key: cord-0715201-w5h6gke0
authors: Fountain, Daniel M; Piper, Rory J; Poon, Michael T C; Solomou, Georgios; Brennan, Paul M; Chowdhury, Yasir A; Colombo, Francesca; Elmoslemany, Tarek; Ewbank, Frederick G; Grundy, Paul L; Hasan, Md Tanvir; Hilling, Molly; Hutchinson, Peter J; Karabatsou, Konstantina; Kolias, Angelos G; McSorley, Nathan J; Millward, Christopher P; Phang, Isaac; Plaha, Puneet; Price, Stephen J; Rominiyi, Ola; Sage, William; Shumon, Syed; Silva, Ines L; Smith, Stuart J; Surash, Surash; Thomson, Simon; Lau, Jun Yi; Watts, Colin; Jenkinson, Michael D
title: CovidNeuroOnc: a UK multi-centre, prospective cohort study of the impact of the COVID-19 pandemic on the neuro-oncology service
date: 2021-01-28
journal: Neurooncol Adv
DOI: 10.1093/noajnl/vdab014
sha: a9a3df72c66a8b7154ac1e915df07d0be3a9943f
doc_id: 715201
cord_uid: w5h6gke0

BACKGROUND: The COVID-19 pandemic has profoundly affected cancer services. Our objective was to determine the effect of the COVID-19 pandemic on decision making and the resulting outcomes for patients with newly diagnosed or recurrent intracranial tumours. METHODS: We performed a multi-centre prospective study of all adult patients discussed in weekly neuro-oncology and skull base multidisciplinary team meetings who had a newly diagnosed or recurrent intracranial (excluding pituitary) tumour between 01 April and 31 May 2020. All patients had at least 30-day follow-up data. Descriptive statistical reporting was used. RESULTS: There were 1357 referrals for newly diagnosed or recurrent intracranial tumours across fifteen neuro-oncology centres. Of centres with all intracranial tumours, a change in initial management was reported in 8.6% of cases (n=104/1210). Decisions to change the management plan reduced over time from a peak of 19% referrals at the start of the study to 0% by the end of the study period. Changes in management were reported in 16% (n=75/466) of cases previously recommended for surgery and 28% of cases previously recommended for chemotherapy (n=20/72). The reported SARS-CoV-2 infection rate was similar in surgical and non-surgical patients (2.6% vs. 2.4%, p>0.9). CONCLUSIONS: Disruption to neuro-oncology services in the UK caused by the COVID-19 pandemic was most marked in the first month, affecting all diagnoses. Patients considered for chemotherapy were most affected. In those recommended surgical treatment this was successfully completed. Longer-term outcome data will evaluate oncological treatments received by these patients and overall survival.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (COVID-19) pandemic has caused an unprecedented impact on the UK National Health Service (NHS). Major restrictions on resources and capacity have affected provision of both medical and surgical cancer therapies. 1 Cancer Research UK documented a 60% reduction in cancer surgery and an international study by the CovidSurg Collaborative reported that an estimated 2.3 million elective cancer cases had been cancelled worldwide. 2, 3 Delays to cancer surgery can impact on overall survival. A three-month delay across all stage 1 to 3 cancers is estimated to cause >4,700 attributable deaths per year in England alone. 4 Delaying surgical treatment of a brain tumour can lead to irreversible neurological impairment and be rapidly life-threatening because of the risk of raised intracranial pressure and coma. Two international reports from over 90 countries, reported cancellation rates of up to 57.5% for neurosurgical operations and clinics across the globe. 5, 6 Furthermore, there are considerable risks from surgery for patients with SARS-CoV-2 infection. An international pan-specialty study by the CovidSurg Collaborative showed that in 1128 patients with a perioperative SARS-CoV-2 infection, the mortality rate was 24% and 51% had pulmonary complications. 7 Guidance set out by the British Neuro-Oncology Society (BNOS) and the Society of British Neurological Surgeons (SBNS) on the 19 th March 2020 made several recommendations for surgical and oncological practice during the COVID-19 pandemic, 8 including giving high surgical and oncology priority to patients with:

 Malignant gliomas suitable for surgery and adjuvant therapies  Posterior fossa tumours causing symptoms or hydrocephalus  Meningiomas causing major mass effect or neurological deficit A c c e p t e d M a n u s c r i p t 9  Brain metastases suitable for surgery and supratentorial, or suitable for stereotactic radiosurgery or whole brain radiotherapy Conversely, low surgical and oncology priority were designated to patients with:

