Abstract
Purpose
Determine the ocular complications in school-age children and adolescents surviving at least one year following allogeneic bone marrow transplantation.
Design
Retrospective cohort study
Methods
In our institutional study, 162 (7–18 years old) patients met our inclusion criteria with a mean age of 13.4 years at bone marrow transplantation. Follow up ranged from 13 months to 12 years (mean 4 years, median 3.2 years). Patient charts were screened for cataract formation, dry eye, and other anterior and posterior segment disease.
Results
Cataract formation was noted in 57 patients. Univariate analysis showed that fractionated total body irradiation, race, and use of cytarabine significantly increased the incidence of cataracts forming (P < 0.05). Multivariate analysis of significant variables showed that total body irradiation was a risk factor for cataract formation. Of the 57 (97 eyes) that developed cataracts after bone marrow transplantation, 4 (6 eyes) required cataract surgery. After surgery, all patients had visual acuities of 20/20–20/25.
Of the 162 patients, 51 developed dry eyes. Univariate analysis showed that age at transplantation, steroid use, chronic graft-versus-host disease, fludarabine use, melphalan use, thiotepa use, and receiving no pre-transplant conditioning regimen prior to bone marrow transplant significantly increased the risk of dry eye syndrome (p < 0.05). In multivariate analysis, chronic graft-versus-host disease was a significant risk factor for dry eye syndrome.
Conclusion
Due to the high incidence of cataract formation and dry eye disease in this population, we propose screening exams by a pediatric or general ophthalmologist at least every year.
Table of Contents Statement:
After bone marrow transplantation in 162 patients aged 7–18 years, 57 developed cataracts and 51 developed dry eyes. Younger age at transplantation and cytomegalovirus recipient-positive status was protective against cataract formation, whereas total body irradiation was a risk factor for cataract formation. For dry eye syndrome, chronic graft-versus-host disease was a significant risk factor and cytarabine treatment was protective. Yearly eye exams are recommended in these patients.
Introduction
Allogeneic bone marrow transplantation (BMT) is a widely used treatment for a variety of childhood hematologic malignancies, some solid tumors as well as several nonmalignant diseases. With ever-increasing survival rates these children will require long-term follow up for transplantation-related complications. This long-term follow up may not take place at transplant centers but instead in or near the local communities of these children.
Ocular complications following BMT have been well described in adults,1,2 but few studies have reported these complications in school-age children and adolescents. We have previously reported outcomes in children less than 7 years of age,3 but wished to consider older children. Therefore, we studied a group of children aged from 7 to 18 years who survived at least one year following BMT to determine risk factors for ocular complications following transplantation.
Patients and Methods
A St. Jude Children’s Research Hospital Internal Review Board-approved and Health Insurance Portability and Accountability Act-compliant retrospective cohort study was performed on data from children aged 7 to 18 years who underwent BMT between 1995 and 2009 at St. Jude. Patients were included if they had an ophthalmic examination before BMT and a follow-up examination one year after transplantation. A total of 365 patients received allogeneic BMT during the study period, of whom 162 (44.4%) patients met the inclusion criteria. The remaining 203 (55.6%) patients were excluded from the study as they did not have a pre-BMT evaluation or a follow up one year after BMT. Of those excluded, 158 (43.3%) patients died before their one-year post-BMT examination.
All patients underwent transplantation per treatment protocols used by the Department of Bone Marrow Transplantation and Cellular Therapy at St. Jude. Pre-transplant conditioning regimens were determined by the patients’ specific diagnoses and treatment protocols. Ocular examinations were performed by an ophthalmologist at St. Jude and included measurement of visual acuity, biomicroscopic or penlight examination of the anterior segment, and dilated fundus examination. Schirmer’s testing with topical anesthetic was attempted but many of the younger patients could not cooperate fully and thus was not routinely performed.
Data collected included patient demographics, diagnosis and treatment protocol, age at transplantation, dose of ablative radiation received (total and dose per fraction), site of irradiation, history of irradiation before pre-transplant conditioning regimen, specific pre-transplant conditioning regimen, presence of graft-versus-host disease (GVHD), use of calcineurin inhibitors for both prophylaxis and treatment of GVHD, pre- and post-transplant ophthalmic examinations, post-transplant ocular complications, and treatment outcomes. Ocular examinations were reviewed for adverse events including the development of cataract, dry eye syndrome (DES) and any other ocular disease documented in the ophthalmologist’s notes. DES was diagnosed if there was either the presence of superficial punctate keratopathy (SPK) on slit lamp examination and/or Schirmer’s test results of less than 5 mm.
