and antibiotics, she resolved with scarring and vascularization. The learning here is that bacterial co-infection can occur in fungal keratitis in 5 to 25% of cases. Gram positive cocci are usually equally associated with either yeast or filamentous fungus. Gram negative rods have a propensity to affect yeast or candida keratitis. Gram positive rods are usually associated with filamentous fungi, and Gram negative with yeasts. This happens because yeasts have a biofilm, and the biofilm attracts the bacteria. Sometimes this is a mutual synergistic mechanism. Where each organism helps the other to grow and create a superinfection. Sometimes the patient’s own microflora can co-infect a healing or resolving fungal keratitis. All these scenarios exist. The demographics of such kind of keratitis is very similar. That means that fungal keratitis without co-infection and those with bacterial co-infections are usually very similar to start with. But we really do not know: What are the factors that are responsible for certain patients having a bacterial co-infection? But what we do know is that if there are Gram positive cocci, there is a fair chance it could either be a yeast infection or a filamentous fungal infection. Whereas Gram positive rods are usually more commonly associated with filamentous fungi and Gram negative rods are more commonly associated with candida. The typical clinical features that the clinician would expect is a sudden worsening after improvement of fungal keratitis. With this background, let’s have this question about: Are all true of bacterial co-infection in fungal keratitis? Thank you. So the correct answer is C. The demographics are very similar. Irrespective of whether there is a co-infection or not. In the background of bacterial — fungal keratitis. Because co-infections do occur, and that’s true. There is increased severity of the keratitis, because of the co-infection. That’s true. It occurs because of adherence of the bacteria to the fungal biofilm. That’s also true. The demographics are not different. The demographics are exactly the same. Irrespective of whether the co-infection occurs or not. The demographics are very similar. Irrespective of whether it’s just a fungal infection or a fungal infection with a bacterial co-infection. Our next case is a very interesting case of a farmer who presented with a five month history of watering, pain, and redness in his left eye. To begin with, there was some areas of scarring of the cornea. And if we look at the superficial cornea, there is just scarring with pigmentation in the inferonasal, paracentral, peripheral cornea. Whereas the 12:00 cornea has some very interesting lesions, which were deep — mid to deep stromal corneal infiltrates, and some keratic precipitates with endothelial blots. This typically had a history of waxing and waning. And likewise, we made a provisional diagnosis of HSV immune stromal keratitis. It’s primarily a clinical diagnosis. There is typically a history of waxing and waning. There is a long duration of treatment, and the corneal sensations could be impaired. In addition, this patient also had some areas of peripheral scarring and also vascularization. All this fit into the clinical picture of HSV immune stromal keratitis. So when the patient was started with topical corticosteroids, and oral acyclovir, 400 milligrams, five times a day, the patient started worsening. When there is worsening in such cases, immediately think of two possibilities. Number one, are we dealing with acanthamoeba? Or are we dealing with fungus here? So the patient was taken for a corneal biopsy. This was so because the lesions were in the midstroma of the cornea. And we noted fungal filaments and high copy numbers of HSV1 DNA by quantitative PCR. This is how a corneal biopsy should be taken. Usually corneal biopsy is performed for mid to deep stromal lesions. The lesions should be in the peripheral cornea, because this process can lead to scarring, and therefore this process is not possible in lesions that are in the central cornea. After raising a flap from the deeper corneal tissues, we scraped some material, and also took one small amount of corneal stroma, and subjected it to histopathology and microbiology. The superficial flap is sutured back with interrupted 10-0 nylon sutures. The microbiology in GMS stain and in Gram stain yielded septate filamentous fungus. We also subjected it to quantitative PCR, because we strongly suspected from clinical signs that this was a case of HSV, immune stromal keratitis, and we got high copy numbers of HSV1 DNA. The high copy numbers of HSV1 DNA was suggested that we are dealing with an active replication of HSV. So this patient was unusual, because it had infection with HSV as well as fungus. The patient continued to worsen and was taken out for a therapeutic penetrating keratoplasty. We understand from this review that nearly 100% of individuals above the age of 60 and most middle aged persons harbor HSV in the trigeminal ganglion. Conventional PCR and immunohistochemistry are both sensitive for the detection