The locally advanced non-small cell lung cancer (LANSCLC) stage groupings include stage III patients (any T N2-N3M0, T4N0M0, and T3N1M0) and those stage II patients (T2B-T3NOM0 and T1-2N1M0 who cannot undergo a definitive resection due to medical inoperability. Locally advanced NSCLC makes up about 30-35% of the NSCLC patient population.¹

Stage III NSCLC patients appropriate for combined modality therapy should have good performance status (ECOG 0 or 1) and minimum weight loss (less than 5% in the preceding three months). The importance of these simple criteria is that when patients are chosen that violate these rules, analysis of outcome shows that they do not benefit as much from intensive therapy. Additionally, patients with poor performance status and weight loss have a high risk of serious complications including treatment related death from toxicity.  In carefully selected patients with ECOG PS 2 or weight loss up to 10%, concurrent chemoradiation may be a consideration if the attending consultants are mutually agreed that this is appropriate. Patient-related and tumour-related factors can influence the balance of risks vs benefits; patient preferences should also play a significant role.

Because of the intensive nature of the treatment for potentially curable locally advanced lung cancer, staging must be precise to avoid aggressive combined modality treatment in a patient that is incurable with metastases. In addition to CT scanning, positron emission tomography (PET) should be done routinely to exclude distant metastases and help define the disease in the thorax requiring radiotherapy. A favourable distribution of thoracic disease for definitive local therapy (radiotherapy/surgery) is a major factor.¹ Additionally, many patients will require pathologic assessment of the lymph node stations by endobronchial ultrasound biopsy (EBUS) or mediastinoscopy to minimize false positive or negative CT and PET scan results.

Patients with locally advanced NSCLC have high risk disease but when treatment can be given with curative intent, it is necessary to organize investigation and treatment in a timely fashion. Rapid disease progression may occur with delays as has been documented by Mohammed et al.² Using serial PET and CT scans, the authors showed 3% of untreated patients developed distant metastases at 4 weeks and 13% at 8 weeks. Restaging should be considered if there is a delay of 8 weeks from the most recent CT scan. Thus, delay not only impacts on resource utilization but directly on patient outcome.

Although age is not a clear selection criteria, generally speaking, patients over 75 years of age must be very fit and motivated to receive multi-modality therapy.  Co-morbid conditions impact treatment selection for all disciplines. Definitive local therapy with surgery and thoracic irradiation requires adequate pulmonary function. Standard chemotherapy regimens require adequate hematologic, hepatic and renal function.

It is crucial that all treating oncologists (thoracic surgeons, radiation oncologists and medical oncologists) mutually agree on the proposed treatment plan and the suitability of the individual patient for such therapy before any component of treatment has commenced. Such cases should be presented at a multi-disciplinary lung conference.  Of all patients with locally advanced NSCLC, a minority rather than a majority will be suitable for aggressive treatment with curative intent.

Active patient participation in decision making respects the fundamental ethical and legal doctrine of autonomy, while exploring the concerns of these individuals, including functional limitations, symptoms and side effects/risks of therapy. Patients should have some idea of the quantity of median and long-term survival gain expected with intensive therapy to make an informed choice on the pros and cons of a conservative versus a more aggressive treatment program.  Unfortunately, the majority of locally advanced NSCLC patients will experience an incurable relapse and it is unfair for them to go into a complex and toxic treatment program under the false impression that the probability of cure is high. On the other hand, the gain in median survival and the increase in the proportion of long-term survivors with multi-modality therapy are sufficient that this arduous treatment should be recommended for appropriately selected patients.

In patients with symptomatic infiltrative stage III (N2, N3) NSCLC and either performance status 3 or 4, comorbidities, or disease too extensive to treat with curative intent, palliative radiotherapy is recommended. The fractionation pattern should be chosen based on the physician’s judgment and patient needs.

