**3. Results**

Data from a total of 114 patients were analyzed in this study. Two patients were excluded due to a substantial amount of missing information. Of the included 112 patients, most were diagnosed with Relapsing Remitting MS (RRMS). About 73% of NH-White, 92% of NH-Black, and 95% of Hispanic patients had RRMS, whereas only 18% of NH-White, 5% of NH-Black, and 2.5% of Hispanic patients were diagnosed with Primary Progressive MS (PPMS) (Table 1). One Hispanic patient had a diagnosis of SPMS. There were no significant di fferences among the groups with regard to MS type at diagnosis (*p* = 0.1859), or smoking status (*p* = 0.3079). All groups had a similar female to male ratio, with a greater proportion of female MS patients (Table 1, *p* = 0.3675). Average age at diagnosis (*p* = 0.9918) and mean time to diagnosis (*p* = 0.9934) were also similar across all groups (Table 1). Interestingly, between the groups, we found significant di fferences in the percentage of patients who were adherent or experienced relapse while on escalation or high e fficacy therapies. Specifically, 63.2% of NH-White, 73% of NH-Black, and 61.8% of Hispanic patients were adherent to escalation therapy (Table 1, *p* = 0.0252). 100% of NH-White, 84.2% of NH-Black, and 50% of Hispanic patients were adherent to high e fficacy therapy (Table 1, *p* = 0.0252). 26.3% of NH-White, 31.1% of NH-Black, and 36.4% of Hispanic patients relapsed while on escalation therapy (Table 1, *p* = 0.000151). 0% of NH-White, 10.5% of NH-Black, and 0% of Hispanic patients relapsed while on high e fficacy therapy (Table 1, *p* = 0.00015). Of note, one of the reasons for relapse includes non-adherence; thus interpretation of relapse data must consider the adherence percentages presented.

Notably, only 28% of the NH-Black population had received an evaluation by a neurologist at symptom onset, whereas 53% of Hispanic and 45% of NH-White patients had, although this was not statistically significant (Table 1, *p* = 0.1778). In this cohort, there were no statistically significant di fferences in receipt of a medical evaluation at symptom onset; 63–70% of patients from all groups were able to access medical evaluation. Additionally, NH-White, NH-Black and Hispanic patients exhibited no di fferences in symptoms at diagnosis or mean EDSS score at diagnosis and last encounter (Table 2). There was a significant di fference in the percentage of patients with severe disability (EDSS score > 4.5) at diagnosis and at last encounter; 14.3% of NH-White MS patients had severe disability at diagnosis compared to 50% of NH-Black and 31.6% of Hispanic patients (Table 2, *p* < 0.001). This was also true at last encounter with 32.5% of NH-White, 45.5% of NH-Black and 41% of Hispanic MS patients with severe disability at their most recent clinical visit (Table 2, *p* < 0.001).

Assessment of degree of brain atrophy and progression over time revealed that NH-White, NH-Black and Hispanic patients in this cohort had a similar degree of brain atrophy at diagnosis and over time (Figure 1). Enumeration of T1, T2, and gadolinium-enhancing brain lesions at diagnosis also showed no significant di fferences between the groups (data not shown). Spinal atrophy and quantity of T2 and gadolinium-enhancing lesions in the spine at diagnosis and at last presentation were also similar between groups (Figure 2).


**Table 1.** Diagnosis Characteristics of patients with MS stratified by race/ethnicity.

EDSS score at diagnosis and EDSS score at last clinical visit were compared within each group. Standard deviation is shown in parentheses. *p* = 0.4253 (NH-Black), *p* = 0.1757 (Hispanic), *p* = 0.0324 (NH-White), (paired sample *t*-test). For adherence and relapse data, chi-squared test and Fisher's-exact test were used respectively. Escalation therapies included Glatiramer Acetate, Interferons, Teriflunomide, Dimethyl Fumarate and Fingolimod. High efficacy therapies included Rituximab, Ocrelizumab, Alemtuzumab and Natalizumab.

> **Table 2.** Clinical characteristics of MS patients by race/ethnicity.


**Figure 1.** Degree of brain atrophy at diagnosis and worsening of brain atrophy from diagnosis to most recent MRI scan. Presented as the total percentage of each group, is the proportion of patients who had none (**A**), mild (**B**), or moderate (**C**) brain atrophy at the time of diagnosis, as well as the proportion of patients who had increased brain atrophy in their most recent MRI scan compared to diagnosis (**D**). Only patients who had MRI scans on file were included in this analysis. *p* = 0.5155 for comparison between degree of brain atrophy (none, mild, moderate) (Fisher's exact). *p* = 0.3387 for comparison of total percentage of patients who had worsening brain atrophy on most recent MRI compared to diagnosis (Fisher's exact).

