Introduction (by Myrna R. Rosenfeld, M.D., Ph.D.)

The observation in the early 1990’s that patients with newly diagnosed anaplastic oligodendroglioma (AO) often showed responses to the combination of procarbazine, lomustine, vincristine (PCV) led to two prospective, randomized, phase III trials aimed at evaluating the impact of neoadjuvant (RTOG 9402) or adjuvant (EORTC 26951) PCV in conjunction with standard radiation therapy (RT) for patients with newly diagnosed AO. At the time enrollment started it was unlikely that the investigators would have predicted that the story of these trials would continue to unfold over 17 years.

The initial results were published in 2006 in back to back articles in the Journal of Clinical Oncology [1, 2]. After a median follow-up of 5 years the results demonstrated no overall survival benefit with either regimen. However, both studies demonstrated a difference in the time to tumor progression that strongly favored the subjects who received chemotherapy. The use of PCV was associated with a significant rate of hematological toxicity with grade 3 or 4 toxicity reported in up to 65 % of patients. Due to the uncertain benefit and added toxicity of PCV neither of these regimens became a generally accepted standard of care.

Both studies validated the prognostic value of 1p/19q co-deletion [3]. The studies had been conceived before the importance of the co-deletion was known and thus were amended when it was shown that 1p/19q co-deletion was a marker for response and improved outcome in this population. For example, in RTOG 9402 patients with 1p/19q co-deletion had median survival times of 7 years compared to less than 3 years for those without the co-deletion (P ≤ 0.001). While neither study had a chemotherapy-only arm, these results led some practitioners to delay radiation until progression in patients with AO and 1p/19q co-deletion with the goal of avoiding the late cognitive effects associated with radiation. In contrast, the poor-survival of patients with non-deleted tumors which was similar to that seen in glioblastoma was impetus for some to treat this cohort with the radiation and temozolomide regimen shown in 2005 to benefit patients with newly diagnosed glioblastoma [4].

Neuro-oncologists must now reconsider these approaches in light of the data derived from updated analyses of the RTOG 9402 and EORTC 26951 studies. Now, with a median follow up of 11.3 years, RTOG 9402 data shows that the overall survival of AO patients with 1p/19q co-deletions who received neoadjuvant PCV was significantly longer (14.7 years) than those who received only radiation (7.3 years, P = 0.03). In contrast, the addition of chemotherapy to radiation had no impact on the overall survival of those subjects with non-deleted tumors (2.6 vs 2.7 years, P = 0.39) [5]. Similar long-term outcome data were reported for adjuvant PCV in EORTC 26951 [6].

These data strongly support a new standard of care for 1p/19q co-deleted AO. The question being widely asked is whether temozolomide can substitute for the more toxic PCV regimen. An answer was to come from the intergroup study: “Newly diagnosed AA/AO/AO with 1p/19q codeletion (CODEL, NCCTG N0577).” This study opened before the new analyses of RTOG 9402 and EORTC 26951 were available. Patients with anaplastic gliomas were to be randomized after surgery to radiation with concurrent and adjuvant temozolomide, radiation alone, or temozolomide alone. The study was appropriately put on hold when the long term analyses demonstrated the inferiority of the radiation only arm. Whether the study will reopen with a PCV plus radiation arm is under consideration.

What we do know is that it will be a long wait for these data to be forthcoming leaving open the question of what is a neuro-oncologist to do? There are those who feel that adjuvant or neoadjuvant PCV should be offered to all AO patients with 1p/19q co-deletions. For others this is not so clear. Perhaps PCV feels like a step back. Newly trained neuro-oncologists likely have little experience with the regimen and for all, PCV toxicity is a consideration. But can or should temozolomide replace PCV based on currently available information? In this point and counterpoint, Dr. John Villano of the Markey Cancer Center, University of Kentucky advocates for PCV by noting the strength of the RTOG and EORTC data. He reminds us that multi-agent and often highly toxic chemotherapy is commonly used in systemic cancers especially when there is potential for durable long-term benefit. Dr. Patrick Wen and colleagues from the Center for Neuro-oncology, Dana-Farber Cancer Institute, present the counterpoint, noting among other points that there is data suggesting that PCV is not necessarily an optimal regimen.

Point (by J. Lee Villano, M.D., Ph.D.)

Based on the recent results of the pivotal EORTC 26951 and RTOG 9402 randomized studies, administration of PCV combination chemotherapy should be accepted as the standard of care for newly diagnosed AOs with 1p/19q heterozygous co-deletions [1, 2, 5, 6]. These two studies provide remarkably similar outcomes that have important implications for neuro-oncology [1, 2].

