When I was a medical student, the question about lung cancer that was always asked on "the Boards" had to do with the difference between stage IIIA and stage IIIB non-small cell lung cancer (NSCLC). The reason this question was always asked is because patients with stage IIIA NSCLC might be considered for surgery, whereas patients with stage IIIB NSCLC would not be considered for surgery and instead would be treated with chemotherapy and radiation. The idea is that young doctors should be able to make that distinction and to direct patients to the appropriate specialist/treatment. While I guess it makes a good test question, this distinction is too simplistic and doesn't really give anyone a good understanding of the complexities of managing stage III lung cancer. And, in reality, all patients with suspected stage III lung cancer should be evaluated by a multidisciplinary team that includes thoracic surgeons, radiation oncologists, pulmonologists and medical oncologists. If the Medical Board would write a test question aimed at getting across this important principle, I'd breathe a big sigh of relief for lung cancer patients.
First, what makes a lung cancer stage III (as opposed to I, II or IV)? Let's start with a lesson in anatomy and staging. The staging of lung cancer is determined by the size and invasiveness of the primary tumor (T), the location of any lymph nodes involved (N) and whether there is any evidence of metastasis to other organs, including the lining of the lung (pleura) or the opposite lung (M). The hilum is the highway into the lung - it contains the airways and blood vessels that enter the lung and branch off into the various lobes and segments of the lung. The trachea (windpipe) branches off into the right and left mainstem bronchi and these enter each lung through the hilum. The blood vessels that supply the lung also enter at the hilum, as do the lymphatic channels. Lymph nodes at the hilum can be involved with a lung cancer. These are considered N1 nodes, and most cases of lung cancer with just N1 involvement are stage II.
Most commonly, patients are classified as stage III because of involvement of lymph nodes in the mediastinum (a space in the middle of the chest between the two lungs). If the involved mediastinal lymph nodes are on the same side of the chest as the lung tumor (ipsilateral nodes), this is classified as N2 and is considered stage IIIA disease. If the involved lymph nodes are on the opposite side of the mediastinum (contralateral nodes), these are considered N3 nodes, and the cancer is stage IIIB. If a patient has involvement of supraclavicular lymph nodes (just above the collar bone, or clavicle) on either side, these are considered N3 lymph nodes, and the stage is likewise IIIB.
The assignment of lymph nodes to N1, N2 or N3 based on location is not arbitrary: the prognosis changes depending on the site of lymph node involvement. In general, lung cancers with the best prognosis are N0 (no lymph node involvement). As we go from N1 to N3, the risk of recurrence of the cancer increases. A patient may also be classified as having stage III lung cancer even if no lymph nodes are involved, if the tumor invades another organ in the chest such as the heart, large blood vessels, esophagus or bones of the spine (T4).
There are a variety of ways to establish the diagnosis of locally advanced NSCLC. Most patients with lung cancer have a CT of the chest and we may see enlarged lymph nodes on the CT. Alternatively, lymph nodes in the chest may not be particularly enlarged but may light up on a PET scan. Although abnormal lymph nodes on PET are suspicious, other processes such as infection or inflammation can cause lymph nodes to light up on a scan and most experts recommend that suspicious mediastinal lymph nodes on PET get biopsied to confirm that the patient has stage III disease. This is usually accomplished through a mediastinoscopy, EBUS or EUS.
So let's say that a mediastinoscopy is performed for PET-positive lymph nodes in the mediastinum, and multiple lymph nodes come back with cancer in them. This patient has stage IIIA NSCLC. Most lung cancer experts would recommend radiation and chemotherapy given at the same time (concurrent therapy) for a patient with multiple mediastinal lymph nodes involved. Why? Patients with stage III NSCLC have a high risk of recurrence, both locally in the chest and at distant sites, such as bone, liver, etc. Although radiation can address the tumor in the chest, it cannot kill off micrometastases that are already floating around.
Several clinical trials have examined whether chemotherapy improves survival when added to radiotherapy for stage III NSCLC - the answer is yes! So why not do chemotherapy and radiation sequentially rather than concurrently? Again, the answer comes from clinical trials: giving therapy concurrently results in better outcomes for patients. Chemotherapy acts as a radiosensitizer so in addition to its potential benefit in reducing metastases at distant sites, it also helps radiation to work better. The downside of concurrent therapy is that it is definitely more toxic than giving chemo and radiation sequentially, particularly with respect to inflammation of the esophagus (esophagitis), which is associated with difficulty swallowing.
If you are uncertain about which chemotherapy regimen should be given with radiation for stage III NSCLC, you are not alone. Why not just give the same drugs we use in stage IV NSCLC with radiation? First, chemotherapy drugs differ in how much they sensitize cancer cells (and normal tissues) to radiation. Some drugs, like cisplatin and etoposide, can be given at full doses and provide the right amount of radiosensitization. Taxol (paclitaxel) and Taxotere (docetaxel) can be given with radiation, but the doses must be reduced substantially, to a level that would not be considered as effective at fighting micrometastases as full doses. Gemcitabine is an incredibly potent radiosensitizer and is not considered safe in combination with radiation for lung cancer.
