- Email: [email protected]
Spontaneous Vitreous Hemorrhage in Children
Spontaneous Vitreous Hemorrhage in Children ADITYA SUDHALKAR, JAY CHHABLANI, SUBHADRA JALALI, ANNIE MATHAI, AND AVINASH PATHENGAY PURPOSE: To determine the clinical profile, causes, and outcomes of ‘‘spontaneous’’ vitreous hemorrhage in children (<18 years). DESIGN: Retrospective computer-assisted chart review. METHODS: Charts of 124 eyes of 76 children who presented with nontraumatic, nonsurgical vitreous hemorrhage between 2002 and 2012 were reviewed. All children underwent an appropriate ocular and systemic examination. Data collected included demographics, visual acuity, cause of ‘‘spontaneous’’ vitreous hemorrhage, ocular and systemic findings at presentation and at last follow-up, investigations, management details, and visual outcomes. RESULTS: The median age was 153.45 ± 56.19 months. There were 39 female and 37 male patients. Forty-eight patients had bilateral vitreous hemorrhage. The most common presenting complaints were diminished vision (96.45%) and behavioral changes (87.24%). The mean baseline BCVA in logMAR was 2.25 ± 1.11. The most common causes included vasculitis (34.6%) and hematologic disorders (27.4%). Patients were given medical therapy (topical and/or systemic) or underwent laser photocoagulation (29%) and/or surgery (55.6%). Twenty-nine eyes (23.3%) did not require any intervention. The mean number of surgeries was 1.89 ± 1.45 (range 1-4 surgeries). The mean final visual acuity was 0.76 ± 0.58 logMAR and was significantly better than the baseline (P < .001). The best anatomic and visual outcomes were seen in vasculitis, whereas congenital disorders such as retinoschisis had the worst. CONCLUSIONS: Spontaneous pediatric vitreous hemorrhage has a diverse etiology, vasculitis being the most common cause in our series. A comprehensive evaluation (systemic and ocular) is required to ensure that visionand life-threatening conditions are not missed. (Am J Ophthalmol 2013;156:1267–1271. Ó 2013 by Elsevier Inc. All rights reserved.)
V
ITREOUS HEMORRHAGE, BY ITSELF, IS AN INSUFFI-
cient diagnosis and is an indicator of the need for a systematic analysis of its etiology. The diverse etiology can range from purely ocular conditions, such as posterior vitreous detachment, to diseases with an Accepted for publication May 31, 2013. From the L.V. Prasad Eye Institute, Kallam Anji Reddy Campus, Hyderabad, India (A.S., J.C., S.J., A.M.); and L.V. Prasad Eye Institute, GMR Varalakshmi Campus, Visakhapatnam, India (A.P.). Inquiries to Dr Jay Chhablani, L.V. Prasad Eye Institute, Kallam Anji Reddy Campus, L.V. Prasad Marg, Banjara Hills, Hyderabad - 500 034, India; e-mail: [email protected] 0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.05.042
Ó
2013 BY
underlying systemic etiology, such as hypertension, diabetes mellitus, or vasculitis. Concurrent pathology is often the main determinant of the anatomic, visual, and functional outcome in any patient with vitreous hemorrhage. Vitreous hemorrhage, its underlying causes, and their management are well described in adults.1 Literature on vitreous hemorrhage in children is very limited, including mostly case reports and case series.2–9 A large series by Spirn and associates10 has added considerably to our understanding of the factors causing and the determination of the etiology-specific outcomes of vitreous hemorrhage in children. Trauma, expectedly, was found to be the dominant cause of vitreous hemorrhage in children.10 Certain other causes leading to so-called ‘‘spontaneous vitreous hemorrhage,’’ some of which can be life threatening, merit further discussion. These potentially life-threatening conditions include retinoblastoma or hematologic disorders. It is often very difficult to establish a correct diagnosis in children, as history and elucidation of signs and symptoms is often not forthcoming. Considerable regional variation can be expected when analyzing a condition with such varied etiology, given that each place is socioeconomically, ethnically, and culturally diverse. It follows, thus, that appropriate documentation of characteristics that define or are common to vitreous hemorrhage secondary to such conditions is essential, as an aid to establishing the correct diagnosis. We thus aimed to determine the causes, clinical presentation, and anatomic and functional outcomes of ‘‘spontaneous’’ vitreous hemorrhage in the pediatric population as noted at a tertiary eye care center in India.
METHODS A RETROSPECTIVE COMPUTER-ASSISTED DATABASE SEARCH
and chart review of all patients aged less than 18 years with vitreous hemorrhage who presented to L.V. Prasad Eye Institute, Kallam Anji Reddy campus, Hyderabad between January 1, 2002 and May 31, 2012 was conducted. All procedures conformed to the Declaration of Helsinki for research involving human subjects. The Institutional Review Board of L.V. Prasad Eye Institute, Hyderabad, India approved this retrospective review of these patients’ records for the study. Written informed consent for all subjects was taken from their guardians/parents for the diagnostic and treatment procedures conducted. Cases of vitreous hemorrhage secondary to retinopathy of prematurity (ROP) or trauma (surgical or nonsurgical) and those with incomplete records or diagnosis and/or a follow-up of less than 1 month
ELSEVIER INC. ALL
RIGHTS RESERVED.
