BEVACIZUMAB LEVELS IN BREAST MILK AFTER LONG-TERM INTRAVITREAL INJECTIONS TREVOR J. MCFARLAND, BS,* ANDREW D. RHOADS, BA,* MATTHEW HARTZELL, BS,† GEOFFREY G. EMERSON, MD,‡ ABDHISH R. BHAVSAR, MD,‡ J. TIMOTHY STOUT, MD, PHD, MBA† Purpose: The purpose of this study is to determine whether bevacizumab is detectable in the breast milk of nursing mothers. Methods: Breast milk samples were collected from 2 patients receiving monthly intravitreal bevacizumab injections for choroidal neovascularization over the course of 16 months. Enzyme-linked immunosorbent assay and Western blot analysis was used to determine the levels of bevacizumab in the milk samples. Results: An enzyme-linked immunosorbent assay was developed using antibodies specific to bevacizumab in which the sensitivity threshold was 3 ng/mL. All breast milk samples assayed from the two patients actively undergoing treatment did not have detectable levels of bevacizumab. Samples collected 1.5 hours and 7 hours after an injection and 2 randomly chosen samples were negative by Western blot analysis. Conclusion: A sensitive assay to detect bevacizumab in breast milk samples assayed suggests that intravitreal injections do not result in detectable bevacizumab in breast milk. RETINA 35:1670–1673, 2015
A
however, no statistical analysis of significance or comparison with positive and negative control milk samples were provided. Additionally, no consideration was given to data obtained from the patient’s samples, explaining the gradual increase in vascular endothelial growth factor A over the course of the study. Although it is claimed that the enzyme-linked immunosorbent assay did not detect any free bevacizumab within the patient’s breast milk, there is no indication of the sensitivity of the bevacizumab assay, leading to concerns regarding the detection of the protein at lower levels.1 Furthermore, without the aforementioned controls, there is no accurate indication of the natural variation of vascular endothelial growth factor A within human breast milk. There have been studies indicating that bevacizumab can traverse the blood–retina barrier in a mouse model to the placenta through the Fc receptor of the neonate and to the fetus during gestation in a rat model, which could predispose to serious developmental complications.2–4 Studies have postulated that the relative size of the IgG antibody should prevent transfer of bevacizumab into human breast milk, and thus intravitreal treatment should be safe.5 Several cases of spontaneous loss of pregnancy have been observed in pregnant women who have received intravitreal bevacizumab, but evidence supporting a cause–effect relationship is
growing number of patients are treated with intravitreal bevacizumab for choroidal neovascularization (CNV) and retinal vascular disease. Although the majority of these patients are older than 50 years, this therapy also benefits younger patients, including women of childbearing age. Bevacizumab is a pregnancy Class C drug, and patients are warned to avoid the medication during pregnancy or breastfeeding because of potential harmful effects on the developing fetus or infant. Specifically, bevacizumab is an angiogenesis inhibitor that could impact growth and development. Little is known about the degree to which a dose of intravitreal bevacizumab injected into the mother’s vitreous is transferred into breast milk. Some studies have attempted to accurately measure serum and breast milk levels of bevacizumab after intravitreal injection. One study suggested a 35% drop from 13.3 ng/mL to 8.6 ng/mL in vascular endothelial growth factor A concentration in breast milk 2 weeks after injection;
From the *Casey Eye Institute, Oregon Health and Science University, Portland, Oregon; †Cullen Eye Institute, Baylor College of Medicine; and ‡Retina Center of Minnesota, Minneapolis, Minnesota. Supported by the Clayton Foundation for Research. None of the authors have any conflicting interests to disclose. Reprint requests: J. Timothy Stout, MD, PhD, MBA, Cullen Eye Institute, Baylor College of Medicine, 6565 Fannin NC-205, Houston, TX 77030; e-mail: [email protected]
1670
BEVACIZUMAB LEVELS IN BREAST MILK MCFARLAND ET AL
lacking.6 Thus, there is a need for more comprehensive research regarding the potential transfer of bevacizumab into breast milk. This study sought to more accurately quantitate the levels of bevacizumab, over a lengthy time course and with multiple sampling time points, in two patients receiving frequent intravitreal injections of bevacizumab. This information will aid counseling of lactating mothers who wish to minimize exposure and risk to their breastfeeding infants while they are receiving intravitreal bevacizumab treatment for a vision-threatening condition. Patients A 28-year-old mother presented with 20/60 blurred vision in her right eye. She was treated with intravitreal bevacizumab for CNV secondary to histoplasmosis, and her vision improved to 20/20. Treatment was withheld when she became pregnant with her second child, but the CNV recurred at 7-week gestation, and bevacizumab was resumed at 2-month intervals for the duration of her pregnancy. Treatment continued after a normal healthy delivery, and breast milk samples were obtained 1 day before and 1 week after intravitreal bevacizumab injections over a 6-month period. Visual acuity measured 20/16 3 years after presentation. A 25-year-old woman was treated with bevacizumab injections for subfoveal CNV because of histoplasmosis beginning in 2007. She was treated with a “treat-and-extend” method and opted to discontinue treatment after several injections; after approximately 9 months, there was recurrent CNV and bevacizumab injections were resumed. After several injections, she decided to discontinue treatment because of pregnancy.
