Impact of a Virtual Multidisciplinary Sarcoma Case Conference on Treatment Plan and Survival in a Large Integrated Healthcare System

PURPOSE: Quantifying the impact of a multidisciplinary cancer case conference on patient outcome and care quality remains challenging. PATIENTS AND METHODS: We prospectively investigated the impact of our virtual multidisciplinary sarcoma case conference (VMSCC) on treatment plan in patients presented to the VMSCC from July to October 2020 (prospective cohort) and retrospectively in patients with metastatic or locally advanced high-grade soft-tissue sarcoma (STS) reviewed in the VMSCC in 2016 and 2017 (high-grade STS cohort). We also investigated the factors related to the nonadherence to the VMSCC-recommended plan in both cohorts. RESULTS: In both cohorts, approximately 28% of the patients were referred to the VMSCC for review without a treatment plan. In significantly more cases, referring physicians outside of the sarcoma group did not have a plan formulated before the VMSCC review compared with the referring physicians within the sarcoma group. In 28.2% (prospective cohort) and 19.5% (high-grade STS cohort) of the patients, VMSCC recommended a different plan. The adherence to the VMSCC-recommended plan was 87.9% and 83.1%, respectively. The causes of the nonadherence were primarily due to disease progression or patient’s decision against recommended therapy. The median overall survival for the high-grade STS cohort was 26 months. CONCLUSION: VMSCC affected the treatment plan in approximately 50% of the patients in both cohorts. The median overall survival of the patients with high-grade STS reviewed by the VMSCC in our cohort is comparable with the literature.


METHODS:
We investigated the impact of VMSCC on treatment plan in two cohorts: a prospective cohort with all the patients reviewed at the VMSCC from July to October 2020 and a retrospective cohort of patients with high-grade STS reviewed at the VMSCC from 2016 to 2017. We chose only the patients with highgrade STS from the retrospective cohort to minimize the heterogeneity for survival assessment and to compare if the impact of the VMSCC on treatment plan would be different compared with the prospective cohort. We studied the treatment plan formulated by referring physicians from both within and outside of the sarcoma group. We explored the rate of adherence and the factors associated with nonadherence (Table).
BIAS, CONFOUNDING FACTORS(S), DRAWBACKS: The populations of the two cohorts are very different, making it challenging to compare. The overall survival of the patients with high-grade STS from the retrospective cohort is associated with selection bias and the fact that the cohort is single-arm and singleinstitution only.

REAL-LIFE IMPLICATIONS:
Quantifying the impact of a multidisciplinary case conference (MCC) is inherently challenging but important in understanding the costeffectiveness of various multidisciplinary care settings, especially in today's challenging healthcare environment with limited resources and increasing cost. Establishing an impactful MCC requires meaningful investment for organizations; however, significantly improved care quality and outcomes may lead to cost savings and hence potentially reduced cost. Individual physicians with the specialty expertise can still benefit significantly from MCC review, which can be important for the practice not only in the community settings but also in the academic and other settings and is likely relevant to the noncancer MCC settings as well. Our study may help multidisciplinary teams of various organizations in optimizing their care model to further enhance the value of their multidisciplinary approach.

PATIENTS AND METHODS
We prospectively investigated the impact of our virtual multidisciplinary sarcoma case conference (VMSCC) on treatment plan in patients presented to the VMSCC from July to October 2020 (prospective cohort) and retrospectively in patients with metastatic or locally advanced high-grade soft-tissue sarcoma (STS) reviewed in the VMSCC in 2016 and 2017 (high-grade STS cohort). We also investigated the factors related to the nonadherence to the VMSCC-recommended plan in both cohorts.

RESULTS
In both cohorts, approximately 28% of the patients were referred to the VMSCC for review without a treatment plan. In significantly more cases, referring physicians outside of the sarcoma group did not have a plan formulated before the VMSCC review compared with the referring physicians within the sarcoma group. In 28.2% (prospective cohort) and 19.5% (high-grade STS cohort) of the patients, VMSCC recommended a different plan. The adherence to the VMSCC-recommended plan was 87.9% and 83.1%, respectively. The causes of the nonadherence were primarily due to disease progression or patient's decision against recommended therapy. The median overall survival for the high-grade STS cohort was 26 months.
CONCLUSION VMSCC affected the treatment plan in approximately 50% of the patients in both cohorts. The median overall survival of the patients with high-grade STS reviewed by the VMSCC in our cohort is comparable with the literature.

