Myeloma clinical outcomes following the first wave of COVID-19: results from the Thames Valley Cancer Alliance (UK).

Abstract

Concern for higher mortality in myeloma patients was reflected in the response to the first wave of the novel coronavirus SARS-CoV-2 (COVID-19) pandemic. Modifications were made to patients’ treatment during the first wave of this pandemic, which were guided by pragmatic recommendations from both national and international professional organisations. Key guiding principles were as follows: to limit hospital visits, using oral agents as much as possible and/or allowing treatment breaks.1 Our recently published paper emphasised the importance of formally assessing the impact of myeloma care adaptations and treatment modifications to inform practice in future waves of the pandemic.2 Here, we describe the outcomes of 219 myeloma patients treated within the Thames Valley Cancer Alliance during the first wave of COVID-19, which we defined between the 1st of March and the 30th of June 2020. The primary outcome was occurrence of COVID-19, recorded as the incidence rate of COVID-19 in myeloma patients on active treatment, compared to the incidence rate of all cases within the Thames Valley Region (Oxfordshire and Buckinghamshire), obtained from the Office for National Statistics.3 The secondary outcome was haematological response (HR) at three months. We audited our Thames Valley region chemotherapy database for patients with multiple myeloma on active therapy. Patients were excluded if they were not on chemotherapy, and where detailed treatment data were incomplete. All myeloma patients were discussed in a virtual multidisciplinary team (MDT) meeting comprising myeloma consultants, pharmacists and specialist nurses. The following were used to stratify a patient’s risk of COVID-19 mortality: age, co-morbidities, degree of immunosuppression/immunoparesis (reflected by lymphopenia/neutropenic and hypogammaglobulinaemia), the number of previous lines of therapy and whether the regimen was parenteral. This was weighed against the risk of myeloma disease progression assessed by: baseline cytogenetics, immunoglobulin subtype and the current disease response status. Outcomes from the MDT regarding the decision on disease management during the first wave of COVID-19 of every individual patient were based on the above-mentioned factors. MDT decisions were discussed with all patients to explain the reasoning during an unprecedented pandemic situation. Lack of evidence was also highlighted. MDT decisions were supported by all of our patients, who gave their consent. The treatment modification was categorised as: (i) no modification (treatment continued in the same regimen/dosing/intensity as planned before the COVID-19 pandemic); (ii) treatment continued, but interval extended; (iii) dose reduction; (iv) treatment paused for less than an eight-week interval; (v) treatment paused for longer than eight weeks; (vi) treatment cessation; (vii) switched to oral regimen; (viii) autologous stem cell transplant withheld; and (ix) other. HR to ongoing therapy was assessed at three months, defined according to International Myeloma Working Group (IMWG) criteria and also categorised as: stable, deepening of response and disease progression.4 A three-month timepoint was chosen to capture the impact of treatment modification during the COVID-19 period. Patients showing symptoms raising suspicion of a COVID-19 infection underwent a SARS-CoV-2 PCR on a nasal or throat swab, to ascertain whether they were COVID-19-positive or -negative. At the start of the first wave, there were no routine SARS-CoV-2 PCR tests available for asymptomatic patients. If a patient tested positive, the outcome of COVID-19 infection was also recorded as: recovery or death. Deaths not related to COVID-19 were also recorded, including the cause of death, where available. All patients consented to retrospective analysis of their records at the point of treatment, and all patient records were anonymised at the point of analysis. Service evaluation approval