1267
were excluded. The chart review adhered to previously established guidelines, described in earlier publications.11 Data collected included demographics; laterality; presenting symptoms and signs (if the child was old enough to cooperate for appropriate history taking; if not, information about the onset of symptoms, possible causes, and previously existing conditions, if any, was obtained from parents); the presenting corrected distance visual acuity; details of the examination procedure; additional investigations and systemic examination, if conducted; the cause, if evident; the treatment offered; the duration of follow-up; and the final anatomic and visual outcomes. Visual acuity was measured using the ETDRS (Early Treatment Diabetic Retinopathy Study) chart and the results subsequently converted to logarithm of minimal angle of resolution (logMAR) chart and noted in logMAR units. For children in the age range of 6 months to 4 years, Teller acuity charts were used and similarly converted to logMAR units. A comprehensive ocular examination was performed in all cases, including under short general anesthesia if the child was not cooperative. If media clarity was insufficient to permit adequate fundal examination, or when additional information, such as choroidal thickness or orbital status, was required, an ultrasound B-scan was performed (OPKO OTI Pvt Ltd., Miami Florida, USA). Further investigations were decided based on the clinical impression. Investigations included ocular coherence tomography (Carl Zeiss Meditec Pvt Ltd., Erlangen, Germany); fluorescein angiography (including under anesthesia, if indicated); visual evoked potential; radiologic imaging (such as computerized tomography scans); blood and urine examinations such as complete blood count, erythrocyte sedimentation rate, C reactive protein, Quantiferon tuberculosis (TB) test, serum calcium, serum angiotensin converting enzyme level, rheumatoid arthritis factor, antinuclear antibody, and urine gross and microscopic examination; and systemic and laboratory tests for specific vasculitis. Treatment, both ocular and systemic, was planned on the basis of diagnosis and can be broadly classified as surgical or nonsurgical. Surgical intervention ranged from vitreous biopsy as a diagnostic aid to combined buckling and vitreoretinal procedures for associated vitreoretinal pathology. Nonsurgical therapy consisted of observation, topical or systemic medical therapy, and/or laser photocoagulation. Frequent cycloplegic refractions and amblyopia therapy by part-time occlusion was part of the management protocol. Statistical analysis was performed using SPSS (v. 16; SPSS, Inc, Chicago, Illinois, USA), with special emphasis on the aforementioned information. Statistical significance was set at P < .05.
RESULTS A TOTAL OF 619 PEDIATRIC PATIENTS WITH AN ENTERED
diagnosis of vitreous hemorrhage in our medical records 1268
department database were identified. Analysis after exclusion of acute ROP cases (n ¼ 34 patients) and those with an incorrect diagnosis (n ¼ 27 patients) or coding errors (n ¼ 13) was performed on 545 patients. A total of 656 eyes with any vitreous hemorrhage (as the diagnosis) were available for analysis. We identified and included 76 patients (124 eyes; 18.09% of the total number of pediatric patients (619) with vitreous hemorrhage who presented to our institute in the stated period) of nontraumatic and nonsurgical ‘‘spontaneous’’ vitreous hemorrhage and performed the analysis. Six patients had an identified cause of spontaneous vitreous hemorrhage; a follow-up of less than 1 month led to their exclusion. Their baseline characteristics did not differ significantly from those included in the study. The median age of the patients was 12.78 6 4.68 (standard deviation) years (range 4 months to 17 years). There was no significant sex bias (39 female vs 37 male patients). Forty-eight patients presented with bilateral vitreous hemorrhage; Table 1 shows the proportion of patients having a specific cause in different age groups. All patients with an identified cause for the vitreous hemorrhage had local and/or systemic comorbidities, which provided the underlying pathologic mechanism of hemorrhage. The mean follow-up was 28 6 18.38 months. The most common presenting complaint in children aged more than 3 years was diminished vision (96.45%). The most common presenting complaint in children aged more than 3 years was diminished vision (119 of 124 patients, 95.96%). The most common presenting complaint in children aged less than 3 years (as reported by the parents) was behavioral changes (110 of 124 patients, 88.7%), such as not being able to recognize the mother, or a recent-onset lack of interest in surroundings. The mean baseline corrected distance visual acuity in logMAR was 2.25 6 1.11. In addition to vitreous hemorrhage, 76 eyes (61.3%) had sufficient view of the fundus and showed typical features of the underlying disease on fundus examination. A total of 51 eyes required an ultrasonography for evaluation of the posterior segment of the eye. Fifty patients required hematologic, immunologic, or radiologic investigations. Patients were given medical therapy (topical and/or systemic) or underwent laser photocoagulation (36 eyes) and/or surgery (69 eyes). Overall, 69 eyes (55.64%) required some form of surgical intervention, such as scleral buckling, vitrectomy, lensectomy, plaque placement, enucleation, or diagnostic vitreous biopsy. Table 2 shows various strategies for management in different age groups. The mean number of surgeries was 1.89 6 1.45, with a range of 1-4 surgeries. None of the operated cases developed endophthalmitis or secondary glaucoma. A total of 19 eyes underwent surgery for tractional or rhegmatogenous retinal detachment, with 3 eyes having a recurrent retinal detachment at final follow-up. Two of these eyes had familial exudative vitreoretinopathy (FEVR) and 1 eye had retinoschisis.
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
TABLE 1. Causes of Spontaneous Vitreous Hemorrhage in Different Age Groups in Children <1 Year Age
>1-10 Years Age
>10 Years Age
Causes (n ¼ Eyes)
(14 Eyes)
(41 Eyes)
(69 Eyes)
Vasculitis Hematologic disorders Ocular tumors/vascular malformations Familial exudative vitreoretinopathy Retinoschisis Retinal breaks/detachment Diabetic retinopathy Pars planitis Hepatorenal dysfunction Sepsis Unknown etiology
3
9 13 4
34 21 1
3
8
2
-
2 2
3 11
2 -
1
4 -
1 5
2
1
The mean final corrected distance visual acuity was 0.76 6 0.58 logMAR. The difference was statistically significant, in that most patients had improved considerably in terms of visual acuity at final follow-up (P < .001). Two patients with FEVR presented with vitreous hemorrhage and inoperable retinal detachment. Two patients with FEVR, 1 patient with retinoschisis, and 2 with vasculitis had a recurrence of retinal detachment at final followup and refused further intervention. Three patients required enucleation for retinoblastoma. One eye with choroidal hemangioma that had undergone multiple surgeries eventually progressed to phthisis. Eleven patients with vasculitis, 6 with FEVR, and 2 with retinoschisis required some surgical/ablative intervention in the other fellow eye, and regular follow-up possibly prevented a more advanced disease occurrence, including vitreous hemorrhage, in the fellow eye, through the last follow-up. We did not come across any case of vitreous hemorrhage as a late consequence of ROP in the present series.