Summary statistics for patients who did or did not develop cataract or DES were compared by the Pearson’s chi-squared test or Fisher’s exact test for categorical variables and by the two-sample t-test or Wilcoxon rank sum test for quantitative variables, depending on the normality test of data. Cumulative incidence of cataract or DES was estimated by the Kalbfleisch-Prentice method4,5 and compared by the Gray’s test. When estimating cumulative incidence, deaths due to any cause were considered as competing events. The Fine and Gray regression model was used to evaluate associations between cumulative incidence and all other covariates. A hazard ratio and the corresponding 95% confidence interval were calculated by the Fine and Gray model. Covariates with P values < 0.15 in univariate analyses were assessed in the respective multivariate Fine and Gray regression analyses. All P values reported are two-sided and considered significant if less than 0.05. Statistical analyses were performed with SAS version 9.4 and R version 3.2.5.
Results
Of 162 patients who met the inclusion criteria, 67 (41%) were female; 110 (68%) were white, 28 (17%) were black, and 24 (15%) were of other ethnic origins. The follow-up period ranged from 1 to 12 years (mean 4 years). Median follow up was 3.2 years (38 months). Thirteen patients died during the follow-up period from complications of their disease. Table 1 presents diagnoses and treatment data for all patients meeting inclusion criteria. The mean age of patients at the time of BMT was 13.4 years, and mean age of patients who developed cataracts was 12.2 years. The mean age at diagnosis for patients receiving BMT and for those who developed cataracts were 10.2 and 9.3 years, respectively.
Table 1.
Patient Characteristics and Development of Cataract and/or Dry Eye Syndrome
| Variable | All | Cataract | No Cataract | Pa | Dry Eye | No Dry Eye | P |
|---|---|---|---|---|---|---|---|
| Age at Transplant (yrs) | |||||||
| 0.0023 | 0.0067 | ||||||
| Mean (SDb) | 13.41 (3.74) | 14.07 (3.61) | 12.21 (3.7) | 12.93 (3.68) | 14.46 (3.67) | ||
| Median (range) | 13.57 (7.06–25.16) | 14.4 (7.38–25.16) | 12.52 (7.06–21.17) | 12.94 (7.06–25.16) | 15.21 (7.17–21.07) | ||
| Age at Cancer Diagnosis (yrs) | |||||||
| 0.034 | 0.42 | ||||||
| Mean (SD) | 10.24 (5.45) | 10.77 (5.75) | 9.25 (4.75) | 10.12 (5.2) | 10.48 (6.02) | ||
| Median (range) | 11.8 (0–18.8) | 12.82 (0–18.8) | 9 (0–18.44) | 10.66 (0–18.8) | 12.63 (0–18.14) | ||
| Total Body Irradiation Dosec (cGy) | |||||||
| <0.001 | 0.29 | ||||||
| Mean (SD) | 782.72 (606.96) | 600 (615.19) | 1119.3 (422.34) | 807.21 (605.99) | 729.41 (611.65) | ||
| Median (range) | 1200 (0–1400) | 200 (0–1400) | 1200 (0–1400) | 1200 (0–1400) | 1200 (0–1400) | ||
| Head/Neck Radiation Dose (cGy) | |||||||
| 0.097 | 0.95 | ||||||
| Mean (SD) | 137.96 (634.83) | 110.86 (656.94) | 187.89 (594.36) | 115.41 (467.17) | 187.06 (902.21) | ||
| Median (range) | 0 (0–5940) | 0 (0–5940) | 0 (0–2550) | 0 (0–2550) | 0 (0–5940) | ||
| Non-Head/Neck Radiation Dose (cGy) | |||||||
| 0.71 | 0.82 | ||||||
| Mean (SD) | 130.56 (336.6) | 119.05 (304.48) | 151.75 (390.97) | 132.88 (349.84) | 125.49 (309.09) | ||
| Median (range) | 0 (0–2100) | 0 (0–2100) | 0 (0–1950) | 0 (0–2100) | 0 (0–1500) | ||
