Response of Retinoblastoma With Vitreous Tumor Seeding to Adenovirus-Mediated Delivery of Thymidine Kinase Followed by Ganciclovir
http://www.100md.com
《临床肿瘤学》
the Departments of Pathology, Ophthalmology, and Pediatrics, Texas Children's Cancer Center
the Center for Cell and Gene Therapy, Baylor College of Medicine, Texas Children's Hospital
the Methodist Hospital, Houston, TX
ABSTRACT
PURPOSE: To evaluate the feasibility and safety of adenovirus-mediated gene therapy as a treatment for tumor seeds in the vitreous of children with retinoblastoma.
PATIENTS AND METHODS: An Institutional Biosafety Committee–, Institutional Review Board–, Recombinant DNA Advisory Committee–, and US Food and Drug Administration–approved phase I study used intrapatient dose escalation of adenoviral vector containing a herpes simplex thymidine kinase gene (AdV-TK) followed by systemic administration of ganciclovir to treat bilateral retinoblastoma with vitreous tumor seeding refractory to standard therapies. Vitreous tumor seeds were treated by intravitreous injection of AdV-TK adjacent to disease sites. Each injection was followed by ganciclovir delivered intravenously every 12 hours for 7 days.
RESULTS: Eight patients with vitreous tumor seeds were enrolled. One patient who was treated with 108 viral particles (vp) had resolution of the tumor seeds around the injection site. The seven patients who were treated with doses 1010 vp had resolution of their vitreous tumor seeds documented by fundoscopy. Toxicity included mild inflammation at 1010 vp and moderate inflammation, corneal edema, and increased intraocular pressure at 1011 vp. One patient was free of active vitreous tumor seeds 38 months after therapy. There has been no evidence of extraocular spread of tumor along the needle tract in any patient.
CONCLUSION: AdV-TK followed by ganciclovir can be administered safely to children with retinoblastoma. Suicide gene therapy may contribute to the treatment of children with retinoblastoma tumor seeds in the vitreous, a resistant complication of retinoblastoma.
INTRODUCTION
Retinoblastoma, a tumor arising from the retina, is the most common primary intraocular tumor of childhood and affects between 1 in every 15,000 and 1 in 20,000 live births in the United States.1 Retinoblastoma most often presents as leukocoria.2 This disease is caused by mutations of the RB1 gene on the 13th chromosome and occurs in both sporadic (unilateral, unifocal) or germline (multifocal, usually bilateral) forms.3-5 The vast majority of cases can be cured by enucleation of the eye when treated before invasion of the extraocular tissues, ocular coats, or the optic nerve.6 Recently, attempts to salvage the eye and vision for patients with bilateral disease have centered on the use of systemic chemotherapy and radiotherapy to shrink tumors sufficiently to allow ablation with local measures including laser and/or cryotherapy.7,8 However, radiotherapy often leads to second malignancies of the orbit, especially in children with a germline mutation,9,10 and systemic chemotherapy carries an increased risk of second malignancy as well as short-term morbidity.10 Moreover, when vitreous seeds (small pieces of avascular tumor floating in vitreous) are present, these therapies are usually unsuccessful.11 There is a need, therefore, to develop a new therapeutic modality that is effective for vitreous tumor seeds and that can be safely combined with local therapies to eradicate residual tumor.
Suicide gene transfer using an adenoviral vector to locally deliver the herpes simplex thymidine kinase gene (AdV-TK) followed by systemic administration of ganciclovir has been shown to be safe for the treatment of a number of different human malignancies.12-18 In preclinical studies, using this approach has been shown to shrink retinoblastoma in animal models.19 The present study demonstrates that this therapy is both safe and possibly effective for the treatment of children with refractory tumor seeding in the vitreous, a complication of retinoblastoma.
PATIENTS AND METHODS
Patient Eligibility
The patients were enrolled in a phase I trial. All of them had bilateral retinoblastoma and had experienced treatment failure with previous therapies including systemic chemotherapy (at least six cycles of vincristine, etoposide, and carboplatin), external-beam radiotherapy, and local treatments including cryotherapy and laser therapy, but they had remaining visual function and therefore could potentially benefit from additional therapy. In addition to the above, one patient had received intensive chemotherapy followed by a stem cell rescue and both intraocular and systemic phase I chemotherapy before enrollment. Patients with metastatic disease were excluded. In addition, all patients reported in this article had active tumor seeds in the vitreous refractory to all standard therapies and therefore were facing imminent enucleation. Six children were Hispanic, one was black, and one was white. There were five girls and three boys between the ages of 1 and 7 years at the time of enrollment.
Before gene therapy, all patients had a complete medical history recorded and physical examination including a visual acuity test, hematologic, hepatic, and renal function laboratory testing, magnetic resonance imaging of the brain and orbits, lumbar puncture, and examination of the eye under anesthesia (EUA). To determine if the intraocular injection of an adenoviral vector resulted in circulation of the vector or a systemic immune response, the presence of serum-neutralizing antibodies to wild-type adenovirus and the presence of adenovirus in the blood, urine, and nasal mucosa as measured by polymerase chain reaction (PCR) and viral culture were determined before the first treatment and 6 to 8 weeks after the last injection.
AdV-TK Vector
Construction of the replication-deficient group C serotype 5 adenoviral vector used in this trial has been described.12,20 The adenovirus E1A region was replaced with a herpes simplex thymidine kinase minigene under the control of the Rous sarcoma virus promoter and propagated in human embryonic kidney HEK293 cells. The vector was produced by the Baylor College of Medicine Gene Vector Laboratory and fully certified before use. The vector preparation was free of replication-competent adenovirus at less than 1 per 3 x 1010 viral particles (vp). The vp/infectious-unit ratio was determined to be 20:1.
Injection of the AdV-TK Vector
An anterior chamber paracentesis through clear cornea opposite the selected injection site was performed to reduce intraocular pressure before injection. The aqueous humor that was obtained underwent cytologic examination. Intraocular administration of the AdV-TK vector was achieved by using a transcorneal approach21 to minimize the risk of tumor spread during the injection. While using a microscope or indirect ophthalmoscopy, a 30-gauge needle was inserted through the peripheral cornea, iris, and zonules opposite the paracentesis site, avoiding the lens. The needle was positioned adjacent to the vitreous tumor seeds, and the viral vector was injected. The dose was divided between areas of the vitreous when more than one area was affected. When more than one site was treated, the needle was repositioned without removing it from the vitreous cavity. After removal of the needle, cryotherapy was administered to the needle tract and the area was subjected to copious irrigation. In addition, active primary tumors not adjacent to the vitreous seeds were treated with laser or cryotherapy specifically directed to the retinal tumors at the time of vector injection. Postoperatively, the patients received ciprofloxacin (0.3%), scopolamine (0.25%), and prednisolone (1%) eye drops at a dose of one drop each, four times a day, for 1 week. Patients who had grade 1 or greater ocular inflammation (defined below) received oral prednisone at a dose of 1 mg/kg for up to 4 weeks followed by a short tapering schedule (patients 2, 4, 5, 6, and 7).
Study Design
The study was approved by the Recombinant DNA Advisory Committee, the US Food and Drug Administration, and the Baylor College of Medicine Institutional Review Board and Institutional Biosafety Committee. Although this trial was a phase I toxicity study, an additional end point of the study was efficacy. Therefore, if a patient had resolution of his or her targeted vitreous seeds as determined by fundoscopy, the patient was taken off the study, and standard therapies appropriate for the treatment of the primary tumors were instituted in an attempt to salvage the eye. Patients and their parents were informed that the injection of the viral vector was experimental, and details of the procedures were presented before informed consent was obtained.
