Association of Pulmonary Hypertension and Monoclonal Gammopathy of Undetermined Significance

Objective To determine the prevalence of monoclonal gammopathy of undetermined significance (MGUS) in patients with PH as well as precapillary PH. Methods Olmsted County residents with PH, diagnosed between 1/1/1995 and 9/30/2017, were identified, and age and sex were matched to a normal control group. The PH group and normal control group were then cross-referenced with the Mayo Clinic MGUS database. Charts were reviewed to verify MGUS and PH. Heart catheterization data were then analyzed in these patients for reference to the gold standard for diagnosis. Results There were 3419 patients diagnosed with PH by echocardiography between 1995 and 2017 in Olmsted County that met the criteria of our study. When the PH group (N = 3313) was matched to a normal control group (3313), a diagnosis of MGUS was significantly associated with PH 10.2% (OR = l.84 [95% CI 1.5–2.2], p < 0.001), compared with controls 5.8% based on echo diagnosis. Using heart catheterization data (484 patients), a diagnosis of MGUS was associated with PH 13.0% (OR = 3.94 [95% CI 2.28–6.82], p < 0.001). For pulmonary artery hypertension (N = 222), a diagnosis of MGUS was associated with PH at similar 12.2% (OR = 4.50 [95%CI 1.86–10.90], p < 0.001. Conclusions There is a higher prevalence of MGUS in patients with PH and precapillary PH compared with normal controls. This association cannot be explained fully by other underlying diagnoses associated with PH. Assessing for this in patients with PH of unclear etiology may be reasonable in the workup of patients found to have PH.


Introduction
Pulmonary hypertension (PH) is defned by a mean pulmonary artery pressure (mPAP) > 25 mm·Hg at rest. As PH progresses, it results in right ventricular systolic dysfunction and eventually right ventricular failure. Pulmonary hypertension is classifed as fve groups based on the World Health Organization (WHO) classifcation system: pulmonary arterial hypertension (Group 1), pulmonary hypertension due to left heart disease (Group 2), pulmonary hypertension due to chronic lung disease (Group 3), chronic thromboembolic pulmonary hypertension (Group 4), and pulmonary hypertension due to unclear mechanisms (Group 5) [1][2][3]. Specifcally, Group 5.1 includes hematologic disorders. Updates to the clinical classifcation of PH and the recent 6 th World Symposium on Pulmonary Hypertension suggested an mPAP of > 20 mm·Hg as abnormal. Chronic hemolytic anemia and myeloproliferative disorders are included under Group 5.1 [4]. Tere is not yet any specifc mention of an association in the classifcation with any monoclonal gammopathy such as myeloma, but this would be presumed to be classifed under Group 5.1. More recent work by the Task Force [4] looks at a further breakdown by hemodynamic data. A categorization of precapillary PH mPAP > 20, PAWP < 15, and PVR > 3 included members of clinical groups 1, 3, 4, and 5 above.
Monoclonal gammopathy of undetermined signifcance (MGUS) is the premalignant precursor of multiple myeloma. MGUS progresses to multiple myeloma at a rate of approximately 1% per year [14][15][16]. Te prevalence of MGUS increases with age, and a prior study reported a 5.3% prevalence of MGUS among Olmsted County residents 70 years or older [17].
Although there have been series describing PH associated with MM and other plasma cell dyscrasias, to our knowledge, there has been no prior evaluation of MGUS in PH. In this study, we sought to estimate the prevalence of MGUS in patients with echo confrmed diagnosis of PH, and right heart catheterization confrmed cases and assess whether this prevalence is higher than would be expected for individuals of similar age and sex without PH.

