Outcomes and Risk Factors for Influenza and Respiratory Syncytial Virus Lower Respiratory Tract Infections and Mortality in Patients With Lymphoma or Multiple Myeloma: A 7-Year Retrospective Cohort Study

Abstract

Background

Respiratory viral infection (RVI) is a significant complication in patients with hematologic malignancies. While risk factors of lower respiratory tract infections (LRIs) and mortality have been studied in allogeneic hematopoietic cell transplant recipients, data remain limited for patients with lymphoma and multiple myeloma (MM). We investigated outcomes and risk factors of LRI and mortality secondary to respiratory syncytial virus (RSV) or influenza virus (IFV) infections in these populations.

Methods

We performed a retrospective study in adults with lymphoma or MM with RSV or IFV RVIs between 2016 and 2022. Primary outcomes were LRI and all-cause 30- and 90-day mortality.

Results

We analyzed 440 patients with 490 consecutive viral episodes: 297 (61%) with MM and 193 (39%) with lymphoma, 258 (52%) were IFV-related, and 234 (48%) RSV-related (2 coinfections). At presentation, 62% were diagnosed with upper respiratory tract infection (URI) and 38% with LRI. During follow-up, 57% were hospitalized, 8% required intensive care unit transfer, and 20 (4%) died within 30 days. On multivariable analysis, RSV infection (vs IFV), current/former smoking, steroid exposure, lymphopenia (≤200 cells/mL), and high serum creatinine were associated with LRI. MM (vs lymphoma) diagnosis, current/former smoking, lymphopenia, and nosocomial infection were associated with 30-day mortality, whereas LRI (vs URI), current/former smoking, and lymphopenia were associated with 90-day mortality.

Conclusions

We described a high burden of IFV and RSV infections in patients with lymphoma and MM and found risk factors associated with LRI and mortality. These factors could potentially identify high-risk patients, enabling better and prompt management strategies.

Respiratory viral infections (RVIs) pose a significant challenge for patients with hematologic malignancies (HMs), often leading to substantial morbidity and mortality [1, 2]. Early intervention is critical for patients with HM who have respiratory syncytial virus (RSV) or influenza virus (IFV) infections to prevent progression to lower respiratory tract infection (LRI) and mortality [3, 4]. The disease course and risk factors for RSV and IFV infections have been extensively studied among allogeneic hematopoietic cell transplant (HCT) recipients [5, 6]. However, data on patients with lymphoma and multiple myeloma (MM) are limited to small cohort studies [7–9], and there is no validated scoring index to predict outcomes.

The scarce data on IFV and RSV infections in patients with lymphoma and MM showed LRI progression rates of 27% to 31%; hospitalization rates of 56% to 71%, including 7% to 13% intensive care unit (ICU) admissions; and 30- and 90-day mortality rates of 5% and 8% to 10%, respectively [7, 9, 10]. Within the subpopulation of patients with MM and IFV infections, 1 cohort study reported even worse outcomes, with up to 75% of patients experiencing progression to LRI, 42% admitted to the ICU, and 33% dying during hospitalization [10].

Previously identified risk factors for progression to LRI and mortality in patients with HM and RSV infections, including patients with leukemia in particular, were concurrent lymphopenia and neutropenia and lack of ribavirin treatment [9], hypoalbuminemia, hypoxemia at diagnosis, steroid exposure, elevated creatinine levels, and respiratory coinfections [7]. However, risk factors specific to the development of LRI and mortality due to RVI in lymphoma and MM patients are limited.

In this study, we aimed to identify factors associated with LRI and mortality in a large cohort of patients with lymphoma or MM who had RSV or IFV infections. Identifying such prognostic variables could help identify high-risk patients and guide decision-making to facilitate timely treatment, thereby improving outcomes.