INTRODUCTION
There is increasing evidence that multidisciplinary case conference (MCC) improves the care quality and possibly survival of patients with cancer, despite the lack of prospective randomized trials that are difficult to carry out in clinical practice. 1-3 Multidisciplinary evaluation of patients with cancer is often incorporated into guidelines such as National Comprehensive Cancer Center Network, especially for the rare malignancies. The American College of Surgeons mandates implementation of MCC for the Commission on Cancer accreditation of hospitals. In Canada, Cancer Care Ontario directs all practices to implement MCC for patients with cancer. 4,5 It is conceivable that MCC can improve quality of care for patients with cancer by providing a communication platform for real-time discussion of cases prospectively, but the magnitude of impact may be dependent on many factors including the case conference format and interval, physician composition and efforts, disease types and case volume, resource availability and ancillary support, etc. We previously published our experiences with a virtual multidisciplinary sarcoma case conference (VMSCC) from 2013 to 2019 with nearly 1,600 cases reviewed. 6 Our study showed that the majority of referring physicians felt that the VMSCC improved the quality of care and enhanced their confidence in treating patients with sarcoma. In addition, more than 50% of the responses in the most recent survey indicated that the VMSCC resulted in change of treatment plan very or somewhat frequently. The COVID-19 pandemic since early 2020 has galvanized the utilization of technology and the movement of many aspects of cancer care to virtual platforms, which likely has accelerated the movement of MCC to virtual format globally. 7 Such a movement can especially benefit the Author affiliations and support information (if applicable) appear at the end of this article.
patients who live in underserved rural areas and face care disparities, including the veterans who rely on the Veterans Health Administration System that is dispersed throughout the country. 8,9 However, despite many studies assessing the value and effectiveness of MCC, it remains challenging to quantify the impact of MCC on patient care quality and survival. In addition, the rate of adherence to the MCC recommendations and the factors related to the nonadherence remain to be understood. Here, we attempted to quantify the impact of our VMSCC on the care quality and patient survival both prospectively and retrospectively using two cohorts of patients. The prospective cohort includes all the patients referred to the VMSCC from July to October 2020, whereas the retrospective cohort includes only the patients with high-grade soft-tissue sarcoma (STS) referred to the VMSCC during 2016-2017 to minimize the heterogeneity for determining its overall survival (OS) and to compare with the prospective cohort if the impact of VMSCC on treatment plan was consistent. Our results shall provide insight for further enhancing the multidisciplinary care of cancer.

Patients
Kaiser Permanente Northern California (KPNC) is an integrated healthcare system with approximately 4.5 million members. The VMSCC provides services to KPNC and the other regions of Kaiser Permanente including Mid-Atlantic, Northwest, and Hawaii regions. We prospectively monitored the cases reviewed at the VMSCC from July to October 2020 (prospective cohort) and retrospectively studied the cases with high-grade STS reviewed by the VMSCC in 2016 and 2017 (high-grade STS cohort). Demographics, histology, treatment records, and other types of data were collected from the electronic medical records (Epic). This study was approved by the Institutional Review Board of KPNC.

Assessment of Treatment Plan
We have maintained an Excel file database of all the sarcoma cases reviewed by the VMSCC since its inception in 2013 with patient demographics, sarcoma subtype, specialty of the referring physicians, the referral information, the treatment recommendations of the VMSCC, and other information. This enabled the comparison of the differences between the treatment plan by referring physicians and the plan recommended by the VMSCC. If a treatment plan was absent in the referral, we reviewed the cases electronically to identify the treatment plan formulated by the referring physicians if any. Only the major differences in treatment plans between the referring physician and the VMSCC recommendation were considered truly different. For example, if a referring physician planned to give firstline chemotherapy using single-agent adriamycin, while the VMSCC recommended adriamycin plus olaratumab, it would be considered the same as the original plan.

Assessment of OS
The OS of the high-grade STS cohort was measured from the date of VMSCC review to the date of death.