Systemic toxicity was graded by using the NCI Common Toxicity Criteria version 2.0 (http://ctep.cancer.gov/reporting/CTC-3.html). In addition, ocular toxicity was assessed by using a 0 to 3 grading system. Toxicities that were evaluated included corneal edema, hyphema, ocular inflammation, increased intraocular pressure, retinal damage, or toxicity to the optic nerve or conjunctiva. Grades 1 (mild severity) and 2 (moderate severity) ocular toxicities were assessed by the treating ophthalmologist. Ocular toxicities that could not be managed medically (considered grade 3) or systemic toxicity (grade 4) were considered dose limiting.
Eight patients with bilateral retinoblastoma whose remaining eye contained vitreous seeds resistant to chemotherapy and external-beam radiation therapy were treated on the study. The first patient enrolled received one injection of AdV-TK at a dose of 108 vp. Subsequent patients were enrolled onto an intrapatient dose-escalation study. The patients received up to five injections (100 μL each) of AdV-TK at doses outlined in Table 1. The initial dose of AdV-TK for the escalation study was 109 vp, and the highest dose administered to any patient was 1011 vp. The first two patients on the dose-escalation trial reported in this study received a single dose of 109 vp and then received up to four additional injections of 1010 vp each at 2- to 4-week intervals. The next three patients received a single dose of 1010 vp and then up to four additional injections of 1011 vp each. The next patient received a single injection of 1011 vp. Because two patients experienced increased toxicity that could possibly be related to the gene-therapy injections at the 1011-vp dose, subsequent patients received 2.5 x 1010 vp up to a possible total of four injections. Twenty-four hours after each injection, the patients received the first of 14 doses of ganciclovir (5 mg/kg per dose) delivered intravenously at 12-hour intervals. All children underwent EUA and MRI of the brain and orbits before each AdV-TK injection; every 1 to 4 weeks during and after treatment, an EUA, visual acuity examination, and magnetic resonance imaging (MRI; to assess possible extraocular extension of disease) were performed. If the vitreous tumor seeds showed partial response (decreasing number of vitreous seeds but thought to be viable in appearance) or no response as determined by direct visualization at the time of EUA and the patient had experienced no dose-limiting toxicity, the eye was reinjected with AdV-TK dose escalated by 1 log unit for a total of no more than four additional injections at that dose. If the vitreous tumor seeds resolved (complete disappearance of vitreous seeds or calcified vitreous seeds with no apparent growth over time as determined by comparison with previous fundus photographs), the patients were said to have clinical resolution of their vitreous seeds and were re-examined at a minimum of every 3 months for the next 3 years. Residual primary retinal tumors but not the vitreous seeds were treated with local therapy as appropriate. If the patient was found to have clinical resolution of the vitreous seeds, then he or she was taken off the study, and any residual tumors were treated with standard therapies. Histologic examination of the eye was performed at the time of enucleation.
At each EUA, fundus photographs were obtained by using a RetCam digital camera (Clarity Medical Systems Inc, Pleasanton, CA). Weekly evaluation of kidney, hepatic, and hematologic function was obtained throughout therapy. Eyes were enucleated if at any time there was evidence of progressive disease that could not be controlled by standard therapies or additional gene therapy within the limits of the study.
RESULTS
The primary purpose of this study was to evaluate the feasibility and safety of adenovirus-mediated gene therapy as a treatment for tumor seeds in the vitreous of children with retinoblastoma. All the patients included in this study faced imminent enucleation from their refractory retinoblastoma. None of the patients had gene-therapy–related toxicities that were indications for enucleation, and no dose-limiting toxicities were observed. Vitreous condensation and banding occurred to some degree in all patients and did not seem to be dose related. No toxicity was observed at the time of EUA in the patient treated with 108 vp. Two patients treated with 109 vp followed by 1010 vp showed grade 1 inflammatory changes. One of two patients treated with 2.5 x 1010 vp had mild inflammation observed. Three of four patients treated with 1011 vp showed moderate inflammation (grade 2), and two of these patients showed corneal edema and increased intraocular pressure (Table 1).
All seven patients treated at doses 1010 vp had apparent clinical resolution of their vitreous tumor seeds (as defined in the Methods) (Table 1) as determined by direct visualization. Retinal fundoscopic photographs before and after gene therapy (Fig 1) show representative findings. Most patients enrolled had fluffy tumor seeds (Figs 1A, 1C, 1E) in the vitreous and had evidence of regressed, often calcified, retinal tumors not affecting the optic nerve (Fig 1E). After gene therapy, clinical observation by the treating ophthalmologist indicated the absence and/or calcification of vitreous tumor seeds (Fig 1B, 1D, 1F). The vitreous was hazy and banded (Figs 1B, 1F) without well-formed tumor seeds. Thirty months after patient 3 received the initial three injections of the gene vector (Fig 1D), vitreous tumor seeds were stable and totally calcified, and the retinal tumor was also calcified and regressed. The patient's visual acuity was 20/30 at this time. The patient initially presented with a mild, slowly progressing subcapsular cataract that was probably secondary to external-beam radiation delivered before enrollment onto the study and is the cause of the hazy appearance of the retinal photograph in Figure 1D.
All patients enrolled have received enucleations because of apparent progression of their primary tumor not treated with gene therapy. None of the patients have had gene-therapy–related toxicities that were indications for enucleation. Histopathologic findings of two enucleated eyes are shown in Figure 2. Gross examination of enucleated eyes shows either vitreous contraction (patient 2, Fig 2A) or vitreous bands (patient 4, Fig 2E) similar to that seen in the eyes from other patients. Figure 2A shows a gross photograph of the enucleated eye with active retinal tumors (T) but a clear, partially empty vitreous cavity with a condensed vitreous behind the lens. The microscopic features of the same eye demonstrate the absence of tumor seeds in the condensed vitreous (Fig 2B), an active retinal tumor (T) with superficial necrosis and choroidal invasion (IT) (Fig 2C), and a well-healed wound with fibroblasts and histiocytes surrounding areas of embedded cryomaterial deep in the corneal stroma (Fig 2D). The corneal endothelial cells seen were unremarkable, and tumor cells were absent. The eye of patient 4 was enucleated for apparent progressive disease after five injections of the vector and 5 months after the beginning of treatment. The vitreous was condensed, forming bands from the regressed tumor and the optic nerve to the base of the vitreous and contained few foci of "seeds" (Fig 2E). The microscopic findings of the clinically interpreted vitreous "seeds" are foci of cellular debris and macrophages without viable tumor cells (Fig 2F). Immunohistochemistry staining of the peripheral retina and vitreous shows groups of histiocytes labeled by CD68 antibody (stained brown) and small plasma cells (Fig 2G). Tumor cells were not detected.
Patient 5 received two separate courses of gene therapy 9 months apart for the treatment of vitreous seeds associated with different tumors and had apparent resolution of her vitreous seeds on both occasions. The initial response to the intravitreous injections was dissolution of the tumor seeds and mild haziness of the vitreous. Later, the vitreous either cleared completely or condensed and contracted into bands. Patient 3 received local therapy to treat an edge recurrence of her original retinal tumor but remained free of active vitreous seeds more than 3 years after gene therapy.
The duration of response of the treated vitreous tumor seeds was persistent to the time of enucleation (13 to 176 weeks) in all but patient 8, who had a suspected recurrence of vitreous tumor seeds noted 2 months after completion of gene therapy. Histopathologic examination demonstrated lack of vitreous tumor seeds in the central vitreous in all patients treated at a dose 1010 vp. Vitreous tumor seeds were only observed immediately adjacent to active retinal tumors.