METHODS

Study Design

We performed a retrospective cohort study involving consecutive adult patients with lymphoma or MM who had IFV or RSV infections between 1 January 2016 and 31 December 2022, at our institution. Patients with respiratory symptoms were tested for respiratory viruses in either the ambulatory or inpatient setting and at the discretion of their clinical providers. The diagnosis of RVI in symptomatic patients was established using the BioFire Film Array Respiratory Panel 2.1 (BioFire Diagnostics, Salt Lake City, Utah) on nasal washes or swabs. We included multiple viral infection episodes for the same patient if they occurred at least 1 month apart and if the patient experienced clinical recovery between them.

Patient data were collected through electronic medical records. Our primary outcomes of interest were LRI at any time point (from presentation or progression) and all-cause mortality at 30 and 90 days. Secondary outcomes included hospitalization, admission to the ICU, oxygen requirements, mechanical ventilation, and mortality attributed to RVI within 30 days. We also investigated RVI trends over time and seasonality, especially in relation to the coronavirus disease 2019 (COVID-19) pandemic era. This study received approval from the MD Anderson Institutional Review Board (IRB# PA15-0002), and a waiver of consent was granted.

Definitions

Upper respiratory tract infection (URI) was determined when RSV/IFV was detected in samples obtained from the mucosal surfaces of the upper airways, accompanied by upper respiratory tract symptoms such as nasal congestion, cough, rhinorrhea, sinusitis, and pharyngitis and the absence of clinical or radiologic evidence of LRI [11]. LRI was determined when RSV/IFV was detected in samples from the upper or lower airways, along with new or progressive pulmonary infiltrates suggestive of viral infection and at least 1 lower respiratory tract symptom such as cough, sputum production, fever, hypoxia, shortness of breath, and pleuritic chest pain [12]. Probable LRI was defined as above and in addition to RSV/IFV polymerase chain reaction (PCR) detected only in a nasopharyngeal sample, while laboratory-confirmed LRI required the presence of RSV/IFV PCR in bronchoalveolar lavage fluid [11]. Nosocomial infection was defined as a new-onset infection that occurred >48 hours after admission for IFV or >5 days after admission for RSV.

Statistical Analysis

The χ² or Fisher exact test was used to compare categorical data, as appropriate. The Student t test or Mann-Whitney U test was used to compare continuous variables, depending on whether the data followed a normal distribution. Logistic regression analysis was used to identify the independent predictors of the primary outcomes, including LRI and 30- and 90-day all-cause mortality. In detail, first, univariable logistic regression analysis was performed. Next, variables with P values ≤.15 from their univariable analyses were selected to construct an initial multivariable logistic regression model, and then the full model was reduced to the final model using the backward elimination procedure, ensuring that all the variables remaining in the final model had P values <.05. Survival analysis was conducted using Kaplan-Meier curves to depict 90-day mortality, and the log-rank test was used to compare the curves. All the tests were 2-sided with a significance level of .05. The statistical analyses were performed using IBM SPSS Statistics version 25 (IBM Corporation, Armonk, New York) and SAS version 9.4 (SAS Institute, Cary, North Carolina) software.

RESULTS

Study Population

We analyzed 440 patients with 490 consecutive viral episodes; 60.6% had multiple myeloma and 39.4% lymphoma. Table 1 depicts the baseline characteristics of patients with HM and RVI by the site of infection. In brief, more patients with LRI, when compared to URI, were older, former smokers, with active malignancy, infected with RSV, had lymphopenia (<200 cells/mL), had neutropenia (<500 cells/mL), and had higher 30- and 90-day all-cause mortality. At presentation, 187 episodes (38.2%) were classified as LRI. Among the 303 (61.8%) URI episodes, 19 (6.3%) progressed to LRI during 30-day follow-up (Supplementary Table 1). Most of the LRI episodes (82.5%) were categorized as probable.

Patients with RSV infections had a total of 234 viral episodes, including 2 episodes with concurrent IFV infection. Ribavirin was administered to 142 (60.7%) patients, mainly using the oral formulation (97.9%), for a median duration of 6.5 days. Among the 142 patients treated with ribavirin, 57 (40.1%) presented with URI, and 7 (12.3%) progressed to LRI. However, among untreated patients, 4 of 74 (5.4%) patients with URI progressed to LRI (P = .208).