Definition of Sarcoma Group
Our sarcoma group (referring to physicians only in this manuscript) was composed of the physicians who were regular participants of the VMSCC with special focus on the diagnosis and treatment of sarcoma, including medical oncologists, surgical oncologists, musculoskeletal oncologists, radiation oncologists, musculoskeletal radiologists, pathologists, nuclear medicine physician, and pediatric oncologists, with the support of nurse care coordinators and geneticists. 6

Statistical Methods
We used chi square for analyzing the differences in treatment plans between the VMSCC and the referring physicians and the referral patterns between the referring physicians from outside of the sarcoma group and within the sarcoma group. We used Kaplan-Meier plot to display OS of the patients with metastatic and locally advanced high-grade STS reviewed by the VMSCC in 2016 and 2017. All statistical analysis was performed using MedCalc software (Belgium).  Abbreviations: LMS, leiomyosarcoma; N/A, not available; STS, soft-tissue sarcoma; UPS, undifferentiated pleomorphic sarcoma; VMSCC, virtual multidisciplinary sarcoma case conference. and in 19.5% of the patients, VMSCC recommended a different treatment plan (nonapplicable to the other 28.6% of the patients where there was no treatment plan formulated before VMSCC review). The adherence rate to the VMSCC-recommended treatment plan was 83.1%, and the nonadherence was 16.9%. 3. Differences in treatment plan between the physicians from the sarcoma group and nonsarcoma group. We compared the differences in treatment plan formulated by the referring physicians within the sarcoma group and the referring physicians outside of the sarcoma group (nonsarcoma group) and evaluated how the VMSCC affected the treatment plan of patients referred from the two groups. In both prospective and high-grade STS cohorts, significantly more cases without a treatment plan formulated before the VMSCC review came from the referring physicians outside of the sarcoma group compared with the referring physicians within the sarcoma group ( Table 2, 38.1% v 6.8% in the prospective cohort and 42.8% v 11.4% in the high-grade STS cohort). In addition, in significantly more cases, the VMSCC-recommended treatment plan did not differ from the referring physicians within the sarcoma group, compared with the referring physicians outside of the sarcoma group (Table 2). However, in 25% and 17.1% of the patients in both cohorts, respectively, the VMSCC-recommended treatment plan differed from the plan by the referring physicians within the sarcoma group, not significantly different from that of the referring physicians outside of the sarcoma group (Table 2), indicating that all referring physicians including the experts in the core sarcoma group benefitted from the VMSCC review. 4. Factors related to the nonadherence to the VMSCCrecommended treatment plan: Nonadherence because of rapidly progressive disease or declining physical performance occurred in four of 17 patients in the prospective cohort and in six of 13 patients in the highgrade STS cohort. Nonadherence because of patient's declining to follow the VMSCC-recommended treatment plan occurred in 10 of 17 patients in the prospective cohort and four of 13 patients in the high-grade STS cohort (Table 3, in the majority of cases, patients declined to consider chemotherapy or surgery that was recommended). In three patients in the prospective cohort, the referring physicians did not adhere to the VMSCC-recommended treatment plan (in two patients, it was due to conflicting recommendations from a different institution). In one patient of the high-grade STS cohort, the VMSCC-recommended plan was not followed likely due to the treating physician not aware of the VMSCC recommendation at the time (   Patient did not adhere to VMSCC recommendation 10 4

Prospective
Referring physician did not adhere to VMSCC recommendation 3 1 Insurance coverage changed 0 2 Abbreviations: PS, performance status; STS, soft-tissue sarcoma; VMSCC, virtual multidisciplinary sarcoma case conference. received observation, palliative radiation, or hospice care. The median OS for the entire cohort of 77 patients with metastatic or locoregionally advanced highgrade STS presented to the VMSCC in 2016 and 2017 was 26 months (Fig 1, 95% CI, 24.2 to 33.6), with a median follow-up of 26 months.