1671
She had no recurrence of CNV during the uncomplicated pregnancy or afterwards until 7 months after delivery of her first healthy child. Because of concerns about bevacizumab in a nursing mother, she opted for photodynamic therapy for recurrent CNV. Two months later, she resumed bevacizumab injections. She underwent several injections and decided to discontinue after 6 months as she desired to become pregnant. She developed recurrent CNV 18 months later, 3 months after her second healthy child was born. At that time, bevacizumab injections were given at monthly intervals with some extension to 8-week intervals. Breast milk samples were obtained 1 day before and 1 week after bevacizumab injections for 16 months. Additional samples after several injections were obtained 1 day after bevacizumab injection. Visual acuity was maintained at approximately 20/40 after 5 years of intermittent treatment.
Methods Bevacizumab Enzyme-Linked Immunosorbent Assay Breast milk samples were centrifuged at 4,000 rpm for 5 minutes to separate the fatty components of the milk. A 96-well plate was coated with 2.5 mg/mL of antibody to bevacizumab (Abd Serotec HCA182; Kidlington, United Kingdom) in buffer overnight at 4°C. The plate was washed and blocked with SuperBlock buffer (Pierce, Rockford, IL). One hundred microliter of samples, controls, and standards in duplicate were incubated at 37°C for 90 minutes. A horseradish peroxidaseconjugated antibody to bevacizumab (AbD Serotec HCA184P) was added and incubated at 37°C for 90 minutes. The plate was washed, and substrate was added
Fig. 1. Injection sample and collection timeline. Patient 1 (red) samples were collected the day before injection and 1 week after injection. Patient 2 (blue) samples were collected 1 day before injection, day of injection, and 1 week after injection. One collection was also taken 1.5 hours and 7 hours after injection.
1672 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 8
Fig. 2. Enzyme-linked immunosorbent assay diagram and standard graph. Assay sensitivity limit is 3 ng/mL (absorbance: 450 nm = 0.294). All samples measured below the assay’s limit of detection. Sample absorbance values measured between 0.068 and 0.166 (mean = 0.112). HRP, horseradish peroxidase.
(hydrogen peroxide and TMB [3,39,5,59-tetramethylbenzidine 1:1]). The reaction was terminated after 25 minutes with 25 mL of 2 M hydrogen sulfate per well. The absorbance was read with a plate reader at 450 nm. Bevacizumab Western Blot Assay Western blot analysis was performed to detect the presence of bevacizumab in breast milk from recently treated samples (day before, 1.5 hours after injection, and 7 hours after injection) and 2 randomly treated samples and a sample of breast milk obtained from a nontreated control patient. Standard Western blot protocols were used using 45 mL of each sample. The membrane was incubated and washed at room temperature for 1-hour intervals using the following antibodies: antibevacizumab (Abd Serotec HCA182) containing V5 tag (1:500), mouse anti-V5 tag (Invitrogen, Carlsbad, CA) (1:5,000), and rabbit anti-mouse (Li-Cor, Lincoln, NE) containing an 800-nm IRDye (1:5,000). The blot was visualized using the Li-Cor Odyssey IRDye scanner. Results We developed an assay to detect bevacizumab in breast milk samples obtained from 2 patients receiving injections on a monthly basis over the course of a year (Figure 1). The results of the study demonstrated that the level of detection by an enzyme-linked immunosorbent assay–based system was sensitive to 3 ng/mL
(Figure 2). All samples assayed were negative for the presence of bevacizumab at this level of detection. In addition to enzyme-linked immunosorbent assay testing, Western blot analysis was performed to observe whether samples collected after a recent injection (1.5 and 7 hours) contained low levels of bevacizumab. This analysis revealed no detectable level of bevacizumab in milk samples (Figure 3). Given the fact that this assay is sensitive to microgram levels, the results suggest that the transfer of bevacizumab to breast milk after intravitreal injection does not occur at least in a detectable quantity. Discussion The determination of whether a neutralizing antibody to vascular endothelial growth factor A is present in breast milk is important, as it has been documented that intravitreally delivered bevacizumab is capable of crossing the retinal–blood barrier and can lead to reduced levels of vascular endothelial growth factor in serum.7,8 Infants whom have been treated with bevacizumab exhibit lower vascular endothelial growth factor in their serum.9 The effects of such a compound delivered to a neonate through ingestion or through amniotic fluid are unknown but could have important developmental consequences. However, the mechanisms by which a molecule travels from the vitreous to the breast milk or amniotic fluid likely differ, and we have no evidence of bevacizumab in the patient’s amniotic fluid.