Patients with IFV infections had a total of 258 episodes, including 2 episodes with concurrent RSV infection, as mentioned above. Oseltamivir was administered to 242 (93.8%) patients for a median duration of 5 days. In 172 (66.7%) episodes of URI at presentation, 160 (93.0%) were treated with oseltamivir, and only 8 (4.7%) progressed to LRI within 30 days of follow-up. The timing of oseltamivir within or beyond 48 hours from symptom onset in patients with URI was associated with 6 (8.0%) versus 2 (2.4%) progressions to LRI, respectively (P = .148). All-cause mortality at day 30 was also similar between the 2 groups (3.6% vs 4.5%, respectively; P = .759). Of 58 (22.7%) patients with IFV infections who received the IFV vaccine during the relevant season, 15 (25.9%) had LRI compared to 77 (38.9%) of the nonvaccinated patients (P = .069).

Table 2 compares the characteristics and clinical outcomes of patients with HM stratified by virus (RSV/IFV). When compared to patients with IFV infections, more patients with RSV infections were female, had active malignancy at diagnosis, presented with or progressed to LRI, and had nosocomial infection.

Risk Factors for LRI and Mortality

In a multivariable analysis, factors independently associated with LRI were RSV infection (vs IFV; adjusted odds ratio [aOR], 1.77 [95% confidence interval {CI}, 1.20–2.61]), current/former smoking (aOR, 1.64 [95% CI, 1.10–2.45]), recent steroid exposure (aOR, 2.14 [95% CI, 1.43–3.22]), lymphopenia (aOR, 3.82 [95% CI, 2.02–7.22]), and elevated creatinine level (≥1.2 mg/dL) at presentation (aOR, 2.07 [95% CI, 1.36–3.15]) (Table 3).

Twenty patients, 10 with RSV and 10 with IFV, died by day 30, accounting for a 4.1% all-cause mortality rate; 19 of the 20 deaths were RVI-related. The 90-day all-cause mortality rate was 6.5% (32/490). The comparisons of the characteristics of 30- and 90-day survivors and nonsurvivors are presented in Supplementary Tables 2 and 3, respectively. Notably, nineteen (95.0%) of the 20 patients who died by day 30 presented with LRI, while 1 patient experienced progression to LRI during follow-up. In a multivariable analysis, baseline MM diagnosis (vs lymphoma; aOR, 4.08 [95% CI, 1.03–16.15]), current/former smoking (aOR, 2.97 [95% CI, 1.02–8.63]), nosocomial infection (aOR, 12.8 [95% CI, 3.10–52.76]), and lymphopenia (aOR, 20.12 [95% CI, 6.79–59.60]) were independently associated with 30-day all-cause mortality (Table 3). Thirty-two patients died by day 90 (Supplementary Table 2). In a multivariable analysis, independent risk factors of the 90-day mortality were current/former smoking (aOR, 2.48 [95% CI, 1.10–5.59]), lymphopenia at presentation (aOR, 9.05 [95% CI, 3.96–20.66]), and LRI (vs URI; aOR, 5.06 [95% CI, 1.83–13.95]) (Table 3). In Kaplan-Meier survival analysis, patients with LRI (vs URI), patients with laboratory-confirmed LRI (vs probable LRI), and patients with MM and laboratory-confirmed LRI (vs patients with MM and probable LRI) were all associated with lower 90-day survival (all log-rank P < .001) (Figure 1  A–D).

Table 4 describes patients by their underlying malignancy. When compared to patients with lymphoma, more patients with MM were older, were Black, had active malignancy at RVI diagnosis, were on steroids, received autologous transplantation, were admitted for RVIs, required oxygen supplementation, and presented with or progressed to LRI. To account for the dissimilarities between patients with lymphoma and MM, separate multivariable analyses for each group were performed. In patients with lymphoma, the risk factors for LRI were older age, previous chest radiotherapy, RSV infection (rather than IFV), respiratory viral coinfection, and lymphopenia (Supplementary Table 4A). The risk factors for 90-day mortality were lymphopenia (aOR, 15.62 [95% CI, 3.86–63.18]; P < .001) and LRI (aOR, 13.20 [95% CI, 1.56–111.60]; P = .018). A 30-day mortality analysis was not performed due to the small number of events (n = 5).