 Low-grade glioma where active monitoring is a reasonable option  Skull base tumours where the patient was already planned for elective surgery  Radiotherapy for atypical / recurrent meningioma Guidelines regarding the overall surgical and adjuvant therapies for high-grade gliomas have also been published by an international consensus group. 9 The COVID-19 pandemic presented several problems including how to maintain a safe surgical neuro-oncology service, the risks posed to patients undergoing treatment, and how the decisions made by the neuro-oncology multi-disciplinary teams (MDT, a team of professionals including neurooncologists, radiologists, neuropathologists, specialist nurses, and neurosurgeons facilitating shared decision making between specialties -known as tumor board in North America) were affected. We therefore conducted the CovidNeuroOnc multi-centre, prospective cohort study to assess the impact of the COVID-19 pandemic on the UK neuro-oncology service for patients with newly diagnosed or recurrent brain tumours.

CovidNeuroOnc is a national, multi-centre, prospective observational study in the UK. We invited all adult neurosurgical units in the UK to collaborate on this study and 15 of 32 participated. The study A c c e p t e d M a n u s c r i p t 10

Consecutive patients were identified from weekly neuro-oncology and skull base MDT (tumor board) meetings between 1 st April and 31 st May 2020 in participating units. All patients aged ≥ 16 years were included if they were found to have a newly diagnosed or recurrent intracranial tumour (including low-grade glioma, high-grade glioma, primary central nervous system lymphoma, meningioma, vestibular schwannoma or metastases) based on either computed tomography (CT) or magnetic resonance imaging (MRI)). Pituitary tumours were excluded from this study due to the possible endocrinological management and the separate MDT management of these tumours.

De-identified data were collected using a secure, online data collection tool (www.castoredc.com).

Each local collaborator was given a unique account to facilitate an accurate audit trail. Data fields included: age, sex, Eastern Cooperative Oncology Group (ECOG) performance status, date of MDT and radiological diagnosis. We asked collaborators to record: (i) the "pre-COVID-19" MDT decision (the hypothetical decision of "usual" first line management that the MDT would have made without the influence of COVID-19), and (ii) the "post-COVID-19" MDT decision, which is the first line management offered during the COVID-19 period. Sites recorded a single management option from the following: surgery (biopsy or resection), chemotherapy, fractionated radiotherapy, stereotactic radiosurgery, active monitoring (watch and wait), no treatment required or best supportive care. A decision to delay or defer treatment was included when asking sites for "post-COVID-19" MDT decisions. Data were also collected on the types and dates of treatments administered (surgery, chemotherapy, radiotherapy, radiosurgery, active monitoring), and date of confirmed COVID-19 status (if applicable). Extent of resection was confirmed on post-operative MRI where it occurred over the course of the study period. Date of death was also recorded for patients with suspected highgrade glioma based on MRI or confirmed high-grade glioma after surgery. SARS-CoV-2 infection was determined either by viral RNA detection (nose and throat swab) or by CT chest imaging as per A c c e p t e d M a n u s c r i p t 11 the diagnostic process during the study period. Data collection was finalised on the 30 th of June 2020 to allow 30-day follow up following the index MDT within the study period. All participating units attained local departmental approval as a service evaluation prior to anonymised data collection and submission such that individual consent was not required. Additional daily COVID-19 confirmed cases were retrieved for temporal analysis from the UK government. 11

The primary objective was to determine whether the COVID-19 pandemic changed the management of patients with either newly diagnosed or recurrent intracranial tumours, compared to usual care.

Secondary objectives were to determine i) how many patients did not receive surgery, despite this being the MDT recommendation, and ii) how many patients contracted a SARS-CoV-2 infection, and iii) how many patients with high-grade glioma died during the study period up to the first data lock on 30 th June 2020.

Categorical variables were reported as percentages. Continuous variables were reported as median and interquartile range (IQR) or mean and standard deviation based on tests for normality with the Shapiro-Wilk test. Univariable categorical statistical tests were performed with Chi-square testing unless small samples sizes where Fisher"s exact testing was utilised. Odds ratios and 95% confidence intervals were computed using the Wald test. A threshold p-value of <0.05 was set to denote statistical significance. All analyses, tables and graphics including Sankey diagrams were completed using the tidyverse, gtsummary, epitools, RColorBrewer and riverplot packages in R v 3.6.0. [12] [13] [14] [15] [16] M a n u s c r i p t 12

There were 1357 consecutive referrals for newly diagnosed or recurrent intracranial tumours across fifteen regional neurosurgical units in the United Kingdom between 1 st April and 31 st May 2020.