The primary tumor in the eye of patient 7 was treated with brachytherapy 4 months after gene therapy. Six months after brachytherapy, tumor progression was noted, and the eye was enucleated. Histologic examination showed that his primary retinal tumor, which was not treated with gene therapy, had extended along an emissary vessel of the sclera adjacent to the brachytherapy scar into the episclera, predicting metastatic potential. The gene-therapy corneal-injection site on the opposite side of the eye was free of retinoblastoma. In view of the episcleral invasion, this patient received orbital radiation and chemotherapy but developed metastatic disease 14 months after enucleation.
All but two of the patients had immunoglobulin G antiadenoviral titers ranging from 1:8 to 1:128 before treatment. None of the patients had immunoglobulin M antiadenoviral titers before or after treatment, nor did any of the patients have a rise in antibody titer after therapy. Two patients did not have measurable titers before or after therapy. Culture and PCR analysis of blood, urine, or nasal washings showed no evidence of adenovirus in any of the patients.
DISCUSSION
Although surgical intervention remains the definitive therapy for unilateral retinoblastoma6 and has close to a 90% cure rate in the absence of metastasis,22 there is clearly a need to develop therapies for bilateral retinoblastoma that spare the eye and any visual function it may retain. Local control using cryotherapy or laser photocoagulation may meet this need but can only be used when the tumor is small or when vitreous tumor seeds are absent.23,24 External-beam radiation therapy is useful occasionally in the treatment of vitreous tumor seeds but is accompanied by high morbidity including radiation retinitis, closed-angle glaucoma, and second malignancies in the field of radiation in patients with constitutional mutations of the retinoblastoma gene.
Preclinical gene-therapy studies in the ocular environment have shown considerable promise for the treatment of oncologic,19 retinal degenerative,25 and neovascular26 diseases. These studies have paved the way for clinical trials such as the one described in here. In this phase I study, eight patients with vitreous tumor seeds were treated with a total of 21 injections of AdV-TK. All the injections were tolerated well. The patients all presented with extensive vitreous tumor seeds away from the main tumor, a finding that predicts failure of standard therapy. These patients therefore would be classified as group Vb of the Reese-Ellsworth classification (Ia-Vb) for conservative treatment of retinoblastoma (radiotherapy)27 or groups D and E of the newly proposed international retinoblastoma classification (A-E), which aims to correlate tumor characteristics with response to chemotherapy and local therapies.28 Patients who present with disease at these stages usually encounter treatment failure with current therapeutic modalities. The vitreous tumor seeds of all eight of the patients responded to the gene-therapy treatment, and all seven patients treated at a dose 1010 vp had apparent resolution of vitreous tumor seeds as determined by indirect ophthalmoscopy. At the time of histopathologic evaluation of the enucleated eyes, all but patient 4 had vitreous seeds directly overlying active tumors. Whether these vitreous seeds were a recurrence of previously treated seeds or new extensions from active primary tumors that occurred after gene therapy cannot be determined. At the time of this writing, all the children were alive and without metastatic disease or systemic complications that resulted from the adenoviral vector administration. The one child who developed metastatic disease had extension of disease from a primary tumor through the emissary vessel adjacent to a brachytherapy scar, treatment that was used after gene therapy in an attempt to eradicate remaining retinal disease and preserve the eye. The needle track and the injection site of the viral vector in this patient were free of disease, suggesting that the disease extension was not secondary to gene-therapy administration. The gene-therapy protocol required that all selected patients had experienced treatment failure with all previous therapies and were facing imminent enucleation. This patient was 6 years old at the time of gene-therapy treatment and had received treatment for his retinoblastoma for 4 years before being enrolled onto the gene-therapy protocol. The primary reason that the patients had therapy failure was the presence of vitreous seeds. Although this protocol was a phase I study, the vitreous seeds responded to the therapy, which allowed the ophthalmologist to attempt additional standard therapies to control the primary tumor. Such therapies in these extensively treated patients must be used with extreme caution.
Our results indicate that transfer of the herpes simplex thymidine kinase gene followed by treatment with ganciclovir is both safe and feasible. The unanticipated clinical responses that were observed suggest that this approach may have value as a supplement to current therapeutic approaches. This therapy did not exhibit cross resistance in patients who had experienced chemotherapy and radiotherapy failure. Because the injected adenoviral vector cannot diffuse through vitreous, each focal area of vitreous tumor seeds had to be injected separately. This was accomplished by repositioning the needle without removing it from the vitreous to minimize the chance of tumor spread. Patient 2, for example, who had vitreous seeds throughout the eye, had the dose of viral vector divided equally into each quadrant of the eye.
The injection technique that was used in this trial is a modification of the transcorneal approach that was developed to biopsy intraocular tumors.21 In this study reported by Shields et al, this technique was used safely in more than 150 patients with different intraocular lesions, including retinoblastoma, and never resulted in a case of metastatic disease by dissemination of tumor cells through the needle tract. None of our patients had a tumor biopsy. The transcorneal approach was used only to inject the adenoviral vector, and no tumor seeding along the needle tract was detected.
The standard diagnostic criteria for retinoblastoma are based on indirect ophthalmoscopic examination. Although there are other diseases such as Toxocara canis infection and Coat's disease that can mimic retinoblastoma, all the patients reported in this study had typical findings of retinoblastoma complicated by vitreous seeds that had been confirmed by multiple examinations by different ophthalmologists over several months. Patient 4, who had classical findings of a regressed primary tumor after chemotherapy and radiation treatment and multiple large, fluffy vitreous seeds before gene-therapy treatment, had an enucleation because of the apparent persistence of vitreous seeds and an imaging report (MRI) consistent with recurrence of her primary tumor with invasion into the optic nerve head. The appearance of her vitreous seeds had changed from her classical initial presentation. Histopathologic examination revealed an inflammatory response unique in our experience and probably related to the gene-therapy treatment. Only a benign retinocytoma consistent with response to prior therapy and mild edema of the optic nerve head was observed.
Because gene therapy for cancer relies on the destructive capacity of the transgene, persistence is not a desired attribute for the vector. Instead, one seeks high-level transduction (from a high-titer vector) with high-level gene expression. For this purpose, first-generation adenoviral vectors remain unsurpassed and therefore were used in this trial. Multiple previous studies have documented that AdV-TK followed by ganciclovir can kill retinoblastoma cells and decrease the size of retinoblastomas in vivo in a murine xenograft model.19 For successful treatment, both the administration of the viral vector and ganciclovir were necessary, suggesting that delivery of active herpes thymidine kinase to the tumor had occurred. One advantage of this therapeutic regimen is that it produces a pronounced bystander effect. Phosphorylated ganciclovir, the cytotoxic product of the herpes simplex thymidine kinase reaction, is transported to adjacent cells through gap junctions.29 Thus, not every tumor cell must be transduced by the viral vector for the therapy to be effective. Although suicide gene-therapy treatment has been shown to be safe when used for prostate cancer30 and brain tumors,12 and transgene expression has been clearly documented, clinically significant efficacy was not demonstrated. Given this background, in this study we chose not to measure transgene expression, because it would have required early enucleation after gene transfer. Instead, we elected to directly measure tumor response as a nonsurrogate marker of efficacy.