**Figure 2.** The total percentage of patients in each group who had spinal atrophy, T2, or gadoliniumenhancing lesions in the spine as determined by MRI findings at diagnosis. Only patients who had MRI scans on file were included in this analysis. *p* = 0.6974, *p* = 0.5128, *p* = 0.2957 for comparison of total percentage of patients in each group that had spinal atrophy, spinal T2 lesions, and spinal gadolinium-enhancing lesions respectively (Fisher's exact).

Patient usage of escalation or high efficacy therapies did not significantly impact the patient's likelihood of having an EDSS score > 4.5 at last clinical encounter after adjustment for adherence, smoking, race, age, prior exposure to escalation therapies, and EDSS at diagnosis (Table 3). Active smokers were 2.44 times as likely to have an EDSS score > 4.5 at their last clinical encounter compared to non-smokers after adjustment for age and race (OR: 2.44, 95% CI: 1.36–6.12, *p* = 0.01) (Table 3). Interestingly, after adjustment for race and age, patients who were evaluated by a neurologist at diagnosis had significantly lower adjusted odds of an EDSS score > 4.5 at last presentation compared to patients who were not evaluated by a neurologist (OR: 0.40, 95% CI: 0.16–0.90, *p* = 0.04) (Table 3).


**Table 3.** Association between various patient characteristics and high EDSS score (>4.5) at last presentation.

a adjusted for adherence, smoking, race and age and EDSS at diagnosis; b adjusted for prior exposure to escalation therapies, adherence, smoking, race, age and EDSS at diagnosis; c adjusted for age and race.

Active smokers were 2.79 times as likely to have an EDSS score > 4.5 at diagnosis compared to non-smokers after adjustment for age and race (OR: 2.79, 95% CI: 1.10–7.10, *p* = 0.01) (Table 4). There was no significant association between time to diagnosis and having a high EDSS score at diagnosis (Table 4). There were no significant differences in total relapse occurrence for patients on escalation therapy vs. high efficacy therapy for each racial/ethnic group (data not shown). Of 24 NH-white patients, 19 had ever used escalation therapy, and 5 had used high efficacy therapy. Of 93 NH-Black patients, 74 had used escalation therapy, and 19 had used high efficacy therapy. For the Hispanic patients group of 63 patients, 55 had ever used escalation therapy while 18 had documented high efficacy therapy use. We found no differences between the groups concerning the usage of escalation vs. high efficacy therapies.

**Table 4.** Association between various patient characteristics and high EDSS score at diagnosis (>4.5).


adjusted for age and race.

### **4. Discussion**

The goal of this retrospective cohort study was to describe MS patient characteristics in a multi-ethnic population in Houston and compare findings between racial/ethnic groups. Our study demonstrates several racial/ethnic similarities and a few di fferences in multiple sclerosis presentation and disease course. We found that the groups had a similar mean age at diagnosis, mean EDSS score at diagnosis and last presentation, and a similar degree of brain and spinal atrophy at diagnosis as well. MRI spinal findings were also comparable between NH-White, Black and Hispanic groups. The average time from symptom onset to diagnosis, and overall symptom presentation, were also similar between the groups. The clinic that these patients were treated at is a hub for the underserved and low socioeconomic communities. Thus, we suspect that many of the patients in this cohort were of a similar socioeconomic background, which undoubtedly can influence disease manifestation and outcomes. It is plausible that these similar environmental factors, along with the small sample size may explain the many similarities detected between the groups. However, further studies are required to evaluate this hypothesis.

Interestingly, after adjustment for race and age, patients who were evaluated by a neurologist at diagnosis had 60% lower odds (OR = 0.40, 95% CI: 0.16–0.90) of an EDSS score > 4.5 at last presentation compared to patients who were evaluated by a non-neurology specialist. This suggests a logical protective e ffect of treatment by a neurologist at symptom onset and highlights the importance of access to treatment for all patients. Indeed, a national descriptive study found that people with MS who saw a neurologist were more likely to receive appropriate DMT treatment and see rehabilitation and urologist specialists compared to people who saw other providers [27]. A 2017 study on racial disparities in neurologic health care access revealed that Black patients were 30% less likely to see an outpatient neurologist and were more likely to be cared for in the emergency department compared to their White counterparts [23]. Similarly, Hispanic patients were 40% less likely to see an outpatient neurologist compared to NH-Whites [23].

