Impact of COVID-19 in patients with multiple myeloma based on a global data network

The COVID-19 pandemic has represented a major cause of morbidity/mortality worldwide, overstressing health systems. Multiple myeloma (MM) patients show an increased risk for infections and they are expected to be particularly vulnerable to SARS-CoV-2 infection. Here we have obtained a comprehensive picture of the impact of COVID-19 in MM patients on a local and a global scale using a federated data research network (TriNetX) that provided access to Electronic Medical Records (EMR) from Health Care Organizations (HCO) all over the world. Through propensity score matched analyses we found that the number of new diagnoses of MM was reduced in 2020 compared to 2019 (RR 0.86, 95%CI 0.76–0.96) and the survival of newly diagnosed MM cases decreased similarly (HR 0.61, 0.38–0.81). MM patients showed higher risk of SARS-CoV-2 infection (RR 2.09, 1.58–2.76) and a higher excess mortality in 2020 (difference in excess mortality 9%, 4.4–13.2) than non-MM patients. By interrogating large EMR datasets from HCO in Europe and globally, we confirmed that MM patients have been more severely impacted by COVID-19 pandemic than non-MM patients. This study highlights the necessity of extending preventive measures worlwide to protect vulnerable patients from SARS-CoV-2 infection by promoting social distancing and an intensive vaccination strategies.

Multiple myeloma (MM) patients are known to have variable degrees of multifactorial immunodeficiency related to the disease itself and the administered therapies. Therefore the risk of infection is increased in MM patients, and it is a major cause of morbidity and mortality [13][14][15][16]. In particular, recent studies have shown that viruses represent a frequent etiology of infections in MM patients [17,18]. Multicenter and international clinical studies have documented that patients with MM are a vulnerable population at high risk of hospitalization and death following a COVID-19 infection [19][20][21][22][23]. However, there are no data evaluating whether MM patients specifically are at increased risk of infection by SARS-CoV-2 compared with the overall population.
We hypothesized that the COVID-19 pandemic might have decreased or delayed the diagnosis of MM for many patients, in comparison with the pre-COVID-19 era [24,25], in the context of overburdened NHS, movement restrictions, and other measures implemented to control the spread of the infection. Additionally, we tried to confirm that MM patients were disproportionately impacted by COVID-19 compared with non-MM patients in terms of incidence of COVID-19 infection and survival outcomes.
Since geographical differences may exist between countries such as the incidence of COVID-19, the organization of healthcare delivery, the availability of health resources, the government strategy to combat the pandemic, and many other factors, it seemed appropriate to draw on a global data network to obtain a more accurate and comprehensive picture of the impact of COVID-19 in MM patients. Large network data platforms based on electronic medical records (EMRs) provide capabilities for comparing cohorts across time periods and clinical profiles. For this study, we used TriNetX, a global health research network, to test our hypothesis in an attempt to overcome some of the constraints of previous studies which used local or regional data sources. This platform has previously been shown to be a useful tool for answering research questions in different settings of diseases [26][27][28][29][30].
Thus, the general objective of this study was to provide comparative data on a local, regional and global scale on the impact of COVID-19 on MM patients. The specific aims of our study were to compare the occurrence of new MM diagnoses and the survival of MM patients in 2020 and 2019, and to compare the proportion of COVID-19 cases and the excess mortality in MM and non-MM patients.

Database network and patients selection
This study was conducted with data obtained from TriNetX, LLC ("TriNetX") a global federated health research network that provides access to EMRs from healthcare organizations ("HCOs") all over the world. TriNetX provides access to data containing diagnoses, procedures, medications, laboratory values, genomic information from approximately 90 000 MM patients from over 66 HCOs from the 68 that are part of the network. The analyses were conducted utilizing three networks to confirm our hypotheses: the Hospital 12 de Octubre network (H12O), with 930 000 patients; the TriNetX EMEA Collaborative Network ("EMEA"), with 9,800,000 patients from 15 HCOs (including Hospital 12 de Octubre); and the TriNetX Global Collaborative Network ("Global"), with 82 000 000 patients from 68 Healthcare Organizations (including EMEA HCOs). All data collection, processing, and transmission were performed in compliance with all Data Protection laws applicable to the contributing HCOs, including the EU Data Protection Law Regulation 2016/679, the General Data Protection Regulation on the protection of natural persons with regard to the processing of personal data and the Health Insurance Portability and Accountability Act ("HIPAA"), the US federal law which protects the privacy and security of healthcare data. The TriNetX EMEA and Global Collaborative Networks are distributed networks, and analytics are performed on anonymized or pseudonymized/de-idenfied (per HIPAA) data housed at the HCOs, with only aggregate results being returned to the TriNetX platform. Individual personal data does not leave the HCO. TriNetX is ISO 27001:2013 certified and maintains a robust IT security program that protects both personal data and health care data.
MM patients were identified by the presence of the parent ICD-10-CM code for MM (C90.0) or any of the specific codes (C90.00, C90.01 and C90.02) in their EMR. In addition, we included patients whose MM was miscoded initially as monoclonal gammopathy by identifying patients who had the diagnosis code D47.2 (monoclonal gammopathy) and had been treated with one of the following MM treatments: thalidomide, bortezomib, lenalidomide, daratumumab, melphalan, ixazomib or carfilzomib. For the overall survival analysis, the first day of MM was either the diagnosis or the start date of the treatment, whatever happened first. The infection of COVID-19 was identified as either a positive PCR test or an ICD-10-CM diagnosis of COVID-19 in 2020, ICD codes U07.1 (COVID-19, Virus identified), U07.2 (COVID-19, Virus not identified), or B97.29 (Other coronavirus as the cause of diseases classified elsewhere). Finally, included patients were ≥25 years years old.