of HSV1, but a combination, meaning if there is a positivity on conventional PCR and immunohistochemistry, the specificity of diagnosis of herpes simplex keratitis increases to 97%. We also recently published a paper where we found that compared to conventional PCR, realtime PCR is a much faster, more sensitive, and specific method of detecting HSV1 DNA. Quantitative PCR not only determines the presence of the copy numbers, but it also provides us the exact viral load. And it distinguishes that whether there is an infection with HSV1 or HSV2 — therefore it helps in assessing the response to treatment. Patients who have high viral loads, meaning larger amounts of copy numbers, suggest that the disease is increasing in severity, and probably it may have a poorer prognosis. Dual infection with HSV1 and 2 have also been described. And the kind of virus which has the lower copy number is usually competed out. Co-infection can occur, as we saw in this case. With either fungi or bacteria. And when such a scenario occurs, then one needs to treat for them in addition to treatment for the HSV. Our next case is that of a 38-year-old homemaker. She presented with 10 day history of watering, pain, and redness. There was a dry appearing, raised slough in the central cornea. There was a peripheral guttering. There were ill defined margins. No obvious satellites. But it certainly had dot-like infiltrates around the entire central lesion. This picture is very characteristic, and should alert the surgeon about a possible Pythia insidiosum keratitis. The smear and culture yielded Pythium. Pythium is often confused with fungal keratitis. This is because they have septate filamentous banks, which are seen in the Calcofluor-white KOH wet mounts, seen in this figure. However, they are scantily septate or usually aseptate. They have ribbon-like folds and often they have 90 degree branchings. So a flat ribbon-like scantily septate or aseptate filaments with 90 degree branchings should alert the physician about a possible Pythium infection. It grows in chocolate agar as flat glabrous or oily colorless colonies. The definitive diagnosis for Pythium is by demonstrating zoospore formation or DNA analysis by PCR. Using genes which are for the internal region of the ribosomal DNA. So the patient was started with linezolid and azithromycin. The patient continued to worsen on this treatment and ultimately she had a large central corneal perforation. The patient was taken up for therapeutic penetrating keratoplasty. Keratoplasty is very challenging in the scenario of Pythium insidiosum keratitis. This is because there is a very high incidence of recurrence of Pythium after keratoplasty. Which is because the lesions can actually extend much farther into the peripheral cornea. Than what can be observed by the naked eye. What is recommended, and with our experience, we have also known that one needs to take a large margin clearance of at least 1 millimeter all around, so that we do not inadvertently include any tissue that might be infected with the infiltrates, and thereby leaving behind an area of infection. Which may lead to recurrence, postkeratoplasty. Fortunately, there was no recurrence, because this was entirely a corneal lesion, and we could take a very healthy margin of 1 millimeter on either side, and she ended up with a clear graft that remained clear for six months, until she had an episode of acute graft rejection, following which the graft failed. After six months, very recently, about two weeks earlier, we performed optical penetrating keratoplasty and cataract extraction with IOL implantation for this patient. After two weeks of the second transplant, the patient is doing well with 20/80 vision. Our understanding of Pythium keratitis comes from the paper that was published earlier from our group. Typically, Pythium has tentacle, dot-like, plaque-like presentations with the peripheral gutter. Scraping shows broad sparsely septate filaments and glabrous flat colonies on chocolate agar. DNA sequencing to the internal transcribed spacer region of ribosomal DNA and zoospore formation are confirmatory of Pythium. The response is better to antibacterials, such as TG, azithromycin, linezolid, or macrolides. There is a consistent poor response to therapeutic penetrating keratoplasty, and often the grafts fail, or worse, they may have recurrence, which may lead to evisceration or phthisis. Our next case is that of a 55-year-old farmer, who presented with a three week history of watering, pain, and redness. The scrapings were negative, as he was perforating, we performed a tissue adhesive and bandage contact lens. Exudates were also taken, and the patient was started on broad spectrum antibiotics. The AC exudates yielded septate fungal filaments, but the realtime PCR showed there were 8239 copy numbers of HSV1 DNA. This was a very rare scenario, with active co-infection with HSV and fungus. The patient was started with oral acyclovir, 400 milligrams 5 times a day for 8 weeks, and hourly natamycin. Eventually over a two month period, the infiltrates