| Gender (number, %) | |||||||
| 0.41 | 0.61 | ||||||
| Female | 67 (41.36) | 46 (43.81) | 21 (36.84) | 44 (39.64) | 23 (45.1) | ||
| Male | 95 (58.64) | 59 (56.19) | 36 (63.16) | 67(60.36) | 28 (54.9) | ||
| Race (number, %) | |||||||
| 0.0049 | 0.49 | ||||||
| Black | 28 (17.28) | 25 (23.81) | 3 (5.26) | 19 (17.12) | 9 (17.65) | ||
| Other | 24 (14.81) | 16 (15.24) | 8 (14.04) | 14 (12.61) | 10 (19.61) | ||
| White | 110 (67.9) | 64 (60.95) | 46 (80.7) | 78 (70.27) | 32 (62.75) | ||
| Diagnosis (number, %) | |||||||
| <0.001 | 1 | ||||||
| Hematologic malignancy | 121 (74.69) | 68 (64.76) | 53 (92.98) | 83 (74.77) | 38 (74.51) | ||
| Nonmalignant disease | 41 (25.31) | 37 (35.24) | 4 (7.02) | 28 (25.23) | 13 (25.49) | ||
| Donor Type (number, %) | |||||||
| 0.053 | 0.16 | ||||||
| Others | 27 (16.67) | 21 (20) | 6 (10.53) | 16 (14.41) | 11 (21.57) | ||
| Sibling | 67 (41.36) | 47 (44.76) | 20 (35.09) | 43 (38.74) | 24 (47.06) | ||
| Unrelated | 68 (41.98) | 37 (35.24) | 31 (54.39) | 52 (46.85) | 16 (31.37) | ||
| Survival (number, %) | |||||||
| 1 | 0.55 | ||||||
| Alive | 149 (91.98) | 96 (91.43) | 53 (92.98) | 103 (92.79) | 46 (90.2) | ||
| Expired | 13 (8.02) | 9 (8.57) | 4 (7.02) | 8 (7.21) | 5 (9.8) | ||
| Product Type (number, %) | |||||||
| 0.15 | 0.11 | ||||||
| Apheresis | 31 (19.14) | 24 (22.86) | 7 (12.28) | 17 (15.32) | 14 (27.45) | ||
| Cord blood | 2 (1.23) | 2 (1.9) | 0 (0) | 1 (0.9) | 1 (1.96) | ||
| Marrow | 129 (79.63) | 79 (75.24) | 50 (87.72) | 93 (83.78) | 36 (70.59) | ||
| Cytomegalovirus Status of Recipient (number, %) | |||||||
| 0.0082 | 0.74 | ||||||
| Negative | 81 (50) | 44 (41.9) | 37 (64.91) | 57 (51.35) | 24 (47.06) | ||
| Positive | 81 (50) | 61 (58.1) | 20 (35.09) | 54 (48.65) | 27 (52.94) | ||
| Steroid Use (number, %) | |||||||
| 0.048 | 0.026 | ||||||
| No | 115 (70.99) | 69 (65.71) | 46 (80.7) | 85 (76.58) | 30 (58.82) | ||
| Yes | 47 (29.01) | 36 (34.29) | 11 (19.3) | 26 (23.42) | 21 (41.18) | ||
| Calcineurin Inhibitor Use (number, %) | |||||||
| 0.63 | 26 (23.42) | 21 (41.18) | 0.61 | ||||
| No | 21 (12.96) | 15 (14.29) | 6 (10.53) | 16 (14.41) | 5 (9.8) | ||
| Yes | 141 (87.04) | 90 (85.71) | 51 (89.47) | 95 (85.59) | 46 (90.2) | ||
| Presence of Acute Graft-versus-host Disease (number, %) | |||||||
| 0.41 | 0.4 | ||||||
| No | 91 (56.17) | 56 (53.33) | 35 (61.4) | 65 (58.56) | 26 (50.98) | ||
| Yes | 71 (43.83) | 49 (46.67) | 22 (38.6) | 46 (41.44) | 25 (49.02) | ||
| Presence of Chronic Graft-versus-host Disease (number, %) | |||||||
| 0.33 | 0.00046 | ||||||
| No | 125 (77.16) | 78 (74.29) | 47 (82.46) | 95 (85.59) | 30 (58.82) | ||
| Yes | 37 (22.84) | 27 (25.71) | 10 (17.54) | 16 (14.41) | 21 (41.18) | ||
| Use of Total Body Irradiation (number, %) | |||||||
| <0.001 | 0.6 | ||||||
| No | 58 (35.8) | 52 (49.52) | 6 (10.53) | 38 (34.23) | 20 (39.22) | ||
| Yes | 104 (64.2) | 53 (50.48) | 51 (89.47) | 73 (65.77) | 31 (60.78) | ||
| Busulfan Use (number, %) | |||||||
| 0.3 | 0.59 | ||||||
| No | 144 (88.89) | 91 (86.67) | 53 (92.98) | 100 (90.09) | 44 (86.27) | ||
| Yes | 18 (11.11) | 14 (13.33) | 4(7.02) | 11 (9.91) | 7 (13.73) | ||
| Cyclophosphamide Use (number, %) | |||||||
| 0.039 | 0.16 | ||||||
| No | 25 (15.43) | 21 (20) | 4 (7.02) | 14 (12.61) | 11 (21.57) | ||
