Up to a few years ago, whole brain radiation therapy (WBRT) was the clear standard intervention for patients with lung cancer and brain metastases. Although median survival after WBR has been typically measured as a several months, that clearly exceeded the survival without treatment (typically in the range of weeks). There are several schedules, but the most common one has been 30 Gray (Gy) delivered in 10 three-Gy fractions over a two week period (Mon-Fri x 2 weeks) (abstract here). In addition to improving survival, WBRT also has a good likelihood of improving symptoms. Generally, about 60% of patients experience a complete or partial response to WBRT in terms of tumor shrinkage and/or improvement in neurologic symptoms. The dose is not likely to be enough to definitely control disease in the brain long-term, but the dose and schedule evolved as a compromise of trying to reduce the duration, minimize side effects, and control lung cancer in the brain for reasonably well. All too often, disease outside of the brain is progressing as well, so pushing the brain metastases onto the back burner has been the main achievement. It has remained the cornerstone of treatment for multiple brain mets, and particularly so for SCLC, where patients are particularly likely to have multiple lesions in the brain.
The other setting where WBRT is commonly recommended is after either surgery or radiosurgery (“gamma knife”) for one or a few lesions, in order to reduce the risk for further brain metastases. Studies have shown that on MRI patients with a brain metastasis have an 80% likelihood of an additional lesion, 50% have more than three, and over 70% of patients will develop recurrence in the brain if WBRT is not included after resection or radiosurgery (reference here).
The downside is that there is a real risk of developing significant symptoms from WBRT itself, including dementia or even, rarely, death (abstract here). It has actually been quite hard to assess the degree of neurocognitive decline in patients who receive WBRT, because detailed cognitive testing has often demonstrated a subtle deficits before starting WBRT, presumably as a function of their brain metastases. One study shows that memory (delayed recall) initially declines by three months after WBRT but then improves, even beyond pre-WBRT baseline (abstract here):
One other way to assess the potential long-term detrimental effects of brian radiation is to assess memory and other cognitive functions in patients who have undergone prophylactic cranial irradiation, or PCI, which is part of the standard treatment for SCLC (reviewed in in prior post here), particularly limited but sometimes also extensive SCLC, and is also being studied carefully in locally advanced NSCLC (prior post on the subject here). A study of PCI in 75 patients with stage III NSCLC (abstract here) looked at long-term neurocognitive problems with detailed testing and found that, among the 15 patients who were alive and agreed to testing 3 years after treatment, patients who received PCI but also those who did not receive PCI had subtle deficits, and both groups had some MRI changes in the “white matter” (the area of the brain that has the insulation around the neurons of the brain), although more extensive in recipients of PCI. This same trial saw clear improvements in the risk of subsequent brain failure and also a benefit in overall survival with PCI that was felt to overshadow the potential detrimental cognitive effects of PCI, according to the authors. Small trial, though, so it’s hard to say anything definitive.
In fact, reviewing these studies, it’s fair to say that there is a lot of variability in the reports. Most of the work in this area suggests that there is a small (1-5%) but real risk of significant and permanent cognitive problems, and that there are common but much more subtle deficits, which are found more or less frequently depending on how hard you look for them. Moreover, deficits are clearly NOT limited just to people who receive brain radiation, and many people (myself included) feel that the risk of uncontrolled brain metastases and all of the problems this causes often far exceed the risk of WBRT itself. This was clearly less of a “devil’s bargain” when patients were quite unlikely to live more than a few months after a diagnosis of brain metastases, but now we have the fortunate problem of an ever-growing proportion of patients with brain metastases who are living longer and longer due to better treatments of brian disease (including gamma knife and surgery in some case) and better treatment for disease outside of the brain. Because of this, we need to weigh more carefully the risks of WBRT vs. the anticipated advantages. It still remains the standard for patients with multiple NSCLC lesions (at least more than 2-3) and nearly all patients with brain metastases from SCLC (where risk of multiple lesions is especially high), but it’s not risk-free, and not everyone will want to accept the risks of serious or even mild deficits.
Obviously, it would be very helpful to have therapies that could protect cognitive function, and I’ll turn to potential neuroprotective therapies next.