started resolving. Until the patient scarred. Once the scarring started and we were sure that there were no residual infiltrates, topical corticosteroids — it was 1% prednisolone acetate in this case. Fungal keratitis with HSV co-infection is very rare, unless one has a very high index of suspicion, one is usually going to miss such cases. In last lesions, like in this case, clinical features might be undistinguishable. Quantitative PCR where available is confirmatory, and combination therapy usually for long duration. By long, I mean at least 8 weeks or more of treatment is required. There is a need for topical steroids to control the inflammation, but one should be very cautious in its use, after ensuring that all the fungal infection has subsided. Or it will lead to recurrence in such cases. Our last case for the evening is that of a 61-year-old homemaker. Who presented with a 5 day history of watering, pain, and redness. She was a patient who was having a typical feature of fungal keratitis. Ill defined margins, dry raised slough, appearance of a dry texture of these lesions. And we certainly were expecting fungus. From the 10% KOH wet mount. And she resolved rapidly with 5% natamycin. After one month of treatment, she presented with a lot of pain and watering in her eyes. When we everted the lids, the upper and lower conjunctiva was like this. There were a lot of papillary responses, a lot of congestion, and at the same time, the patient was beautifully responding. The infiltrate was just about over. There was just a bit of pain and scarring in the cornea. What’s happening here? What’s wrong? Clearly these are signs of drug toxicity. Meaning that the toxicity could be due to the preservative in the drugs. Or this could be due to the active drug molecule, which was natamycin. It could be the cyclopentolate or the atropine as well. It was important to understand that at this point of time, there is no infection. And we should stop the medications. And a decision was made to stop all drugs, and the patient was just maintained on lubricants, and this was the clinical picture after two weeks of stopping all treatment. And one can see the remarkable change in the upper and lower tarsal conjunctiva. This was before and this was two weeks after, of lubricants. And one can see there is a dramatic change, and the patient was very comfortable. The initial corneal lesion had healed by now. So all these are pointers that drug toxicity can also occur, and they may confuse the picture of microbial keratitis. Therefore… The learning from these cases is that: When there are preservatives, they may cause decrease in cell viability, they may cause increase in the cellular permeability, with impaired wound healing, they’re affected by the dose, duration, concentration, and the presence of preservatives in the drugs. The topical drugs. There was a very interesting paper that was published some time back by Dua et al.. Specifically pointed out: What should be the clinical pointers that should alert the clinician about a possible drug toxicity. There should be a clinical response of worsening after initial improvement. The moment the frequency of the drug increases, there is increase in symptomatology of the patient. There is mild to intense forniceal congestion, and if we compare the upper and lower halves of the bulbar conjunctiva, it is worse in the lower, because all the drugs are gravitating to the inferior fornix. If one everts the lid, one can see a florid papillary response in the conjunctiva. So with this, we’ll come to the last question. What are the factors that are responsible for ocular toxicity? That’s correct. So allergy is not a factor. It could be because of the concentration, duration, or frequency, but not allergy. So with this, we end our discussion. And we’ll take a few questions. And then we’ll take up any other interesting experiences that you might want to share with us. So we’ll take the questions as they keep coming. Our first question was by Dr. Shaikh, who asked us: Should we patch the eye during microbial infection? No. We should not patch the eye at all. We should keep it open. Perhaps we would advise dark glasses, as the patients are usually photophobic. The second question is: If microbial keratitis leads to neovascularization, how should we treat it? Usually the microbial keratitis — the neovascularization resolves, once the ulcer heals with scarring and gross scars are left behind. But microcautery can be used, as well as injection of anti-VEGF to the cornea. They have been tried by various groups with various successes. Is lactophenol blue stain better than KOH? I’m not really sure. Lactophenol blue is equally effective in identifying fungi. The next question is that… Can a patient with microbial keratitis continue to wear contact lenses? Can we recommend scleral lenses? We typically recommend the patients to be off the contact lens. During an active episode of microbial keratitis. And in fact, if a patient with — a contact lens wearer presents with a corneal ulcer, it’s recommended to send the contact