In patients with MM, recent steroid exposure, lymphopenia, and elevated creatinine levels were associated with LRI (Supplementary Table 4B). The risk factors for 30-day mortality were lymphopenia (aOR, 13.78 [95% CI, 4.29–44.22]; P = .001) and nosocomial infection (aOR, 12.37 [95% CI, 2.21–69.26]; P = .004), whereas the risk factors for 90-day mortality were lymphopenia (aOR, 6.43 [95% CI, 2.34–17.68]; P < .001), LRI (aOR, 3.95 [95% CI, 1.20–13.04]; P = .024), and nosocomial infection (aOR, 8.12 [95% CI, 1.63–40.40]; P = .011) (Supplementary Table 4B).

IFV and RSV Seasonality

This study included 7 years of RVI episodes; 77% occurred before the COVID-19 pandemic (January 2016–February 2020) (Figures 2 and 3), when the seasonality of RVIs was predictable, with cases peaking during the winter months (October–March). From March 2020, when the COVID-19 pandemic started in the United States, until June 2021, there were no documented RSV/IFV infections in our cohort of patients with lymphoma and MM, and until February 2022, there were no IFV cases. Furthermore, since March 2020, more IFV/RSV cases have been diagnosed in patients with MM compared to patients with lymphoma (26.9% vs 17.6% of cases, P < .017) (Table 4). Total LRI and 30-day mortality rates among infected individuals were like those in the pre-COVID-19 era (42.1% vs 42.0% LRIs and 4.4% vs 4.0% 30-day mortality rates, respectively).

DISCUSSION

We identified a high burden of IFV and RSV infections among patients with lymphoma and MM at our center from 2016 to 2022. In this cohort, we found a high rate of RVI-related LRI (42%), with an RVI-related mortality of 3.9% at 30 days. Additionally, patients were often admitted to the hospital, with 32.5% requiring oxygen supplementation and 7.8% requiring ICU admission. We identified multiple factors associated with RVI-related LRI, including smoking status, lymphopenia, elevated creatinine level, and steroid use; additionally, we determined that almost half of the RSV infections were associated with LRI. There were significant associations between smoking status, lymphopenia, type of underlying malignancy (MM), nosocomial infection, and 30-day mortality.

Patients with HM are at significant risk for RVI complications, including progression to LRI and mortality, due to multiple host-related factors, such as immunosuppressive therapy and underlying disease. Yet there are only limited data on RVIs other than COVID-19 in patients with lymphoma and MM [13, 14]. Previous studies have reported higher rates of LRI and RVI-related mortality in patients with MM than in those with lymphoma [9, 10]. This difference was previously attributed to impaired lymphocyte function, more aggressive chemotherapy, higher autologous transplant rate, and hypogammaglobulinemia in MM patients [8, 15]. In our cohort, lymphopenia and steroid exposure were associated with LRI and 30- and 90-day mortality; additionally, patients with MM were at greater 30-day mortality risk.

Prior studies in cancer patients have reported high rates of LRI and mortality associated with IFV and RSV [2, 5, 16]. In patients with HM, reported progression rates to LRI range from 20% to 60%, depending on the degree of immunosuppression, history of HCT, and antiviral treatment [3, 7, 16–18]; among patients with lymphoma and MM, LRI rates can be as high as 75% [4, 7, 9, 10]. In our study, RSV infection was more strongly associated with LRI than IFV. The most significant difference was noted in patients with lymphoma. Both viral infections had similar 30- and 90-day mortality rates, around 4.3% and 6.5% for RSV and 3.9% and 6.6% for IFV, respectively.