Patient flowchart and characteristics
A total of 855 patients with MM criteria were identified in H12O network, 7265 in EMEA, and 83,550 in Global. Control cohorts of non-MM patients were identified by propensity score matching 1:1 on age and gender from a pool of individuals without any diagnoses of MM or monoclonal gammopathy and without any record of receipt of MM treatments (Table 1). For the non-MM patients, the cohort was limited to patients thathad a hospital visit in the last 5 years. The three networks showed a higher probability (p < 0.005) of hypertension, other forms of heart diseases (different from ischemic heart diseases), chronic kidney disease and diseases of the musculoskeletal system, disorders of bone density for patients with MM. MM patients in the EMEA and Global comparisons also showed a significantly higher probability of diabetes mellitus, other dorsopathies, injuries and lower mean value of glomerular filtration rate (MDRD) compared with non-MM patients. In addition, a significantly higher percentage of MM patients in the global cohort had ischemic heart diseases.
The MM cohort in each of the networks was further stratified to run the different analyses. One set of sub-cohorts included the patients with a new diagnosis of MM from 2019Q1 to 2021Q2 (Fig. 1A). They were used to evaluate the impact of the COVID-19 pandemic in the number of new MM diagnoses (Fig. 2) and in the overall survival (OS) of newly diagnosed MM patients (Fig. 3). Another set ofsub-cohorts included MM and non-MM patients that had an emergency visit or hospitalization in 2019 or 2020 (Fig. 1B); they were used to assess the incidence of COVID-19 and the excess mortality in 2020. A third set of cohorts was built with COVID-19 patients, MM and non-MM, to analyse the evolution of the OS in the period of January to May, 2020, compared to the period of June to December 2020 (Fig. 1C). In all settings, non-MM cohorts included patients with sufficient information in their EMR to run the analysis: patients without any diagnosis or with an EMR trajectory shorter than 3 months were excluded.

Statistical analysis
All analyses were generated with TriNetX platform software (TriNetX, Cambridge, MA) in August of 2021 [31]. We compared the incidence (new cases) of MM diagnosis and the survival of MM patients in 2020 and 2019. We also compared the incidence (case count) of COVID-19 and the excess mortality of MM and non-MM patients. Finally, we compared the survival of COVID-19 MM and non-MM cases over two time periods in 2020.
The number of new cases of MM and of COVID-19 cases were compared with risk ratios and 95% confidence intervals (95% CI). Kaplan-Meier analysis was used to estimate survival probabilities, and the difference between groups was tested using the log-rank test and quantified with hazard ratios (95% CI), calculated with TriNetX Analytics features. All the cohorts were propensity score-matched on age and gender. For the survival analysis of MM and non-MM patients and the comparisons between 2020 and 2019, the cohorts were also matched on mortality risk factors for COVID-19 patients: I10-I16 (Hypertensive diseases), E08-E13 (Diabetes mellitus), N18 (Chronic kidney disease), I30-I52 (Other forms of heart disease), M00-M99 (Diseases of the musculoskeletal system and connective tissue), J40-J47 (Chronic lower respiratory diseases) and I20-I25 (Ischemic heart diseases) [32][33][34]. Global was not used in the survival analysis because of network data limitations on out-of-hospital mortality.    The final analysis compared the survival of COVID-19 MM and non-MM patients diagnosed in two periods, January to May 2020 (first period) and June to December 2020 (second period) (Fig. 1). After balancing cohorts with propensity score matching, the analysis showed no statistically significant differences for the survival of COVID-19 MM patients: in H12O, 84.6% (11) vs 76.9%,

