Do treatments that help the body’s immune system fight cancer cells (immunotherapy) make patients with non-small cell lung cancer (NSCLC) who have had surgery or radiotherapy aimed at a cure, live longer?
Many people with NSCLC, who have had surgery or radiotherapy to cure their cancer, eventually die because the cancer comes back, either in the chest, or somewhere else in the body. There have been a number of clinical trials over the years that have looked at whether immunotherapy helps patients live longer. Some seemed to show a benefit, others did not.
We searched four computerised databases and five trial registers to 20 January 2017. We looked for all trials that randomly allocated participants to one treatment or another (randomised controlled trials (RCTs)), and included adults (aged 18 years or older) with early non-small cell lung cancer (stages I to III), confirmed by laboratory testing of a sample of the tumour. We found nine RCTs, which included nearly 5000 participants who had received surgery or curative radiotherapy, and were randomly allocated to receive either immunotherapy or no further treatment.
We found that giving immunotherapy, mainly vaccine-based (aiming to activate the host immune system to induce human immune response to tumour-specific antigens), after surgery or radiotherapy did not, on average, make people live longer. We did not find any results that could tell us whether the addition of immunotherapy improved the quality of life, but it seemed that those who were given vaccine-based immunotherapy may have experienced, on average, more side effects. At the moment, there is no evidence to support or refute giving immunotherapy (mainly vaccine-based) to people with localized NSCLC (stages I to III). RCTs are in progress that are testing new, more promising immunotherapy drugs (e.g. checkpoint inhibitors).
Quality of the evidence
The evidence we found about overall survival and progression-free survival was high quality. When we looked for evidence about how many patients lived to one, two, three, or five years, it was only moderate or low quality, because the RCTs were not very well done, and their results did not agree with each other.
The current literature does not provide evidence that suggests a survival benefit from adding immunotherapy (excluding checkpoint inhibitors) to conventional curative surgery or radiotherapy, for patients with localised NSCLC (stages I to III). The addition of vaccine-based immunotherapy might increase the risk of adverse events. Several ongoing trials with immune checkpoints inhibitors (PD-1/PD-L1) might bring new insights for role of immunotherapy for patients with stages I to III NSCLC.
Non-small cell lung cancer (NSCLC) is the most common lung cancer, accounting for approximately 80% to 85% of all cases. For patients with localised NSCLC (stages I to III), it has been speculated that immunotherapy may be helpful for reducing postoperative recurrence rates, or improving the clinical outcomes of current treatment for unresectable tumours. While several new agents have now entered phase III clinical trials, we felt a systematic review was needed to address the question of the effectiveness and safety of immunotherapy in patients with stages I to III NSCLC.
To evaluate the effectiveness and safety of immunotherapy (excluding checkpoint inhibitors) in patients with localised NSCLC (stages I to III) who received surgery or radiotherapy with curative intent.
We searched the following databases (from inception to 20 January 2017): CENTRAL, MEDLINE, Embase, and CINAHL, and five trial registers. We also manually checked abstracts or reports from relevant conference proceedings and the reference lists of included trials.
We searched for randomised controlled trials (RCTs) in adults (≥ 18 years) with histologically-confirmed early-stage (stages I to III) NSCLC after surgical resection, and those with unresectable locally advanced stage III NSCLC who had received radiotherapy with curative intent. For patients who had received primary surgical treatment, postoperative radiotherapy or chemoradiotherapy was allowed if it was used for both experimental and control groups.
Two review authors independently selected eligible trials, assessed risk of bias, and extracted data. We used survival analysis to pool time-to-event data, expressing the intervention effect as a hazard ratio (HR). We calculated risk ratios (RR) for dichotomous data, and mean differences for continuous data, with 95% confidence intervals (CI). Due to clinical heterogeneity (immunotherapeutic agents with different underlying mechanisms), we used random-effects models for our meta-analyses.
We identified nine eligible trials that randomised 4940 participants, who had received surgical resection or curative radiotherapy, to either an immunotherapy group or a control group. Included immunological interventions were active immunotherapy (i.e. Bacillus Calmette-Guérin (BCG)), adoptive cell transfer (i.e. transfer factor (TF), tumour-infiltrating lymphocytes (TIL), dendritic cell-cytokine induced killer (DC-CIK), and antigen-specific cancer vaccines (melanoma-associated antigen 3 (MAGE-A3) and L-BLP25). Except for one small trial, which provided insufficient information for risk assessment, we assessed five studies at high risk of bias for at least one of the seven biases studied; we considered the risk of bias in the other three trials to be low. We included data from seven of the nine trials in the meta-analyses (4695 participants). We pooled data from 3693 participants from the three high quality RCTs to evaluate overall survival (OS) and progression-free survival (PFS). We found a small, but not statistically significant, improvement in OS (HR 0.94, 95% CI 0.83 to 1.06; P = 0.35), and PFS (HR 0.93, 95% CI 0.81 to 1.07; P = 0.19; high-quality evidence). The addition of immunotherapy resulted in a small, but not statistically significant, increased risk of having any adverse event (RR 1.15, 95% CI 0.97 to 1.37; P = 0.11, three trials, 3955 evaluated participants, moderate-quality evidence), or severe adverse events (RR 1.10, 95% CI 0.88 to 1.39; four trials, 4362 evaluated participants; low-quality evidence).
We analysed data from six studies for one-, two-, and three-year survival rates (4265 participants), and from six studies for five-year survival rates (4234 participants). We observed no clear between-group differences (low-quality evidence for one- and two-year survival rates, and moderate-quality evidence for three- and five-year survival rate).
No trial reported the overall response rates; only one trial provided health-related quality of life results.