The higher rate of RSV LRI in our cohort may be due to the lack of effective antiviral therapy and vaccinations (until recently) to treat or prevent RSV infection complications. Neuraminidase inhibitors such as oseltamivir are widely available to treat IFV-related URI or LRI [19]. Early treatment of IFV infections with oseltamivir has been associated with decreased rates of IFV LRI in HCT recipients [20], and annual IFV vaccination among patients with HM was associated with improved outcomes [19, 21, 22]. For instance, Kumar et al found that the current-season IFV vaccination reduced LRI risk by 66% and ICU admission risk by 51% among posttransplant patients infected with IFV [23]. In patients with lymphoma, IFV vaccination resulted in a good immune response [24]. In our cohort, oseltamivir was administered in almost all influenza cases (93.8%); however, we did note a low vaccination rate (26.5%). By comparison, no approved treatment for RSV is available, and the newly approved RSV vaccines were not commercially available during the study period [25, 26]. Furthermore, RSV vaccines are recommended for older adults, and their efficacy in immunocompromised patients has yet to be determined. Ribavirin therapy for RSV infections in high-risk patients with HM has been used, but there is a lack of high-quality clinical efficacy data [4, 27, 28]. Therefore, well-designed clinical trials of antiviral agents and vaccines for immunocompromised patients infected with respiratory viruses such as IFV and RSV are needed as the burden of LRI and mortality is substantial.

IFV and RSV seasonality was constant and predictable for decades until the COVID-19 pandemic began in March 2020. The circulation of severe acute respiratory syndrome coronavirus 2 resulted in significant changes in the presence and timing of RSV and IFV circulation, as reported in numerous epidemiologic studies among the general population [29–31]. According to the Centers for Disease Control and Prevention, RSV infections peaked earlier than usual during the 2022–2023 respiratory virus season, in October and November [30], and IFV also peaked earlier, during December and January [29]. Evaluating the seasonality of those viruses is crucial for vaccination timing and diagnostic approaches.

Our study has several limitations. First, our cohort included 2 distinct subpopulations: patients with lymphoma and MM, whose diseases have different pathogenesis, antineoplastic treatment regimens, and prognosis. By analyzing the risk factors and outcomes as 1 group, we might draw conclusions that do not reflect 1 of the subpopulations. However, we elected to combine those groups because of their underlying B- and T-cell deficiencies that distinguish them from other patients, such as HCT recipients. To account for these dissimilarities, we performed a separate multivariable analysis for each group. Second, few patients in our cohort experienced progression from URI to LRI during the study period (rather than being diagnosed with LRI at presentation); thus, a multivariate analysis to identify specific risk factors for progression to LRI was not attempted (Supplementary Table 1). Third, our cohort included only patients tested at our institution, and we probably missed infected patients diagnosed in outside facilities or mildly symptomatic patients who were not tested at all. Furthermore, the study describes a single-center experience that could affect the generalizability of the results. However, to our knowledge, this is the largest cohort that includes patients with lymphoma and MM and RVIs, and most of the results align with data from previous cohorts. Last, the study's retrospective nature affects the ability to account for potential confounders. We tried to adjust for the most clinically important ones using multivariable analyses.

In summary, our study describes the high burden of IFV and RSV infections in patients with lymphoma and MM. We also identified unique risk factors associated with LRI and mortality in this population. Of importance, we note that RSV may be associated with worse outcomes in patients with lymphoma and MM compared to IFV. This highlights the need for prospective studies to measure the rates of RSV- or IFV-related complications in patients with HM, investigating the effects of vaccination and antiviral therapy on significant clinical outcomes.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author contributions. D. R.-M., D. V.-C., M. V. B., and R. F. C. conceived and designed the study. D. R.-M. and T. S. extracted, collected, and validated the data. Y. J. and T. S. performed the analysis. Y. J. supervised the analysis. D. R.-M., D. V.-C., M. V. B., T. S., F. K., and R. F. C. interpreted the data. D. R.-M. and T. S. wrote the original draft. F. K., Y. J., A. S., E. A. H., D. V.-C., S. A., M. B. and R. F. C. reviewed and edited the manuscript. R. F. C. supervised the work. All authors reviewed the manuscript and approved the final version of the manuscript.

Acknowledgments. The authors thank Amy Ninetto of the Research Medical Library at The University of Texas MD Anderson Cancer Center for editing the manuscript.

References

Author notes