Regulatory requirements for this study necessitated that all patients enrolled must have experienced treatment failure with extensive systemic and local therapies including external-beam radiation therapy. Although these therapies disrupt the blood-retina barrier, presumably violating the immune privilege status of the eye, only mild to moderate degrees of local inflammation were observed in our patients after the intraocular injections. The inflammation was controlled by treatment with local and systemic steroids. None of the patients developed a measurable humoral or cellular immune response against the vector, and adenovirus was not detected by culture or PCR in the saliva, blood, or urine of treated patients up to 6 weeks after the completion of gene therapy. Two patients developed corneal edema that resulted in significant loss of vision and increased intraocular pressure. These toxicities were observed only at the highest dose level (1011 vp). Because we saw efficacy and less toxicity with patients treated at 2.5 x 1010 vp, this dose will be used in future studies. Some degree of condensation of vitreous resulting in loss of vision was observed in all patients. The degree of condensation seemed to be related to the amount of vitreous seeds that were being treated rather than the dose of viral vector that was being administered. Vision seemed to improve over time. Patients 1 and 3 had better vision after therapy. None of the patients had damage to the retina. Therefore, if the vitreous opacities could be removed (ie, vitrectomy), the patient's vision could be improved. This procedure should only be attempted in patients who have no evidence of active retinoblastoma for at least 2 years because of the risk of metastasis. Less toxicity may be observed in patients with an earlier presentation or less severe disease and in patients who have not received therapies that would disrupt the blood-retina barrier. In addition to expanded trials that will further study the utility of gene therapy to treat vitreous tumor seeds, future clinical studies are being developed that will define the toxicity and efficacy of direct injection into the primary retinal tumor.
In this phase I trial we have shown that adenoviral vectors used to deliver the herpes simplex thymidine kinase gene can be administered to children's eyes containing retinoblastoma with an acceptable safety profile. All the patients ultimately required enucleation, emphasizing that gene therapy administered in this trial could not be used as a stand-alone treatment for cure of retinoblastoma. These results suggest that gene therapy may be a useful adjuvant to other standard therapies for the treatment of children with vitreous seeds that are complicating retinoblastoma. Future studies will be designed to further develop the role that gene therapy may play in the therapy of this disease.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank Drs Célia Antoneli, Juan Carlos Bernini, Raul Cordero, Doris Hadjistilianou, Carlos Leal, and Paul Salmonsen for patient referrals; Marlen Dinu, Olga Arroyo, Jeri Gates, Sheila Love, and Shanna Noel for assistance with patient management; and Bambi Grilley for assistance with regulatory compliance.
NOTES
National Institutes of Health Grant No. CA97762 defrayed costs of the viral vector and together with the Foundation for Research and Retina Research Foundation provided personnel support. National Institutes of Health Grant No. RR00188 and funding from the Texas Children's Hospital covered patient care expenses, and funding from the Fleming-Davenport Fund defrayed patient-transportation costs.
Both P.C.-B. and M.C. contributed equally to this work.
Presented in part at the annual Association for Research in Vision and Ophthalmology meeting April 25-29, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Devesa S: The incidence of retinoblastoma. Am J Ophthalmol 80:263-265, 1975
Margo C, Hidayet A, Kopelman A, et al: Retinocytoma: A benign variant of retinoblastoma. Arch Ophthalmol 101:1519-1531, 1983
Knudson AG Jr: Mutation and cancer: Statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820-823, 1971
Dryja TP, Rapaport JM, Weichselbaum R, et al: Chromosome 13 restriction fragment length polymorphisms. Hum Genet 65:320-324, 1984
Harbour JW: Overview of RB gene mutations in patients with retinoblastoma: Implications for clinical genetic screening. Ophthalmology 105:1442-1447, 1998
Shields CL, Shields JA: Recent developments in the management of retinoblastoma. J Pediatr Ophthalmol Strabismus 36:8-18, 1999
Gallie BH, Budning A, DeBoer G, et al: Chemotherapy with focal therapy can cure intraocular retinoblastoma without radiotherapy. Arch Ophthalmol 114:1321-1328, 1996 [Erratum: Arch Ophthalmol 115:525, 1997]
Murphree AL, Villablanca JG, Deegan WF III, et al: Chemotherapy plus local treatment in the management of intraocular retinoblastoma. Arch Ophthalmol 114:1348-1356, 1996
Roarty JD, McLean IW, Zimmerman LE: Incidence of second neoplasms in patients with bilateral retinoblastoma. Ophthalmology 95:1583-1587, 1988
Abramson DH, Beaverson K, Sangani P, et al: Screening for retinoblastoma: Presenting signs as prognosticators of patient and ocular survival. Pediatrics 112:1248-1255, 2003
Shields CL, de Potter P, Himelstein BP, et al: Chemoreduction in the initial management of intraocular retinoblastoma. Arch Ophthalmol 114:1330-1338, 1996
Trask TW, Trask RP, Aguilar-Cordova E, et al: Phase I study of adenoviral delivery of the HSV-tk gene and ganciclovir administration in patients with current malignant brain tumors. Mol Ther 1:195-203, 2000
Teh BS, Ayala G, Aguilar L, et al: Phase I-II trial evaluating combined intensity-modulated radiotherapy and in situ gene therapy with or without hormonal therapy in treatment of prostate cancer: Interim report on PSA response and biopsy data. Int J Radiat Oncol Biol Phys 58:1520-1529, 2004
Havlik R, Jiao LR, Nicholls J, et al: Gene therapy for liver metastases. Semin Oncol 29:202-208, 2002
Banerjee D: Technology evaluation: Gene therapy (mesothelioma), NCI. Curr Opin Mol Ther 1:517-520, 1999
Sung MW, Yeh HC, Thung SN, et al: Intratumoral adenovirus-mediated suicide gene transfer for hepatic metastases from colorectal adenocarcinoma: Results of a phase I clinical trial. Mol Ther 4:182-191, 2001
Alvarez RD, Gomez-Navarro J, Wang M, et al: Adenoviral-mediated suicide gene therapy for ovarian cancer. Mol Ther 2:524-530, 2000
Morris JC, Wildner O: Therapy of head and neck squamous cell carcinoma with an oncolytic adenovirus expressing HSV-tk. Mol Ther 1:56-62, 2000
Hurwitz MY, Marcus KT, Chevez-Barrios P, et al: Suicide gene therapy for treatment of retinoblastoma in a murine model. Hum Gene Ther 10:441-448, 1999
Chen SH, Shine HD, Goodman JC, et al: Gene therapy for brain tumors: Regression of experimental gliomas by adenovirus-mediated gene transfer in vivo. Proc Natl Acad Sci USA 91:3054-3057, 1994
Shields CL, Shields JA, Baez KA, et al: Choroidal invasion of retinoblastoma: Metastatic potential and clinical risk factors. Br J Ophthalmol 77:544-548, 1993
Tamboli A, Podgor MJ, Horm JW: The incidence of retinoblastoma in the United States. Arch Ophthalmol 108:128-132, 1990
Shields JA: The expanding role of laser photocoagulation for intraocular tumors. The 1993 H. Christian Zweng Memorial Lecture. Retina 14:310-322, 1994
Shields CL, Santos MC, Diniz W, et al: Thermotherapy for retinoblastoma. Arch Ophthalmol 117:885-893, 1999
Acland GM, Aguirre GD, Ray J, et al: Gene therapy restores vision in a canine model of childhood blindness. Nat Genet 28:92-95, 2001
Mori K, Gehlbach P, Ando A, et al: Regression of ocular neovascularization in response to increased expression of pigment epithelium-derived factor. Invest Ophthalmol Vis Sci 43:2428-2434, 2002
Reese AB, Ellsworth RM: Management of retinoblastoma. Ann N Y Acad Sci 114:958-962, 1964
Shields CL, Mashayekhi A, Demirci H, et al: Practical approach to management of retinoblastoma. Arch Ophthalmol 122:729-735, 2004
Elshami AA, Saavedra A, Zhang H, et al: Gap junctions play a role in the "bystander effect" of the herpes simplex virus thymidine kinase/ganciclovir system in vitro. Gene Ther 3:85-92, 1996
Herman JR, Adler HL, Aguilar-Cordova E, et al: In situ gene therapy for adenocarcinoma of the prostate: A phase I clinical trial. Hum Gene Ther 10:1239-1249, 1999(Patricia Chévez-Barrios, )
the Center for Cell and Gene Therapy, Baylor College of Medicine, Texas Children's Hospital
the Methodist Hospital, Houston, TX
ABSTRACT
PURPOSE: To evaluate the feasibility and safety of adenovirus-mediated gene therapy as a treatment for tumor seeds in the vitreous of children with retinoblastoma.