The studies demonstrate the superiority of PCV chemotherapy administered peri-radiation therapy for patients with the 1p/19q heterozygous co-deletions. While no benefit was seen in patients lacking the co-deletions, one-fourth to nearly one-half of patients with AO are expected to have the molecular findings of 1p/19q heterozygous co-deletions. In the RTOG 9402 trial, the survival benefit for patients with PCV administration before RT, versus RT alone, was more than 7 years [5]. In the EORTC 26951 study, median overall survival has not yet been reached in the RT + PCV arm with follow-up over 12 years; the control RT arm has an overall survival of 9.4 years. Survival improvement with chemotherapy measured in several years is rare in our field [6].

Although both studies were randomized to early PCV with RT versus RT alone, the studies more correctly tested the timing of PCV administration as the majority of patients in the control arms received PCV at progression. Early therapy therefore with PCV and RT is essential in AO with 1p/19q co-deletion as salvage treatment is inferior. This is reminiscent of the seminal EORTC 22981/26981–NCIC CE.3 (EORTC/NCIC) study of temozolomide for treatment of glioblastoma [4]. The majority of patients in the control arm (RT alone) also received temozolomide as salvage therapy and the outcomes were inferior compared to the investigation arm of concurrent RT and temozolomide followed by six cycles of maintenance temozolomide.

A second major point is the concern of increased hematologic toxicity with PCV treatment. It is important to note that despite the fact that more patients did not receive the planned number of cycles the PCV arm did better for 1p/19q co-deletions. The correlation between number of cycles received and outcome was not presented and data regarding late toxicity and quality of life (QOL) are planned. In the EORTC 26951 study, patients in the experimental arm received a median of three out of six planned cycles of PCV; only 30 % received the full six cycles. In RTOG 9402, only 42 % of patients tolerated the full four cycles of PCV. These studies leave unanswered how many cycles of treatment are needed.

The remarkable findings of the two studies are clear and confirmatory: PCV + RT is the standard of care in AO with 1p/19q co-deletion. PCV administration schedules differed in the two trials: PCV was given as an intensive regimen for four cycles prior to RT in RTOG 9402, and PCV was administered in six cycles after RT in EORTC 26951. The RTOG 9402 study had more hematologic toxicity with 64 % of patients experiencing grades III–V toxicities (including two deaths secondary to treatment). In EORTC study 46 % of subjects had grade III or IV toxicities with no treatment associated deaths. Therefore, the EORTC dosing schedule will likely dominate.

It is possible that the earlier publication of both studies in 2006 stating progression-free survival benefit but not overall survival with PCV has negatively affected the field. However, these two studies have now moved 1p/19q co-deletion from a prognostic marker for improved clinical outcomes to a predictive marker for improved survival when PCV is added to radiation.

Although there are arguments such as vincristine not crossing the blood–brain barrier (BBB), or that the toxicities of administrating PCV are excessive compared to temozolomide, these should not deter use [7]. Note that the BBB is not entirely intact in AO [8, 9] and that toxicities from temozolomide including aplastic anemia can be significant and fatal [10]. While PCV is relatively more toxic when compared to other regimens used in CNS tumors in the context of treating other solid tumors PCV toxicities are not extreme. It is also noteworthy that use of multi-agent cytotoxic chemotherapy is the standard for treating chemotherapy-sensitive neoplasms. This includes germ cell, small cell lung cancer, colon cancer, leukemia, lymphoma, and multiple myeloma among others.

Another potential barrier for adoption is the study design. The inception of both studies was prior to identification of 1p/19q co-deletion as a prognostic marker for improved outcomes and sensitivity to treatment in AO. The complementary findings presented by both studies makes the initial unplanned statistical analyses less of a concern, and it is important to remember it is the reason why we do these large studies—to provide an answer and move forward.

In summary, EORTC 26951 and RTOG 9402 are pivotal studies that demonstrate remarkable improved outcomes in patients with 1p/19q co-deleted AO treated with RT and PCV as initial therapy. Except for rare patients having neuropathy, bone marrow suppression, or concerns of receiving radiotherapy, the majority of patients should receive RT with PCV as the standard of care [11]. The CODEL study will investigate the role of temozolomide in AO with 1p/19q co-deletion and a PCV arm is under consideration; regardless results will take well-over a decade and until then one cannot assume equivalency between PCV and temozolomide. The good news is we have positive results now from two large randomized phase III trials that give our patients the best chance to be alive the longest.

Counterpoint (by Patrick Y. Wen, M.D., Eudocia Q. Lee, M.D., M.P.H., Lakshmi Nayak, M.D. and David A. Reardon, M.D.)

Until recently the treatment for AOs involved a wide range of options including RT alone, chemotherapy alone (mainly with temozolomide), or combined RT and chemotherapy [12]. The recent phase III studies from RTOG and EORTC conclusively showed that the addition of PCV chemotherapy to RT significantly increased survival compared to RT alone and established combination therapy as the standard of care for AOs [5, 6]. While PCV is occasionally used in Europe, the majority of neuro-oncologists in North America have been using temozolomide chemotherapy [13]. However, since PCV was the chemotherapy regimen used in both the RTOG and EORTC studies, an increasing number of neuro-oncologists are abandoning temozolomide and adopting the PCV regimen. The CODEL study will directly compare the PCV regimen to temozolomide in patients with newly-diagnosed AOs and hopefully address the relative effectiveness of the two therapies, but these results will not be available for many years. In the meantime the increasing use of PCV potentially exposes patients to a more toxic regimen that may not necessarily be more effective.