| Yes | 137 (84.57) | 84 (80) | 53 (92.98) | 97 (87.39) | 40 (78.43) | ||
| Cytarabine Use (number, %) | |||||||
| <0.001 | 0.46 | ||||||
| No | 113 (69.75) | 87 (82.86) | 26 (45.61) | 75 (67.57) | 38 (74.51) | ||
| Yes | 49 (30.25) | 18 (17.14) | 31 (54.39) | 36 (32.43) | 13 (25.49) | ||
| Fludarabine Use (number, %) | |||||||
| 0.0037 | 0.12 | ||||||
| No | 135 (83.33) | 81 (77.14) | 54 (94.74) | 96 (86.49) | 39 (76.47) | ||
| Yes | 27 (16.67) | 24 (22.86) | 3 (5.26) | 15 (13.51) | 12 (23.53) | ||
| Melphalan Use (number, %) | |||||||
| 0.02 | 0.12 | ||||||
| No | 143 (88.27) | 88 (83.81) | 55 (96.49) | 101(90.99) | 42 (82.35) | ||
| Yes | 19 (11.73) | 17 (16.19) | 2 (3.51) | 10(9.01) | 9 (17.65) | ||
| Thiotepa Use (number, %) | |||||||
| 0.091 | 0.49 | ||||||
| No | 102 (62.96) | 61 (58.1) | 41 (71.93) | 72 (64.86) | 30 (58.82) | ||
| Yes | 60 (37.04) | 44 (41.9) | 16 (28.07) | 39 (35.14) | 21 (41.18) | ||
| Conditioning Treatment (number, %) | |||||||
| <0.001 | 0.72 | ||||||
| Chemotherapy/radiation | 104 (64.2) | 53 (50.48) | 51 (89.47) | 73 (65.77) | 31 (60.78) | ||
| None | 14 (8.64) | 13 (12.38) | 1 (1.75) | 10 (9.01) | 4 (7.84) | ||
| Chemotherapy only | 44(27.16) | 39 (37.14) | 5 (8.77) | 28 (25.23) | 16 (31.37) | ||
Abbreviations: SD, standard deviation.
Statistically significant values are in bold.
A value of 0 was given for those patients who did not receive radiation.
The most common indications for BMT were acute lymphocytic leukemia (34 patients, 21%), chronic myeloid leukemia (21 patients, 13%), and aplastic anemia (17 patients, 10%). Other indications for BMT are listed on Table 2.
Table 2.
Other Indications for Bone Marrow Transplantation
| • Erythroleukemia |
| • Mixed phenotype acute leukemia |
| • Acute monocytic leukemia |
| • Acute myelocytic leukemia |
| • Acute myelomonocytic leukemia |
| • Acute promyelocytic leukemia |
| • Aplastic anemia |
| • Chronic myelocytic leukemia |
| • Duncan syndrome (X-linked lymphoproliferative disorder) |
| • Hodgkin lymphoma |
| • Hyperimmunoglobulin M syndrome |
| • Non-Hodgkin lymphoma |
| • Myelodysplastic syndrome |
| • Paroxysmal nocturnal hemoglobinuria |
| • Sickle cell disease |
| • Hb S/beta thalassemia |
| • T-cell large granular lymphocytic leukemia |
| • Wiskott-Aldrich syndrome |
Of the 162 patients included in our study, 57 (35%, 97 eyes) developed cataracts after BMT. Of these, four patients (2%, six eyes) required cataract surgery. Following surgery, all patients had visual acuities of 20/25 or better. Univariate analysis of these patients’ outcomes showed that fractionated total body irradiation (TBI) dose, race, TBI, and use of cytarabine significantly increased the incidence of cataracts forming (Table 3). In contrast, a significantly lower incidence of cataracts was associated with older age at diagnosis of cancer and transplantation, non-hematologic malignancy, cytomegalovirus (CMV) recipient-positive status, fludarabine use, melphalan use, and all pre-transplant conditioning regimens that did not include radiation. Multivariate analysis of significant variables showed that younger age at transplantation and CMV recipient-positive status reduced the risk of cataract formation, whereas TBI was a risk factor for cataract formation.
Table 3.