Study Population
2.1.1. PH Cases. Te study was approved by the Mayo Clinic Institutional Review Board (IRB Protocol 12004026). A computerized review of the electronic medical records was performed to identify Olmsted County residents seen at Mayo Clinic Rochester between 1/1/1995 and 9/30/2017 with one or more codes consistent with a potential diagnosis of PH (HTCDA codes: 04261110, 04261120, 04261210, 04262410, 04262430, and 04262431; ICD-9 codes: 416.0 and 416.8). Te medical records of the identifed patients were then manually reviewed to ascertain whether or not the patient met diagnostic criteria for PH, and if so, the date of the earliest PH diagnosis. PH was suggested if the right ventricular systolic pressure (RVSP) on transthoracic echocardiography was > 40 mm·Hg. RVSP was calculated using the modifed Bernoulli equation (4v 2 plus estimated right atrial pressure, where v = tricuspid regurgitant peak velocity by continuous wave Doppler) as previously demonstrated [18]. If the patient underwent a hemodynamic right heart catheterization procedure, then PH was defned by a mean pulmonary artery pressure (mPAP) > 25 mm·Hg at rest [1,18,19]. Precapillary PH was defned as an mPAP >25 mm·Hg and a PAWP of < or equal to 15 mm·Hg and a PVR of > or equal to 3. Clinical documentation was reviewed for relevant information regarding the diagnosis of PH as well as additional review of the echocardiographic and heart catheterization data (MAL, ERF). Features such as peak TR velocity, eccentricity index, basal LV/RV ratio RVOT acceleration time and midsystolic notching, early diastolic PR velocity, RA area, and IVC size were assessed in borderline cases.

Controls.
For each PH case, a computerized search of electronic medical records was used to create a pool of potential controls by identifying all Olmsted County residents of the same sex and age who had a medical visit at Mayo Clinic Rochester within 3 months of the diagnosis date of the PH case. From these pools of potential controls, a single control was chosen at random for each PH case (1:1 matching). For each PH case, the index date was defned as the date of their PH diagnosis, and for the corresponding control, the index date was defned as the date of their medical visit that occurred during the same time frame.

Monoclonal Gammopathy of Undetermined Signifcance (MGUS).
In order to identify PH cases and controls who met diagnostic criteria for MGUS, a prospectively maintained Mayo Clinic database was used. We compared the hematology database with our ICD-9 obtained diagnoses and conducted chart reviews to verify accuracy. Serum protein electrophoresis was checked to be present if this test was specifcally drawn for any reason. It was not done routinely on all patients. Tis was true in the pulmonary hypertension group as well as in the control group.

Statistical Analysis.
Te number and percentage of PH cases and controls diagnosed with MGUS are presented overall and according to decades of age. Since MGUS is asymptomatic and routine screening is not performed, patients may have MGUS for years before it is diagnosed. For this reason, several time intervals were used to identify patients with MGUS when assessing the association between PH and MGUS. For the primary analysis, patients were categorized as having MGUS if the MGUS diagnosis occurred prior to or within 30 days after the index date. Secondary analyses were performed with MGUS defned using time intervals that extended 5 and 10 years following the index date and also using any MGUS diagnosis through 12/31/2017 (the fnal date for which MGUS diagnosis information was available). In all cases, analyses were performed using conditional logistic regression taking into account the 1:1 matched case-control study design. Initial analyses were performed which included interaction terms to assess whether the association between MGUS and PH difered according to age or sex. Since a signifcant age-by-MGUS interaction efect was detected, supplemental analyses were performed separately for subgroups defned according to decades of age. In all cases, the results of the logistic regression analyses are summarized using the odds ratio (OR) and corresponding 95% confdence interval with odds ratios >1.0 indicating a positive association between MGUS and PH.