In patients with stage III NSCLC with performance status of 0-1 and weight loss < 5%, radiotherapy alone is not recommended. Should the patient refuse multi-modality therapy, definitive radiotherapy alone with curative intent may be an alternative plan associated with a lower probability of long-term survival. Similarly, patients with co-morbid conditions, performance status of 2 and/or weight loss of 5-10% that are felt unsuitable for multi-modality therapy may be considered for definitive radiotherapy alone according to the radiation oncologist’s judgment.  

Conventional thoracic radiotherapy alone in unresectable stage III NSCLC using conventional (2 Gy or equivalent to a total dose of 60 Gy) has been associated with a median survival of about 1 year and a 5-year survival of 5-7%.³

With platinum containing chemotherapy regimens, investigators have improved survival by combining chemotherapy with thoracic irradiation in patients with good performance status and minimal weight loss. Three meta-analyses reviewing more than 50 trials confirm the survival benefit of combined platin-based chemotherapy with radiotherapy over radiotherapy alone in locally advanced, unresectable NSCLC.^4,5,6

Chemotherapy has been combined with radiotherapy in different ways (chemotherapy followed sequentially by radiotherapy, concurrent chemoradiation, induction chemotherapy followed by concurrent chemoradiation, or concurrent chemoradiation followed by consolidation chemotherapy). For curative intent treatment of locally advanced NSCLC, concurrent chemoradiation is recommended because it improves local control and overall survival compared with sequential chemotherapy followed by radiation or radiation alone. In multiple randomized trials and a meta-analysis (Auperin A. J Clin Oncol 28; 2181-2190) induction chemotherapy followed by radiotherapy (sequential chemoradiation) was compared to concurrent chemoradiation. Concurrent chemoradiation was associated with superior 5-year survival and better local control at the cost of a higher rate of reversible esophagitis.

The standard dose fractionation of radiation with concurrent chemotherapy is 60 Gy given in fractions of 2 Gy once per day over 6 weeks.^3,11,12 Dose escalation beyond 60 Gy with conventional fractionation has not been demonstrated to be of benefit.¹³ In RTOG 0617,¹³ 74Gy was compared with 60Gy and was not found to be superior. Although the RTOG dose escalation study did not have toxicity as the primary endpoint, analysis of side effects suggests that planning radiotherapy with IMRT compared with 3DCRT allowed treatment of larger tumours with a lower risk of severe pneumonitis.

Surgical recommendations depend on category of N2 disease and the performance status of the patient. For the first group with infiltrative N2 disease, surgery is not recommended. Surgery plays a role in the second and third group of patients. 

When involved N2 lymph nodes are identified pre-operatively by invasive mediastinal staging, surgery alone is not a recommended treatment option. Pre-operative chemotherapy followed by surgery has been investigated in randomized trials. A meta-analysis showed the absolute survival benefit at 5-years was 6% and downstaging to pT0 in the surgical specimen by chemotherapy was only 6-7%.¹⁶ Most patients were given thoracic irradiation after surgery because of unfavourable pathology reports.  Protracted sequential deployment of chemotherapy followed by surgery followed by radiotherapy is not recommended.

A number of trials have examined pre-operative chemotherapy and radiotherapy after surgery.^17,18,19 The results have not always showed benefit for tri-modality therapy versus chemoradiation alone, particularly when a series of neo-adjuvant chemotherapy treatments were given before pre-operative thoracic irradiation.^18,19  

There remains a persistent interest in the scientific community not to abandon surgery in stage III NSCLC because of the sub-optimal local control of the primary lung tumour achievable with thoracic radiotherapy or chemo-radiotherapy.   Additionally, surgery provides the most definitive staging.

