While many of the most exciting questions about the initial formation of galaxy clusters must wait for the next-generation X-ray missions, the complementarity between SZ surveys and current X-ray observatories can lay an important foundation now by studying clusters in the 1 < z < 2 range, where to date only a dozen of the most massive systems have been observed. I will present the results of a joint analysis of Chandra X-ray and South Pole Telescope (SPT) SZ observations targeting the first sample of clusters selected to be the progenitors of well-studied nearby systems such as Perseus at 0.3 < z < 1.3. We developed a new procedure in order to tackle the analysis challenge that is to estimate the Intra-Cluster Medium (ICM) properties of low-mass and high-redshift clusters with about 150 X-ray counts. One of the dominant sources of uncertainty on the ICM density profile estimated with a standard X-ray analysis with such shallow X-ray data is due to the systematic uncertainty associated with the ICM temperature obtained through the analysis of the background-dominated X-ray spectrum. We showed that we can decrease the uncertainty on the density profile by a factor 5 with a joint deprojection of the X-ray surface brightness profile measured by Chandra and the SZ integrated Compton parameter available in the SPT cluster catalog. This technique has been appplied to the whole sample of 67 clusters in order to track the evolution of the ICM core density during cluster growth. We confirmed that the evolution of the gas density profile is well modeled by the combination of a fixed core and a self-similarly evolving non-cool core profile. We further showed that the fraction of cool-cores in this sample is remarkably stable with redshift although clusters have gained a factor 4 in total mass over the past 9 Gyr. This implies that cool core disruption by mergers has to be compensated by cool core restoration mechanisms in timescales that are shorter than the Hubble time. This new sample combined with our new X-ray/SZ analysis procedure and an extensive multi-wavelength coverage will allow us to address fundamental shortcomings in our current understanding of cluster formation and evolution at z > 1.