PATIENTS AND METHODS: An Institutional Biosafety Committee–, Institutional Review Board–, Recombinant DNA Advisory Committee–, and US Food and Drug Administration–approved phase I study used intrapatient dose escalation of adenoviral vector containing a herpes simplex thymidine kinase gene (AdV-TK) followed by systemic administration of ganciclovir to treat bilateral retinoblastoma with vitreous tumor seeding refractory to standard therapies. Vitreous tumor seeds were treated by intravitreous injection of AdV-TK adjacent to disease sites. Each injection was followed by ganciclovir delivered intravenously every 12 hours for 7 days.
RESULTS: Eight patients with vitreous tumor seeds were enrolled. One patient who was treated with 108 viral particles (vp) had resolution of the tumor seeds around the injection site. The seven patients who were treated with doses 1010 vp had resolution of their vitreous tumor seeds documented by fundoscopy. Toxicity included mild inflammation at 1010 vp and moderate inflammation, corneal edema, and increased intraocular pressure at 1011 vp. One patient was free of active vitreous tumor seeds 38 months after therapy. There has been no evidence of extraocular spread of tumor along the needle tract in any patient.
CONCLUSION: AdV-TK followed by ganciclovir can be administered safely to children with retinoblastoma. Suicide gene therapy may contribute to the treatment of children with retinoblastoma tumor seeds in the vitreous, a resistant complication of retinoblastoma.
INTRODUCTION
Retinoblastoma, a tumor arising from the retina, is the most common primary intraocular tumor of childhood and affects between 1 in every 15,000 and 1 in 20,000 live births in the United States.1 Retinoblastoma most often presents as leukocoria.2 This disease is caused by mutations of the RB1 gene on the 13th chromosome and occurs in both sporadic (unilateral, unifocal) or germline (multifocal, usually bilateral) forms.3-5 The vast majority of cases can be cured by enucleation of the eye when treated before invasion of the extraocular tissues, ocular coats, or the optic nerve.6 Recently, attempts to salvage the eye and vision for patients with bilateral disease have centered on the use of systemic chemotherapy and radiotherapy to shrink tumors sufficiently to allow ablation with local measures including laser and/or cryotherapy.7,8 However, radiotherapy often leads to second malignancies of the orbit, especially in children with a germline mutation,9,10 and systemic chemotherapy carries an increased risk of second malignancy as well as short-term morbidity.10 Moreover, when vitreous seeds (small pieces of avascular tumor floating in vitreous) are present, these therapies are usually unsuccessful.11 There is a need, therefore, to develop a new therapeutic modality that is effective for vitreous tumor seeds and that can be safely combined with local therapies to eradicate residual tumor.
Suicide gene transfer using an adenoviral vector to locally deliver the herpes simplex thymidine kinase gene (AdV-TK) followed by systemic administration of ganciclovir has been shown to be safe for the treatment of a number of different human malignancies.12-18 In preclinical studies, using this approach has been shown to shrink retinoblastoma in animal models.19 The present study demonstrates that this therapy is both safe and possibly effective for the treatment of children with refractory tumor seeding in the vitreous, a complication of retinoblastoma.
PATIENTS AND METHODS
Patient Eligibility
The patients were enrolled in a phase I trial. All of them had bilateral retinoblastoma and had experienced treatment failure with previous therapies including systemic chemotherapy (at least six cycles of vincristine, etoposide, and carboplatin), external-beam radiotherapy, and local treatments including cryotherapy and laser therapy, but they had remaining visual function and therefore could potentially benefit from additional therapy. In addition to the above, one patient had received intensive chemotherapy followed by a stem cell rescue and both intraocular and systemic phase I chemotherapy before enrollment. Patients with metastatic disease were excluded. In addition, all patients reported in this article had active tumor seeds in the vitreous refractory to all standard therapies and therefore were facing imminent enucleation. Six children were Hispanic, one was black, and one was white. There were five girls and three boys between the ages of 1 and 7 years at the time of enrollment.
Before gene therapy, all patients had a complete medical history recorded and physical examination including a visual acuity test, hematologic, hepatic, and renal function laboratory testing, magnetic resonance imaging of the brain and orbits, lumbar puncture, and examination of the eye under anesthesia (EUA). To determine if the intraocular injection of an adenoviral vector resulted in circulation of the vector or a systemic immune response, the presence of serum-neutralizing antibodies to wild-type adenovirus and the presence of adenovirus in the blood, urine, and nasal mucosa as measured by polymerase chain reaction (PCR) and viral culture were determined before the first treatment and 6 to 8 weeks after the last injection.
AdV-TK Vector
Construction of the replication-deficient group C serotype 5 adenoviral vector used in this trial has been described.12,20 The adenovirus E1A region was replaced with a herpes simplex thymidine kinase minigene under the control of the Rous sarcoma virus promoter and propagated in human embryonic kidney HEK293 cells. The vector was produced by the Baylor College of Medicine Gene Vector Laboratory and fully certified before use. The vector preparation was free of replication-competent adenovirus at less than 1 per 3 x 1010 viral particles (vp). The vp/infectious-unit ratio was determined to be 20:1.
Injection of the AdV-TK Vector
An anterior chamber paracentesis through clear cornea opposite the selected injection site was performed to reduce intraocular pressure before injection. The aqueous humor that was obtained underwent cytologic examination. Intraocular administration of the AdV-TK vector was achieved by using a transcorneal approach21 to minimize the risk of tumor spread during the injection. While using a microscope or indirect ophthalmoscopy, a 30-gauge needle was inserted through the peripheral cornea, iris, and zonules opposite the paracentesis site, avoiding the lens. The needle was positioned adjacent to the vitreous tumor seeds, and the viral vector was injected. The dose was divided between areas of the vitreous when more than one area was affected. When more than one site was treated, the needle was repositioned without removing it from the vitreous cavity. After removal of the needle, cryotherapy was administered to the needle tract and the area was subjected to copious irrigation. In addition, active primary tumors not adjacent to the vitreous seeds were treated with laser or cryotherapy specifically directed to the retinal tumors at the time of vector injection. Postoperatively, the patients received ciprofloxacin (0.3%), scopolamine (0.25%), and prednisolone (1%) eye drops at a dose of one drop each, four times a day, for 1 week. Patients who had grade 1 or greater ocular inflammation (defined below) received oral prednisone at a dose of 1 mg/kg for up to 4 weeks followed by a short tapering schedule (patients 2, 4, 5, 6, and 7).
Study Design
The study was approved by the Recombinant DNA Advisory Committee, the US Food and Drug Administration, and the Baylor College of Medicine Institutional Review Board and Institutional Biosafety Committee. Although this trial was a phase I toxicity study, an additional end point of the study was efficacy. Therefore, if a patient had resolution of his or her targeted vitreous seeds as determined by fundoscopy, the patient was taken off the study, and standard therapies appropriate for the treatment of the primary tumors were instituted in an attempt to salvage the eye. Patients and their parents were informed that the injection of the viral vector was experimental, and details of the procedures were presented before informed consent was obtained.