DISCUSSION TRAUMA REMAINS THE FIRST AMONG ALL CAUSES OF PEDI-
atric vitreous hemorrhage;9 however, as is seen in our series, a number of conditions can result in ‘‘spontaneous’’ vitreous hemorrhage in children. These conditions account for 18.09% of pediatric vitreous hemorrhage as seen at our institute. This is a significant number as a group, and is a collection of various conditions that often only have vitreous hemorrhage as a common presenting sign. Articles addressing the same issue in adults show that spontaneous hemorrhage is very common,1,12 VOL. 156, NO. 6
TABLE 2. Management Strategies for Children of Various Age Groups Presenting With Spontaneous Vitreous Hemorrhage
None/medical Laser photocoagulation Surgical
<1 Year (14 Eyes)
>1-10 Years (41 Eyes)
>10 Years (69 Eyes)
5 (35.8%) 1 (7%) 8 (57.2%)
5 (12.2%) 7 (17.1%) 29 (70.7%)
9 (13.0%) 28 (40.6%) 32 (46.4%)
in contrast with children, in whom it accounts for a minority of the reported cases.10 We propose an algorithm approach for a child presenting with vitreous hemorrhage (Figure). As can be seen from Table 1, the predominant cause of spontaneous vitreous hemorrhage was vasculitis (43 of 124 eyes [34.67%], 30 patients), of which 17 patients had unilateral vitreous hemorrhage and 11 patients had signs suggestive of vasculitis in the other eye. Timely laser photocoagulation and/or oral immunosuppressive therapy helped prevent a similar occurrence in the fellow eye. It also helped us advise all patients as to the need for adequate, regular, and longterm follow-up. The cause of vasculitis could not be determined accurately in 22 patients who were treated with oral steroids and, wherever appropriate, with laser therapy or surgery. The remaining 8 patients had a positive Mantoux test and Quantiferon gold TB test and responded well to antitubercular treatment supplemented with steroids. Recurrences were noted in 4 patients with idiopathic vasculitis and none of the tuberculosis-positive patients. Patients with vasculitis had the best visual outcomes compared with other causes of spontaneous vitreous hemorrhage. Hematologic disorders constituted the next major group of etiologic diagnosis of vitreous hemorrhage (21 patients, 34 of 124 eyes; 27.4%). These included sickle cell anemia (5 patients), leukemia (9 patients), anemia (3 patients), and thrombocytopenia (4 patients). These were managed appropriately in consultation with the treating hematologist. Clinical features of ocular involvement as noted in our study are similar to past reports.10 Five patients had retinal breaks without retinal detachment; 3 presented with a sudden onset of floaters and 2 presented with a recent onset of flashes. Two patients had high myopia. One patient had an associated lattice. Retinal breaks without retinal detachment (5 eyes) were treated with laser photocoagulation alone; retinal detachment (8 eyes) was treated with scleral buckling (5 eyes), vitrectomy (2 eyes), or a combination of both (1 eye). FEVR (13 of 124 eyes; 10.48%) has been reported to cause vitreous hemorrhage.12 It was managed with vitreoretinal surgical intervention (9 eyes) and/or laser therapy (4 eyes). Retinoschisis patients (5 eyes, 4.03%) with vitreous hemorrhage underwent vitreoretinal surgery with endolaser photocoagulation.
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1269
FIGURE. Approach to a child presenting with vitreous hemorrhage. Management of vitreous hemorrhage related to retinopathy of prematurity or trauma has not been discussed in detail in this algorithm. CT scan, computed tomography scan; RD, retinal detachment; ROP, retinopathy of prematurity; FFA, fundus fluorescein angiography; ACE, angiotensin converting enzyme; ANA, antinuclear antibody; CD, cluster of differentiation.