Variables Affecting Cataract and/or Dry Eye Syndrome Development
| Cataract | Dry Eye | |||||
|---|---|---|---|---|---|---|
| Variable | HRa | 95% CI | Pb | HR | 95% CI | P |
| Age at Transplantation | 0.9 | 0.8361–0.9654 | 0.0035 | 1.2 | 1.074–1.267 | 0.00026 |
| Age at Cancer Diagnosis | 0.96 | 0.9185–0.9992 | 0.046 | 1 | 0.9803–1.106 | 0.19 |
| Dose of Total Body Irradiation | 1 | 1.001–1.002 | <0.001 | 1 | 0.9989–0.9998 | 0.0029 |
| Dose of Head/Neck Radiation | 1 | 0.9998–1.001 | 0.33 | 1 | 0.9996–1.001 | 0.51 |
| Dose of Non-Head/Neck Radiation | 1 | 0.9996–1.001 | 0.28 | 1 | 0.9994–1.001 | 0.68 |
| Gender: Male | 1.3 | 0.7797–2.257 | 0.3 | 0.94 | 0.5491–1.609 | 0.82 |
| Race: Other | 3.5 | 0.9293–13.1 | 0.064 | 0.96 | 0.4574–2.009 | 0.91 |
| Race: White | 4.9 | 1.556–15.43 | 0.0066 | 0.75 | 0.3716–1.507 | 0.42 |
| Nonmalignant Disease | 0.19 | 0.06523–0.5448 | 0.0021 | 1.3 | 0.6975–2.325 | 0.43 |
| Sibling Donor | 1.2 | 0.492–3.044 | 0.66 | 0.52 | 0.2486–1.097 | 0.086 |
| Unrelated Donor | 2.3 | 0.9423–5.406 | 0.068 | 0.34 | 0.1492–0.789 | 0.012 |
| Survival: Deceased | 0.95 | 0.3017–2.998 | 0.93 | 1.2 | 0.3684–3.941 | 0.76 |
| Product Type: Cord blood | 7.00E-05 | 1.45E-05–0.000337 | 0 | 0.93 | 0.08214–10.61 | 0.96 |
| Product Type: Marrow | 1.8 | 0.796–3.881 | 0.16 | 0.35 | 0.1967–0.6276 | 0.00041 |
| Cytomegalovirus-Positive Recipient | 0.45 | 0.2632–0.7828 | 0.0045 | 1.1 | 0.6348–1.906 | 0.73 |
| Steroid Use | 0.61 | 0.3065–1.215 | 0.16 | 3.5 | 1.974–6.132 | <0.001 |
| Calcineurin Inhibitor Use | 1.2 | 0.4844–2.78 | 0.74 | 1.3 | 0.5418–3.256 | 0.53 |
| Acute Graft-versus-Host Disease | 0.83 | 0.4841–1.412 | 0.49 | 1.6 | 0.9487–2.826 | 0.077 |
| Chronic Graft-versus-Host Disease | 0.72 | 0.3537–1.469 | 0.37 | 3.5 | 2.045–5.967 | <0.001 |
| Total Body Irradiation | 5.6 | 2.302–13.41 | 0.00014 | 0.48 | 0.2723–0.8371 | 0.0099 |
| Busulfan Use | 0.6 | 0.1993–1.797 | 0.36 | 1.2 | 0.6129–2.416 | 0.57 |
| Cyclophosphamide Use | 2.5 | 0.8398–7.18 | 0.1 | 0.28 | 0.1388–0.5851 | 0.00063 |
| Cytarabine Use | 3 | 1.812–5.045 | <0.001 | 0.25 | 0.1346–0.4694 | <0.001 |
| Fludarabine Use | 0.26 | 0.07598–0.8755 | 0.03 | 3.4 | 1.777–6.539 | 0.00023 |
| Melphalan Use | 0.26 | 0.05969–1.148 | 0.076 | 7.2 | 3.526–14.58 | <0.001 |
| Thiotepa Use | 0.68 | 0.3813–1.206 | 0.19 | 3 | 1.681–5.477 | 0.00023 |
| Conditioning Treatment: Chemotherapy only | 0.1 | 0.01379–0.7778 | 0.028 | 0.99 | 0.3716–2.643 | 0.99 |
| Conditioning Treatment: None | 0.21 | 0.08074–0.5519 | 0.0015 | 2.9 | 1.573–5.468 | 0.00071 |
Abbreviations: CI, confidence interval; HR, hazard ratio
Statistically significant values are in bold.
Use of TBI was linked to many variables in this analysis, due to the specific treatment protocols that require TBI. As this correlation had to be considered statistically, univariate analysis was performed only on data from the 104 patients who received TBI. In the subset of patients who received TBI, cytarabine use was associated with an increased incidence of cataracts while older age at diagnosis and transplantation and CMV-positive status of the recipient was associated with a lower incidence of cataracts. Patients who received busulfan and melphalan were excluded from this additional analysis, because none of our patients received either of these agents combined with TBI.
Of the 162 patients, 51 (31%) developed DES. Univariate analysis showed that age at transplantation, corticosteroid use, chronic GVHD, fludarabine use, melphalan use, thiotepa use, and receiving no pre-transplant conditioning regimen significantly increased the risk of DES. On the other hand, dose of TBI, unrelated donor, bone marrow product, TBI, cyclophosphamide use, and cytarabine use significantly decreased the risk of developing DES (Table 3). For DES, chronic GVHD was a significant risk factor and cytarabine use reduced the risk. Table 4 shows the results of the multivariate analysis.
Table 4.
