Discussion

Osteosarcoma remains one of the most challenging pediatric cancers to treat. Despite therapy advancements in the 1970s to 1980s, survival rates have plateaued over the past four decades with a 5-year overall survival rates remaining around 63%.9,10 Advancements continue to prove challenging, in part due to the rarity of the disease and very complicated biology. Tumor necrosis has long been known to be a strong prognostic factor, but its utility in informing changes in adjuvant regimens remains elusive. A large, international randomized controlled trial, EURAMOS-1, was conducted to investigate whether intensified MAP plus ifosfamide and etoposide (MAPIE) therapy in patients with poor response (tumor necrosis less than 90%) would result in improved outcome. They also investigated whether maintenance therapy with pegylated interferon alfa-2b in patients with good histologic response (tumor necrosis greater than or equal to 90%) was superior to MAP alone. These studies failed to demonstrate improvement in long-term outcomes when stratifying treatments based on histologic response.12,13The results from our study revealed that minor aberrations from the SOC schedule did not correlate with lower tumor necrosis. This is consistent with the hypothesis that tumor necrosis is likely a biologic marker of a specific tumor’s sensitivity to chemotherapy and is thus not improved with small alterations of the existing neoadjuvant regimen. It should be noted, however, that in our patient population, all timing aberrations were relatively minor, with the longest neoadjuvant chemotherapy to surgery interval being 33 days. It is reasonable to hypothesize that longer delays could result in tumor regrowth, and thus a lower percentage of tumor necrosis. This study did, however, demonstrate that delays in LC may confer worse outcome, as defined by 5-year EFS. Additionally, longer neoadjuvant chemotherapy to surgery intervals in general correlate with increased risk of having an event in patients with localized disease at the time of diagnosis after adjusting for relevant potential confounders. This is clinically intuitive, as patients with metastatic disease or very large primary tumors at the time of diagnosis face poor outcomes regardless of treatment. Though it is unclear if there is an optimal interval length between neoadjuvant treatment and surgery, this study suggests that attempts to reduce delays to surgery may improve outcomes in children and young adults with osteosarcoma. This requires early surgical planning, close communication with surgeons regarding treatment delays, and ample coordination with outside institutions when medical and surgical management is occurring at different locations. Notably, a recent study conducted at King Hussein Cancer Center in Jordan examining patients under the age of 18 with non-metastatic osteosarcoma demonstrated that a time from diagnosis to LC greater than or equal to 18 weeks was associated with worse event free survival and overall survival.20 This is consistent with our findings that treatment delay generally is associated with worse outcomes. It is important to note that intervals from diagnosis to LC greater than 18 weeks are uncommon in our clinical setting. Only one patient had a time from the initiation of chemotherapy to LC interval greater than 18 weeks. While this does not include time from diagnosis to initiation of chemotherapy, delays prior to chemotherapy initiation are exceedingly uncommon. This study adds context that more minor delays, particularly after completion of neoadjuvant chemotherapy, also may confer worse outcomes. There are limitations to our retrospective study. It does not allow us to fully control covariate differences between the exposure and control groups, introducing a greater degree of potential confounders than a randomized clinical control study. This study can only suggest correlation, and the possibility of an unmeasured confounder, such as socioeconomic barriers contributing to treatment delays and contributing to worse outcomes, cannot be excluded. There are limitations to assessing current patient status based on accessibility of information. Attempts to mitigate this included obtaining follow up information from outside institutions, as is standard practice, when possible and censoring all follow up information at 5-years. Since OS is such a highly aggressive cancer, most patients who will relapse do so in the first five years following completion of chemotherapy. Notably some patients (2019-2022) do not yet have five years of follow up information and this is a limitation of this study. Additionally, though tumor necrosis is a universally established measurement for osteosarcoma, there may be a certain degree of variability in the exact measurements between pathologists. Another limitation of this study is the fairly small sample size; results such as the correlation between interval length and tumor necrosis were not statistically significant but may require more data to yield sufficient precision before meaningful correlations can be ruled out. However, this is a substantial sample size for a rare disease like OS. Relatedly, interval length was treated as a linear effect in all models due to limited modeling degrees of freedom but may have a non-linear effect on EFS. Finally, it is important to note that we are reducing highly individualized chemotherapy and disease courses into measurable data. Many factors affect a patient’s therapeutic course. While the neoadjuvant regimen largely remains unchanged, toxicities in the adjuvant setting may lead to non-optimal regimens, altering outcomes but not being captured in the data. In conclusion, our study demonstrates that delays in local control are not strongly associated with lower tumor necrosis. However, surgical delays from completion of induction chemotherapy correlate to worse outcome in localized disease. Therefore, minimizing the delay from the completion of neoadjuvant chemotherapy to LC should continue to be weighed against the potential risks of infection and wound healing complications associated with decreasing this interval.