Fig. 3. Bevacizumab Western diagram and blot. Patient 2 samples (May) were taken 1 day before injection (P2.a), 1 hour after injection (P2.b), and 7 hours after injection (P2.c). Patient 1 samples (January) were taken 1 day before injection (P1.a) and 1 week after injection (P1.b). Control samples contained 200 pg/mL of bevacizumab (A.), 2.3 mg/mL of bevacizumab (B.), and the breast milk from an uninjected female (C.). The light and heavy chains (arrow) of bevacizumab can be observed in the 2.3 mg/mL control sample. Protein molecular weight marker (M) measured in kilodaltons.
BEVACIZUMAB LEVELS IN BREAST MILK MCFARLAND ET AL
In addition, children process antibodies from their mothers during infancy, so it is reasonable to assume that they could absorb an anti-vascular endothelial growth factor antibody where one to be ingested through breast milk.10,11 Our study confirms that bevacizumab is not transferred from mother to child through breast milk after many months of intravitreal injections. Therefore, we have no indication of oral ingestion of an anti-vascular endothelial growth factor molecule by infants. Given that a routine intravitreal dose of bevacizumab is 1.25 mg per 0.05 mL and our level of detection is approximately 3 ng/mL, it is reasonable to conclude that the presence of bevacizumab in breast milk after treatment does not occur above a 10 million-fold dilution range. These results can provide useful information for patients and ophthalmologists when recommending treatment for CNV in women who are also breastfeeding infants. Key words: bevacizumab, breast milk, choroidal neovascularization, ELISA, Western blot. References 1. Ehlken C, Martin G, Stahl A, Agostini HT. Reduction of vascular endothelial growth factor A in human breast milk after intravitreal injection of bevacizumab but not ranibizumab. Arch Ophthalmol 2012;130:1226–1227.
1673
2. Kim H, Robinson SB, Csaky KG. FcRn receptor-mediated pharmacokinetics of therapeutic IgG in the eye. Mol Vis 2009;15:2803–2812. 3. Chaparro M, Gisbert JP. Transplacental transfer of immunosuppressants and biologics used for the treatment of inflammatory bowel disease. Curr Pharm Biotechnol 2011;12:765–773. 4. Thorn M, Piche-Nicholas N, Stedman D, et al. Embryo-fetal transfer of bevacizumab (Avastin) in the rat over the course of gestation and the impact of neonatal Fc receptor (FcRn) binding. Birth Defects Res B Dev Reprod Toxicol 2012;95:363–375. 5. Pistilli B, Bellettini G, Giovannetti E, et al. Chemotherapy, targeted agents, antiemetics and growth factors in human milk: How should we counsel cancer patients about breastfeeding? Cancer Treat Rev 2013;39:207–211. 6. Petrou P, Georgalas I, Giavaras G, et al. Early loss of pregnancy after intravitreal bevacizumab injection. Acta Ophthalmol 2010;88:136. 7. Qian J, Lu Q, Tao Y, Jiang YR. Vitreous and plasma concentrations of apelin and vascular endothelial growth factor after intravitreal bevacizumab in eyes with proliferative diabetic retinopathy. Retina 2011;31:161–168. 8. Matsuyama K, Ogata N, Matsuoka M, et al. Plasma levels of vascular endothelial growth factor and pigment epitheliumderived factor before and after intravitreal injection of bevacizumab. Br J Ophthalmol 2010;94:1215–1218. 9. Sato T, Wada K, Arahori H, et al. Serum concentrations of bevacizumab and vascular endothelial growth factor in infants with retinopathy of prematurity. Am J Ophthalmol 2012;153:327–333. 10. Hanson L. Breastfeeding provides passive and likely longlasting active immunity. Ann Allergy Asthma Immunol 1998;81;523–537. 11. Hanson L, Korotkova M, Lundin S, et al. Transfer of immunity from mother to child. Ann N Y Acad Sci 2003;987:199–206.