was obtained locally to conduct this analysis. Categorical data were tabulated and summarised using percentages and continuous data were summarised with medians and ranges. Pearson’s chi-square test was used to assess the likelihood of pausing/stopping according to myeloma response categories, and according to different myeloma treatment regimen. A total of 219 myeloma patients were identified as being on active treatment at the start of the first wave of the COVID19 pandemic. Of these, 125 patients continued treatment and 94 patients underwent treatment modifications. The consort diagram (Figure S1) shows the flow of patients. The baseline patient, disease and treatment characteristics of both cohorts of patients, including their HR achieved at the time of the MDT discussion are outlined in Supplementary Table SI. With regard to the primary outcome, a total of six patients (6/219, 3%) developed COVID-19, based on a SARS-CoV-2 PCR test. This gives a COVID-19 incidence rate of 0·027 (n = 6/219) over the three-month period March to June. In this same period, the Thames Valley region had 2,894 confirmed COVID-19 cases (Buckinghamshire 947, Oxfordshire 1,974). This gives a COVID-19 incidence rate of 0·0023 cases over the same three-month period. The characteristics of these patients are outlined in Table I. The characteristics of those confirmed COVID-19-positive, compared to those who tested negative, or had no symptoms and were therefore not tested, are outlined in Table SII. A total of 52 (n = 52/219, 24%) patients progressed and 15 died (n = 15/219, 6·8%) during the follow-up period. In the 52 patients who progressed, 34 (n = 34/52 = 65%) remained on treatment and in 18 (n = 18/52 = 35%) no treatment modifications were made. In the 15 patients who died, nine remained on therapy, (n = 9/18, 50%), five (n = 5/18, 27·8%) had their treatment paused for up to eight weeks, one patient had his treatment stopped altogether (n = 1/18, 5·6%) and the exact modification was missing in the remaining three cases (n = 3/18 = 16·7%). The cause of death was available in 13 cases (n = 13/15, 87%): multiple myeloma (n = 8/13, 87%) and COVID-19 (n = 5/13, 48%). Although one patient died with COVID-19, the cause of death was listed as gastric volvulus. One COVID-19-positive patient recovered and survived. The case fatality rate for COVID-19 was 66·7%. The corresponding death rate within our region was 175·4 in Buckinghamshire and 287·1 in Oxfordshire respectively. Table II outlines the HR at the time of MDT discussion and at three months, compared to expected response rates, obtained from recent trial data (Table II). The HR in the patients alive at three months was as follows: stable disease (n = 110/204, 54%), deepening of response (n = 46/204, 23%) and disease progression (n = 52/204, 25%). Treatment was more likely to be paused/stopped in those with a deep remission (CR 48% vs 17%, Pearson’s chi-square 0·00) and in patients on parenteral therapy (DVd- 16% vs 8%; daratumumab monotherapy 14% vs 7%; and carfilzomib 12% vs 2% for patients who received modified versus continued therapy respectively, Pearson’s chi-square 0·03). A very small number of patients tested positive for COVID-19 (n = 6/219, 3%). It is therefore difficult to conclude whether patients with myeloma have a low risk of contracting the virus, whether the methods outlined in this study reduced this risk, or whether myeloma patients within our region simply adopted the recommended fastidious shielding methods. Despite a low incidence rate of infection, the case fatality was high in those who contracted the virus, confirmed by other recent real-world data.5 Even with the small number of positive cases, the majority of patients who tested positive tended to be lymphopenic (83% vs 48%), had low immunoglobulins (83% vs 62%) and were more likely to have poor disease control (SD/PD 83% vs 35%) compared to those who did not contract the virus, but there did not appear to be a trend for a specific treatment regimen. The other notable finding was the reduced HR in patients on