Fourteen units provided data on all intracranial tumours, while one unit provided data on malignant gliomas only (n=147). Data from this unit were excluded from total cohort summative statistic and included for specific analysis of malignant gliomas to optimise external validity. Descriptive statistics for the remaining 1210 referrals are presented in Tables 1-2. The majority of referrals were for newly diagnosed intracranial tumours (n=950, 79%) and included patients aged 50-80 years old (n=858, 71%) who were ECOG performance status 0 or 1 (n=862/1210, 71%). The most common referral of a new intracranial tumour was for metastasis (n=344, 36%) or high-grade glioma (n=295, 31%), whereas the most common recurrence was for glioma (n=130/260, 50%, Table 2 ).

Overall, 8.6% of cases had a documented change in MDT decision compared to usual care (n=104/1210). Figure 1 shows the trends in weekly COVID-19 cases and number of referrals to the neuro-oncology MDT stratified by change in management. Changes in MDT decision were more likely in recurrent than newly diagnosed tumours (OR 1.8 95% CI 1.2-2.8, p=0.010). Over the study period, there was a significant reduction in the number of patients where COVID-19 resulted in a change in management plan at the MDT. In the first week of the study, a change in MDT decision was seen in 19% (n=23/120) of referrals, and this reduced to 0% (n=0/82) by the end of May 2020.

The majority of referrals with a change in MDT decision occurred in the first four weeks of the study period, which corresponded to the peak of the pandemic in the UK (n=78/104, 75%).

The pre-COVID-19 and post-COVID-19 MDT decision are depicted in the Sankey diagram in Figure   2 . The most common pre-COVID-19 management in all cases was surgery (n=466, 39%). While a A c c e p t e d M a n u s c r i p t 13 small proportion of patients were subject to a delay or deferral of treatment (n=16, 1%), there was a larger proportion of patients where the MDT decision changed from surgical intervention. Of the 466 patients considered for surgery in pre-COVID-19 "usual care" decisions, 75 (16%) patients were instead offered alternative management plans including active monitoring (n=28, 37%), radiotherapy (n=18, 24%), best supportive care (n=17, 23%), and a delay in treatment (n=9, 12%). Of the 72 patients considered for chemotherapy in pre-COVID-19 decisions, 20 (28%) were subsequently offered alternatives, most commonly best supportive care (n=13, 65%).

Given the large proportion of patients who would have been offered no treatment or best supportive care in a pre-COVID-19 situation (n=235, 19%), these were excluded for the purpose of identifying factors resulting in a change in management as a result of COVID-19. Comparative descriptive statistics of the remaining 975 referrals are shown in Table 3 . There was no significant difference in age, sex or ECOG, but patients presenting with a recurrence and in particular recurrent glioma were more likely to have a change in management plans (OR 3.3 95% CI 1.5-7.9 p=0.003). Patients referred to the MDT with a suspected SARS-CoV-2 infection at the time of MRI diagnosis were no more likely to be offered a change in management plan (p=0.4).

Descriptive statistics of all 354 patients who underwent surgery up to 30 th June 2020 are presented in the Supplementary Material (Table S1 ). Of the 391 patients with a plan for surgery following the MDT, 345 (88%) were recorded to have undergone surgery, with 368 operations performed in total.

The majority of surgery performed was for glioma (newly diagnosed n=180/313, 58%, recurrent n=21/41, 51%) and metastasis (newly diagnosed n=72/313, 23%, recurrent n=10/41, 24%). Surgical and histopathological data is provided in the Supplementary Material (Table S2) Figure S1 ). Overall, 23 (7%) patients with newly diagnosed high-grade glioma had a change in management as a result of COVID-19. Comparing with those without a change of management, patients offered an alternative management were more likely to be ECOG 2 (p=0.017) but there was no difference in age (p=0.6) or sex (p=0.2). Of the 202 patients who would have been offered surgery as "usual care" before COVID-19, 11 (5%) were instead offered best supportive care, and 9 (4%) were offered fractionated radiotherapy without the need for a diagnostic biopsy. Of all 157 patients referred with recurrent glioma, 26 (17%) had a change in MDT decision because of COVID-19. There was no significant difference in age (p=0. 

Including all fifteen neuro-oncology units, 395 patients were referred with suspected cerebral metastases, of which, 314 (79%) were referred with a known primary cancer. MDT decisions are provided in the supplementary material ( Figure S3 ). Fifteen (8%) patients with newly diagnosed metastasis had a change in management. There was no significant difference in age (p=0.9), sex (p=0.5) or ECOG (p=0.5) in the cohort of patients where a change in management plan was made.

Stereotactic radiosurgery was the most common pre-COVID-19 management plan (n=109/395, 28%) whereas best supportive care was most common for post-COVID-19 management (n=113/395, 29%).

