The Synergy Between Radiotherapy and Molecularly Targeted Agents

Division Of Clinical Sciences - Nci

Investigators

Linked publications & trials

Abstract

Recently we reported on our Phase II trial of Valproic Acid/ RT/TMZ in patients with newly diagnosed GBM. Briefly, 37 patients with newly diagnosed GBM were enrolled and administered VPA 25 mg/kg orally divided into two daily doses concurrent with RT and TMZ. The first dose of VPA was given 1 week before the first day of RT at 10 to 15 mg/kg/day and subsequently increased up to 25 mg/kg/day over the week prior to radiation. We used this one week run in period to escalate the VPA dose as the final VPA dose could have had pancreatic/hepatic toxicity. VPA and TMZ related acute toxicities were evaluated using the Cancer Therapy and Evaluation Program Common Toxicity Criteria (CTC) Version 3.0 for toxicity and adverse event reporting as well as the RTOG/EORTC Radiation Morbidity Scoring Scheme. Although VPA has been used for 40 years as an anti-convulsant, often in combination with RT/TMZ, it is typically used at a much lower dose (5-10mg/kg). Overall, 67% of patients completed the VPA/RT/TMZ according to protocol. The patient's median overall survival (OS) was 29.6 months (21-63.8m) and median progression free survival (PFS) was 10.5m (6.8-51.2m). Additionally, the combination was well tolerated; the most common grade 3/4 acute toxicities of VPA/RT/TMZ were blood/bone marrow toxicity (32%), neurological (11%), and metabolic/laboratory (8%). Interestingly, the serum VPA levels were not correlated with the measured grade 3/4 toxicities and most of the toxicities could be attributed to the concurrently administered TMZ. Overall, there were 16 (43%) acute grade 3-5 toxicities, significantly lower than the 6 other Phase II sensitizer/RT/TMZ studies (reviewed in Krauze 2015), which had grade 3-5 toxicities rates ranging from 62-142%. As the pattern of recurrence for GBM is 85% local, the use of re-RT is commonly prescribed. However, the appropriate features of the tumor, patient, and previous radiation doses that might impact on the success of re-irradiation are not well understood. To develop this understanding, we pursued two lines of investigation: the first was surveying Radiation Oncologists that primarily treat patients with recurrent brain tumors. We asked 13 experts in brain tumor re-irradiation (8 US and 5 European) to rate 11 clinical factors impacting their decision to offer re-RT. From this, we found a heterogeneous treatment pattern including the prescribed doses and volumes, but the clinical factors used most often to decide on the appropriateness of offering re-irradiation were: time since previous radiation therapy (92%), recurrent tumor volume (85%), previously administered dose to organs at risk (77%) and patient performance status (69%). These factors have been previously mentioned in multiple publications but this is the first time a ranking of the various clinical factors has been published. Additional findings included a split in the fractionation schedules most often recommended with 25% offering SRS or standard fractionation (2Gy) and 50% offering hypo-fractionated regimens. There was also a difference in the clinical practice of these respondents for the time required between the two courses of radiotherapy with 50% requiring 6m and 50% requiring 12m before they would consider offering re-RT. An additional common perception was that if re-RT is given earlier in the re-treatment schema the patients responded better. Our second investigation was a report of the results for the re-irradiation of patients with brain tumors performed within the ROB. To properly evaluate every patient, primary treatment plans and re-treatment plans we loaded into the current Eclipse software to create cumulative tumor doses and cumulative doses to the organs at risk (OAR). We found a 15% variance in the OAR volumes when comparing the original treatment plan volumes to those generated from the combination of treatment plans. In the 31 patients that had both plans available, the median cumulative Biological Effective Dose 2Gy (BED) combining the two treatments was 96Gy (range 72-112Gy). Using the doses from the combined plans and the mathematical model, Normal Tissue Complication Probability (NTCP), we would have expected complication rates of 25% for the chiasm, 21% for the right optic nerve and 59% for the brainstem. Thus, 25-60% of our patients should have had a long term neurological toxicity. Yet with a median overall survival of 6m (range 0.7-103m) we had no documented grade 3-5 toxicities in any OAR nor evidence of necrosis in the re-RT treatment fields. This included 13 patients that had concurrent chemotherapy with the re-RT course. This underscores the disconnect that we currently have between the mathematical models of neurological toxicity (NTCP or QUANTEC) and the toxicity that is observed in the clinic. Our two publications highlight some of the current issues in GBM re-RT: what is the appropriate total dose and fractionation; the optimal interval between irradiation courses; the role of concomitant therapy with re-RT; and the potential toxicity of the re-RT?