ixazomib, lenalidomide and dexamethasone (IRD) (n = 34): overall response rate (ORR) 53% compared to expected ORR of 78%. Treatment modifications were made in 12 patients (13%) on IRD. We acknowledge that this study has limitations. Firstly, it is not a randomised study. The decision to continue or pause therapy was made by clinicians who knew the patients, risking bias. The incidence rate only reflects those who attended hospital and were tested for COVID-19, missing those with asymptomatic/mild symptoms and those who shielded at home. These limitations do reflect reality, and the government is likely to use methods of self-isolation/shielding of high-risk individuals to reduce infection rates during future waves of infection. Finally, HR rate was assessed at three months, in patients at various different timepoints in their disease, and compared to ORR quoted in the literature. The rationale for this methodology was the critical need for an early assessment to evaluate the impact of treatment modifications. Despite these limitations, we hope the data presented here highlight how knowledge from the first wave of the COVID-19 pandemic must be used to inform and guide clinicians. Immunosuppression was initially feared as a major risk factor for contracting COVID-19, and myeloma physicians were advised to omit or modify myeloma patients’ glucocorticoid doses. Published guidelines must now be revised, given that steroid may improve survival in COVID-19. The patients’ perspective is a critical and yet overlooked aspect of the pandemic. Future studies should aim to include qualitative data to encapsulate this. In conclusion, this descriptive study of 219 MM patients illustrates how a systematic approach to a patient’s risk of infection and individualised treatment modifications are possible and may have contributed to a low incidence rate of COVID-19 infection. Yet despite this coordinated approach, the high mortality rate warns of a substantial risk of fatality should myeloma patients contract COVID-19. We would like to thank all the patients, clinical and nursing staff at the Oxford University Hospitals. The authors have no conflict of interest to disclose. FAS, FD and KR conceived the study. FAS, JL, AH, BT and HL collected the data. FAS analysed the data. FAS, FD and KR wrote the manuscript. All the authors contributed to the manuscript and provided critical input. All the authors reviewed the final version of the manuscript. Not required. 6 CR – 5 (17) VGPR – 4 (13) PR – 9 (30) SD – 3 (10) PD – 9 (30) Missing – 4 ORR – 18 (60) CR – 10 (32) VGPR – 10 (32) PR – 4 (13) SD – 2 (6) PD – 5 (16) Missing – 3 ORR – 24 (77) ≥CR – 10 ≥VGPR – 55-65 Overall response rate (ORR) – 80–90 (Moreau et al.6) CR – 10 (40) VGPR – 3 (12) PR – 4 (16) SD – 5 (20) PD – 3 (12) Missing – 0 ORR – 17 (68) CR – 12 (50) VGPR – 3 (13) PR – 1 (4) SD – 1 (4) PD – 7 (29) Missing/died – 1 ORR –1 6 (67) ≥CR – 19 ≥VGPR – 59 ORR – 83 (Palumbo et al.7) CR – 9 (31) VGPR – 6 (21) PR – 7 (24) SD – 4 (14) PD – 3 (10) Missing – 2 ORR – 22 (76) CR – 7 (23) VGPR – 6 (20) PR – 3 (10) SD – 7 (23) PD – 7 (23) Missing – 1 ORR – 16 (53) ≥CR – 12 ≥VGPR – 48 ORR – 78 (Moreau et al.8) CR – 3 (10) VGPR – 4 (14) PR – 8 (28) SD – 8 (28) PD – 6 (21) Missing – 2 ORR – 15 (52) CR – 3 (11) VGPR – 7 (26) PR – 4 (15) SD – 7 (26) PD – 6 (22) Missing or died – 4 ORR – 14 (52) ≥CR – 8 ≥VGPR – 23 ORR – 72 (Durie et al.9) CR – 0 VGPR – 0 PR – 5 (20) SD – 3 (12) PD – 5 (20) Missing – 0 ORR – 3(12) CR – 0 VGPR – 0 PR – 3 (14) SD – 3 (14) PD – 4 (18) Missing – 3 ORR – 3 (14) ≥CR – 0 ORR – 30 (Richardson et al.10) Table S1. Baseline characteristics of patients who continued myeloma treatment (as per pre-COVID-19 plan) and those who had their myeloma disease management modified as a result of the COVID-19 pandemic (including stopping therapy), and their haematological response at the time and treatment regimen. Table SII. Comparison of myeloma patients who developed COVID-19 infection compared to those who remained free of COVID-19 infection. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.