Including 89 cases of recurrent cerebral metastasis, data on oncological treatment was available for 484 patients (Supp Table 4 ). Of these, the majority of patients underwent a single treatment (SRS n=103/132 78%, surgery n=43/80 54%, radiotherapy n=43/61 70%, chemotherapy n=27/54 50%).

The most common combination therapy was surgery and SRS (n=14), followed by surgery and radiotherapy (n=10). Patients with a recurrent metastasis were significantly more likely to receive chemotherapy for their systemic disease (21% vs. 9%, p=0.002).

For patients with a radiological diagnosis of meningioma (n=157, 17%), the MDT decision changed because of COVID-19 for 16 (10%) patients. The most common management plan for patients with suspected meningioma was for active monitoring (n=84, 54%), followed by surgery (n=50, 32%). Of those, 37 (74%) were recommended surgery post-COVID-19 with 9 (18%) patients recommended A c c e p t e d M a n u s c r i p t 16 active monitoring and 4 (8%) patients recommended a delay in treatment. Within the study period 30 (81%) patients with meningioma had successfully undergone surgery.

Results were similar for patients with newly diagnosed low-grade glioma (n=60, 6%). MDT decisions changed due to COVID-19 for 10 (17%) patients. The most common pre-COVID-19 MDT plans were active monitoring (n=29, 48%) and surgery (n=28, 47%). Ultimately 19 patients with suspected lowgrade glioma were offered surgery post-COVID-19, 7 (12%) were instead offered active monitoring and 2 (3%) patients were subject to a delay in planned treatment. Of available data for 18 patients, 16 had undergone surgery within the study period (89%).

Suspected and confirmed SARS-CoV-2 infection data were available for 1184/1210 patients (98%).

The overall infection rate was 2.4% (29/1184). Of the 28 patients where mortality data was available, eight patients died (29%), with five deaths directly attributed to SARS-CoV-2 infection. Diagnosis of SARS-CoV-2 was made using a swab in 25 cases, whereas in 3 cases diagnosis was made radiologically and in one case it was unknown. Of the 348 patients undergoing surgery, nine (2.6%) developed a confirmed SARS-CoV-2 infection (Supplementary Table S1 ). Eight cases were diagnosed pre-operatively, and the rate of infection was not significantly different to the cohort of patients developing SARS-CoV-2 not undergoing an operation (n=20/826, 2.4%, OR 1.1 95% CI 0.5-2.3, p=0.852). Of the overall deaths in patients with high-grade glioma, five cases had a confirmed diagnosis of SARS-CoV-2, none of whom underwent surgery during the study period. In three of these cases SARS-CoV-2 was documented as the primary cause of death.

A c c e p t e d M a n u s c r i p t 17

This prospective, multi-centre study reveals that during the height of the COVID-19 pandemic in the United Kingdom, a change in MDT decision making compared to "usual care" was recorded in 8.6% Overall, 16% patients who would have been offered surgery as "usual care" were given a different recommendation and 24% patients who would have previously been offered chemotherapy had a change in recommendation. In the overall cohort we did not find that age or sex had a significant impact on MDT decision making. Although performance status was not associated with a change in management plan in the overall cohort, a poorer performance status did influence management recommendations in patients with a new diagnosis of a high-grade glioma. There was wide variation in the number of referrals received across units but this was due to a varying catchment population; all included units were offering regional services for patients with brain tumours.

There are two possible explanations to explain the reduction in changes in MDT decision making due to COVID-19 over the course of the study. It may be that the disruption to the neuro-oncology services caused by the COVID-19 pandemic was decreasing, or alternatively that neuro-oncology services adapted to provide a service despite COVID-19 restrictions. It is notable that of the oncology treatment reported, a third of patients who underwent an operation have not been treated with adjuvant chemotherapy or radiotherapy at the time of writing. Despite the reported pre-COVID-19 recommendation being surgery, 9% of patients with a radiologically defined, newly diagnosed highgrade glioma were offered best supportive care or fractionated radiotherapy without a tissue diagnosis.

Comparing the practice observed to the published guidelines in March 2020, there has been a sustained delivery of surgical services for newly diagnosed high-grade glioma and metastasis with appropriate changes in MDT decisions to active monitoring for patients with low-grade glioma and meningioma. 8 Of the 391 patients who were recommended surgery, 345 (88%) underwent their operation in the two months of greatest disruption due to COVID-19 in the UK. Furthermore, a very small minority (2.6%) of patients treated surgically developed SARS-CoV-2 and of these all but one A c c e p t e d M a n u s c r i p t 19 was diagnosed pre-operatively. This is a significant deviation from estimates and data published from the CovidSurg Collaborative where rates of cancellation in the 12 weeks of peak disruption were forecasted to be 37.7%, while rates of overall pre-operative SARS-CoV-2 infection were 26% with a mortality rate of 18.4% for neurosurgical procedures (overall mortality 26%). In our study where mortality of patients with high-grade glioma was recorded, three patients died as a result of SARS-CoV-2 none of whom received surgical treatment. The data presented in this study is encouraging with regards to the continued delivery of surgical neuro-oncology services in the UK. 2, 7 This is particularly important given the key role of surgical resection.