The diagnosis of retinoblastoma was possible from an ultrasound examination in all patients with the disease. Three patients were advised to have an additional computerized axial tomography scan to determine if there was any disease extension and the presence of calcification. Retinoblastoma (5 eyes, 4.03%) was treated with chemotherapy, radiotherapy, plaque therapy, or enucleation, as per the disease stage in a given patient. Patients with vasoproliferative tumors (3 eyes, 2.43%), such as capillary hemangiomas, underwent laser photocoagulation, surgery, or both. The causes of vitreous hemorrhage, as seen in the series by Spirn and associates,10 while similar in some aspects, are quite different from the etiology noted in our series. Whereas uncommon conditions like retinoschisis were found in our series, there were hardly any reported cases of vasculitis by Spirn and associates. Pars planitis is known to cause vitreous hemorrhage in children as well.13 One possible explanation is that immunologic reactions to infectious agents such as tuberculosis were more likely to occur in a developing country, for approximately one fourth of our vasculitis cases had concurrent evidence of tuberculosis. It is possible that the remaining ‘‘idiopathic’’ cases were secondary to an immune reaction to other agents. Spirn and associates reported many cases of vitreous hemorrhage as late sequelae of untreated retinopathy of prematurity; however, we did not come across any such patients in our study. We did not come across spontaneous vitreous hemorrhage attributable to Norrie disease or Coats disease in our retrospective review. These differences could reflect practice 1270
patterns, terminology usage, and referral influences. One finding that stands in stark contrast to Western literature was the lack of any patient’s being diagnosed with battered baby syndrome in the entire series, traumatic or otherwise; we believe this could probably be attributable to underreporting, though the syndrome may be quite rare in our population; we can only speculate that cultural and socioeconomic circumstances are different in different regions, accounting for no cases of battered baby syndrome. This difference also holds true when we consider bilateral vitreous hemorrhage, for which the major cause reported in Western literature is battered baby syndrome.14 In contrast, vasculitis followed by hematologic disease emerged as the major cause of bilateral spontaneous vitreous hemorrhage. Our study shows significantly improved visual outcomes at final follow-up; an important point to be kept in mind while interpreting visual outcomes is that visual acuity assessment in very young children has limitations. When analyzed separately, cases of FEVR and retinoschisis had very poor visual and anatomic outcomes; these conditions can affect the retina early and severely and can hinder proper anatomic and functional visual development. Spontaneous vitreous hemorrhage in the first year of life, especially that which requires intervention, is uncommon, and the reported etiologies9 thus far are quite different from the ones noticed in our series. We also noted in our series that the number of infants with a vitreous hemorrhage of an unknown etiology was more than twice the number of children in the 1-10 year age group and 5 times that in the older age group. This points
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
to the difficulty in eliciting appropriate history and clinical signs and symptoms in very young children. Four of these patients were lost to follow-up prior to appropriate review of investigations. When considering vitreous hemorrhage in infants, a previous study shows that in many cases the exact cause was unknown, similar to what we have seen in our series.14 However unlike past literature, we did not come across any infant with vitreous hemorrhage secondary to hematologic or coagulative disorders.14 Additionally, as already noted, we did not come across even a single case of battered baby syndrome, a not-uncommon cause of vitreous hemorrhage and other injuries in infants.15 We did not notice any signs of preretinal or intracranial hemorrhage in such infants. In our series only 2 neonates were evaluated with vitreous hemorrhage, but the cause remained unknown. In all probability, it was secondary to birth trauma, but verification was not forthcoming and the patients were lost to follow-up after first presentation.
Thus, going by the data in the series as a whole, the importance of bilateral dilated fundus examination, under anesthesia if required, cannot be overemphasized. Appropriate intervention, such as diagnostic vitreous biopsies, proved useful in patients not only in ruling out infection etiology in suspicious cases, but also via histopathologic confirmation of inflammatory disorders, such as tuberculosis. To conclude, spontaneous vitreous hemorrhage in children is a condition with a diverse etiology that requires a comprehensive ocular and systemic examination, often aided by additional investigations, to be able to reach an appropriate diagnosis. Outcomes relate to the underlying cause, but the hemorrhage can be a pointer to more sinister disease, such as malignancy, both ocular and systemic. There are significant differences in the etiology and presentation of pediatric vitreous hemorrhage, as seen in different geographic regions, a point that needs to be noted while formulating differential diagnosis and proceeding with management for the same.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. The authors indicate no funding support. Contributions of authors: conception and design (A.S., J.C.); analysis and interpretation (A.S., J.C., S.J., A.N.M., A.P.); writing the article (A.S., J.C., S.J.); critical revision of the article (A.S., J.C., S.J., A.N.M., A.P.); final approval of the article (A.S., J.C., S.J., A.N.M., A.P.); data collection (A.S.); provision of materials, patients, or resources (J.C.); statistical expertise (A.S.); obtaining funding (none); literature search (A.S., J.C.); and administrative, technical, or logistic support (A.N.M., A.P.).
REFERENCES 1. Spraul CW, Grossniklaus HE. Vitreous hemorrhage. Surv Ophthalmol 1997;42(1):3–39. 2. Ferrone PJ, de Juan E Jr. Vitreous hemorrhage in infants. Arch Ophthalmol 1994;112(9):1185–1189. 3. Braendstrup P. Vitreous haemorrhage in the newborn. A rare type of neonatal intraocular haemorrhage. Acta Ophthalmol (Copenh) 1969;47(3):502–513. 4. Cekic O, Totan Y, Batman C. Traumatic vitreous hemorrhage from a persistent hyaloid artery. J Pediatr Ophthalmol Strabismus 2000;37(2):117–118. 5. Graham EM, Stanford MR, Shilling JS, Sanders MD. Neovascularisation associated with posterior uveitis. Br J Ophthalmol 1987;71(11):826–833. 6. Hutcheson KA, Nguyen AT, Preslan MW, Ellish NJ, Steidl SM. Vitreous hemorrhage in patients with high-risk retinopathy of prematurity. Am J Ophthalmol 2003;136(2):258–263. 7. Marshman WE, Adams GG, Ohri R. Bilateral vitreous hemorrhages in an infant with low fibrinogen levels. J AAPOS 1999; 3(4):255–256. 8. Schulman J, Peyman GA, Jednock N, Larson B. Indications for vitrectomy in congenital retinoschisis. Br J Ophthalmol 1985;69(7):482–486.
VOL. 156, NO. 6
9. Saxena S, Jalali S, Verma L, Pathengay A. Management of vitreous haemorrhage. Indian J Ophthalmol 2003;51(2): 189–196. 10. Spirn MJ, Lynn MJ, Hubbard GB 3rd. Vitreous hemorrhage in children. Ophthalmology 2006;113(5): 848–852. 11. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27(3): 305–308. 12. Dana MR, Werner MS, Viana MA, Shapiro MJ. Spontaneous and traumatic vitreous hemorrhage. Ophthalmology 1993; 100(9):1377–1383. 13. Lauer AK, Smith JR, Robertson JE, Rosenbaum JT. Vitreous hemorrhage is a common complication of pediatric pars planitis. Ophthalmology 2002;109(1):95–98. 14. Simon J, Sood S, Yoon MK, et al. Vitrectomy for dense vitreous hemorrhage in infancy. J Pediatr Ophthalmol Strabismus 2005;42(1):18–22. 15. Green MA, Lieberman G, Milroy CM, Parsons MA. Ocular and cerebral trauma in non-accidental injury in infancy: underlying mechanisms and implications for paediatric practice. Br J Ophthalmol 1996;80(4): 282–287.