A
however, no statistical analysis of significance or comparison with positive and negative control milk samples were provided. Additionally, no consideration was given to data obtained from the patient’s samples, explaining the gradual increase in vascular endothelial growth factor A over the course of the study. Although it is claimed that the enzyme-linked immunosorbent assay did not detect any free bevacizumab within the patient’s breast milk, there is no indication of the sensitivity of the bevacizumab assay, leading to concerns regarding the detection of the protein at lower levels.1 Furthermore, without the aforementioned controls, there is no accurate indication of the natural variation of vascular endothelial growth factor A within human breast milk. There have been studies indicating that bevacizumab can traverse the blood–retina barrier in a mouse model to the placenta through the Fc receptor of the neonate and to the fetus during gestation in a rat model, which could predispose to serious developmental complications.2–4 Studies have postulated that the relative size of the IgG antibody should prevent transfer of bevacizumab into human breast milk, and thus intravitreal treatment should be safe.5 Several cases of spontaneous loss of pregnancy have been observed in pregnant women who have received intravitreal bevacizumab, but evidence supporting a cause–effect relationship is
growing number of patients are treated with intravitreal bevacizumab for choroidal neovascularization (CNV) and retinal vascular disease. Although the majority of these patients are older than 50 years, this therapy also benefits younger patients, including women of childbearing age. Bevacizumab is a pregnancy Class C drug, and patients are warned to avoid the medication during pregnancy or breastfeeding because of potential harmful effects on the developing fetus or infant. Specifically, bevacizumab is an angiogenesis inhibitor that could impact growth and development. Little is known about the degree to which a dose of intravitreal bevacizumab injected into the mother’s vitreous is transferred into breast milk. Some studies have attempted to accurately measure serum and breast milk levels of bevacizumab after intravitreal injection. One study suggested a 35% drop from 13.3 ng/mL to 8.6 ng/mL in vascular endothelial growth factor A concentration in breast milk 2 weeks after injection;
From the *Casey Eye Institute, Oregon Health and Science University, Portland, Oregon; †Cullen Eye Institute, Baylor College of Medicine; and ‡Retina Center of Minnesota, Minneapolis, Minnesota. Supported by the Clayton Foundation for Research. None of the authors have any conflicting interests to disclose. Reprint requests: J. Timothy Stout, MD, PhD, MBA, Cullen Eye Institute, Baylor College of Medicine, 6565 Fannin NC-205, Houston, TX 77030; e-mail: [email protected]
1670
BEVACIZUMAB LEVELS IN BREAST MILK MCFARLAND ET AL
lacking.6 Thus, there is a need for more comprehensive research regarding the potential transfer of bevacizumab into breast milk. This study sought to more accurately quantitate the levels of bevacizumab, over a lengthy time course and with multiple sampling time points, in two patients receiving frequent intravitreal injections of bevacizumab. This information will aid counseling of lactating mothers who wish to minimize exposure and risk to their breastfeeding infants while they are receiving intravitreal bevacizumab treatment for a vision-threatening condition. Patients A 28-year-old mother presented with 20/60 blurred vision in her right eye. She was treated with intravitreal bevacizumab for CNV secondary to histoplasmosis, and her vision improved to 20/20. Treatment was withheld when she became pregnant with her second child, but the CNV recurred at 7-week gestation, and bevacizumab was resumed at 2-month intervals for the duration of her pregnancy. Treatment continued after a normal healthy delivery, and breast milk samples were obtained 1 day before and 1 week after intravitreal bevacizumab injections over a 6-month period. Visual acuity measured 20/16 3 years after presentation. A 25-year-old woman was treated with bevacizumab injections for subfoveal CNV because of histoplasmosis beginning in 2007. She was treated with a “treat-and-extend” method and opted to discontinue treatment after several injections; after approximately 9 months, there was recurrent CNV and bevacizumab injections were resumed. After several injections, she decided to discontinue treatment because of pregnancy.