Systemic toxicity was graded by using the NCI Common Toxicity Criteria version 2.0 (http://ctep.cancer.gov/reporting/CTC-3.html). In addition, ocular toxicity was assessed by using a 0 to 3 grading system. Toxicities that were evaluated included corneal edema, hyphema, ocular inflammation, increased intraocular pressure, retinal damage, or toxicity to the optic nerve or conjunctiva. Grades 1 (mild severity) and 2 (moderate severity) ocular toxicities were assessed by the treating ophthalmologist. Ocular toxicities that could not be managed medically (considered grade 3) or systemic toxicity (grade 4) were considered dose limiting.
Eight patients with bilateral retinoblastoma whose remaining eye contained vitreous seeds resistant to chemotherapy and external-beam radiation therapy were treated on the study. The first patient enrolled received one injection of AdV-TK at a dose of 108 vp. Subsequent patients were enrolled onto an intrapatient dose-escalation study. The patients received up to five injections (100 μL each) of AdV-TK at doses outlined in Table 1. The initial dose of AdV-TK for the escalation study was 109 vp, and the highest dose administered to any patient was 1011 vp. The first two patients on the dose-escalation trial reported in this study received a single dose of 109 vp and then received up to four additional injections of 1010 vp each at 2- to 4-week intervals. The next three patients received a single dose of 1010 vp and then up to four additional injections of 1011 vp each. The next patient received a single injection of 1011 vp. Because two patients experienced increased toxicity that could possibly be related to the gene-therapy injections at the 1011-vp dose, subsequent patients received 2.5 x 1010 vp up to a possible total of four injections. Twenty-four hours after each injection, the patients received the first of 14 doses of ganciclovir (5 mg/kg per dose) delivered intravenously at 12-hour intervals. All children underwent EUA and MRI of the brain and orbits before each AdV-TK injection; every 1 to 4 weeks during and after treatment, an EUA, visual acuity examination, and magnetic resonance imaging (MRI; to assess possible extraocular extension of disease) were performed. If the vitreous tumor seeds showed partial response (decreasing number of vitreous seeds but thought to be viable in appearance) or no response as determined by direct visualization at the time of EUA and the patient had experienced no dose-limiting toxicity, the eye was reinjected with AdV-TK dose escalated by 1 log unit for a total of no more than four additional injections at that dose. If the vitreous tumor seeds resolved (complete disappearance of vitreous seeds or calcified vitreous seeds with no apparent growth over time as determined by comparison with previous fundus photographs), the patients were said to have clinical resolution of their vitreous seeds and were re-examined at a minimum of every 3 months for the next 3 years. Residual primary retinal tumors but not the vitreous seeds were treated with local therapy as appropriate. If the patient was found to have clinical resolution of the vitreous seeds, then he or she was taken off the study, and any residual tumors were treated with standard therapies. Histologic examination of the eye was performed at the time of enucleation.
At each EUA, fundus photographs were obtained by using a RetCam digital camera (Clarity Medical Systems Inc, Pleasanton, CA). Weekly evaluation of kidney, hepatic, and hematologic function was obtained throughout therapy. Eyes were enucleated if at any time there was evidence of progressive disease that could not be controlled by standard therapies or additional gene therapy within the limits of the study.
RESULTS
The primary purpose of this study was to evaluate the feasibility and safety of adenovirus-mediated gene therapy as a treatment for tumor seeds in the vitreous of children with retinoblastoma. All the patients included in this study faced imminent enucleation from their refractory retinoblastoma. None of the patients had gene-therapy–related toxicities that were indications for enucleation, and no dose-limiting toxicities were observed. Vitreous condensation and banding occurred to some degree in all patients and did not seem to be dose related. No toxicity was observed at the time of EUA in the patient treated with 108 vp. Two patients treated with 109 vp followed by 1010 vp showed grade 1 inflammatory changes. One of two patients treated with 2.5 x 1010 vp had mild inflammation observed. Three of four patients treated with 1011 vp showed moderate inflammation (grade 2), and two of these patients showed corneal edema and increased intraocular pressure (Table 1).
All seven patients treated at doses 1010 vp had apparent clinical resolution of their vitreous tumor seeds (as defined in the Methods) (Table 1) as determined by direct visualization. Retinal fundoscopic photographs before and after gene therapy (Fig 1) show representative findings. Most patients enrolled had fluffy tumor seeds (Figs 1A, 1C, 1E) in the vitreous and had evidence of regressed, often calcified, retinal tumors not affecting the optic nerve (Fig 1E). After gene therapy, clinical observation by the treating ophthalmologist indicated the absence and/or calcification of vitreous tumor seeds (Fig 1B, 1D, 1F). The vitreous was hazy and banded (Figs 1B, 1F) without well-formed tumor seeds. Thirty months after patient 3 received the initial three injections of the gene vector (Fig 1D), vitreous tumor seeds were stable and totally calcified, and the retinal tumor was also calcified and regressed. The patient's visual acuity was 20/30 at this time. The patient initially presented with a mild, slowly progressing subcapsular cataract that was probably secondary to external-beam radiation delivered before enrollment onto the study and is the cause of the hazy appearance of the retinal photograph in Figure 1D.
All patients enrolled have received enucleations because of apparent progression of their primary tumor not treated with gene therapy. None of the patients have had gene-therapy–related toxicities that were indications for enucleation. Histopathologic findings of two enucleated eyes are shown in Figure 2. Gross examination of enucleated eyes shows either vitreous contraction (patient 2, Fig 2A) or vitreous bands (patient 4, Fig 2E) similar to that seen in the eyes from other patients. Figure 2A shows a gross photograph of the enucleated eye with active retinal tumors (T) but a clear, partially empty vitreous cavity with a condensed vitreous behind the lens. The microscopic features of the same eye demonstrate the absence of tumor seeds in the condensed vitreous (Fig 2B), an active retinal tumor (T) with superficial necrosis and choroidal invasion (IT) (Fig 2C), and a well-healed wound with fibroblasts and histiocytes surrounding areas of embedded cryomaterial deep in the corneal stroma (Fig 2D). The corneal endothelial cells seen were unremarkable, and tumor cells were absent. The eye of patient 4 was enucleated for apparent progressive disease after five injections of the vector and 5 months after the beginning of treatment. The vitreous was condensed, forming bands from the regressed tumor and the optic nerve to the base of the vitreous and contained few foci of "seeds" (Fig 2E). The microscopic findings of the clinically interpreted vitreous "seeds" are foci of cellular debris and macrophages without viable tumor cells (Fig 2F). Immunohistochemistry staining of the peripheral retina and vitreous shows groups of histiocytes labeled by CD68 antibody (stained brown) and small plasma cells (Fig 2G). Tumor cells were not detected.
Patient 5 received two separate courses of gene therapy 9 months apart for the treatment of vitreous seeds associated with different tumors and had apparent resolution of her vitreous seeds on both occasions. The initial response to the intravitreous injections was dissolution of the tumor seeds and mild haziness of the vitreous. Later, the vitreous either cleared completely or condensed and contracted into bands. Patient 3 received local therapy to treat an edge recurrence of her original retinal tumor but remained free of active vitreous seeds more than 3 years after gene therapy.
The duration of response of the treated vitreous tumor seeds was persistent to the time of enucleation (13 to 176 weeks) in all but patient 8, who had a suspected recurrence of vitreous tumor seeds noted 2 months after completion of gene therapy. Histopathologic examination demonstrated lack of vitreous tumor seeds in the central vitreous in all patients treated at a dose 1010 vp. Vitreous tumor seeds were only observed immediately adjacent to active retinal tumors.
