It is well known that neutron stars can undergo a phase transition under a certain class of Scalar Tensor Theories of gravity (STT's) where a new order parameter, the scalar charge, appears within the star. This is the well known phenomenon of spontaneous scalarization (SC) discovered by Damour and Esposito-Farèse in 1993. Under such mechanism neutron stars can afford in principle a maximum mass larger than in general relativity (GR) for a given equation of state without taking into account additional observational constraints (e.g. binary systems). This opens the possibility that neutron stars might be formed with masses as large as ∼ 2M⊙ without the need of stiff, or more exotic, equations of state for the nuclear matter. Thus, STT's through SC may account for compact objects with large masses observed recently in the sky in the form of pulsars (PSR J0348+0432 with M = 2.01M±0.04⊙ observed in 2013, PSR J1614-2230 with M = 1.97 ± 0.04M⊙ observed in 2010 or J0740+6620 M = 2.14 +0.10 −0.09 M⊙ observed in 2019). However, we argue that even if that was possible such maximum mass models within STT cannot be formed solely from the dynamic transition of an initial "isolated" unscalarized neutron star whose mass cannot exceed the maximum mass in GR. This is because SC, being an energetic-preferred configuration, produces a final static star with a mass lower that the initial one with a fixed baryon mass. The mass difference between the initial and final configurations is radiated away in the form of a scalar-field. Thus, maximum mass models of scalarized neutron stars, if present in nature, must have formed by a different process, perhaps of cosmological origin or by the subsequent accretion of additional scalar charge and mass.