Evaluation of Planned versus Unplanned Soft-Tissue Sarcoma Resection Using PROMIS Measures

Background The Patient Reported Outcomes Measurement Information System (PROMIS) is a tool developed by the National Institutes of Health that allows comparisons across conditions or even the United States (U.S.) general population. Objectives Our purpose was to compare PROMIS outcomes between patients who underwent a planned resection to those who underwent an initial unplanned excision of their sarcoma followed by a definitive oncologic resection. We then compared these groups to the U.S. general population. Methods Eighty-five patients were included and were divided into those who underwent an initial planned resection (67) and unplanned excision (18). These patients were then further categorized based on the length of follow-up since their last surgery, either early (<12 months) or late (>12 months). Results We evaluated seven PROMIS domains and found no differences between patients who underwent planned resection versus those who underwent an initial unplanned excision followed by a wide resection of the previous wound bed. When compared to the U.S. population, both cohorts demonstrated significantly improved scores in several emotional health domains. Conclusions Patients who undergo an unplanned excision followed by a definitive oncologic procedure have similar PROMIS scores compared to patients who undergo an initial planned resection.


Introduction
e Patient Reported Outcomes Measurement Information System (PROMIS) is an outcome tool that was developed by the United States (U.S.) National Institutes of Health. It is a patient-reported tool that categorizes responses into health domains. ese domains cover aspects of physical, mental, and social health. A unique and attractive feature of the PROMIS system is the ability to standardize responses. In converting individual patient responses to T-scores, a researcher can evaluate the health impact a rare entity such as a sarcoma has on a patient's life and compare this to a more common disease process, or even the U.S. general population [1]. PROMIS accomplishes this by converting scores from the United States general population to a T-score of 50, with a standard deviation of 10. In evaluating outcomes with the PROMIS tool, a lower score signifies less of the tested function; for example, if a patient scored 40 in the physical function domain, they would have a lower physical function level compared to the U.S. general population. Conversely, a score of 40 in the depression domain would indicate that the patient has a lower level of depression. e desired score is therefore dependent on the domain being tested [2]. e ability to compare across disease entities and to the U.S. general population allows physicians to more accurately explain the impact a certain diagnosis or treatment will have on a patient's life and allow the physician to explain this in a way that the patient is more apt to understand. Despite the advantages of the PROMIS system over previous outcome tools, such as the Musculoskeletal Tumor Society Scoring System (MSTS) or Toronto Extremity Salvage Score (TESS), which are oncology-specific and do not allow such broad comparisons, few orthopedic oncology studies have utilized the PROMIS system [3][4][5]. Our purpose in this study, therefore, was to compare patients who had undergone an unplanned excision of a nonmetastatic sarcoma and required a definitive oncologic resection to those who had undergone a planned initial resection. We attempted to answer the following questions: (1) Is there a difference in PROMIS scores between patients who underwent a planned versus unplanned resection of a nonmetastatic sarcoma? (2) Are these scores significantly different than the U.S. general population? In addition to answering these questions, we also provide normative values for future comparison studies.

Methods
We collected PROMIS data on all orthopedic oncology patient visits beginning September 1, 2016, as our standard of care. Prior to the visit with the physician, the patient was led to the clinic room, and a nurse loaded the questionnaire onto a computer for the patient to complete. e patient was given adequate time to complete the questionnaire before the physician entered the room. If the patient had difficulty completing the questionnaire, the nurse was available to assist.
Following approval from our institutional review board (IRB), we queried this database from September 1, 2016, to December 31, 2016. Six hundred four patients had completed a PROMIS questionnaire during the study period. We excluded patients with benign disease, those with nononcologic diagnoses, metastatic lesions, and those who did not undergo the final surgery at our institution. We additionally excluded patients with osseous lesions. Eighty-five patients were included in the final analysis. We subdivided these patients into those who underwent a planned resection and those who underwent an unplanned excision and subsequently required an additional resection for definitive management. Finally, the cohorts were then divided into early, defined as less than 12 months from their last surgical procedure, and late, defined as greater than 12 months from their last surgical procedure ( Figure 1).
We obtained demographic data and pathologic diagnoses from review of the patient's chart. Follow-up was determined based on the last surgical procedure rather than the initial resection to allow appropriate recovery time for patients who underwent multiple surgical procedures or wound complications. We used the latest PROMIS questionnaire in the analysis. e PROMIS 43 Profile was used to collect data. is consists of short forms for seven health domains. e health domains include physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate, and pain interference. Physical function measures patient perception of their physical function and ability to participate in activities of daily living. Anxiety, depression, fatigue, sleep disturbance, and pain interference evaluate the difficulties with each in their respective domains. Finally, ability to participate evaluates the patient's perception of their ability to participate in normal social activities [6]. Raw scores were converted to T-scores in order to allow for comparison with the United States general population.

Statistical Methods.
e Profile 43 PROMIS questionnaire was completed by patients during routine clinic visits and stored in the electronic medical record. Incomplete entries were assessed per the PROMIS guidelines; briefly, if more than 50% of the modality entries were completed, then the raw score was calculated and adjusted for the number of missing entries. All raw scores were used to determine T-scores using the standard PROMIS T-score scales for adults. Mean T-scores were compared using the 2-tailed ttest where equal variance was not assumed. Significance was set at less than 0.05.