Results
A total of 5,440 Olmsted County residents seen at Mayo Clinic Rochester between 1/1/1995 and 9/30/2017 were identifed as having one or more codes consistent with a potential diagnosis of PH. After a manual review of the electronic medical charts, 3419 of these patients met the echo diagnostic criteria for PH. Of these PH patients, 17 were excluded because the earliest date of PH diagnosis was prior to the start of the study period and 88 were excluded because they were less than 18 years of age at the time of diagnosis. Of the 3314 echo confrmed PH patients (cases), 3313 were agematched and sex-matched to an Olmsted County resident who did not have PH. Tere was one male diagnosed with PH at the age of 106 years that an age-matched control could not be found for, so he was excluded.
Of the 3313 PH cases included in the analysis, 1377 (41.6%) were males and 1936 (58.4%) females. Te mean age (±SD) on the index date was 72.4 ± 14.8 years (range 18 to 104 years). Compared to the controls, a signifcantly higher percentage of the PH cases had a diagnosis of MGUS diagnosis prior to or within 30 days following their index date (6.3% vs. 4.0%, OR � l.62 [95% C.I. 1.29 to 2.03], P < 0.001). Te association between MGUS and PH was not found to difer between males and females (sex-by-MGUS interaction P � 0.728). However, the association between MGUS and PH was found to be dependent upon age (age-by-MGUS interaction, P � 0.047) with a stronger association found in those diagnosed with PH at a younger age (Table 1). Similar fndings were observed when MGUS was defned as any MGUS diagnosis prior to the end of the study period (overall frequency: 10.2% vs. 5.8%, OR � l.84 [95% C 0.l. 1.53 to 2.22], P < 0.001; sex-by-MGUS interaction P � 0.971, age-by-MGUS interaction P � 0.002). For the primary analysis, which looked for a diagnosis of MGUS prior to 30 days following the index date, we performed a sensitivity analysis where we only included cases and controls who had undergone SPEP testing at some point prior to 30 days following their index date. For this analysis, we did fnd that the percentage of individuals who had undergone testing prior to 30 days following their index date was higher for PH cases compared to controls (1588/3313 � 48% for cases versus 1319/3313 � 40% for controls; P < 0.001). However, if the analysis is restricted to only those who were tested, we still fnd a signifcantly higher frequency of MGUS diagnosed among cases compared to controls (OR � 1.5, 95% C.I. (1.19, 1.90), P < 0.001).
Similar methods were used for the heart catheterization data. We cross-referenced our list with patients that had right heart catheterization (Tables 2 and 3). 484 patients were identifed by this means with confrmed PH with an mPAP of >25, and of this group, 63 were found to have MGUS. Te odds ratio was 3.94 (95% CI of 2.28-6.82) with a P < 0.001 (see Table 2). Within this subgroup, 222 patients had precapillary pulmonary hypertension of which 27 were found to have MGUS (Table 3). Subanalysis of this was carried out in a similar fashion looking at MGUS diagnosed within 30 days, 5 years, 10 years or ever diagnosis. Te odds ratio in this group was 4.50 (95% CI (1.86-10.90 with P � 0.001). We again see a stronger association in younger age groups whether in PH or precapillary PH patients.
Tere was a subanalysis of the 36 patients with PH that did not meet the criteria for precapillary pulmonary hypertension which showed the following. Tere were 3 patients with myeloma, 2 with amyloidosis, and 3 with CLL. Tere were 13 females and 23 males. Analysis of the 27 patients with precapillary PH and a positive monoclonal study revealed that 5 had amyloidosis, 3 CLL, 1 myeloma, 1 Hodgkin's, 1 non-Hodgkin's, 2 myelodysplasia, and 1 AIDS. Also of those with a positive MGUS, there were 13 females and 14 males.