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1271
Biosketch Aditya Sudhalkar completed his Masters in Ophthalmology from the M & J Western Regional Institute Of Ophthalmology, Ahmedabad, India. He is currently a clinical fellow in vitreoretinal surgery at the LV Prasad Eye Institute, Hyderabad. His areas of interest are surgical retina and endophthalmitis.
1271.e1
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
Biosketch Dr Jay Chhablani completed his clinical vitreo-retina fellowship from Sankara Nethralaya, India and a clinical research fellowship at University of California, San Diego. At present he is serving as Vitreo-retina faculty at L V Prasad Eye Institute, India. His areas of interest are macular disorders and choroidal imaging.
VOL. 156, NO. 6
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1271.e2
V
ITREOUS HEMORRHAGE, BY ITSELF, IS AN INSUFFI-
cient diagnosis and is an indicator of the need for a systematic analysis of its etiology. The diverse etiology can range from purely ocular conditions, such as posterior vitreous detachment, to diseases with an Accepted for publication May 31, 2013. From the L.V. Prasad Eye Institute, Kallam Anji Reddy Campus, Hyderabad, India (A.S., J.C., S.J., A.M.); and L.V. Prasad Eye Institute, GMR Varalakshmi Campus, Visakhapatnam, India (A.P.). Inquiries to Dr Jay Chhablani, L.V. Prasad Eye Institute, Kallam Anji Reddy Campus, L.V. Prasad Marg, Banjara Hills, Hyderabad - 500 034, India; e-mail: [email protected] 0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.05.042
Ó
2013 BY
underlying systemic etiology, such as hypertension, diabetes mellitus, or vasculitis. Concurrent pathology is often the main determinant of the anatomic, visual, and functional outcome in any patient with vitreous hemorrhage. Vitreous hemorrhage, its underlying causes, and their management are well described in adults.1 Literature on vitreous hemorrhage in children is very limited, including mostly case reports and case series.2–9 A large series by Spirn and associates10 has added considerably to our understanding of the factors causing and the determination of the etiology-specific outcomes of vitreous hemorrhage in children. Trauma, expectedly, was found to be the dominant cause of vitreous hemorrhage in children.10 Certain other causes leading to so-called ‘‘spontaneous vitreous hemorrhage,’’ some of which can be life threatening, merit further discussion. These potentially life-threatening conditions include retinoblastoma or hematologic disorders. It is often very difficult to establish a correct diagnosis in children, as history and elucidation of signs and symptoms is often not forthcoming. Considerable regional variation can be expected when analyzing a condition with such varied etiology, given that each place is socioeconomically, ethnically, and culturally diverse. It follows, thus, that appropriate documentation of characteristics that define or are common to vitreous hemorrhage secondary to such conditions is essential, as an aid to establishing the correct diagnosis. We thus aimed to determine the causes, clinical presentation, and anatomic and functional outcomes of ‘‘spontaneous’’ vitreous hemorrhage in the pediatric population as noted at a tertiary eye care center in India.
METHODS A RETROSPECTIVE COMPUTER-ASSISTED DATABASE SEARCH
and chart review of all patients aged less than 18 years with vitreous hemorrhage who presented to L.V. Prasad Eye Institute, Kallam Anji Reddy campus, Hyderabad between January 1, 2002 and May 31, 2012 was conducted. All procedures conformed to the Declaration of Helsinki for research involving human subjects. The Institutional Review Board of L.V. Prasad Eye Institute, Hyderabad, India approved this retrospective review of these patients’ records for the study. Written informed consent for all subjects was taken from their guardians/parents for the diagnostic and treatment procedures conducted. Cases of vitreous hemorrhage secondary to retinopathy of prematurity (ROP) or trauma (surgical or nonsurgical) and those with incomplete records or diagnosis and/or a follow-up of less than 1 month
ELSEVIER INC. ALL
RIGHTS RESERVED.
1267
were excluded. The chart review adhered to previously established guidelines, described in earlier publications.11 Data collected included demographics; laterality; presenting symptoms and signs (if the child was old enough to cooperate for appropriate history taking; if not, information about the onset of symptoms, possible causes, and previously existing conditions, if any, was obtained from parents); the presenting corrected distance visual acuity; details of the examination procedure; additional investigations and systemic examination, if conducted; the cause, if evident; the treatment offered; the duration of follow-up; and the final anatomic and visual outcomes. Visual acuity was measured using the ETDRS (Early Treatment Diabetic Retinopathy Study) chart and the results subsequently converted to logarithm of minimal angle of resolution (logMAR) chart and noted in logMAR units. For children in the age range of 6 months to 4 years, Teller acuity charts were used and similarly converted to logMAR units. A comprehensive ocular examination was performed in all cases, including under short general anesthesia if the child was not cooperative. If media clarity was insufficient to permit adequate fundal examination, or when additional information, such as choroidal thickness or orbital status, was required, an ultrasound B-scan was performed (OPKO OTI Pvt Ltd., Miami Florida, USA). Further investigations were decided based on the clinical impression. Investigations included ocular coherence tomography (Carl Zeiss Meditec Pvt Ltd., Erlangen, Germany); fluorescein angiography (including under anesthesia, if indicated); visual evoked potential; radiologic imaging (such as computerized tomography scans); blood and urine examinations such as complete blood count, erythrocyte sedimentation rate, C reactive protein, Quantiferon tuberculosis (TB) test, serum calcium, serum angiotensin converting enzyme level, rheumatoid arthritis factor, antinuclear antibody, and urine gross and microscopic examination; and systemic and laboratory tests for specific vasculitis. Treatment, both ocular and systemic, was planned on the basis of diagnosis and can be broadly classified as surgical or nonsurgical. Surgical intervention ranged from vitreous biopsy as a diagnostic aid to combined buckling and vitreoretinal procedures for associated vitreoretinal pathology. Nonsurgical therapy consisted of observation, topical or systemic medical therapy, and/or laser photocoagulation. Frequent cycloplegic refractions and amblyopia therapy by part-time occlusion was part of the management protocol. Statistical analysis was performed using SPSS (v. 16; SPSS, Inc, Chicago, Illinois, USA), with special emphasis on the aforementioned information. Statistical significance was set at P < .05.