lens, the case, and the solution for microbiology. Because often the offending organism could be sourced from the contact lens or the solution as well. Does honey avoid this fungus? I’m not sure what this question is. I’m not sure about this question. The next question is that… Are steroids contraindicated in the treatment? If we’re dealing with fungus, definitely steroids are contraindicated. If we’re dealing with bacterial keratitis, the body of evidence points that usually there is no difference, irrespective of whether we give steroids or not. But there are clear cut guidelines when one should give corticosteroids in the setting of microbial keratitis. These guidelines include that the organism should be identified. The organism should be responsive to the drugs that we are giving. The sensitivity profile of the organism has been identified. And we are dealing with an organism that is affecting the central cornea, and there’s a chance that if we do not give steroids, the underlying scarring will threaten the sight. And we are giving steroids when there is a graft infiltrate. And the graft infiltrate is resolving to a bacterial microorganism. And the bacterial microorganism — once it has started resolving, if we do not give steroids, there’s a chance that the graft will fail. So when all these conditions are met, then we should give steroids. Otherwise, it is not really indicated or beneficial in the setting of microbial keratitis. Place of contact lens in enhancing corneal ulcers in microbial keratitis? I’m not sure I understood the question. Would you mean that… Should we use contact lenses? Perhaps no. We should not use contact lenses. We should encourage the patient to discontinue contact lenses in the setting of microbial keratitis. So atypical mycobacteria — I believe ATB… Intrastromal injections? Intrastromal injections have been tried in cases which are not responding to topical treatment. Especially antifungals, such as voriconazole injection have been tried. But there are studies that have found there is ultimately no difference. But it has been tried. Oral antifungals — the MUTT II trial had definitively found that there is no body of evidence which strongly supports the use of oral antifungals in filamentous fungal keratitis, but it’s not uncommon for many practitioners to use oral antifungals in deep mycosis or mycosis which involves the limbus or sclera, because the underlying hypothesis is that when there is an oral intake of antifungals, then they will be secreted better and therefore will lead to increased therapeutic availability at the level of the sclera, or perhaps in the aqueous, which is why in these settings, some people certainly do prescribe it. But there is no definitive evidence. Can we use topical NSAID instead of steroids? Certainly we can use that. But we can also use something like immunomodulators, such as cyclosporine, which is actually used in the setting of one eyed patients, or patients where there’s a graft and we cannot give steroids because we’re considering perhaps we are having a fungal etiology here. So it can be given. What is the role of crosslinking in fungal keratitis? This is a very good question. I think all of us should deliberate something more about it. Crosslinking has recently been studied a lot in the setting of intractable microbial keratitis. And the evidence that has been generated is that usually it can work in end stage microbial keratitis. Which is completely refractory to all treatment. Having said that, unfortunately it does not work in the setting of fungal keratitis. Rather, it has detrimental effects or no effect when crosslinking has been done or not done. So studies such as the CLARE that have been recently published in Ophthalmology have found that crosslinking is not recommended in fungal keratitis. The major treatment for bacterial keratitis — it is completely based on the microbiology. So one has to start with empirical broad spectrum treatment. And that treatment should cover both Gram positive and Gram negative organisms. But once that has been achieved, then one should switch over to a type of antibiotic against which a sensitivity profile has been mapped, and specifically one should treat the patient for those antibiotics to which the organism is sensitive. Amniotic membrane has been tried. Yes, it has been tried. I would not use oral antifungals in fungal cases. I typically do not. The MUTT II trial showed very conclusive evidence that there is no role of oral antifungals in this context. But then as we discussed, in certain deep mycosis, it can be used. When there is a fungal scleritis or scleral involvement, certainly one can use oral antifungals. I’m not sure whether the question is the best topical steroid, because it’s usually the most potent one — is prednisolone acetate 1%. And dexamethasone. But if one is trying to use a low potency steroid, especially if there is a chance that the patient may be a steroid responder, or might have IOP shoot, then one can use a low potency steroid such as fluorometholone or perhaps loteprednol. Are there