DISCUSSION
Quantifying the impact of an MCC may provide helpful insight into understanding the cost-effectiveness of various multidisciplinary care settings, which is especially important in today's challenging healthcare environment with limited resources and increasing cost. Establishing an impactful MCC requires meaningful investment for organizations; however, significantly improved care quality and outcomes may lead to cost savings. A few studies have attempted to quantify the impact by assessing the rate of differences on treatment plan formulated by the MCC and by the referring providers. Most of the studies have shown that in approximately 20%-50% of cases, an MCCrecommended treatment plan differed from the plan by the referring providers. For example, Schmidt et al 10 from Virginia Mason Medical Center at Seattle investigated a thoracic oncology MCC on the impact of lung and esophageal cancer and found that their MCC recommendations changed the treatment plans in 26%-40% of patients, with 97% adherence to the MCC-recommended plan. However, a retrospective study by Osarogiagbon et al 11 showed nonadherence in one third of the patients in a thoracic oncology MCC and those patients with lung cancer whose treatment did not follow the MCCrecommended plan had worse progression-free survival. In other disease types, van Hagen et al 12 from the Netherlands examined an MCC on upper GI malignancies and found that the MCC altered workup or treatment plan in 34.5% of the cases. A study on a head and neck cancer MCC from University of North Carolina Hospitals showed that in 27% of patients, there were changes made by the MCC. 13 A study by the University of Alabama showed that an MCC altered treatment plan in approximately 20% of patients with gynecologic cancer. 14 Studies from South Korea, Lebanon, and other countries have also shown similar impact of oncology MCCs on treatment plan. [15][16][17] In our previous study, we found that approximately 20%-50% of the referring physicians responded that our VMSCC led to change of treatment plan very or somewhat frequently, based on the surveys. 6 Consistent with this finding, our current study showed that the VMSCC made impact on approximately 50% of the patients, including approximately 28% of the patients referred without a treatment plan in both prospective and high-grade STS cohorts, as well as 28.2% and 19.5% of the patients in each cohort with a plan. Importantly, for the patients referred by the physicians who were the regular members of the sarcoma group, the VMSCC changed the treatment plan in approximately 25% and 17.1% of the patients, indicating that the VMSCC benefited the patients referred by the expert physicians as well. This further demonstrates the significance of an MCC and the power of prospective multidisciplinary discussion in real time in formulating a treatment plan especially for patients with uncommon malignancies.
The adherence to the VMSCC-recommended plan was more than 83% in both cohorts. The causes of the nonadherence to the VMSCC-recommended plan were primarily due to disease progression or patient's decision not to follow the recommended plan, most of these patients had poor outcome primarily because of disease progression. The still-intact primacy of the referring provider and the patient themselves in ultimately having all the clinical facts to bear when making a decision as to whether or not to enact the VMSCC's advice was always advised and respected by the VMSCC.
The median OS of 26 months in the patients with highgrade STS in our cohort from 2016 to 2017 is comparable with the OS reported in the literature by majority of retrospective studies and prospective clinical trials. In the phase III study by Demetri et al 18 on the efficacy of trabectedin for advanced liposarcoma and LMS who had progressed on an anthracycline and at least another line of therapy, the median OS was approximately 12-13 months. In the pivotal phase III study on the efficacy of eribulin in patients with advanced liposarcoma who progressed on at least two lines of therapy, the eribulin arm showed a median OS of approximately 15 months. 19 In the phase II open-label trial by Maki et al 20 (SARC002) in patients with advanced STS who had received zero to three lines of chemotherapy treated with gemcitabine alone or in combination with docetaxel, the median OS was 11.5 and 17.9 months, respectively. In our recent study examining the patients with metastatic or locally advanced STS and bone sarcoma treated with gemcitabine and docetaxel, the median OS was approximately 12-15 months. 21 One of the factors related to the long median OS with our cohort could be that some of the patients had low disease burden and obtained long and durable remission after local therapy (surgical resection, metastatectomy, or stereotactic radiation) without requiring systemic therapy or after the resection of responded locoregionally advanced disease, and another factor could be related to that a majority of patients in our cohort were chemotherapy-naïve. For example, one patient with newly diagnosed metastatic epithelioid hemangioendothelioma of stomach to liver who received gemcitabine-docetaxel chemotherapy with a minor response and went on to have both the primary tumor and liver metastasis resected has remained free of disease for more than 4 years. Another patient with a massive abdominal mass initially diagnosed with smooth muscle tumor of uncertain malignant potential was recommended by the VMSCC to repeat the biopsy as the clinical course did not appear consistent with the diagnosis of smooth muscle tumor of uncertain malignant potential. This patient was found to have LMS on the repeat biopsy and obtained a partial response to gemcitabine/ docetaxel chemotherapy followed by complete resection of the disease and has remained free of disease for nearly 4 years.
The strength of our study is that we have investigated two cohorts of patients, one in prospective manner and the other retrospectively to quantify the impact of our VMSCC and that the results from both cohorts are very consistent. In addition, we investigated the differential impact of the VMSCC on the patients referred by the physicians from within the sarcoma group and outside of the sarcoma group, which shall provide further insight. The limitations of our study are that it is an observational and single institutional study and that OS of the patients in the high-grade STS cohort was affected by selection bias because of the exclusion of some histology types in the effort to minimize heterogeneity.
In conclusion, our study shows that the VMSCC made an impact on treatment plan in approximately 50% of the patients and this impact was significant not only for the patients referred by the physicians outside of the sarcoma group but also the physicians within the sarcoma group. The median OS of the patients with high-grade STS in our cohort is comparable with the literature.