I’m going to cover the general concepts of management of brain metastases, a subject that is still evolving because of our growing technology, particularly with stereotactic radiosurgery (SRS), commonly referred to as gamma knife. In many cases, our practice has moved a bit ahead of the data. We’ll start with some general issues and then, over several posts, cover issues from surgery to radiation to medical therapy.
Brain metastases are very common, somewhere in the range of 150,000 - 170,000 patients with cancer who develop brain metastases each year in the US, which makes it the most common complication of cancer (abstract here). Lung cancer is the most common underlying cause, accounting for something in the range of half or a bit more of those cases (abstract here). In SCLC, more than 10% of patients have brain metastases at the time of diagnosis, and more than 50% of those patients with SCLC who aren’t treated with prophylactic cranial irradiation (PCI) will develop brain mets later, generally within two years of diagnosis (abstract here). In NSCLC, PCI has also been considered due to the fact that up to up to half of the total number of people with NSCLC develop brain metastases as well at some point; this has become a common “sanctuary site” for disease to recur first or recur only after treatment for stage I - III disease.
The location where brain metastases end up is proportional to where blood flows throughout the brain. That is, 80% of brain metasatasis are deposited in the large cerebral hemispheres, 15% in the cerebellum that sits at the base of the head, and 5% are in the brainstem (abstract here). People develop symptoms when the tumor cells grow and lead to inflammation and surrounding edema (swelling) in the area of the brain, increasing pressure. The symptoms people develop relate to the neurologic function of the area where the metastasis has developed, so if it’s an area of motor control for the left arm, that’s what people will notice, while the cerebellum controls balance, so metastases there lead to gait unsteadiness and falls. The most common symptom related to new brain metastases is a headache, which occurs in around 25-50% of patients, often due to the increased pressure in the brain/skull from the swelling (abstract here). It’s also common to see confusion and/or forgetfulness, focal weakness, such as in an extremity, seizures, imbalance with walking, vision changes or less, difficulty with articulating speech, and sensation changes (numbness, etc.).
On exam, more than half of patients have areas of detectable weakness or mental status changes, but the changes can be quite subtle, and it has been estimated that up to 30% of cases may not be detected (abstract here).
Head/brain scans with CT or MRI are the way that the vast majority of brain metastases are detected (or confirmed if strongly suspected). Given the high frequency of brain metastases in SCLC in particular, current guidelines for SCLC or a tumor that is a mix of SCLC and NSCLC include an MRI of the brain as part of the initial workup. Recommendations for head imaging in NSCLC are more variable and controversial, with more consensus that stage III, possibly stage II NSCLC patients being considered for curative therapy undergo head imaging before pursuing aggressive curative therapy. There are not clear recommendations favoring brain imaging for earlier stage patients who have no symptoms suggestive of neurologic problems. Although head imaging for patients with metastatic disease else where has not been routine practice previously, standards are now changing for patients in whom avastin is being considered, because this agent is not recommended for patients with brain metastases due to concern for increased risk for intracranial (inside the skull) bleeding around brain metastastases, which is often fatal. Unsuspected brain metastases can be seen in somewhere in the range of 10-15% of asymptomatic patients with metastatic NSCLC, so most oncologists will check rather than presume an asymptomatic patient has no brain metastases.
There is little to no question that head MRIs are superior to CT scans in detecting brain metastases. MRI scans detect a greater number of lesions and defines the location more readily, and they are also better at detecting spread to the meninges, the lining around the brain (and spinal cord). Here is the appearace of a couple of MRI images of a patient of mine who was having increasing confusion and difficulty with his balance (and golf game, which is what he really cared about):
PET scans are also incorporated routinely in the initial staging of patients with lung cancer. However, PET scans are not ideal for detecting brain metastases because the brain has a high baseline uptake of glucose (we hope). One study comparing PET scans to head MRIs showed that PET scans detected only 61% of the lesions detected by MRI (abstract here). Therefore, while a very large proportion of patients with lung cancer will undergo a PET or combined PET/CT scan in their staging workup, we don’t feel confident that this addresses the question of whether there are brain metastases, and an MRI to answer the question is generally recommended if the level of suspicion for brain metastases is high enough.
