Good questions. Easy answer is that it depends (sigh...), on several things.
1) Is there a specific percentage of soft tissue tumors that orthopaedic surgeons handles?
Soft tissue sarcomas (STS) are treated by general surgeons in some programs and locations, determined in part by history and the presence or absence of an interested surgeon. But there is no doubt that orthopaedic oncology owns bone sarcomas, and in most/many locations either have the advantage or share equally in soft tissue sarcomas. Simple lipomas are often handled by many different people, but rare is the general surgeon these days that wants to operate deep in the thigh, forearm, etc. Retroperitoneal sarcomas, in contrast, are largely the domain of general surgeons due to the frequent involvement of the kidneys, bladder, etc.
In residency the orthopaedic oncologists I worked with, at an institution where gensurg did zero STS, the breakdown of bone vs soft tissue was about 50/50 if you include benign bone disease. STS is far more common than bone cancer, so of malignant cases it is closer to 30/70 or even 80/20.
2) What are the bread and butter cases ortho onc guys handles? Biopsies of suspect masses? Surgical treatment of metastatic disease from other primary tumors?
At an academic center, bread and butter cases include straight-forward soft tissue sarcomas, some open biopsies, and some metastatic work. But a nice thing about orthopaedic oncology is the diversity of cases, so there is less 'bread and butter work' than the rest of orthopaedics.
If you are orthopaedic oncology trained but take a job in the community, you will largely do general orthopaedics (total knee, total hip, hip fractures, ankle fractures) unless you do two fellowships. Big time orthopaedic oncology where patients need radiation therapy and some chemotherapy is largely the domain of academic centers.
3) Does it differ from general orthopaedics in terms of having to know more about medicine as well?
Ortho onc folks seem to, and like to, knowing a lot about pathology, medicine, etc. Ideally orthopaedic oncologists are well versed in surgery and orthopaedics, so they are flexible in how they treat things in the operating room. And it is a field that is very academic and a lot of research happens. So, compared to most orthopaedists, it is fair to say 'yes' to this question. Within orthopaedics, onc is more cerebral.
4. How competitive are the fellowships?
This is the key. If you really want to do ortho onc, you probably want to do academics. And there aren't a lot of orthopaedic oncology academic jobs in any given year. So although getting 'a' fellowship isn't so bad, there are WAY more fellowships than there are jobs opening up - which is what leads to having onc-trained folks around that are doing more joints than anything else. Getting into a top 3 fellowship, then, is important for finding a position. About half of people doing oncology know where they are going back to either before they start fellowship or even before applying. Some fellowships are less keen to take applicants that don't have a job already lined up, because they don't want to train someone who won't get to use those skills in an academic environment.
in reply to 'I think these people do a lot of joints on the side since true ortho onc cases are so rare and the fellowship is a lot of arthroplasty training.' - this depends a lot on location. My residency had three orthopaedic oncologists who did 100% tumor, no primary joints (except for tumor), and their fellows didn't do any primary joints in fellowship. But, what I wrote above applies - there are probably too many ortho onc fellowships/fellows, compared to the true number of true onc positions available.