1671
She had no recurrence of CNV during the uncomplicated pregnancy or afterwards until 7 months after delivery of her first healthy child. Because of concerns about bevacizumab in a nursing mother, she opted for photodynamic therapy for recurrent CNV. Two months later, she resumed bevacizumab injections. She underwent several injections and decided to discontinue after 6 months as she desired to become pregnant. She developed recurrent CNV 18 months later, 3 months after her second healthy child was born. At that time, bevacizumab injections were given at monthly intervals with some extension to 8-week intervals. Breast milk samples were obtained 1 day before and 1 week after bevacizumab injections for 16 months. Additional samples after several injections were obtained 1 day after bevacizumab injection. Visual acuity was maintained at approximately 20/40 after 5 years of intermittent treatment.
Methods Bevacizumab Enzyme-Linked Immunosorbent Assay Breast milk samples were centrifuged at 4,000 rpm for 5 minutes to separate the fatty components of the milk. A 96-well plate was coated with 2.5 mg/mL of antibody to bevacizumab (Abd Serotec HCA182; Kidlington, United Kingdom) in buffer overnight at 4°C. The plate was washed and blocked with SuperBlock buffer (Pierce, Rockford, IL). One hundred microliter of samples, controls, and standards in duplicate were incubated at 37°C for 90 minutes. A horseradish peroxidaseconjugated antibody to bevacizumab (AbD Serotec HCA184P) was added and incubated at 37°C for 90 minutes. The plate was washed, and substrate was added
Fig. 1. Injection sample and collection timeline. Patient 1 (red) samples were collected the day before injection and 1 week after injection. Patient 2 (blue) samples were collected 1 day before injection, day of injection, and 1 week after injection. One collection was also taken 1.5 hours and 7 hours after injection.
1672 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 8
Fig. 2. Enzyme-linked immunosorbent assay diagram and standard graph. Assay sensitivity limit is 3 ng/mL (absorbance: 450 nm = 0.294). All samples measured below the assay’s limit of detection. Sample absorbance values measured between 0.068 and 0.166 (mean = 0.112). HRP, horseradish peroxidase.
(hydrogen peroxide and TMB [3,39,5,59-tetramethylbenzidine 1:1]). The reaction was terminated after 25 minutes with 25 mL of 2 M hydrogen sulfate per well. The absorbance was read with a plate reader at 450 nm. Bevacizumab Western Blot Assay Western blot analysis was performed to detect the presence of bevacizumab in breast milk from recently treated samples (day before, 1.5 hours after injection, and 7 hours after injection) and 2 randomly treated samples and a sample of breast milk obtained from a nontreated control patient. Standard Western blot protocols were used using 45 mL of each sample. The membrane was incubated and washed at room temperature for 1-hour intervals using the following antibodies: antibevacizumab (Abd Serotec HCA182) containing V5 tag (1:500), mouse anti-V5 tag (Invitrogen, Carlsbad, CA) (1:5,000), and rabbit anti-mouse (Li-Cor, Lincoln, NE) containing an 800-nm IRDye (1:5,000). The blot was visualized using the Li-Cor Odyssey IRDye scanner. Results We developed an assay to detect bevacizumab in breast milk samples obtained from 2 patients receiving injections on a monthly basis over the course of a year (Figure 1). The results of the study demonstrated that the level of detection by an enzyme-linked immunosorbent assay–based system was sensitive to 3 ng/mL
(Figure 2). All samples assayed were negative for the presence of bevacizumab at this level of detection. In addition to enzyme-linked immunosorbent assay testing, Western blot analysis was performed to observe whether samples collected after a recent injection (1.5 and 7 hours) contained low levels of bevacizumab. This analysis revealed no detectable level of bevacizumab in milk samples (Figure 3). Given the fact that this assay is sensitive to microgram levels, the results suggest that the transfer of bevacizumab to breast milk after intravitreal injection does not occur at least in a detectable quantity. Discussion The determination of whether a neutralizing antibody to vascular endothelial growth factor A is present in breast milk is important, as it has been documented that intravitreally delivered bevacizumab is capable of crossing the retinal–blood barrier and can lead to reduced levels of vascular endothelial growth factor in serum.7,8 Infants whom have been treated with bevacizumab exhibit lower vascular endothelial growth factor in their serum.9 The effects of such a compound delivered to a neonate through ingestion or through amniotic fluid are unknown but could have important developmental consequences. However, the mechanisms by which a molecule travels from the vitreous to the breast milk or amniotic fluid likely differ, and we have no evidence of bevacizumab in the patient’s amniotic fluid.
