HS-PCI in Locally Advanced Adenocarcinoma of the Lung

Patients with locally advanced non-small cell lung cancer (LA-NSCLC) have an increased risk of developing central nervous system (CNS) metastases during the course of their disease. The brain is the most common site of failure after first-line therapies (independent of combined sequential or consolidation chemotherapy). Recent studies employing multimodal therapy have reported overall brain metastasis rates ranging from 22% to 55%, and the rates for brain as first site of relapse range from 16% to 43%.

Prophylactic cranial irradiation (PCI) results in a 2-3 fold lower incidence of brain metastasis. However, randomized studies have failed to demonstrate improved overall survival (OS) after PCI.

One of the major weaknesses of these trials is the unselected mixed pool of stage III patients (stage III A and B, lymph node status N0 to N2, squamous and non-squamous histology etc.). A broad variety of studies have shown that a certain subset of patients with NSCLC (e.g. cancers with adenocarcinoma histology, multilevel nodal involvement) are at highest risk for brain metastases. Furthermore, the risk for brain metastases appears to be specifically higher in younger patients (age <60 years), although this collective commonly undergoes more frequently chemotherapy and/or more aggressive regimens than elderly patients.

Prevention of CNS metastases, even for LA-NSCLC patients with other sites of failure, will improve quality of life and, for patients controlled extracranially, will improve survival. Meta analyses performed on data from several Radiation Treatment Oncology Group (RTOG) studies have shown that longer survival for patients with LA-NSCLC treated with either radiation alone or radiochemotherapy is associated with an increased incidence of CNS metastases. Although the addition of chemotherapy to radiation therapy reduces extracranial distant metastases and improves survival it does not alter brain relapse rates. Even though the addition of modern targeted therapy using small-molecules or antibodies may further improve the outcome, the CNS remains the most common site of failure under targeted therapy, although no evidence for resistance in histological workups of metastases has been found. This emphasizes the urgent need for treatment directed at chemotherapeutically inaccessible (or dormant) micrometastases that are a priori dispersed within the brain. As the median time for relapse in the CNS is approximately 6 months after first-line therapy, the treatment of micrometastases should be meaningfully initiated even during or shortly after first-line therapy.

Irradiation of the brain does not only bear the risk of inducing acute (partially mass-associated) side effects such as nausea, vomiting and fatigue, but also causes long-term neurocognitive deficits. Although neurocognitive disorders after PCI/Whole brain radiotherapy (WBRT) also have a multifactorial etiology based on a patient's individual medical history (preceding chemotherapy, pre-existing vascular damage e.g. from smoking, local reactions/edema), it is currently believed that they are mostly caused by a loss of neural stem cells in the hippocampal areas. Multipotent and self-renewing neural stem cells are found in the subgranular zone of the adult hippocampus and in the subventricular zone of the lateral ventricles. The hippocampus plays an important role in memory consolidation and emotional learning (contextual fear conditioning). The disruption of neurogenesis in the subgranular zone or damage to the hippocampus can lead to impaired short- and long-term memory, learning and contextual fear conditioning. In line with this, irradiating the brain decreases neurogenesis in the hippocampus which leads to impaired hippocampal-dependent learning and memory.

To prevent radiation-induced loss of neuronal stem cells, hippocampus-sparing (HS) radiation techniques have been developed and efficacy has been demonstrated in the recently published phase II RTOG 0933 study. The trial included patients with brain metastases and a Karnofsky Performance Scale (KPS) of 70%. Following HS-WBRT the patients showed a relative neurocognitive function (NCF) decline of 7% four months after HS-WBRT, which is more than four times less than observed in studies with conventional WBRT (30%; p<0.001).

Since the study was a one-arm study without a control, the reported hippocampal failure rate of 4.5% remains controversially discussed. Multiple studies described the hippocampus and limbic circuit to be a generally rare site of brain metastases in many cancers. NSCLC shows a specifically low rate of hippocampal brain metastasis (2.8% of all brain metastases) and risk modelling revealed a only slightly increased absolute risk (+0.2%) after HS-WBRT. Thus, since in the treatment of NSCLC, the efficient prevention of BM and potential CNS micrometastases is currently outweighed by the associated neurotoxicity and a lack of survival benefit, HS-WBRT may provide a possibility to tip the scale toward prophylactic WBRT, at least for a specific subgroup at high risk.

Radiation regimens for PCI that have influenced patterns of CNS failures in NSCLC have included total doses of 20-36 Gy and fraction sizes of 2-3 Gy. A fraction size of 2 Gy and a total dose of 30 Gy was chosen for the RTOG 0214 study of PCI in LA-NSCLC. This regimen has been shown to decrease CNS metastases from 54% to 13% with no difference in NCF decline in PCI versus non-PCI patients at 4 years. In addition, although there is paucity of clinical data from WBRT with doses in the EQD2 (equivalent dose in 2 Gy fractions) range of 10-20 Gy, a dose-response curve providing a 'best fit' model suggests an only minimal benefit from doses above 30 Gy (EQD2).