RESULTS A TOTAL OF 619 PEDIATRIC PATIENTS WITH AN ENTERED
diagnosis of vitreous hemorrhage in our medical records 1268
department database were identified. Analysis after exclusion of acute ROP cases (n ¼ 34 patients) and those with an incorrect diagnosis (n ¼ 27 patients) or coding errors (n ¼ 13) was performed on 545 patients. A total of 656 eyes with any vitreous hemorrhage (as the diagnosis) were available for analysis. We identified and included 76 patients (124 eyes; 18.09% of the total number of pediatric patients (619) with vitreous hemorrhage who presented to our institute in the stated period) of nontraumatic and nonsurgical ‘‘spontaneous’’ vitreous hemorrhage and performed the analysis. Six patients had an identified cause of spontaneous vitreous hemorrhage; a follow-up of less than 1 month led to their exclusion. Their baseline characteristics did not differ significantly from those included in the study. The median age of the patients was 12.78 6 4.68 (standard deviation) years (range 4 months to 17 years). There was no significant sex bias (39 female vs 37 male patients). Forty-eight patients presented with bilateral vitreous hemorrhage; Table 1 shows the proportion of patients having a specific cause in different age groups. All patients with an identified cause for the vitreous hemorrhage had local and/or systemic comorbidities, which provided the underlying pathologic mechanism of hemorrhage. The mean follow-up was 28 6 18.38 months. The most common presenting complaint in children aged more than 3 years was diminished vision (96.45%). The most common presenting complaint in children aged more than 3 years was diminished vision (119 of 124 patients, 95.96%). The most common presenting complaint in children aged less than 3 years (as reported by the parents) was behavioral changes (110 of 124 patients, 88.7%), such as not being able to recognize the mother, or a recent-onset lack of interest in surroundings. The mean baseline corrected distance visual acuity in logMAR was 2.25 6 1.11. In addition to vitreous hemorrhage, 76 eyes (61.3%) had sufficient view of the fundus and showed typical features of the underlying disease on fundus examination. A total of 51 eyes required an ultrasonography for evaluation of the posterior segment of the eye. Fifty patients required hematologic, immunologic, or radiologic investigations. Patients were given medical therapy (topical and/or systemic) or underwent laser photocoagulation (36 eyes) and/or surgery (69 eyes). Overall, 69 eyes (55.64%) required some form of surgical intervention, such as scleral buckling, vitrectomy, lensectomy, plaque placement, enucleation, or diagnostic vitreous biopsy. Table 2 shows various strategies for management in different age groups. The mean number of surgeries was 1.89 6 1.45, with a range of 1-4 surgeries. None of the operated cases developed endophthalmitis or secondary glaucoma. A total of 19 eyes underwent surgery for tractional or rhegmatogenous retinal detachment, with 3 eyes having a recurrent retinal detachment at final follow-up. Two of these eyes had familial exudative vitreoretinopathy (FEVR) and 1 eye had retinoschisis.
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
TABLE 1. Causes of Spontaneous Vitreous Hemorrhage in Different Age Groups in Children <1 Year Age
>1-10 Years Age
>10 Years Age
Causes (n ¼ Eyes)
(14 Eyes)
(41 Eyes)
(69 Eyes)
Vasculitis Hematologic disorders Ocular tumors/vascular malformations Familial exudative vitreoretinopathy Retinoschisis Retinal breaks/detachment Diabetic retinopathy Pars planitis Hepatorenal dysfunction Sepsis Unknown etiology
3
9 13 4
34 21 1
3
8
2
-
2 2
3 11
2 -
1
4 -
1 5
2
1
The mean final corrected distance visual acuity was 0.76 6 0.58 logMAR. The difference was statistically significant, in that most patients had improved considerably in terms of visual acuity at final follow-up (P < .001). Two patients with FEVR presented with vitreous hemorrhage and inoperable retinal detachment. Two patients with FEVR, 1 patient with retinoschisis, and 2 with vasculitis had a recurrence of retinal detachment at final followup and refused further intervention. Three patients required enucleation for retinoblastoma. One eye with choroidal hemangioma that had undergone multiple surgeries eventually progressed to phthisis. Eleven patients with vasculitis, 6 with FEVR, and 2 with retinoschisis required some surgical/ablative intervention in the other fellow eye, and regular follow-up possibly prevented a more advanced disease occurrence, including vitreous hemorrhage, in the fellow eye, through the last follow-up. We did not come across any case of vitreous hemorrhage as a late consequence of ROP in the present series.
