Forecasting the occurrence of heatwaves constitutes a challenging issue, yet of major societal stake, because extreme events are not often observed and (very) costly to simulate from physics-driven numerical models. The present work aims to explore the use of Deep Learning architectures as alternative strategies to predict extreme heatwaves occurrences from a very limited amount of available relevant climate data. This implies addressing issues such as the aggregation of climate data of different natures, the class-size imbalance that is intrinsically associated with rare event prediction, and the potential benefits of transfer learning to address the nested nature of extreme events (naturally included in less extreme ones). Using 1000 years of state-of-the-art PlaSim Planete Simulator Climate Model data, it is shown that Convolutional Neural Network-based Deep Learning frameworks, with large-class undersampling and transfer learning achieve significant performance in forecasting the occurrence of extreme heatwaves, at three different levels of intensity, and as early as 15 days in advance from the restricted observation, for a single time (single snapshoot) of only two spatial fields of climate data, surface temperature and geopotential height.