Discussion
Pulmonary hypertension is defned as mPAP > 25 mm·Hg at rest, assessed by right heart catheterization [2]. If PH is suspected from clinical presentation and/or physical examination, a screening transthoracic echocardiogram (TTE) is appropriate. Te right ventricular systolic pressure (RVSP) is equal to the pulmonary artery systolic pressure in the absence of right ventricular outfow tract obstruction or pulmonary stenosis. Tis was estimated by utilizing the tricuspid regurgitant peak velocity Doppler signal, including an estimate of the right atrial pressure [18,20,21]. When RVSP is> 40 mm·Hg on the TTE, the remainder of the echocardiographic examination typically includes a detailed assessment of right ventricular size and systolic function, the presence and severity of tricuspid regurgitation, and also a comprehensive evaluation of left-sided cardiac structures. Tis helps identify appropriate patients for hemodynamic right heart catheterization to confrm the PH diagnosis, classify the PH group, stratify risk, and determine candidates for PH-directed therapies [22][23][24]. Other tests are needed to help determine the type of PH, including pulmonary function testing to evaluate for underlying lung disease, ventilation-perfusion lung scintigraphy to assess for chronic thromboembolic PH, and laboratory evaluation. An important component of the diagnostic algorithm for PH is the laboratory evaluation, which includes screening for an underlying connective tissue disorder, chronic liver disease, human immunodefciency virus (HIV), and myeloproliferative disorders [25]. Te WHO classifcation system for PH includes chronic hemolytic anemia and myeloproliferative disorders in Group 5.1 [2]. Te assumption would be that plasma dyscrasias should also be included in this group. Tere is an established association between plasma dyscrasias and PH, particularly POEMS syndrome. Recently, there has been increased awareness of PH in POEMS syndrome, with varying reports of prevalence up to 48% [7,8,26]. It is thought that the increased release of vasoactive cytokines, including VEGF, plays a role in the development of PH that has also been described in multiple myeloma. A recent case report describes two cases of PH in smoldering MM. Right heart Advances in Hematology    1 0 (0.0) 1 0 (0.0) * For the primary analysis, MGUS diagnoses made prior to the index date + 30 days were included. Since screening for MGUS was not performed routinely, sensitivity analyses were performed which used various extended time intervals when identifying patients with a diagnosis of MGUS. Since MGUS diagnoses were only available through 12/31/2017, the duration of the extended intervals was truncated for some patients. † Number of PH cases and controls in the given age category. ‡ Any MGUS diagnosis prior to 12/31/2017. * * Conditional logistic regression taking into account the 1:1 matched set study design. catheterization in both patients showed an rnPAP of > 25 mm·Hg with normal pulmonary arterial wedge pressures (PAWP) < 15 mm·Hg and elevated pulmonary vascular resistance (PVR) > 3 Woods units, indicating the presence of precapillary PH. Autologous stern cell transplantations were performed in both patients, and the pulmonary hypertension resolved after complete remission was achieved [9]. Tere have been other case reports as well describing a reversible form of precapillary PH associated with MM [10,11,27]. A recent retrospective study of 359 patients with MM found that of the 123 patients who underwent echocardiography, 39 (32%) had elevated RVSP suggesting PH [28]. No association was found between the presence of PH and specifc myeloma features, but this study further expanded on the suggestion that PH is prevalent in MM.
Although the link between PH and POEMS syndrome and active multiple myeloma has been established, it is worthwhile to delve further into the association between PH and smoldering multiple myeloma and MGUS. Te International Myeloma Working Group (IMWG) defnes the diagnostic criteria for MGUS as less than 3 g/dL of M-protein in the serum, less than 10% monoclonal plasma cells in the bone marrow, and no CRAB criteria (elevated calcium, renal insufciency, anemia, or lytic bone lesions) [14]. Patients with MGUS can be stratifed into low-risk, intermediate-risk, and high-risk groups, based on the risk of progression to active myeloma, which in general is 1% per year [15]. Smoldering myeloma is the asymptomatic stage between MGUS and active myeloma.
Smoldering myeloma is associated with an M-protein > 3 g/dL and/or > 10% plasma cells in the bone marrow [14]. Given smoldering myeloma and MGUS are not associated with end-organ damage by defnition, management is typically focused on surveillance rather than active treatment.
As mentioned above, there have only been a few case reports describing PH in the setting of smoldering multiple myeloma (9), and to our knowledge, no study has investigated a possible association between MGUS and PH. Terefore, our study is the frst to establish a clear association between MGUS and PH. We found that approximately, 10% (echo data) and 13% (heart catheterization data) of patients with PH had a concomitant diagnosis of MGUS, and there was a statistically signifcant association between PH and MGUS when the cohort was compared with a wellestablished age-matched and sex-matched control group without a diagnosis of PH. Tis study expands on the previous work regarding an association between plasma cell dyscrasias and PH, highlighting a signifcant higher prevalence of MGUS in patients with PH. Ratios of lgG, IgM, IgA, and light chains were similar to values noted previously [17].
Te underlying mechanism of PH in this setting is still unclear, but the pathophysiology could involve a cytokine or humoral process since anecdotally PH has been reversible with myeloma treatment [29,30]. Tis study raises the question of whether the presence of PH in smoldering myeloma and MGUS is a higher risk feature, associated with worse prognosis, and it might indicate a need for closer monitoring or the initiation of earlier treatment of the hematologic condition. Further studies regarding outcomes in patients with PH and plasma cell dyscrasias will be needed. Given that laboratory evaluation is needed when there is a new diagnosis of PH, it seems reasonable to routinely screen for a monoclonal gammopathy with serum protein electrophoresis (SPEP) with immunofxation in all patients in Group 1 (3%) or Group 5 (15%) where the cause of the PH is unknown. Adding this testing, just as adding the other blood testing recommended, add to the cost and may lead to additional studies. Based on the reversible nature of the PH in several case studies of smoldering myeloma and multiple myeloma, a SPEP appears warranted. An SPEP also is needed for case fnding of underlying conditions including POEMS, amyloidosis, lymphoma's and leukemias. Further study assessing the type of monocolonal protein would only apply in those with a 2 gm/dL M-protein.
Costs and the morbidity of bone marrow biopsy, skeletal surveys, and additional workup are the risks associated with this. Of patients with MGUS, this equates to a small percentage of the total which reaches this level, with a reasonable yield to determine underlying diagnoses.
Since there are cases of pulmonary hypertension that have reversible features when treating their monoclonal gammopathy as seen in reference 9, the hope is more cases may be present that can also beneft.
Tis study does have limitations, one being the retrospective nature of the study. Another is detection bias, that is, the more times a patient is seen, the more likelihood testing will occur. We tried to exclude or reduce this by the comparison at the end of the result section, by looking at only those who had testing. An analysis of the patient who had right heart catheterization also confrmed the association found with our echo-related diagnoses. Also, the majority of the PH patients and the control group were Caucasian based on the population in Olmsted County thus limiting the ability to broadly apply the results to other ethnic groups. However, Mayo Clinic is a tertiary/quaternary referral center with broad patient population seeking second and third opinions from across the United States and internationally.

