We previously reported in Science Translational Medicine (STM) unprecedented results (1) from gene replacement experiments in MTM1-mutant animal models of X-linked myotubular myopathy (XLMTM, OMIM 310400)(2). This devastating congenital muscle disorder results from deficiency of myotubularin, a phosphatidylinositol-3-phosphatase required for skeletal muscle function, growth and ultrastructural organization (3-6). The overall median survival of XLMTM patients is only 29 months (7) and the majority of patients who survive beyond two years require ventilator support (8). While intensive medical support extends the life of young patients, there is no effective treatment. Of therapeutic approaches tested in animal models of XLMTM, gene replacement targets the underlying cause of this monogenic disorder (9). Analogous to human XLMTM patients, mice engineered with a targeted deletion in the Mtm1 gene demonstrate marked muscle weakness and shortened lifespan (4). A naturally occurring MTM1 missense mutation in dogs results in a similar phenotype with reduction of muscle strength and lifespan (10). As we reported in STM, myotubularindeficient mice and dogs both responded to systemic and quasi-systemic administration of a recombinant serotype 8 adeno-associated virus (AAV8) carrying full-length murine Mtm1 or canine MTM1 coding sequence. Both mouse and dog models responded within a few short weeks with rapid and sustained increases to near normal levels of muscle strength and prolonged survival. As sometimes occurs in science, initial failures can lead to subsequent successes. In our case, systemic AAV8-Mtm1 vector injections in myotubularin deficient mice resulted in body-wide muscular improvement. This result encouraged us to test first the functionality of an AAV8-MTM1 vector by intramuscular delivery in three MTM1-mutant dogs. We first noted robust and rapid muscle size enlargement and strength gains in the injected muscles. We next administered the vector using regional limb perfusion to three additional MTM1-mutant dogs. The idea was to isolate, via a tourniquet, venous blood flow to the entire hind limb musculature, infuse the AAV8-MTM1 vector and then compare findings with the non-infused opposite limb. The first MTM1-mutant dog responded as predicted: large gains of strength and muscle mass only in the infused limb. The non-treated contralateral limb remained in a weakened state. Subsequently, we attempted to replicate these findings using regional hindlimb infusion in two other MTM1-mutant dogs. The result was an “experimental failure” but a “therapeutic success”. Regional limb infusion resulted in systemic distribution of the vector in skeletal musculature with subsequent body-wide expression of the MTM1 transgene. These two treated MTM1-mutant dogs remain in the canine colony today and appear healthy and robust 3 years after AAV-MTM1 vector infusion. To our knowledge, ours was the first report of successful body-wide gene replacement in a congenital muscle disease that resulted in amelioration of severe muscle pathology and prolongation of life in both rodent and canine models. Why did this gene replacement approach rescue severe muscle pathology and prolong survival in myotubularindeficient animals when similar approaches in dystrophindeficient animal models failed to achieved such dramatic improvement? A number of characteristics of XLMTM