The primary tumor in the eye of patient 7 was treated with brachytherapy 4 months after gene therapy. Six months after brachytherapy, tumor progression was noted, and the eye was enucleated. Histologic examination showed that his primary retinal tumor, which was not treated with gene therapy, had extended along an emissary vessel of the sclera adjacent to the brachytherapy scar into the episclera, predicting metastatic potential. The gene-therapy corneal-injection site on the opposite side of the eye was free of retinoblastoma. In view of the episcleral invasion, this patient received orbital radiation and chemotherapy but developed metastatic disease 14 months after enucleation.
All but two of the patients had immunoglobulin G antiadenoviral titers ranging from 1:8 to 1:128 before treatment. None of the patients had immunoglobulin M antiadenoviral titers before or after treatment, nor did any of the patients have a rise in antibody titer after therapy. Two patients did not have measurable titers before or after therapy. Culture and PCR analysis of blood, urine, or nasal washings showed no evidence of adenovirus in any of the patients.
DISCUSSION
Although surgical intervention remains the definitive therapy for unilateral retinoblastoma6 and has close to a 90% cure rate in the absence of metastasis,22 there is clearly a need to develop therapies for bilateral retinoblastoma that spare the eye and any visual function it may retain. Local control using cryotherapy or laser photocoagulation may meet this need but can only be used when the tumor is small or when vitreous tumor seeds are absent.23,24 External-beam radiation therapy is useful occasionally in the treatment of vitreous tumor seeds but is accompanied by high morbidity including radiation retinitis, closed-angle glaucoma, and second malignancies in the field of radiation in patients with constitutional mutations of the retinoblastoma gene.
Preclinical gene-therapy studies in the ocular environment have shown considerable promise for the treatment of oncologic,19 retinal degenerative,25 and neovascular26 diseases. These studies have paved the way for clinical trials such as the one described in here. In this phase I study, eight patients with vitreous tumor seeds were treated with a total of 21 injections of AdV-TK. All the injections were tolerated well. The patients all presented with extensive vitreous tumor seeds away from the main tumor, a finding that predicts failure of standard therapy. These patients therefore would be classified as group Vb of the Reese-Ellsworth classification (Ia-Vb) for conservative treatment of retinoblastoma (radiotherapy)27 or groups D and E of the newly proposed international retinoblastoma classification (A-E), which aims to correlate tumor characteristics with response to chemotherapy and local therapies.28 Patients who present with disease at these stages usually encounter treatment failure with current therapeutic modalities. The vitreous tumor seeds of all eight of the patients responded to the gene-therapy treatment, and all seven patients treated at a dose 1010 vp had apparent resolution of vitreous tumor seeds as determined by indirect ophthalmoscopy. At the time of histopathologic evaluation of the enucleated eyes, all but patient 4 had vitreous seeds directly overlying active tumors. Whether these vitreous seeds were a recurrence of previously treated seeds or new extensions from active primary tumors that occurred after gene therapy cannot be determined. At the time of this writing, all the children were alive and without metastatic disease or systemic complications that resulted from the adenoviral vector administration. The one child who developed metastatic disease had extension of disease from a primary tumor through the emissary vessel adjacent to a brachytherapy scar, treatment that was used after gene therapy in an attempt to eradicate remaining retinal disease and preserve the eye. The needle track and the injection site of the viral vector in this patient were free of disease, suggesting that the disease extension was not secondary to gene-therapy administration. The gene-therapy protocol required that all selected patients had experienced treatment failure with all previous therapies and were facing imminent enucleation. This patient was 6 years old at the time of gene-therapy treatment and had received treatment for his retinoblastoma for 4 years before being enrolled onto the gene-therapy protocol. The primary reason that the patients had therapy failure was the presence of vitreous seeds. Although this protocol was a phase I study, the vitreous seeds responded to the therapy, which allowed the ophthalmologist to attempt additional standard therapies to control the primary tumor. Such therapies in these extensively treated patients must be used with extreme caution.
Our results indicate that transfer of the herpes simplex thymidine kinase gene followed by treatment with ganciclovir is both safe and feasible. The unanticipated clinical responses that were observed suggest that this approach may have value as a supplement to current therapeutic approaches. This therapy did not exhibit cross resistance in patients who had experienced chemotherapy and radiotherapy failure. Because the injected adenoviral vector cannot diffuse through vitreous, each focal area of vitreous tumor seeds had to be injected separately. This was accomplished by repositioning the needle without removing it from the vitreous to minimize the chance of tumor spread. Patient 2, for example, who had vitreous seeds throughout the eye, had the dose of viral vector divided equally into each quadrant of the eye.
The injection technique that was used in this trial is a modification of the transcorneal approach that was developed to biopsy intraocular tumors.21 In this study reported by Shields et al, this technique was used safely in more than 150 patients with different intraocular lesions, including retinoblastoma, and never resulted in a case of metastatic disease by dissemination of tumor cells through the needle tract. None of our patients had a tumor biopsy. The transcorneal approach was used only to inject the adenoviral vector, and no tumor seeding along the needle tract was detected.
The standard diagnostic criteria for retinoblastoma are based on indirect ophthalmoscopic examination. Although there are other diseases such as Toxocara canis infection and Coat's disease that can mimic retinoblastoma, all the patients reported in this study had typical findings of retinoblastoma complicated by vitreous seeds that had been confirmed by multiple examinations by different ophthalmologists over several months. Patient 4, who had classical findings of a regressed primary tumor after chemotherapy and radiation treatment and multiple large, fluffy vitreous seeds before gene-therapy treatment, had an enucleation because of the apparent persistence of vitreous seeds and an imaging report (MRI) consistent with recurrence of her primary tumor with invasion into the optic nerve head. The appearance of her vitreous seeds had changed from her classical initial presentation. Histopathologic examination revealed an inflammatory response unique in our experience and probably related to the gene-therapy treatment. Only a benign retinocytoma consistent with response to prior therapy and mild edema of the optic nerve head was observed.
Because gene therapy for cancer relies on the destructive capacity of the transgene, persistence is not a desired attribute for the vector. Instead, one seeks high-level transduction (from a high-titer vector) with high-level gene expression. For this purpose, first-generation adenoviral vectors remain unsurpassed and therefore were used in this trial. Multiple previous studies have documented that AdV-TK followed by ganciclovir can kill retinoblastoma cells and decrease the size of retinoblastomas in vivo in a murine xenograft model.19 For successful treatment, both the administration of the viral vector and ganciclovir were necessary, suggesting that delivery of active herpes thymidine kinase to the tumor had occurred. One advantage of this therapeutic regimen is that it produces a pronounced bystander effect. Phosphorylated ganciclovir, the cytotoxic product of the herpes simplex thymidine kinase reaction, is transported to adjacent cells through gap junctions.29 Thus, not every tumor cell must be transduced by the viral vector for the therapy to be effective. Although suicide gene-therapy treatment has been shown to be safe when used for prostate cancer30 and brain tumors,12 and transgene expression has been clearly documented, clinically significant efficacy was not demonstrated. Given this background, in this study we chose not to measure transgene expression, because it would have required early enucleation after gene transfer. Instead, we elected to directly measure tumor response as a nonsurrogate marker of efficacy.