We found that actively smoking patients were 2.44 times as likely (95% CI: 1.36–6.12) to have severe disability at diagnosis and at the last clinic follow up. A recent systematic review and meta-analysis found evidence supporting the causal involvement of smoking in the development and progression of MS [28]. Altogether, these data sugges<sup>t</sup> that smoking prevention and cessation education programs and early intervention by a neurologist should be implemented to achieve optimal MS care in diverse patient populations.

Consistent with published reports, a greater proportion of NH-Black patients had early severe disability (defined in our study as an estimated EDSS score > 4.5) when compared to NH-White and Hispanic patients [29,30]. In our present study, treatment modality did not impact the risk of having an estimated EDSS score > 4.5 at the last visit. Nonetheless, we observed a trend towards a higher relapse rate in escalation therapies vs. high e fficacy therapies, especially in NH-Blacks. We also observed significant di fferences in adherence between the groups. Interestingly, a greater percentage of NH-Black patients relapsed while on high e fficacy therapy compared to Hispanic patients, despite having greater adherence. Other studies have found that NH-Black patients treated with interferons experienced more relapses and new MS lesions on T2-weighted brain magnetic imaging than NH-Whites [31]. However, further studies on the interaction between race/ethnicity and DMT response for MS are necessary.

Several studies have shown that African Americans have significantly higher CNS lesion burden, more frequent relapses, worse ambulatory disability, worse post-relapse recoveries, and higher overall disability at diagnosis [5,10,19,29]. Overall, our findings did not confirm these prior observations and we believe that the similar socioeconomic background of this patient cohort, along with the small sample size, may have contributed to this. Nevertheless, it is evident that further studies are needed to investigate the various environmental and social factors contributing to divergent MS clinical course outcomes between diverse populations.

Limitations of this study include its retrospective nature, the variable periods of follow-up and the selection of therapy by the treating physician (nonrandomized). The study was also constrained by a small sample size, which could have induced a type 2 error leading to the inability to reject the null hypothesis in some of our comparisons. Additionally, our interpretation of the relapse data is limited because one of the possible reasons for relapse is non-adherence. Thus, relapse data are not corrected for the degree of non-adherence and should be assessed accordingly. Lastly, it is important to note that we did not analyze the imaging data ourselves. Instead, we collected information from MRI reports. Often, the number of lesions was documented as a range, thereby limiting data precision. Moreover, the atrophy measurements were subjective rather than objective quantification, and some patients were missing MRI information at diagnosis (e.g., performed at a di fferent institution). These limitations may have impacted the capability to show radiological di fferences at presentation between groups.

Our study is important because it adds to emerging literature describing disease characteristics in minority populations with MS. The disparities in MS progression, onset, and disease course warrants further study. Of 60,000 published articles on MS, only 113 focused on NH-Black and only 23 focused on Hispanic American patients with MS as of 2014 [10]. This demonstrates a need for studies that are intentionally inclusive of these populations. Since 2014, there has been a modest but steady increase in studies focused on these populations. There is a clear disparity in MS treatment access for patients from di fferent racial and ethnic backgrounds. Drivers of disparity are often comprised of complex interactions among factors such as socioeconomic status, access to healthcare and wellness resources (clinics, hospitals, grocery stores, fitness centers), systemic racism and biases in healthcare, and limited health literacy. This systemic web of disparity can be challenging to disentangle, but understanding it is necessary for improving the care of minority patients with MS.

Future prospective randomized controlled trials in di fferent racial/ethnic groups with MS are essential to better understand the disease progression, managemen<sup>t</sup> and treatment outcomes for diverse patient populations.

**Author Contributions:** Conceptualization, F.X.C. and G.J.H.; methodology D.D.; software, D.D.; validation, F.X.C. and V.M.; formal analysis, D.D.; writing—original draft preparation, V.M., D.D., K.F. and F.X.C.; writing—review and editing, H.M.S and G.J.H.; supervision, H.M.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by U.S. Department of Health and Human Services and Health Resources and Services Administration, gran<sup>t</sup> number D34HP31024.

**Conflicts of Interest:** The authors declare no conflict of interest.
