Although the ch14.18 and 3F8 antibodies both share the same target onneuroblastoma cells, the GD2 antigen, there are significant differencesbetween them. The two antibodies differ in structure, in protocoldesign and availability, in comparisons of study results, and in methodof administration. For the purposes of this article, we will compareand contrast the use of these antibodies in upfront therapy forchildren with high-risk neuroblastoma. Antibody therapy for childrenwith refractory and relapsed disease is outside the scope of thisarticle.
There are other antibodies used against neuroblastoma, includinghumanized antibodies (and more are in development) but so far the useof those antibodies has been limited to small studies for relapsed orrefractory patients. This discussion will center on ch14.18 and 3F8because these have been used as part of frontline therapy in hundredsof patients over the past two decades.
Structure of ch14.18 and 3F8
There is a notable difference in the structure between the twoantibodies. The ch14.18 is a chimeric (combination) antibody. It ispart mouse (about 25%) and part human (about 75%). By contrast 3F8 is a100% murine (mouse) antibody. This is a considerable distinction andit has far reaching impact.
The biggest issue with murine antibodies like 3F8, 14G2a, and 14.18 (amurine version of ch14.18) is that they are extremely immunogenic. Inother words, the human immune system can have a substantial response tothese antibodies. While this strong response can be good if the immunesystem is recruiting cells to attack the neuroblastoma tumor cells,often the response includes the creation of blocking antibodies to theantibody itself. The human anti-mouse antibody (HAMA) response withmurine-derived antibodies has been detected within 7 days of treatmentand can neutralize further treatments with the mouse anti-GD2 antibody.This prompted development of increasingly human antibodies such asch14.18. The ch14.18 antibody retains the Fab fragment of the mouseantibody with its binding specificity attached to the Fc portion of ahuman antibody.
The earliest published results from a phase I using ch14.18 in childrenwith neuroblastoma was a German phase I study published in 1995. Inthis study 5 of 9 patients had complete response, partial response,minor response, or stable disease.
In 1987 Sloan-Kettering published their Phase I study results using 3F8in neuroblastoma and melanoma, and 7 of 17 patients had a responseranging from complete remission to mixed response.
In September of 2004 the GPOH published the results of theirretrospective study in the Journal of Clinical Oncology which statedthat the use of the ch14.18 antibody for consolidation treatment ofstage IV neuroblastoma had no clear impact on the outcome of patientsby comparing two sequential trials NB90 and NB97 (not randomized). Some have speculated that this apparent lack of effectiveness waspossibly due to different dosing and schedule, lack of the use ofcytokines, and or lack of using cis-retinoic acid. However, at theAdvances in Neuroblastoma Research meeting in 2008 this same grouppresented an abstract which showed that after further retrospectiveanalysis it became clear that immunotherapy with ch14.18 asconsolidation after intensive chemotherapy may actually prevent laterelapses. After observing patients for a median of 9.2 years, incontrast to the earlier report, the paired log rank test now clearlydemonstrated an advantage of ch14.18 treatment compared to no antibodytherapy.
On March 19, 2009 the COG announced that after careful review of theearly ANBL0032 results made available through their independent data andSafety Monitoring Committee, that the immunotherapy(experimental) arm of the study ? a combination of ch14.18 antibody,cytokines (IL2 and GMCSF) and Isotretinoin (also called Accutane orcommonly abbreviated as cisRA) ? more ffectively reduces the risk thatneuroblastoma will grow back than treatment with cisRA alone. They havealso determined that the immunotherapy as specifically delivered on COGANBL0032 increases the chance of survival after completion of therapyincluding stem cell transplantation when compared to treatment withcisRA alone.
Method of administration of ch14.18 and 3F8
The administration of these antibodies have some similarities and somedifferences. Both antibodies are given intravenously, and vital signsare carefully monitored during infusions for reactions and pain.
In the COG study, ANBL0032, the ch14.18 antibody is administered in aninpatient setting. The infusion is five hours per day over a period offour days per cycle. The current protocol includes five cycles everyfour weeks. This protocol includes two different cytokines. Cycles 1,3, and 5 include the use of GM-CSF and cycles 2 and 4 include the useof IL-2. Currently it is not known whether GM-CSF or IL-2 has a betterimpact when used with ch14.18.
The study in the SIOP, SIOP-EUROPE-HR-NBL-1, utilizes ch14.18 withoutthe use of cytokines. Ch14.18 is infused via IV over 8 hours on days1-5. Treatment repeats every 28 days for 5 courses.
3F8 is offered in an outpatient setting. The infusions typically lastone hour. Each cycle includes five days of infusions. During thefirst four cycles each cycle is 21 days apart. Assuming the patienthas not experienced a HAMA these cycles will then continue every 49days until either a HAMA is achieved or treatment has lasted twoyears. All cycles for children in first remission utilize GM-CSF.
We have highlighted some similarities and differences between the 3F8and ch14.18 antibodies. While there is ample evidence of anti-tumoreffect in both, proof of improved survival has only been derived fromthe recent randomized study of ch14.18 when the antibody is given aftertransplant and with GM-CSF, Il-2, and cis-retinoic acid..