Conclusion
Tis population-based study found a statistically signifcant higher prevalence of MGUS associated with PH. Since previous case series have shown instances of reversible PH with treating smoldering myeloma and multiple myeloma,further prospective study appears warranted since patients that have MGUS have the potential to represent a reversible cause of PH especially those who have precapillary PH. Also, the case fnding of amyloidosis, as well as underlying lymphomas and leukemias, is enhanced by this. We propose that SPEP with immunofxation be included in the standard laboratory evaluation of patients presenting with a new diagnosis of PH that potentially fall within Group1 (3%) and Group 5 (15%) which have an unclear cause at the time 6 Advances in Hematology of presentation. Following a positive SPEP over time especially in patients with precapillary PH is reasonable for the case fnding of those patients that may have a reversible form of PH with the treatment of their underlying hematologic condition.

Data Availability
Te data that support the fndings of this study are available on request from the corresponding author. Te data are not publicly available due to privacy or ethical restrictions.

Ethical Approval
In accordance with the Declaration of Helsinki, this study was reviewed and approved ( # 12-004026) by the Mayo Clinic Institutional Review Board (IRB). Mayo Clinic IRB approved the waiver of informed consent for all study participants prior to study participation. Participants included in this report approved use of their clinical data for the purposes of research (Minnesota authorization for research). All authors assert that all procedures contributing to this work comply with the ethical standards of the Mayo Clinic.

Disclosure
All other authors declare no support from any organization for the submitted work; no fnancial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; and no other relationships or activities that could appear to have infuenced the submitted work.

Conflicts of Interest
Te authors declare that they have no conficts of interest.