How do you approach treatment of a subtotally resected low grade glioma in a young patient?

The management philosophy for LGG is radically shifting.  For years, the belief was that these were essentially "incurable", slow-growing tumors, and hence the original debate centered around observation versus resection; the observation camp is now squarely in the minority.  The debate then shifted to RT or not and since no RT-associated survival benefit was observed, RT became "optional".  Two recent developments have turned this philosophy around:

1.  Aggressive, supra-radical resection as promulgated by Duffau et al has resulted in a large % of LGG being alive, disease-free, and without malignant transformation, 15+ years later, implying that getting ahead of the tumor, rather than letting the natural history play out is sensible.

2.  The RTOG randomized chemoRt vs RT trial established a categorical survival benefit, and obviously set a new standard, which unfortunately has not been universally adopted for the following reasons:

a.  Many dislike PCV, the backbone in the RTOG trial, and prefer to substitute temzolomide, an area that remains highly controversial.

b.  Some have assumed that since prior RT trials did not show a survival advantage, the benefit seen from the chemoRT combo must be from the chemo, and therefore delaying RT should be just fine; the EORTC trial which attempted to do this, as well as large single institution reports of TMZ first and RT at progression produce median survival results that are around 60-70% of those observed with chemoRT, belying the belief and hope placed in this strategy of TMZ first followed by RT.

c.  Molecular subsetting has been used to try and parse out who should or should not receive this therapy, and the data from this are intriguing, but statistically weak; perhaps the most reasonable of these are the data suggesting a lack of survival benefit with the addition of chemo in the IDH wild-type tumors (which are possibly more like GBM, in their behavior).

Therefore, for me, RT-chemo is the default approach (unless a Duffauesque operation has been achieved), and I have to look for specific reasons why not to recommend this.

Given RTOG9802 showed OS benefit for RT/chemotherapy vs RT alone, which also included those with age <40 and STR, I would favor adjuvant RT AND chemotherapy (Buckner et al, NEJM 2016). Furthermore, there is this new diagnostic criteria based on CIMPACT-NOW consensus that low grade IDH-wildtype astrocytoma with the following 3 molecular features should be now classified as "diffuse astrocytic glioma, IDH-wildtype, with molecular features of GBM", which would require either TERT mutation, EGFR amplification, or combined chromosome 7 gain and 10 loss (Brat et al, Acta Neuropathologica 2018). For these patients, I would treat them just like GBM, ie 60Gy with TMZ. I also think all other IDH-wildtype gliomas still have unfavorable prognosis and would still favor adjuvant RT and TMZ, actually even if GTR. However, for young patients with IDH-mutant oligodendroglioma (with codeletion) or astrocytoma (noncodeleted), if there is minimal residual disease (not all STR are the same) and potential salvage surgery is possible and patient is very motivated to delay RT, I would consider observation as a reasonable alternative. However, I will warn the patient that at time of progression, the RT field will likely be larger and there may be need for additional surgery. Between the two, the data also show oligodendroglioma has a more favorable prognosis, so that may also play into one's calculation. I would treat oligodendroglioma to 50.4Gy and IDH-mutant astrocytoma to 54Gy at 1.8Gy/fr.

In our tumor boards, we sometimes debate about whether we should start treatment at radiological progression or symptomatic progression. I tend to argue for radiological progression instead of symptomatic progression. If the tumor is growing, I think earlier treatment is better to prevent the RT field from getting even larger and to hopefully prevent symptoms. 

Excellent discussion below.  Given that the early vs. delayed XRT shows improvement in seizure-free survival, I think that earlier XRT is particularly beneficial for patients with seizures or large unresected or partially lesions that are already symptomatic. In my experience these large lesions are not "controlled" well by chemotherapy alone.  They will get a few years of chemo, typically with slow growth along the way, until symptoms or growth get worse.  Unfortunately by that time it is more difficult to reverse much of the damage. A large or symptomatic tumor can cause more profound cognitive deficits and quality of life impairment than XRT. I agree with the notion that it is worth considering the viability of future re-resection and it is best to have an optimal surgery if possible prior to chemoXRT.

For most low grade patients I will use 54 Gy in 1.8 Gy fractions.  Although 45-54 Gy are all accepted and there is no clinical data indicating improved outcomes with the higher dose, my rationale is that many (all I have seen) clinical trial protocols for recurrent high grade gliomas require that patients have received a minimum of 54 Gy.  Since these patients will all ultimately progress, I think it worth considering the few extra fractions to help keep their clinical trial options open later should their tumor transform to higher grade.  A better solution would be to adjust clinical trial protocol criteria for recurrent glioma in general, but until then this has been my practice.

In a young patient with a subtotally resected low grade glioma, I give the options of radiation therapy now vs at the time of symptomatic or imaging progression.  In general, I favour delayed radiation therapy given the results of the European study that showed no change in survival outcomes.  At the time of progression, I do not alter the dose or volume (I generally give 5040 cGym 180 cGy per fraction) with a 1cm PTV) based on the molecular status, with just one major caveat.  If the patient has a very large tumor (particularly if involving the bilateral frontal lobes), I would consider further delaying radiation therapy in favour of recommending temozolomide.  This would particularly be true if the patient has 1p/19q+ or is IDH+. The goal would be to delay radiation therapy, with the hopes of delaying any neurocognitive treatment-related side effects. 

As stated above 9802 RT+PCV demonstrated survival advantage over RT alone which took years to see the OS benefit, implying a durable benefit to upfront treatment. So any high risk feature low grade glioma should be considered for adjuvant RT.

In terms of molecular factors, several excellent articles cited above which will hopefully allow us to better select patients we can and should comfortable observe, and those who need additional treatments. IDH wild type even in low grade glioma particularly with other mutations (ie TERT as mentioned above) would imply prognosis more similar to anaplastic or GBM cohorts.

As far as the poll question, for patients with co-deletion the currently open CODEL trial (https://clinicaltrials.gov/ct2/show/NCT00887146) is open to determine the role of TMZ in co-deleted patients. This would apply to low grade (as well as high grade 1p/19q co-deleted gliomas.) The standard RT arm is 50.4 Gy for grade 2 and 54 Gy for grade 3.