any alternatives? So the first question by Dr. Mathen, we have answered that. The oral antifungals. That’s done. Is there any alternative for surgical therapy? For fungal keratitis, other than PK? Certainly we have a lot of alternatives. There are certain experts who recommend to do a superficial keratectomy. Lamellar keratoplasty or deep anterior lamellar keratoplasty has also been published. And one needs to be very sure that you have excised all the pathology or all the areas of infiltrate, because unless one does that, there is a very high chance of recurrence. Is scraping of ulcer helpful in case of fungal keratitis to enhance penetration? Yes. Scraping or debridement of the ulcer is helpful. To improve the penetration of the antifungal drugs. What is the best way of collecting material for culture? In an impending perforation? It’s better to do that in the setting of the operating theater, because if the cornea perforates, then you should have a tissue adhesive on hand, so that one can apply the tissue adhesive over the area of perforation. What is the best topical antibacterial? Again, it depends on the organism. Can we use a contact lens as a drug delivery system? Certainly this has been tried by different groups. Where contact lens micropatches, microneedles have all been used as drug delivery systems. Which cultures do you suggest as an initial approach? We need to use all the plates. So if you do not have a lot of plates available, then at least a chocolate agar and a Sabouraud dextrose agar should give some basic media. But at least a chocolate agar should be there. A chocolate agar helps both the bacteria as well as the fungi. If you suspect a non-infective PUK, should I start therapy with corticosteroids and antibiotics, or should you do scrapings? Certainly start therapy with corticosteroids. However, if you feel that there are infiltrates which are suspicious, there is no harm in taking a scraping. Before starting the treatment. If we cannot perform culture on clinical examination, I’m not sure I understand the question… But if we can… I think the question is that if we do not have microbiology, then in the setting of Southeast Asia and Africa, there is a scoring system that has been proposed, which can tell us what are the possible features of a fungal keratitis. So we can say that whether we’re dealing with fungus or bacteria… But then if the patient has been treated elsewhere, and the patient presents in an advanced stage of disease, it’s not really clear what would work and what can go wrong either way. So it’s always best that one should scrape and go by the microbiology. For topical antifungal choices natamycin and voriconazole… Yeah, natamycin and voriconazole — the trials have definitively found that natamycin 5% is the better molecule. Having said that, fluconazole is typically reserved for candida keratitis. For filamentous fungal keratitis, one has to prefer natamycin. That is the better molecule. Is there a role for topical Betadine? Not really sure. It’s a disinfectant, but for therapeutic use, it will have a lot of irritation on the ocular surface. So the percentage of microbial keratitis where we get an organism or growth — again, in 30% to 40%, there are going to be microbiologically negative… So one has to deal with either a PCR or a biopsy. One has to just assess the treatment to the… Assess the response to the treatment. Bandage contact lens during treatment is required if a tissue adhesive is placed on the cornea. Because the tissue adhesive will irritate the tarsal conjunctiva, so bandage contact lens certainly will be required. But otherwise a routine bandage contact lens is not required. In fact, we should avoid any contact lens wear. So that there is better access of the drugs to the area of the infiltrate. Can we combine maintenance dose of antivirals? Actually, it’s a therapeutic dose of acyclovir 400 milligrams, five times a day, adult dose, for 2 to 3 weeks. And after that, the prophylactic dose consists of 400 milligrams twice daily for a period of three to six months, after tests are done to make sure there is no underlying toxicity by the oral acyclovir. Can we combine fluconazole with topical antibiotic? It depends. It depends whether there is a combined infection or superinfection — one can certainly combine. Doxycycline certainly helps in reducing corneal melt. And it can be given as an adjuvant treatment. That’s correct. The Gram slides… We don’t have many Gram slides. But the KOH is usually confirmatory of Pythium. And as I showed in my previous slide deck, they are usually flat, ribbon like, with aseptate or scantily septate filaments. They’re usually much larger, broader than the fungal filaments. Which is how we make the distinction between fungus and Pythium. The growth also is very typical. It has flat, colorless colonies. But that’s again not very definitive. What is definitive is demonstration of zoospores by PCR sequencing of the ribosomal transcribed region. Frequency of topical antibiotics — usually we start with a loading dose of hourly drugs for the first 24 