DISCUSSION
In this real-world data analysis, we showed that diagnosis of new MM cases has decreased during the COVID-19 pandemic at a local and global level. Similarly, the survival of MM patients has decreased in 2020 compared to 2019. Our analysis suggests that MM patients have a higher risk of contracting SARS-CoV-2 infection than non-MM patient population. We found also that the excess mortality in 2020 was higher for MM patients than for non-MM patients. These results confirmed the remarkable impact of the COVID-19 pandemic in the management and outcomes of MM patients.
The COVID-19 pandemic has overloaded our health systems globally; this resulted in a lack of attention to other pathologies that require very specialized diagnostic and therapeutic tools.
Some authors hypothesized that the number of newly diagnosed cancer patients has been reduced during pandemic; however, this is the first time in the MM setting that this fact has been demonstrated in a large international multi-site studies with multiinstitutional and multi-national patients [9,24,25,28,[35][36][37]. Our results could mean that globally around 15% of MM patients have not been diagnosed on time, or they have died because of the severe consequences on our health systems and patients. This is consistent with previous analysis on other cancers [28,29,38,39].
There is little information about whether cancer patients have an increased probability of having COVID-19 [37]. In this large series we have confirmed that the likelihood to be diagnosed of COVID-19 is higher in MM patients that in the general population.
Finally, we observed that the survival of MM patients decreased in 2020 compared to 2019. Furthermore, we confirmed that the excess mortality in 2020 was higher in MM population than in non-cancer population. This reflect, the major impact of COVID-19 pandemic in vulnerable populations such as MM patients.
In contrast to the non-MM population, patients with MM did not show any improvement in survival results during the second period of the pandemic (June to December 2020), once the first wave of infection (which was the most devastating and stressing for hospital systems in most European countries and the USA) had passed. This finding could be partially explained by a reduced effectiveness of current COVID-19 therapies in this vulnerable MM population -clinical trials for COVID-19 treatments often excluded cancer patients. The lack of improvement in MM patients survival may also be related to the continued adaptations of the health care systems during the pandemic situation.
Overall, a reduced incidence of hematological malignancies has been reported during the first wave of the pandemic, as we have observed in our study, and some studies are addressing the increased risk of getting COVID-19 in these populations [28]. Prior works have also noticed, in line with our findings in MM, that other malignancies such as chronic lymphocytic leukemia or acute leukemias also associate a higher risk of SARS-CoV-2 infections with fatal complications [40][41][42][43]. By contrast, chronic myeloid leukemia patients may have not resulted so affected by the pandemic [44][45][46]. Interestingly, a decrease in the number of hematology-related diagnostic procedures (bone marrow aspirates, flow cytometry assessments, etc) carried out during pandemic period has been reported [47], which could correlate with our observation of lower number of newly diagnosed MM patients during 2020. Since vaccination against SARS-CoV-2 is less effective in hematological neoplasms, these patients might be at risk even after complete vaccination and a specific management different from the overall population may be required to guarantee a better protection [48][49][50][51]. Future studies assessing the efficacy of vaccines against SARS-CoV-2 in each particular hematological malignancy and evaluating the impact of the different therapies in the acquisition of COVID-19 immunity will help to optimize the vaccination strategy [52].
There are some study limitations worth mentioning. First, EMR data is subject to data entry errors and data gaps; such as the date of MM diagnosis may not be the actual day of MM diagnosis as it was inferred from the EMR record. Mortality data could be incomplete in some organizations or reported with some delay. We excluded from the survival analysis the data network where out-of-hospital deaths are not tracked consistently. Although data were not centrally curated, H12O MM dedicated datasets were used as a reference for controlling the quality of MM data in all cohorts [19,20]. One of the strength of this study is the validation of the results in three different data networks around the world that provided very consistent findings for all impact measures. In addition, this study included a large number of patients in the study population of interest contrary to most most MM studies. The large sample size and the use of propensity score matching allowed for more accurate comparisons through controlling for potential factors with clinical and prognostic relevance in an attempt to minimize the risk of bias.
This real-world data global analysis showed that COVID-19 has severely impacted MM patients at different levels. Diagnosis of MM patients was delayed and survival of MM patients was reduced in 2020 compared to 2019. MM patients were more frequently infected with SARS-CoV-2 and had higher excess mortality in 2020 than non-MM patients visiting hospitals. By interrogating large EMR datasets from HCO in Europe and globally, we confirmed that MM patients have been more severely impacted by COVID-19 pandemic than non-MM patients. This study highlights the necessity of extending preventive measures worlwide to protect vulnerable patients from SARS-CoV-2 infection by promoting social distancing and an intensive COVID-19 vaccination strategies.