Arguments in favor of PCV include the fact that it is the regimen used in the RTOG and EORTC trial for which there is class I evidence, while similar evidence is not available for temozolomide. There was also a large retrospective analysis of therapy for AOs in which patients treated with PCV had an improved outcome compared to patients treated with temozolomide. The median time-to-progression following treatment with PCV was 7.6 years, compared to 3.3 years (P = 0.019) [14]. However, given the retrospective nature of the study, with patients receiving PCV and temozolomide treated in very different eras, it is difficult to know the true validity of these findings.

In contrast, the PCV regimen has significant issues related to increased toxicity and there is no clear data demonstrating greater efficacy compared to temozolomide in other types of high-grade gliomas. The regimen was first proposed by Victor Levin for the treatment of high grade gliomas and subsequently adopted and modified by Cairncross and colleagues [15, 16] for AOs. While the lomustine and procarbazine components probably do achieve adequate concentrations in gliomas, there is increasing evidence that vincristine does not cross the blood brain barrier [7]. If this regimen was developed in the modern era, vincristine would not be included. It is very unlikely to contribute to the efficacy of the regimen but undoubtedly adds to the toxicity.

Most studies suggest that the PCV regimen is more toxic than temozolomide. In the EORTC trial 32 % of patients had grade 3 and 15 % had grade 4 hematologic toxicity [2, 6]. Overall, 38 % of patients in the PCV arm discontinued therapy because of toxicity. In the RTOG study, utilizing an intensive PCV regimen, 65 % of patients experienced grade 3 or 4 toxicities and 2 patients died of treatment-related complications [1, 5]. Moreover, the marrow suppression in some patients can be long lasting and potentially reduce their ability to receive full dose chemotherapy at recurrence. In contrast, in the EORTC/NCIC trial of temozolomide with RT for newly diagnosed glioblastomas, only 16 % of patients experienced grade 3 or 4 hematologic toxicities [4]. The BR12 study comparing temozolomide to PCV in 447 recurrent high-grade glioma patients who were chemotherapy naive, did not show a significant difference in toxicity between PCV and temozolomide but since this study used a lower dose PCV regimen consisting of procarbazine (100 mg/m2) on days 1 through 10, and lomustine (100 mg/m2), and intravenous vincristine (1.5 mg/m2 capped at 2 mg) on day 1, comparison with the other studies is difficult [17]. Overall, the available data suggests standard and intensive PCV regimens are associated with increased toxicity compared to temozolomide and may have an adverse effect on the QOL of these patients. A limitation of the current data is the lack of objective studies comparing QOL measures in anaplastic glioma patients receiving PCV compared to temozolomide. While such studies are available for patients with anaplastic gliomas treated with PCV [18, 19], similar studies are not available for patients treated with temozolomide. The proposed CODEL study will incorporate such QOL measures and provide important information on the relative effects of the two regimens.

There are also questions regarding the efficacy of PCV compared to temozolomide in high-grade gliomas. In the Medical Research Council Brain Tumor Working Group study, PCV chemotherapy produced no survival benefit in newly-diagnosed high-grade gliomas compared to RT alone [20]. While the majority of patients had glioblastomas, 17 % (113 patients) had anaplastic gliomas. PCV did not produce a survival advantage in this subgroup, although the numbers are small. Similarly in the BR12 study comparing temozolomide to PCV in recurrent high-grade glioma patients, including 103 patients with grade III gliomas, there was no survival advantage for PCV [18]. While this study was not formally designed to compare PCV with temozolomide in patients with AOs, they did not show any striking difference in outcome between the two therapies. In contrast the EORTC/NCIC trial showed that the addition of temozolomide to RT in patients with newly-diagnosed glioblastoma increased median survival by 2.6 months and 2-year survival from 10 to 27 % [4]. While the current data suggest temozolomide is likely to be more effective than PCV for glioblastomas, the relative efficacy of the two agents in AOs with 1p/19q deletion is unknown and will be examined in the CODEL trial.

It remains unclear whether patients with AOs would benefit more from PCV than temozolomide. There is no evidence in other types of high-grade gliomas that PCV is more effective than temozolomide. However, there is data suggesting that the vincristine component probably does not pass through the BBB but contributes to neurotoxicity, and there is unequivocal evidence that the PCV regimen as a whole produces significantly more hematologic toxicity. Unless the CODEL trial shows that PCV is superior to temozolomide, patients with AOs could reasonably be treated with RT and temozolomide. At a minimum, PCV should be PC, but to reduce exposing patients to unnecessary toxicity, it would preferably be temozolomide.