While initial guidance from UK 8 and international 9 neuro-oncology experts has been published, it will be important to establish what overall treatment was provided to patients and perhaps consider a strategy for increasing the capacity of surgical neuro-oncology services and ease the pressures faced by local teams. Although nationally we managed to maintain surgical services for malignant brain tumours, a proportion of low-grade gliomas and meningioma were recommended for interval MRI follow-up rather than early surgery.

Our study has several limitations. Firstly, our primary outcome is based on a hypothetical question asked in the neuro-oncology MDTs on what their recommendation for management would have been prior to the COVID-19 pandemic. The exact location of the tumour and presenting symptoms used for priority stratification in guidelines subsequently published during the pandemic was not collected, so it was not possible to exactly compare practice to available guidelines. Therefore the generalisability of our findings is limited by context-specific factors during the COVID-19 pandemic, including healthcare provider resource utilisation (including COVID-19 caseload) and staff sickness and subjective patient perception of safety in proceeding with admission to a hospital for treatment. Most of the change in management for newly diagnosed high-grade glioma occurred in older patients with A c c e p t e d M a n u s c r i p t 20 poorer performance status, where we recommended for either best supportive care or radiotherapy (without tissue diagnosis). Pre-COVID many of these patients would have been offered surgery and radiotherapy +/-chemotherapy, despite the fact that there is often limited benefit from active oncology treatment in terms of overall survival. 19 These data will continue to be collected in this study in preparation for a second report on the longer-term impact of COVID-19. Our patients had a minimum of only 30-days of follow up and it is likely that some patients may have gone on to have surgery outside of this follow-up window and are not captured in our analysis to date. These followup limitations are even more apparent when capturing the data of patients who did or did not receive chemotherapy or radiotherapy. This limitation will be mitigated with a planned second stage of data collection in July 2021 in order to measure this in detail and also to measure survival data. Thirdly, data on the provision of chemotherapy and radiotherapy may have been impaired by the inability to collect data for patients treated outside of their tertiary neuro-oncology centre. Similarly, our study may underestimate the number of patients who contracted the COVID-19 virus during the study period -particularly if they contracted it in a community setting. A factor that we were not able to determine from this study was the change of referral volume to the neuro-oncology MDTs. Results from another unpublished survey of 30 UK neurosurgical units has shown a 27% reduction in the number of patients discussed in the neuro-oncology MDTs. Wide variations in referrals have been reported for other cancers, and a multicentre prospective study from centres in England and Northern Ireland showed that in April 2020, compared to pre-pandemic data, urgent referrals for early cancer diagnoses was down by 70-89%. 1 M a n u s c r i p t 34 A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t 40 Figure 3 

Estimating excess mortality in people with cancer and multimorbidity in the COVID-19 emergency. medRxiv

Elective surgery cancellations due to the COVID-19 pandemic: global predictive modelling to inform surgical recovery plans

Cancer Research UK. Over 2 million people waiting for cancer screening, tests and treatments

Collateral damage: the impact on outcomes from cancer surgery of the COVID-19 pandemic

Letter: The Impact of the Coronavirus (COVID-19) Pandemic on Neurosurgeons Worldwide

The impact of COVID-19 on neurosurgeons and the strategy for triaging non-emergent operations: a global neurosurgery study

Mortality and pulmonary complications in patients undergoing surgery with perioperative SARS-CoV-2 infection: an international cohort study

The British Neuro-Oncology Society (BNOS) and the Society of British Neurological Surgeons (SBNS)

Neuro-oncology management during the COVID-19 pandemic with a focus on WHO grades III and IV gliomas

The British Neurosurgical Trainee Research Collaborative: Five years on

Sankey or Ribbon Plots

Welcome to the {tidyverse}

gtsummary: Presentation-Ready Data Summary and Analytic Result Tables

Neuwirth E. RColorBrewer: ColorBrewer Palettes

Impact of COVID-19 pandemic on surgical neurooncology multi-disciplinary team decision making: a national survey

A c c e p t e d M a n u s c r i p t 22 A c c e p t e d M a n u s c r i p t