The Intergroup 0139 phase III randomized trial compared trimodality therapy (pre-operative concurrent chemoradiation (cisplatin/etoposide, 45 Gy) followed by surgery versus bimodality therapy with chemoradiation (61 Gy) has been influential on BC Cancer Lung Tumour Group Management Guidelines.¹⁷ There was no overall survival advantage to the trimodality arm (median survival 23.6 months versus 22.2 months) but progression-free survival was significantly improved (12.8 months versus 10.5 months). Overall, 88% of patients in the surgical arm were eligible for thoracotomy and 71% had complete surgical excision and only 4.5% had no resection. Notably, pathologic downstaging after chemoradiation showed pT0 in 20% and another 20% had only microscopic residual disease in the previously irradiated volume. Pathologic downstaging with pre-operative chemoradiation appears clearly superior to chemotherapy alone.¹⁶ Unfortunately, in this multicenter trial, experience with trimodality therapy was limited in some centres and serious toxicity problems occurred. Among those requiring pneumonectomy,  post-operative mortality was 26%; it was 50% for complex right pneumonectomy.  In an exploratory unplanned, matched subgroup analysis, patients treated with a lobectomy after induction concurrent chemoradiation had a significantly better survival than those judged suitable for lobectomy but treated non-surgically on the control arm.

While acknowledging the controversies in using trimodality therapy for stage III NSCLC, the BC Cancer Lung Cancer Lung Tumour Group endorses using it for a carefully selected stage III patient population. This is based on the fact that a number of groups^20,21 including the BC Cancer Lung Tumour Group²¹ have shown that carefully selected patients may undergo trimodality therapy including pneumonectomy with low (2%) operative deaths. Our experience with trimodality therapy (cisplatin/etoposide concurrently with thoracic irradiation (45-60 Gy) followed by surgery has yielded median survival of over three years and 5-year survival of 35%. Although the favourable outcome is associated with careful selection as well as the treatment, the results are good enough to continue this policy, particularly when local thoracic surgery expertise has such consistent low morbidity and mortality rates over many years.

In summary, the key to optimal management of discrete N2 disease is concurrent chemotherapy and radiotherapy. Surgical resection has a role in carefully selected patients with limited burden of disease, good performance status, good pulmonary function and patient preference.

With regard to surgical therapy, even within the category of discrete N2 disease, there are subdivisions that have been studied in the literature that carry favourable or unfavourable characteristics. Groups that carry favourable prognosis include patients with complete pathologic response, ypN0 disease even with residual viable tumour in the primary site, patients requiring resection less than pneumonectomy, non-enlarged lymph nodes and single-station N2 disease. Although these groups of patients carry a better prognosis, it is difficult to make decisions on surgical recommendations based on their presence or absence. Many of these factors are not readily identifiable preoperatively, and hence cannot be used to guide decisions about surgery. In general, we do not use these criteria solely to make a decision about candidacy for surgery for discrete N2 disease, rather each piece of data is considered in the context of the individual patient. After discussion at a multidisciplinary lung tumour board, it is appropriate to consider all patients with good performance status and discrete N2 stage IIIA NSCLC for a strategy of pre-operative therapy followed by surgery. This includes patients with multi-station N2 disease and patients that may require pneumonectomy (assuming their pulmonary function allows for such a resection). 

Trimodality therapy requires a dedicated multidisciplinary assessment with an experienced thoracic surgeon, radiation oncologist and medical oncologist.  In cases of bulky locally advanced disease where the surgical treatment is uncertain, the radiation oncologist may prescribe 60 Gy rather than 45 Gy as the pre-operative radiotherapy prescription in case the surgical treatment is not recommended. A crucial step in this process is the surgeon’s assessment of resectability after concurrent chemoradiation. Imaging has typically been with CT.  Repeat PET scanning has been proposed to identify patients where surgery would be futile.²² This has been associated with some controversy because of the high incidence of false positive results when PET is performed after chemoradiation.²³ A comparison across studies suggests that the survival was better in those studies in which patients underwent resection despite the lack of PET response vs those studies in which the patients generally did not undergo resection.  There is insufficient evidence to mandate repeat PET after induction chemoradiation to assess operability but it may be useful at the surgeon’s discretion in cases where there is an unsatisfactory CT response or suspicion of progression.

At this time, there is no evidence that additional chemotherapy or any type of targeted therapy is indicated after surgical resection in trimodality therapy, even if the pathology report is unfavourable.