DISCUSSION TRAUMA REMAINS THE FIRST AMONG ALL CAUSES OF PEDI-
atric vitreous hemorrhage;9 however, as is seen in our series, a number of conditions can result in ‘‘spontaneous’’ vitreous hemorrhage in children. These conditions account for 18.09% of pediatric vitreous hemorrhage as seen at our institute. This is a significant number as a group, and is a collection of various conditions that often only have vitreous hemorrhage as a common presenting sign. Articles addressing the same issue in adults show that spontaneous hemorrhage is very common,1,12 VOL. 156, NO. 6
TABLE 2. Management Strategies for Children of Various Age Groups Presenting With Spontaneous Vitreous Hemorrhage
None/medical Laser photocoagulation Surgical
<1 Year (14 Eyes)
>1-10 Years (41 Eyes)
>10 Years (69 Eyes)
5 (35.8%) 1 (7%) 8 (57.2%)
5 (12.2%) 7 (17.1%) 29 (70.7%)
9 (13.0%) 28 (40.6%) 32 (46.4%)
in contrast with children, in whom it accounts for a minority of the reported cases.10 We propose an algorithm approach for a child presenting with vitreous hemorrhage (Figure). As can be seen from Table 1, the predominant cause of spontaneous vitreous hemorrhage was vasculitis (43 of 124 eyes [34.67%], 30 patients), of which 17 patients had unilateral vitreous hemorrhage and 11 patients had signs suggestive of vasculitis in the other eye. Timely laser photocoagulation and/or oral immunosuppressive therapy helped prevent a similar occurrence in the fellow eye. It also helped us advise all patients as to the need for adequate, regular, and longterm follow-up. The cause of vasculitis could not be determined accurately in 22 patients who were treated with oral steroids and, wherever appropriate, with laser therapy or surgery. The remaining 8 patients had a positive Mantoux test and Quantiferon gold TB test and responded well to antitubercular treatment supplemented with steroids. Recurrences were noted in 4 patients with idiopathic vasculitis and none of the tuberculosis-positive patients. Patients with vasculitis had the best visual outcomes compared with other causes of spontaneous vitreous hemorrhage. Hematologic disorders constituted the next major group of etiologic diagnosis of vitreous hemorrhage (21 patients, 34 of 124 eyes; 27.4%). These included sickle cell anemia (5 patients), leukemia (9 patients), anemia (3 patients), and thrombocytopenia (4 patients). These were managed appropriately in consultation with the treating hematologist. Clinical features of ocular involvement as noted in our study are similar to past reports.10 Five patients had retinal breaks without retinal detachment; 3 presented with a sudden onset of floaters and 2 presented with a recent onset of flashes. Two patients had high myopia. One patient had an associated lattice. Retinal breaks without retinal detachment (5 eyes) were treated with laser photocoagulation alone; retinal detachment (8 eyes) was treated with scleral buckling (5 eyes), vitrectomy (2 eyes), or a combination of both (1 eye). FEVR (13 of 124 eyes; 10.48%) has been reported to cause vitreous hemorrhage.12 It was managed with vitreoretinal surgical intervention (9 eyes) and/or laser therapy (4 eyes). Retinoschisis patients (5 eyes, 4.03%) with vitreous hemorrhage underwent vitreoretinal surgery with endolaser photocoagulation.
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1269
FIGURE. Approach to a child presenting with vitreous hemorrhage. Management of vitreous hemorrhage related to retinopathy of prematurity or trauma has not been discussed in detail in this algorithm. CT scan, computed tomography scan; RD, retinal detachment; ROP, retinopathy of prematurity; FFA, fundus fluorescein angiography; ACE, angiotensin converting enzyme; ANA, antinuclear antibody; CD, cluster of differentiation.
The diagnosis of retinoblastoma was possible from an ultrasound examination in all patients with the disease. Three patients were advised to have an additional computerized axial tomography scan to determine if there was any disease extension and the presence of calcification. Retinoblastoma (5 eyes, 4.03%) was treated with chemotherapy, radiotherapy, plaque therapy, or enucleation, as per the disease stage in a given patient. Patients with vasoproliferative tumors (3 eyes, 2.43%), such as capillary hemangiomas, underwent laser photocoagulation, surgery, or both. The causes of vitreous hemorrhage, as seen in the series by Spirn and associates,10 while similar in some aspects, are quite different from the etiology noted in our series. Whereas uncommon conditions like retinoschisis were found in our series, there were hardly any reported cases of vasculitis by Spirn and associates. Pars planitis is known to cause vitreous hemorrhage in children as well.13 One possible explanation is that immunologic reactions to infectious agents such as tuberculosis were more likely to occur in a developing country, for approximately one fourth of our vasculitis cases had concurrent evidence of tuberculosis. It is possible that the remaining ‘‘idiopathic’’ cases were secondary to an immune reaction to other agents. Spirn and associates reported many cases of vitreous hemorrhage as late sequelae of untreated retinopathy of prematurity; however, we did not come across any such patients in our study. We did not come across spontaneous vitreous hemorrhage attributable to Norrie disease or Coats disease in our retrospective review. These differences could reflect practice 1270
patterns, terminology usage, and referral influences. One finding that stands in stark contrast to Western literature was the lack of any patient’s being diagnosed with battered baby syndrome in the entire series, traumatic or otherwise; we believe this could probably be attributable to underreporting, though the syndrome may be quite rare in our population; we can only speculate that cultural and socioeconomic circumstances are different in different regions, accounting for no cases of battered baby syndrome. This difference also holds true when we consider bilateral vitreous hemorrhage, for which the major cause reported in Western literature is battered baby syndrome.14 In contrast, vasculitis followed by hematologic disease emerged as the major cause of bilateral spontaneous vitreous hemorrhage. Our study shows significantly improved visual outcomes at final follow-up; an important point to be kept in mind while interpreting visual outcomes is that visual acuity assessment in very young children has limitations. When analyzed separately, cases of FEVR and retinoschisis had very poor visual and anatomic outcomes; these conditions can affect the retina early and severely and can hinder proper anatomic and functional visual development. Spontaneous vitreous hemorrhage in the first year of life, especially that which requires intervention, is uncommon, and the reported etiologies9 thus far are quite different from the ones noticed in our series. We also noted in our series that the number of infants with a vitreous hemorrhage of an unknown etiology was more than twice the number of children in the 1-10 year age group and 5 times that in the older age group. This points
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
to the difficulty in eliciting appropriate history and clinical signs and symptoms in very young children. Four of these patients were lost to follow-up prior to appropriate review of investigations. When considering vitreous hemorrhage in infants, a previous study shows that in many cases the exact cause was unknown, similar to what we have seen in our series.14 However unlike past literature, we did not come across any infant with vitreous hemorrhage secondary to hematologic or coagulative disorders.14 Additionally, as already noted, we did not come across even a single case of battered baby syndrome, a not-uncommon cause of vitreous hemorrhage and other injuries in infants.15 We did not notice any signs of preretinal or intracranial hemorrhage in such infants. In our series only 2 neonates were evaluated with vitreous hemorrhage, but the cause remained unknown. In all probability, it was secondary to birth trauma, but verification was not forthcoming and the patients were lost to follow-up after first presentation.