Multivariate Analysis of Significant Variables in Cataract Formation and Dry Eye Syndrome
| Variable | Hazard Ratio | 95% CI | P |
|---|---|---|---|
| Cataract | |||
| Age at transplantation | 0.91 | 0.854–0.9696 | 0.0036 |
| Cytomegalovirus-positive recipient | 0.53 | 0.3049–0.921 | 0.024 |
| Total Body Irradiation | 5.3 | 2.172–13.14 | 0.00026 |
| Dry Eye | |||
| Chronic Graph-versus-Host Disease | 2.8 | 1.596–4.957 | 0.00035 |
| Cytarabine use | 0.3 | 0.1605–0.549 | 0.00011 |
Abbreviations: CI, confidence interval
Other anterior segment complications included development of herpes zoster ophthalmicus in six patients and ocular GVHD in two patients. Reliable statistical analysis of anterior segment complications was not possible due to the insufficient number of patients.
Posterior segment complications included infectious retinitis in six children: two each due to CMV, Candida sp., and Aspergillus sp. Both patients in the Aspergillus group had extensive visual loss. One eye developed endophthalmitis and had to be enucleated. Vision was hand motions to no light perception in the remaining three eyes and both patients died. All patients in the CMV and Candida groups retained 20/30 vision or better, but only one patient in each group survived. Also, seven patients developed papilledema, and two were diagnosed with pseudotumor cerebri, which resolved with acetazolamide. The remaining five patients had abnormal venous drainage or venous sinus thrombosis. All retained 20/30 or better vision, but two died. Additionally, seven patients had retinal hemorrhages (all survived and retained 20/30 or better vision), four had choroidal scars (all retained 20/40 vision or better, but one died), and six had BMT retinopathy. Vision in the BMT retinopathy group ranged from 20/20 to count fingers, but only half of the patients in this group survived. Reliable statistical analysis for posterior segment complications was not possible due to the insufficient number of patients.
Discussion
We present results from a large group of school-age children and adolescents who developed ocular complications after undergoing allogeneic BMT. We searched PubMed using variations of “ocular complications and bone marrow transplant” for English-language publications from 1960 to the present. To our knowledge, this is the first study to exclusively analyze this age group of patients following BMT. As BMT becomes increasingly tailored to individual patients, it is essential to identify possible unique risk factors in this already vulnerable population. Given that pediatric and adult cancers require different treatments and intensities, ocular studies in pediatric and adult patients should also be viewed as separate entities. As expected, the main complications in our children were the development of cataracts and DES. Previous studies of ocular complications in children and adults after BMT reported an 8.5–83.3% incidence of cataracts.3,6–9 The incidence of DES in previous reports range between 17% and 44% in adults and as low as 6–12.5% in children.8,10
In our study, among the children developing cataracts, 7% (4 of 57) underwent surgical intervention. This percentage is much lower than previous studies where cataract surgery was required in up to 21% of patients.3,10,11 In our previous paper describing outcomes in younger children, we had longer follow up which may explain the difference between the two age groups.3 Consistent with previous studies,10,12 visual outcomes were favorable in our study, with all children achieving visual acuities of 20/20–20/25. Since at least 1 in 20 children developing cataracts in our study required surgery, we recommend that these patients be closely followed.
Ionizing radiation is the most recognized risk factor for the development of cataract,4 and we have further confirmed that fractionated TBI is a significant risk factor for cataract development. In our study, 51 of 57 patients who developed cataracts received TBI. Fractionated TBI may reduce not only the incidence of cataracts but also their clinical severity.6 Deeg et al. showed that 50% of patients with cataract who received unfractionated TBI required cataract surgery compared to only 20% of patients who received fractionated TBI.11 In our study, only 7% of patients required cataract surgery after fractionated TBI, which likely confirms the results of Horowitz, et al.13 They postulated that fractionated TBI is more effective at delaying the progression of cataracts than is unfractionated TBI, but it may not change the prevalence overall. In our previous paper we also described the increasing incidence of cataract development with longer survival. We reported the incidence of cataract formation was 58% by 14 years in patients undergoing BMT when less than 7 years old.3 Given these results and those from Horwitz, et al.13 and our previous studies, we recommend that the follow-up period be extended, and patients be monitored for several years, as cataractogenesis is delayed in patients who undergo fractionated TBI.3,13
As our study looked at multiple variables occurring over the course of a child’s treatment, it was essential to establish independent associations with the risk of cataract development. To this end, we looked further into the association of TBI with multiple variables, since TBI is the most known risk factor for cataract development. Our univariate analysis of cataractogenesis revealed multiple drugs and conditions as significant risk factors. However, when we studied only patients who received TBI, the effect of all chemotherapeutic agents was not statistically significant. This demonstrates that most chemotherapeutic agents that are given in conjunction with TBI are not independent risk factors. Univariate analysis of TBI patients revealed that fractionated dose, age at transplantation, recipient CMV status, and cytarabine use remained as correlatives to cataract development. Thus, cytarabine appears to confer an increased risk that is separate from TBI. Cytarabine is known to cause a chemical conjunctivitis and keratitis that necessitates topical steroid use.14 Since steroids are a known risk factor for cataract development, we postulate that the use of steroid eye drops during cytarabine administration increases the rates of cataract development. We have no explanation why CMV-positive recipient status would increase the risk of cataract development. We recommend further study in the future.