hours, after which the patient is maintained for hourly topical antibiotics, for all hours of waking, and after the first 48 hours have been over, with this, we switch over to 8 times a day and assess the response at one week. Is there a role of systemic antibiotics? No. There is no role of that. What are the treatment options in corneal melt? One should apply a tissue adhesive or a patch graft or a penetrating keratoplasty. Depending on the area of melt and the tectonic support of the overlying cornea. What should guide a doctor on how many times the drug should be used? The response to treatment guides the frequency of instillation of drops. Ocular toxicity… Is certainly suitable for MK once the features of toxicity subside. But what it means is that one needs to reduce the frequency of administration of the artificially instilled drugs, because the drugs have preservatives, and that is causing the toxicity. But if there is a role for therapeutic penetrating keratoplasty or optical PK once it has scarred, then certainly one can go ahead and do a PK. But one needs to be sure — one needs to balance the duration and the dosage and the frequency of the drugs. Can an optometrist treat microbial keratitis? Emergency treatment can be given, but it is best referred to an ophthalmologist for treatment of microbial keratitis. Iodine? Again, we have discussed this question. It will not be very helpful. If we do not have access to a lab service, then one needs to look for the scoring system that was discussed. Then that gives an idea of the positive predictive value of each of the signs, which includes the margins of the ulcer, raised slough, dry appearance of the slough, presence of hypopyon, and color of the infiltrates. So this will be true of predicting a fungal keratitis in a region where fungal keratitis is very common, such as Africa, Southeast Asia, or India, or South Florida. Emergency PK needs to be done in the setting of an ulcer that is worsening on maximal treatment. Medical treatment. Or if there is a corneal melt. There is a perforation. There is an extension to the corneal limbus, or involvement of the adjacent sclera. In a progressive PUK would not topical steroids increase risk of perforation? Yes, it would. Therefore we need to understand that this is an immune mediated mechanism. Therefore we need to balance the pro with the antiinflammatory factors over here. And for tectonic support, usually a conjunctival resection with tissue adhesive is done prior to starting steroids. One should watch these cases, because they can get superinfected with bacteria or other organisms. In which case, one has to stop the steroids, do the microbiology, and treat as per the etiology. Traditional treatments are usually not recommended. Therefore I would certainly not recommend this line of treatment. Of applying any other… Either honey or any traditional — because that’s not found to be beneficial. On the contrary, it has led to worsening of such lesions. Usually the first line of treatment for fungus is 5% natamycin, and for bacterial keratitis, one needs to start with a broad spectrum empirical treatment. Which includes fortified aminoglycosides, fortified cefazolin, with a fluoroquinolone. Conjunctival resection is believed to be beneficial at least in two ways. Number one is that in the setting of an immune mediated peripheral ulcerative keratitis… Because it cuts off the blood supply, and second, the tissue adhesive adds to the tectonic support. But having said that, there is a 37 to 46% chance of recurrence in the first to second year of peripheral ulcerative keratitis, immune mediated, after conjunctival resection. So yes, it has benefit. But then it’s not totally foolproof either. I’m not sure about the question of plants fighting the fungus. Role of fortified drugs. If you can achieve a fortified drug, if you can prepare it in your pharmacy, certainly. If not, a broad spectrum fluoroquinolone can be used. But broad spectrum fluoroquinolones also have a lot of underlying resistance. Therefore a lot of organisms may not be responsive to that. Therefore a combination therapy of a fortified cephalosporin or aminoglycoside with a ciprofloxacin usually is the best broad spectrum, because it provides the maximum possible coverage against a range of Gram positive and Gram negative organisms. Okay. Thank you very much. All these experiences are from personal experiences. Those of our mentors, colleagues, and fellows who have learned with us. And the publications that one can find on PubMed. So certainly I have given the citation for almost all the material that I have used here. And I would like to thank Orbis and the managers for this wonderful opportunity of reaching out to all of you. It was a fantastic and very interactive session, and I look forward to coming back again with more of our experiences, and also take some more experiences from your end, and share. Thanks to all for joining in. It was really wonderful interacting with all of you.

Download Slides


February 12, 2021

Last Updated: October 31, 2022

Leave a Comment