Results
Eight-five patients were included in the final analysis, including 43 males (51%) and 42 females (49%). Nineteen patients (22%) had tumors located in the upper extremity compared to 66 (78%) who had tumors located in the lower extremity. e average age was 63 years. We found no significant differences in tumor locations between groups.
We also found no significant differences in the rates of limb salvage between the cohorts. We did observe a significant difference in the rate of adjuvant therapy between the cohorts, with a higher percentage of patients in the planned cohort receiving adjuvant treatment. Finally, there was no significant difference in postoperative complications between cohorts or average resection size (Table 1). e average resection size was based on the pathologic sample obtained during the definitive oncologic procedure. is included the tumor for the planned surgical resection and the wound bed for the unplanned excision cohorts.
Patients were divided into those who had undergone an unplanned excision during their first surgical procedure and patients who had undergone a planned surgical resection. ey were then further subdivided based on the acuity of the last surgical procedure, with those who had undergone a procedure within the last 12 months in the early follow-up group and those who were greater than one year from surgery in the late follow-up cohort. Figure 2 shows the location of the tumor resections. Table 2 lists the most common diagnoses.
In the planned resection group, nine patients (13%) required amputative procedures. Seven of these tumors were located in the lower extremity, and two were located in the upper extremity. In comparison, three patients (17%) in the unplanned excision group required an amputation. All three of these tumors were located in the lower extremity. ere was no significant difference in the rate of amputations between the cohorts (p � 0.728).
Adjuvant therapy, consisting of radiation therapy and/or chemotherapy, was given to 54 patients (81%) who underwent a planned resection. is is compared to ten patients (56%) in the unplanned excision cohort. ere was a significant difference in the rate of adjuvant therapy received between cohorts (p � 0.03).
e majority of planned resection patients received preoperative radiation therapy   (Table 3). e average PROMIS values based on the acuity of the last surgical procedure (early versus late) as well as the initial type of resection (planned versus unplanned) are shown in Table 4. We found a significant difference in depression scores based on the acuity of the surgical procedure, with scores decreasing (improving) in the later follow-up. We did not find a significant difference in the remaining PROMIS values based on the acuity of the surgical procedure. Additionally, we also did not find a significant difference in the PROMIS values based on the type of initial resection performed.
We then compared the patients in the late cohorts (12+months from the last surgical procedure) to the U.S. general population (Table 5). Several significant differences were found. We found a significantly lower physical function score in the planned resection cohort when compared to the U.S. general population (p ≤ 0.001). is was not reproduced in the unplanned excision group (p � 0.708). We also found significantly lower depression and fatigue levels in both the planned and unplanned resection cohorts and significantly lower levels of sleep disturbance in the planned resection cohort compared to the U.S. general population. ese values are represented graphically in Figures 3-5.

Discussion
Several studies have previously evaluated the impact an unplanned excision has on local recurrence and survival, but there is a paucity of data evaluating patient-reported functional outcomes in this population [7][8][9][10][11]. e aim of our study was therefore to compare results in patients who had undergone an unplanned excision and required a repeat resection to those patients who had undergone a planned initial resection. We attempted to answer whether there was a difference in PROMIS scores between these patient cohorts, as well as if these scores differed from the United States general population.    ere are several unavoidable limitations in our study. As these patients were from a single institution and sarcomas are rare, our numbers are limited. Additionally, there is a wide variety in histologic diagnoses as well as tumor locations in our patient population. However, previous research has reported no significant impact on functional results based on anatomic location alone [12]. Future studies will need to independently verify our reported results.
A benefit of the PROMIS scoring system compared to previous systems is the ability to standardize the results. In doing this, we are able to compare our patients to a subset of the United States general population. When this comparison is made, several interesting differences are found. As one might expect, we noted a statistically lower score in the physical function domain in the planned resection cohort compared to the United States general population. is difference was not observed in the unplanned excision population. A reason for this discrepancy is likely due to the small numbers in our study. Alternatively, as Aria et al. suggests, this may also be due to less muscle resection in the unplanned cohort due to a more superficial location in these tumors, although we found no significant difference in the sizes of our resected specimens between groups [13].
Significant differences were also observed in our patients compared to the U.S. general population in several   [14][15][16]. Future studies will need to verify this result. When comparing PROMIS scores between cohorts we found no significant differences for physical function, emotional health, and social health domains based on whether patients underwent an unplanned excision prior to a definitive resection or if they underwent an initial planned resection. is finding is similar to studies that report no difference in functional results between unplanned excision and planned resection cohorts [13,17,18].
In addition to equivalent PROMIS scores between the unplanned excision and planned resection cohorts, we also noted a similar rate in ability to perform a limb salvage procedure between groups. In the planned resection cohort, a limb salvage procedure was performed in 87% of cases.
is is compared to 83% in the unplanned excision cohort (p � 0.728). Other studies have reported similar amputation rates between groups [13,18,19]. As before, Aria proposes that this finding may be due to the tumors in the unplanned cohort typically being smaller and located in a more superficial location compared to those that underwent an initial planned resection [13].

Conclusion
In conclusion, we found no significant difference in PROMIS scores between patients that underwent a planned resection for a nonmetastatic sarcoma compared to those who underwent an initial unplanned excision, followed by a definitive oncologic procedure. Both groups demonstrated improved emotional health scores compared to the U.S. general population.
is result should not condone performing unplanned excisions but may be used to counsel patients who present following such a procedure.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.