The most commonly used regimens are cisplatin and etoposide at full dose, or carboplatin and Taxol given weekly at low doses with radiation. These two regimens have never been compared head to head, but if we look at different studies that have used the two regimens, we see that the studies which used cisplatin typically have longer median survival (approaching the two year mark) compared to the studies that used weekly carbo/taxol. Toxicity is typically greater with full dose cisplatin compared to low-dose carboplatin, and so a patient who is weaker or has significant medical problems (such as kidney dysfunction) may not tolerate a full-dose cisplatin combination. There are certainly many oncologists who favor carbo/taxol weekly with concurrent radiation, but many experts favor cisplatin-based chemo for the subset of stage III NSCLC patients who can tolerate it.
A recent trial reported results incorporating not one but two newer drugs with radiotherapy. CALGB 30407 was a randomized phase II trial looking at two new combinations: carboplatin and Alimta (pemetrexed) plus radiation, with or without the EGFR monoclonal antibody Erbitux (cetuximab). The goal was not to compare these two regimens (the numbers are too small in a randomized phase II study to draw firm comparisons), but rather to explore promising combinations. Both regimens showed a median survival of about 22 months, which compares favorably with historic controls. It is important to note that this trial included patients with squamous cell cancers, a group who we now know does NOT appear to benefit from Alimta. Currently, a randomized phase III study comparing cisplatin/Alimta/radiation to the old standard cisplatin/etoposide/radiation is underway - only patients with non-squamous cancers are being included.
Another area of significant controversy involves what to do after completion of radiation. Patients who have had surgery for NSCLC typically get 4 cycles of chemotherapy as adjuvant treatment. Patients with stage IV NSCLC are typically given 4-6 cycles of a platinum-based combination. We have good data for giving chemotherapy with radiation in stage III disease (typically two cycles are given with radiation) but we are always tempted by the prospect that more is better. Unfortunately, available studies do not support the role of "consolidation" chemotherapy or targeted therapy after completion of concurrent chemoradiation. The Hoosier Oncology Group (HOG) trial randomized patients to either observation or 3 cycles of Taxotere after patients had completed standard cisplatin and etoposide (2 cycles) with radiation. The patients who got Taxotere did not live longer, and they had more treatment-related complications compared to the patients who were observed.
Many made the argument that perhaps Taxotere was just too toxic and that a less toxic consolidation therapy such as one of the oral epidermal groth factor receptor EGFR) inhibitors, Iressa (gefitinib)or Tarceva (erlotinib), would let the glory of consolidation therapy shine through. Unfortunately, we were wrong again.
SWOG S0023 randomized patients with stage III NSCLC to either Iressa or placebo following a challenging regimen of cisplatin/etoposide/radiation, followed by consolidation Taxotere. The trial unexpectedly showed that survival was significantly worse for patients who received Iressa compared to placebo.
Even more surprising was the finding that excess deaths in the Iressa arm were not due to toxicity but rather apparently to lung cancer progression.
A more recent trial comparing placebo to Tarceva after concurrent chemoradiotherapy was more encouraging, with a trend towards improved survival for patients who received consolidation Tarceva.
This difference did not reach statistical significance, and I would say that most lung cancer researchers are not advocating Tarceva after chemoradiation at this time.
Am I forgetting something? Oh, yes - the radiation component. Although medical oncologists would like to take all of the credit for the gradual gains we have made in survival for patients with locally advanced NSCLC, we must acknowledge that tremendous improvement in radiation techniques has occurred over the last 30 years. Most of the trials that established the benefit of concurrent chemoradiation delivered radiation doses of around 60 Gy. With better techniques that can mitigate damage to normal tissues, radiation oncologists are looking to push the envelope by escalating the dose of radiation. A large, multicenter clinical trial is hoping to answer the question of whether higher doses of radiotherapy result in improved outcomes by comparing 60 Gy to 74 Gy concurrent with chemotherapy.
One final issue bears mentioning: relapse in the brain is common for patients with locally advanced NSCLC, occurring in up to 30% of patients. In up to 20%, the brain may be the only site of relapse. Chemotherapy is generally felt to do a poor job of penetrating the blood-brain barrier, so even our most effective regimens are not going to kill off micrometastases in the brain. In small cell lung cancer (SCLC), prophylactic brain radiation for limited stage patients, and more recently even patients with extensive SCLC has been shown to lower rates of brain relapse and improve survival. We sought to answer the question of whether this would hold true for NSCLC but unfortunately the clinical trial aimed at answering this question was closed early because of low accrual. Although brain metastases were reduced in the group who got brain radiation, we could not show any improvement in survival. We will never know if this was because it doesn't improve survival or because we just didn't have enough patients. I think that if this trial had accrued enough patients we would have shown a benefit for those who receive radiation. Nevertheless, brain radiation does come with the risk of neurologic side effects and ideally, a new systemic therapy that crosses the blood brain barrier and works elsewhere in the body would be the way to go. I can dream...
So finally, in summary:
1. Locally advanced, unresectable NSCLC is a common problem, accounting for approximately 1/3 of new lung cancer diagnoses.
2. The optimal way to treat locally advanced NSCLC remains in question though most would advocate chemo and radiation concurrently (without surgery) for patients with bulky lymph nodes or multiple sites of lymph node involvement.
3. The optimal chemotherapy regimen remains in question. Strategies to incorporate more effective therapies are being explored.
4. Concurrent chemoradiation is toxic: strategies to reduce toxicity are needed.
5. We think higher doses of radiation will be better (within reason) and hope to answer this question through an ongoing clinical trials.
6. Relapses in the brain are common and we need better ways of addressing this problem.