Thus, going by the data in the series as a whole, the importance of bilateral dilated fundus examination, under anesthesia if required, cannot be overemphasized. Appropriate intervention, such as diagnostic vitreous biopsies, proved useful in patients not only in ruling out infection etiology in suspicious cases, but also via histopathologic confirmation of inflammatory disorders, such as tuberculosis. To conclude, spontaneous vitreous hemorrhage in children is a condition with a diverse etiology that requires a comprehensive ocular and systemic examination, often aided by additional investigations, to be able to reach an appropriate diagnosis. Outcomes relate to the underlying cause, but the hemorrhage can be a pointer to more sinister disease, such as malignancy, both ocular and systemic. There are significant differences in the etiology and presentation of pediatric vitreous hemorrhage, as seen in different geographic regions, a point that needs to be noted while formulating differential diagnosis and proceeding with management for the same.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. The authors indicate no funding support. Contributions of authors: conception and design (A.S., J.C.); analysis and interpretation (A.S., J.C., S.J., A.N.M., A.P.); writing the article (A.S., J.C., S.J.); critical revision of the article (A.S., J.C., S.J., A.N.M., A.P.); final approval of the article (A.S., J.C., S.J., A.N.M., A.P.); data collection (A.S.); provision of materials, patients, or resources (J.C.); statistical expertise (A.S.); obtaining funding (none); literature search (A.S., J.C.); and administrative, technical, or logistic support (A.N.M., A.P.).
REFERENCES 1. Spraul CW, Grossniklaus HE. Vitreous hemorrhage. Surv Ophthalmol 1997;42(1):3–39. 2. Ferrone PJ, de Juan E Jr. Vitreous hemorrhage in infants. Arch Ophthalmol 1994;112(9):1185–1189. 3. Braendstrup P. Vitreous haemorrhage in the newborn. A rare type of neonatal intraocular haemorrhage. Acta Ophthalmol (Copenh) 1969;47(3):502–513. 4. Cekic O, Totan Y, Batman C. Traumatic vitreous hemorrhage from a persistent hyaloid artery. J Pediatr Ophthalmol Strabismus 2000;37(2):117–118. 5. Graham EM, Stanford MR, Shilling JS, Sanders MD. Neovascularisation associated with posterior uveitis. Br J Ophthalmol 1987;71(11):826–833. 6. Hutcheson KA, Nguyen AT, Preslan MW, Ellish NJ, Steidl SM. Vitreous hemorrhage in patients with high-risk retinopathy of prematurity. Am J Ophthalmol 2003;136(2):258–263. 7. Marshman WE, Adams GG, Ohri R. Bilateral vitreous hemorrhages in an infant with low fibrinogen levels. J AAPOS 1999; 3(4):255–256. 8. Schulman J, Peyman GA, Jednock N, Larson B. Indications for vitrectomy in congenital retinoschisis. Br J Ophthalmol 1985;69(7):482–486.
VOL. 156, NO. 6
9. Saxena S, Jalali S, Verma L, Pathengay A. Management of vitreous haemorrhage. Indian J Ophthalmol 2003;51(2): 189–196. 10. Spirn MJ, Lynn MJ, Hubbard GB 3rd. Vitreous hemorrhage in children. Ophthalmology 2006;113(5): 848–852. 11. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27(3): 305–308. 12. Dana MR, Werner MS, Viana MA, Shapiro MJ. Spontaneous and traumatic vitreous hemorrhage. Ophthalmology 1993; 100(9):1377–1383. 13. Lauer AK, Smith JR, Robertson JE, Rosenbaum JT. Vitreous hemorrhage is a common complication of pediatric pars planitis. Ophthalmology 2002;109(1):95–98. 14. Simon J, Sood S, Yoon MK, et al. Vitrectomy for dense vitreous hemorrhage in infancy. J Pediatr Ophthalmol Strabismus 2005;42(1):18–22. 15. Green MA, Lieberman G, Milroy CM, Parsons MA. Ocular and cerebral trauma in non-accidental injury in infancy: underlying mechanisms and implications for paediatric practice. Br J Ophthalmol 1996;80(4): 282–287.
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1271
Biosketch Aditya Sudhalkar completed his Masters in Ophthalmology from the M & J Western Regional Institute Of Ophthalmology, Ahmedabad, India. He is currently a clinical fellow in vitreoretinal surgery at the LV Prasad Eye Institute, Hyderabad. His areas of interest are surgical retina and endophthalmitis.
1271.e1
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER 2013
Biosketch Dr Jay Chhablani completed his clinical vitreo-retina fellowship from Sankara Nethralaya, India and a clinical research fellowship at University of California, San Diego. At present he is serving as Vitreo-retina faculty at L V Prasad Eye Institute, India. His areas of interest are macular disorders and choroidal imaging.
VOL. 156, NO. 6
PEDIATRIC SPONTANEOUS VITREOUS HEMORRHAGE
1271.e2