Regulatory requirements for this study necessitated that all patients enrolled must have experienced treatment failure with extensive systemic and local therapies including external-beam radiation therapy. Although these therapies disrupt the blood-retina barrier, presumably violating the immune privilege status of the eye, only mild to moderate degrees of local inflammation were observed in our patients after the intraocular injections. The inflammation was controlled by treatment with local and systemic steroids. None of the patients developed a measurable humoral or cellular immune response against the vector, and adenovirus was not detected by culture or PCR in the saliva, blood, or urine of treated patients up to 6 weeks after the completion of gene therapy. Two patients developed corneal edema that resulted in significant loss of vision and increased intraocular pressure. These toxicities were observed only at the highest dose level (1011 vp). Because we saw efficacy and less toxicity with patients treated at 2.5 x 1010 vp, this dose will be used in future studies. Some degree of condensation of vitreous resulting in loss of vision was observed in all patients. The degree of condensation seemed to be related to the amount of vitreous seeds that were being treated rather than the dose of viral vector that was being administered. Vision seemed to improve over time. Patients 1 and 3 had better vision after therapy. None of the patients had damage to the retina. Therefore, if the vitreous opacities could be removed (ie, vitrectomy), the patient's vision could be improved. This procedure should only be attempted in patients who have no evidence of active retinoblastoma for at least 2 years because of the risk of metastasis. Less toxicity may be observed in patients with an earlier presentation or less severe disease and in patients who have not received therapies that would disrupt the blood-retina barrier. In addition to expanded trials that will further study the utility of gene therapy to treat vitreous tumor seeds, future clinical studies are being developed that will define the toxicity and efficacy of direct injection into the primary retinal tumor.
In this phase I trial we have shown that adenoviral vectors used to deliver the herpes simplex thymidine kinase gene can be administered to children's eyes containing retinoblastoma with an acceptable safety profile. All the patients ultimately required enucleation, emphasizing that gene therapy administered in this trial could not be used as a stand-alone treatment for cure of retinoblastoma. These results suggest that gene therapy may be a useful adjuvant to other standard therapies for the treatment of children with vitreous seeds that are complicating retinoblastoma. Future studies will be designed to further develop the role that gene therapy may play in the therapy of this disease.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank Drs Célia Antoneli, Juan Carlos Bernini, Raul Cordero, Doris Hadjistilianou, Carlos Leal, and Paul Salmonsen for patient referrals; Marlen Dinu, Olga Arroyo, Jeri Gates, Sheila Love, and Shanna Noel for assistance with patient management; and Bambi Grilley for assistance with regulatory compliance.
NOTES
National Institutes of Health Grant No. CA97762 defrayed costs of the viral vector and together with the Foundation for Research and Retina Research Foundation provided personnel support. National Institutes of Health Grant No. RR00188 and funding from the Texas Children's Hospital covered patient care expenses, and funding from the Fleming-Davenport Fund defrayed patient-transportation costs.
Both P.C.-B. and M.C. contributed equally to this work.
Presented in part at the annual Association for Research in Vision and Ophthalmology meeting April 25-29, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Devesa S: The incidence of retinoblastoma. Am J Ophthalmol 80:263-265, 1975
Margo C, Hidayet A, Kopelman A, et al: Retinocytoma: A benign variant of retinoblastoma. Arch Ophthalmol 101:1519-1531, 1983
Knudson AG Jr: Mutation and cancer: Statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820-823, 1971
Dryja TP, Rapaport JM, Weichselbaum R, et al: Chromosome 13 restriction fragment length polymorphisms. Hum Genet 65:320-324, 1984
Harbour JW: Overview of RB gene mutations in patients with retinoblastoma: Implications for clinical genetic screening. Ophthalmology 105:1442-1447, 1998
Shields CL, Shields JA: Recent developments in the management of retinoblastoma. J Pediatr Ophthalmol Strabismus 36:8-18, 1999
Gallie BH, Budning A, DeBoer G, et al: Chemotherapy with focal therapy can cure intraocular retinoblastoma without radiotherapy. Arch Ophthalmol 114:1321-1328, 1996 [Erratum: Arch Ophthalmol 115:525, 1997]
Murphree AL, Villablanca JG, Deegan WF III, et al: Chemotherapy plus local treatment in the management of intraocular retinoblastoma. Arch Ophthalmol 114:1348-1356, 1996
Roarty JD, McLean IW, Zimmerman LE: Incidence of second neoplasms in patients with bilateral retinoblastoma. Ophthalmology 95:1583-1587, 1988
Abramson DH, Beaverson K, Sangani P, et al: Screening for retinoblastoma: Presenting signs as prognosticators of patient and ocular survival. Pediatrics 112:1248-1255, 2003
Shields CL, de Potter P, Himelstein BP, et al: Chemoreduction in the initial management of intraocular retinoblastoma. Arch Ophthalmol 114:1330-1338, 1996
Trask TW, Trask RP, Aguilar-Cordova E, et al: Phase I study of adenoviral delivery of the HSV-tk gene and ganciclovir administration in patients with current malignant brain tumors. Mol Ther 1:195-203, 2000
Teh BS, Ayala G, Aguilar L, et al: Phase I-II trial evaluating combined intensity-modulated radiotherapy and in situ gene therapy with or without hormonal therapy in treatment of prostate cancer: Interim report on PSA response and biopsy data. Int J Radiat Oncol Biol Phys 58:1520-1529, 2004
Havlik R, Jiao LR, Nicholls J, et al: Gene therapy for liver metastases. Semin Oncol 29:202-208, 2002
Banerjee D: Technology evaluation: Gene therapy (mesothelioma), NCI. Curr Opin Mol Ther 1:517-520, 1999
Sung MW, Yeh HC, Thung SN, et al: Intratumoral adenovirus-mediated suicide gene transfer for hepatic metastases from colorectal adenocarcinoma: Results of a phase I clinical trial. Mol Ther 4:182-191, 2001
Alvarez RD, Gomez-Navarro J, Wang M, et al: Adenoviral-mediated suicide gene therapy for ovarian cancer. Mol Ther 2:524-530, 2000
Morris JC, Wildner O: Therapy of head and neck squamous cell carcinoma with an oncolytic adenovirus expressing HSV-tk. Mol Ther 1:56-62, 2000
Hurwitz MY, Marcus KT, Chevez-Barrios P, et al: Suicide gene therapy for treatment of retinoblastoma in a murine model. Hum Gene Ther 10:441-448, 1999
Chen SH, Shine HD, Goodman JC, et al: Gene therapy for brain tumors: Regression of experimental gliomas by adenovirus-mediated gene transfer in vivo. Proc Natl Acad Sci USA 91:3054-3057, 1994
Shields CL, Shields JA, Baez KA, et al: Choroidal invasion of retinoblastoma: Metastatic potential and clinical risk factors. Br J Ophthalmol 77:544-548, 1993
Tamboli A, Podgor MJ, Horm JW: The incidence of retinoblastoma in the United States. Arch Ophthalmol 108:128-132, 1990
Shields JA: The expanding role of laser photocoagulation for intraocular tumors. The 1993 H. Christian Zweng Memorial Lecture. Retina 14:310-322, 1994
Shields CL, Santos MC, Diniz W, et al: Thermotherapy for retinoblastoma. Arch Ophthalmol 117:885-893, 1999
Acland GM, Aguirre GD, Ray J, et al: Gene therapy restores vision in a canine model of childhood blindness. Nat Genet 28:92-95, 2001
Mori K, Gehlbach P, Ando A, et al: Regression of ocular neovascularization in response to increased expression of pigment epithelium-derived factor. Invest Ophthalmol Vis Sci 43:2428-2434, 2002
Reese AB, Ellsworth RM: Management of retinoblastoma. Ann N Y Acad Sci 114:958-962, 1964
Shields CL, Mashayekhi A, Demirci H, et al: Practical approach to management of retinoblastoma. Arch Ophthalmol 122:729-735, 2004
Elshami AA, Saavedra A, Zhang H, et al: Gap junctions play a role in the "bystander effect" of the herpes simplex virus thymidine kinase/ganciclovir system in vitro. Gene Ther 3:85-92, 1996
Herman JR, Adler HL, Aguilar-Cordova E, et al: In situ gene therapy for adenocarcinoma of the prostate: A phase I clinical trial. Hum Gene Ther 10:1239-1249, 1999(Patricia Chévez-Barrios, )