In a study of DES by Suh, et al., children (mean age 8.4 years) who underwent BMT had a 12.5% incidence of DES,10 which is lower than that reported in adult studies (17–44%).2,4,5,15,16 We also reported3 that younger children (mean age at transplantation 3.2 years) had an incidence of DES of 14.3%. Suh, et al., postulated that the prevalence they reported was lower because examination of dry eyes in children is often less comprehensive than in adults, and making a diagnosis is more difficult in preverbal children.10 We agree with that rationale as many of the patients in our previous paper were unable to cooperate with slit lamp examination or Schirmer’s testing.3 Our study of school-age children (average age at transplantation was 13.4 years) benefited from the relative maturity of the patients, and enabled reliable diagnosis and assessment of the severity of DES. The prevalence of DES in our study was 31.5%, which is five times more than DeMarco8 and more than twice that reported by Suh, et al.,10 and in our younger cohort.3 It does, however, agree with the prevalence in older populations, such as those studied by Lin et al.2
In patients who have undergone BMT, the known major causes of DES are direct damage to the lacrimal glands and conjunctiva by ionizing radiation, as well as acute and chronic GVHD-induced damage.10 We found risk factors for DES included older age at transplantation, systemic corticosteroid use, chronic GVHD, fludarabine use, thiotepa use, and no pre-transplant conditioning regimen. Apart from GVHD, the other remaining risk factors are new indicators that warrant further examination. Furthermore, older age at transplantation increases the odds of developing DES; thus, lower rates of dry eye in children may be due to younger age and not more difficulty with diagnosis.
Our study also revealed many variables that decreased the prevalence of DES, including unrelated donor type, marrow blood source, TBI, cyclophosphamide use, and cytarabine use. Total body irradiation was significantly protective against DES. However, multivariate analysis revealed only chronic GVHD as an independent risk factor and cytarabine use as protective. We postulate that steroid use to prevent corneal and conjunctival toxicity decreases the risk of DES in patients with a history of cytarabine use. It is well established that decreasing inflammation can improve DES and topical steroids decrease ocular inflammation.18 Perhaps the use of steroids early in the process of DES aborts the inflammatory cascade that leads to DES.
As the prevalence of DES in adolescents and school-age children approaches that of adults, we recommend examinations at least every year to evaluate for DES symptoms and SPK to prevent related corneal complications. Also, we need to inform parents and caregivers that almost a third of patients older than 7 years who receive BMT develop significant symptoms of DES.
One of the rarest but most visually significant occurrences is posterior segment complications. Coskuncan, et al., reported a 2% prevalence of infectious retinitis and endophthalmitis after BMT. Almost 4% of our patients developed infectious retinitis or endophthalmitis, which were associated with significant morbidity and mortality. Although relatively few posterior segment complications were noted in our study, we recommend baseline ophthalmic examinations before BMT to examine pre-existing pathology and serial examinations when patients are immunosuppressed, especially if opportunistic infections are found in other systems, as ocular manifestations of such infections can be devastating.
It should also be noted that the patients in this study received BMTs over a 14-year period. Study protocols and treatment plans have changed as our understanding has improved. This complicates the analysis as no standard pre-transplant conditioning regimen protocol was used.
Conclusion
Ocular complications following allogeneic BMT are well known. However, studies of the ocular complications in children are often difficult to conduct and interpret due to a number of factors including but not limited to variable cooperation of younger children, different treatment protocols based upon diagnosis, missed follow-up examinations due to unanticipated intensive care unit admissions, as well as the heterogenous populations described in previous studies. In this study, we describe a cohort of school-age and adolescent children with ocular complications who survived at least one year following allogeneic BMT. In these patients aged 7–18 years, the most common complications were cataract and DES. Factors associated with the development of cataract were older age at diagnosis and TBI. A reduced risk for cataract was noted in CMV-positive recipients. In addition, DES occurred in nearly one third of all patients in this age group. Development of chronic GVHD was associated with an increased risk of developing DES.
An ever-increasing number of long-term survivors of childhood allogeneic BMT will require ongoing care and follow up in or near their communities. Children with risk factors for development of cataracts and or DES should receive long-term close follow up. Furthermore, yearly ophthalmologic exams are also recommended for all survivors of childhood allogeneic BMT.