Fig. 3. Bevacizumab Western diagram and blot. Patient 2 samples (May) were taken 1 day before injection (P2.a), 1 hour after injection (P2.b), and 7 hours after injection (P2.c). Patient 1 samples (January) were taken 1 day before injection (P1.a) and 1 week after injection (P1.b). Control samples contained 200 pg/mL of bevacizumab (A.), 2.3 mg/mL of bevacizumab (B.), and the breast milk from an uninjected female (C.). The light and heavy chains (arrow) of bevacizumab can be observed in the 2.3 mg/mL control sample. Protein molecular weight marker (M) measured in kilodaltons.
BEVACIZUMAB LEVELS IN BREAST MILK MCFARLAND ET AL
In addition, children process antibodies from their mothers during infancy, so it is reasonable to assume that they could absorb an anti-vascular endothelial growth factor antibody where one to be ingested through breast milk.10,11 Our study confirms that bevacizumab is not transferred from mother to child through breast milk after many months of intravitreal injections. Therefore, we have no indication of oral ingestion of an anti-vascular endothelial growth factor molecule by infants. Given that a routine intravitreal dose of bevacizumab is 1.25 mg per 0.05 mL and our level of detection is approximately 3 ng/mL, it is reasonable to conclude that the presence of bevacizumab in breast milk after treatment does not occur above a 10 million-fold dilution range. These results can provide useful information for patients and ophthalmologists when recommending treatment for CNV in women who are also breastfeeding infants. Key words: bevacizumab, breast milk, choroidal neovascularization, ELISA, Western blot. References 1. Ehlken C, Martin G, Stahl A, Agostini HT. Reduction of vascular endothelial growth factor A in human breast milk after intravitreal injection of bevacizumab but not ranibizumab. Arch Ophthalmol 2012;130:1226–1227.
1673
2. Kim H, Robinson SB, Csaky KG. FcRn receptor-mediated pharmacokinetics of therapeutic IgG in the eye. Mol Vis 2009;15:2803–2812. 3. Chaparro M, Gisbert JP. Transplacental transfer of immunosuppressants and biologics used for the treatment of inflammatory bowel disease. Curr Pharm Biotechnol 2011;12:765–773. 4. Thorn M, Piche-Nicholas N, Stedman D, et al. Embryo-fetal transfer of bevacizumab (Avastin) in the rat over the course of gestation and the impact of neonatal Fc receptor (FcRn) binding. Birth Defects Res B Dev Reprod Toxicol 2012;95:363–375. 5. Pistilli B, Bellettini G, Giovannetti E, et al. Chemotherapy, targeted agents, antiemetics and growth factors in human milk: How should we counsel cancer patients about breastfeeding? Cancer Treat Rev 2013;39:207–211. 6. Petrou P, Georgalas I, Giavaras G, et al. Early loss of pregnancy after intravitreal bevacizumab injection. Acta Ophthalmol 2010;88:136. 7. Qian J, Lu Q, Tao Y, Jiang YR. Vitreous and plasma concentrations of apelin and vascular endothelial growth factor after intravitreal bevacizumab in eyes with proliferative diabetic retinopathy. Retina 2011;31:161–168. 8. Matsuyama K, Ogata N, Matsuoka M, et al. Plasma levels of vascular endothelial growth factor and pigment epitheliumderived factor before and after intravitreal injection of bevacizumab. Br J Ophthalmol 2010;94:1215–1218. 9. Sato T, Wada K, Arahori H, et al. Serum concentrations of bevacizumab and vascular endothelial growth factor in infants with retinopathy of prematurity. Am J Ophthalmol 2012;153:327–333. 10. Hanson L. Breastfeeding provides passive and likely longlasting active immunity. Ann Allergy Asthma Immunol 1998;81;523–537. 11. Hanson L, Korotkova M, Lundin S, et al. Transfer of immunity from mother to child. Ann N Y Acad Sci 2003;987:199–206.