We’ll cover the prognostic and treatment issues in the next few posts. Brain metastases are a big, complex problem, so we’ll need to cover the topic in some depth to really do it justice.
posted by Dr. West @ 5:53 pm link to this post





April 6th, 2007 at 10:45 am
Hi Dr. West,
My husband has a cardiac pacemaker and the info with that device say no MRI’s for him. So far, he has had no evidence of brain mets, but I know that is a possibility down the road. Would a CT be sufficient to pick that up and indicate whether he might need WBR?
Thanks so much!
Tk
April 6th, 2007 at 1:17 pm
I forgot to add that he did have an MRI about 10 months ago when they were doing his original diagnostic workup and it was clear. He got the pacemaker since then.
Thanks again!
April 6th, 2007 at 9:14 pm
That’s not a rare situation. Also, about 10% of the population can’t tolerate an MRI because they experience profound claustrophobia (the MRI machine is very narrow and very loud, and people often compare it to being in a coffin. A lot of people don’t like that feeling — go figure!).
An MRI is a modestly superior test, but a head CT is quite good as well. So if an MRI is an option, that’s generally preferred, but it’s fine to work with a head CT if there’s a reason for that limitation.
-Dr. West
April 9th, 2007 at 6:30 am
At initial in hospital DX my wife had a head CT scan w/o contrast (ordered by Hospitalist doctors) and it didn’t show any brain metastasis. The appointed oncologist ordered a brain MRI and it showed the multiple mets.
Chanwit.
April 9th, 2007 at 2:29 pm
Dr. West,
How often should one get a Brain MRI? My dad received one last year after he blacked out behind the wheel and they thought he had a seizure. It was clear.
Rochelle
April 9th, 2007 at 10:19 pm
Chanwit,
Head CTs will often show brain metastases, but your wife’s case illustrates why we recommend an MRI if possible. I’ve had occasional patients who present with neurologic symptoms to the emergency department, had a clear CT there, but a subsequent MRI showed subtle metastases. A CT isn’t going to miss a 4 cm metastasis with surrounding edema, but the small lesions of a few millimeters and no surrounding edema have a much better chance of being picked up by an MRI.
Rochelle,
There’s no recommendation on that, which is really to say that we don’t repeat them in the absence of new symptoms. It wouldn’t be a wrong thing to do, but we tend to worry so much more about cancer that declares itself that we don’t go out of our way to find progression that is not detected by repeat chest CT or guided by symptoms. And in the setting of metastatic disease, other problems tend to emerge all too often to worry about progression that is too subtle to find through routine mechanisms. The main purpose of the detailed staging studies is to proactively guide treatment plans. Once treatment has occurred, we don’t hunt for disease as much as respond to it as needed.
-Dr. West
May 26th, 2007 at 4:05 pm
Dr. West:
My dad was just diagnosed with SCLC with mets to the liver on 5/3/07. He completed three chemo treatments on 5/7, 5/8 and 5/9 and will continue the cycle for 6 months. I told him to ask his doctor about PCI and the doctor said they probably wouldn’t do that until they had to. I was under the impression that PCI was done to prevent brain mets, not to cure it once it comes about. Can you please clarify that for me? Thank you. Cheryl
May 26th, 2007 at 8:45 pm
Cheryl,
You’re right about the concept behind PCI — it’s done before there’s a problem, in order to reduce the chance of mets developing (which is all too common for SCLC, a tumor that finds the brain to be very fertile soil for its mets to thrive). However, PCI is much more of a standard after completion of treatment for limited disease (although it isn’t recommended as often as many experts think it should be, even in the exact right patients, who are those with a complete or very good partial response after treatment for LD-SCLC). For extensive disease (documented with the liver mets), PCI is certainly a reasonable thing to consider for patients who have a very good response, but it’s much less commonly considered.
Importantly, the results of a trial about PCI for ED-SCLC are being presented at ASCO, our biggest oncology meeting of the year, just a week from now, so I’ll relay whatever is reported from that trial. It should be quite relevant for your father, depending on his response to the treatment he’s received thus far.
-Dr. West
May 28th, 2007 at 10:27 am
Thank you so much for the reply, Dr. West. I will look forward to any information you can share after ASCO!
Cheryl