Acknowledgments
A. Funding Support
This work was supported by an unrestricted grant to the Department of Ophthalmology at the University of Tennessee from Research to Prevent Blindness, Inc., New York, New York. Further grant support was provided by the National Institutes of Health (P30 CA 21765) and by the American Lebanese Syrian Associated Charities.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
B. Financial Disclosures
My coauthors and I have no proprietary or commercial interest in any materials discussed in this article.
References
- 1.Najima Y, Kakihana K, Ohashi K, et al. Incidence, risk factors, and clinical outcomes of cataracts following hematopoietic stem cell transplantation. Am J Hematol 2011; 86(6):508–510. [DOI] [PubMed] [Google Scholar]
- 2.Lin X, Cavanagh HD. Ocular manifestations of graft-versus-host disease: 10 years’ experience. Clin Ophthalmol 2015; 9:1209–1213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hoehn ME, Calderwood J, Gannon E, et al. Ocular complications in a young pediatric population following bone marrow transplantation. J AAPOS 2018; 22(2):102–106e1. https://doi:10.1016/j.jaapos.2017.10.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jack MK, Hicks JD. Ocular complications in high-dose chemoradiotherapy and marrow transplantation. Ann Ophthalmol 1981; 13(6):709–711. [PubMed] [Google Scholar]
- 5.Coskuncan NM, Jabs DA, Dunn JP, et al. The eye in bone marrow transplantation. VI. Retinal complications. Arch Ophthalmol 1994; 112(3):372–379. [DOI] [PubMed] [Google Scholar]
- 6.Livesey SJ, Holmes JA, Whittaker JA. Ocular complications of bone marrow transplantation. Eye (Lond) 1989; 3(Pt. 3):271–276. [DOI] [PubMed] [Google Scholar]
- 7.Belkacemi Y, Ozsahin M, Pene F, et al. Cataractogenesis after total body irradiation. Int J Radiat Oncol Biol Phys 1996; 35(1):53–60. [DOI] [PubMed] [Google Scholar]
- 8.De Marco R, Dassio DA, Vittone P. A retrospective study of ocular side effects in children undergoing bone marrow transplantation. Eur J Ophthalmol 1996; 6(4):436–439. [DOI] [PubMed] [Google Scholar]
- 9.Tichelli A. Late ocular complications after bone marrow transplantation. Nouv Rev Fr Hematol 1994; 36 Suppl 1:S79–82. [PubMed] [Google Scholar]
- 10.Suh DW, Ruttum MS, Stuckenschneider BJ, Mieler WF, Kivlin JD. Ocular findings after bone marrow transplantation in a pediatric population. Ophthalmology 1999; 106(8):1564–1570. [DOI] [PubMed] [Google Scholar]
- 11.Deeg HJ, Flournoy N, Sullivan KM, et al. Cataracts after total body irradiation and marrow transplantation: a sparing effect of dose fractionation. Int J Radiat Oncol Biol Phys 1984; 10(7):957–964. [DOI] [PubMed] [Google Scholar]
- 12.Calissendorff BM, Bolme P. Cataract development and outcome of surgery in bone marrow transplanted children. Br J Ophthalmol 1993; 77(1):36–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Horwitz M, Auquier P, Barlogis V, et al. Incidence and risk factors for cataract after haematopoietic stem cell transplantation for childhood leukaemia: an LEA study. Br J Haematol 2015; 168(4):518–525. Https://doi:10.1111/bjh.13148 [DOI] [PubMed] [Google Scholar]
- 14.Lass JH, Lazarus HM, Reed MD, Herzig RH. Topical corticosteroid therapy for corneal toxicity from systemically administered cytarabine. Am J Ophthalmol 1982; 94(5):617–621. [DOI] [PubMed] [Google Scholar]
- 15.Bray LC, Carey PJ, Proctor SJ, Evans RG, Hamilton PJ. Ocular complications of bone marrow transplantation. Br J Ophthalmol 1991; 75(10):611–614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hirst LW, Jabs DA, Tutschka PJ, Green WR, Santos GW. The eye in bone marrow transplantation. I. Clinical study. Arch Ophthalmol 1983; 101(4):580–584. [DOI] [PubMed] [Google Scholar]
- 17.Jabs DA, Wingard J, Green WR, et al. The eye in bone marrow transplantation. III. Conjunctival graft-vs-host disease. Arch Ophthalmol 1989; 107(9):1343–1348. [DOI] [PubMed] [Google Scholar]
- 18.Jones L, Downie LE, Korb D, et al. TFOS DEWS II: Management and Therapy Report. Ocul Surf 2017;15(3):575–628. https://doi:10.1016/j.jtos.2017.05.006 [DOI] [PubMed] [Google Scholar]