Sage Journals: Discover world-class research

Abstract

Background:

Neurological care and disease-modifying therapies (DMTs) are central to multiple sclerosis (MS) management, but how comorbidities like cancer affect them remains unclear.

Objective:

We assessed the impact of cancer on MS-related healthcare and DMT use.

Methods:

This cohort study accessed population-based data from France (2009–2021) and British Columbia, Canada (1991–2020). Cases were individuals with MS diagnosed with an incident cancer. The 4-year study period encompassed the 2 years before and after the cancer diagnosis. Each case was matched to two cancer-free individuals with MS (i.e. controls). Outcomes were neurologist visit rates, MS hospitalization rates, and percentage of DMT users. Mixed-effects models were used with a knot at cancer diagnosis. Incidence rate ratios (IRRs) were estimated pre- and post-cancer (cases vs. controls) and compared using a slope ratio (SR).

Results:

In total, 6902 cases were matched to 13,804 controls. There was a negative effect of cancer on DMT use (SR = 0.36 [0.19–0.70]), especially among those on chemotherapy (SR = 0.24 [0.14–0.40]). Neurologist visit IRRs did not differ pre- and post-cancer (SR = 0.96 [0.91–1.005]). Similarly, MS hospitalization IRRs did not differ (SR = 1.06 [0.84–1.33]).

Conclusion:

Following cancer, DMT use declined sharply. However, there was no evidence of a strong impact of cancer on neurologist visits and MS hospitalizations.

Keywords

Multiple sclerosis cancer disease-modifying therapies epidemiology healthcare use

Introduction

People with multiple sclerosis (pwMS) are living longer, although their life expectancy remains 5–10 years shorter than the general population.^1
–3 In France and North America, the average age of pwMS has increased, and the majority are aged 50 years or older.^4
–7 Consequently, the prevalence of age-related comorbidity is increasing in this population, particularly cancer. However, how this dual diagnosis is being navigated by health professionals and patients and its impact on MS-related care remains to be determined.

Several studies have examined cancer risk in MS, but little is known about cancer effect on MS-related care.^8
–13 A mixed-methods study looked at the clinical course of MS after a breast cancer diagnosis and found that about half the patients who were on a disease-modifying therapy (DMT) at cancer diagnosis discontinued or changed their DMT. It also reported that about half of the cohort continued a stable MS management routine, whereas 25% switched focus on the cancer, putting MS care aside.¹⁴ However, this study was limited to 43 participants, focused only on one cancer type, and did not include a comparison with a representative cancer-free MS population.

Another smaller study on 16 individuals with MS and central nervous system (CNS) cancer found reduced inflammatory activity in patients undergoing chemotherapy compared with those who did not.¹⁵ In some older studies from the 1990s, radiotherapy for CNS tumors using radiation techniques not typically used at present was linked to MS exacerbations; however, this may not be comparable to conformal or stereotactic radiotherapy techniques.^16,17

Studies on comorbidities in MS have focused on cardiovascular, autoimmune, and mental diseases and failed to include cancer.^18
–21 Moreover, narrative reviews have shown that there is no unified practice regarding MS management after a cancer diagnosis and is highly dependent on the practitioner’s judgment and experience as well as the type, severity, and treatment of cancer.^22
–24 The French MS society recently released guidelines on the topic, highlighting that evidence on MS care pathways after cancer diagnosis remained sparse.²⁴ Research is needed to accurately capture the impact of cancer on MS-related care in order to identify current practices and potential disparities in access to care.

Here, we assessed the impact of cancer on MS-related care and DMT use in a multinational matched population-based retrospective cohort study using administrative data.

Materials and methods

Data source

Population-based health administrative data from France (SNDS) and the province of British Columbia (BC) in Canada were used. Both study regions (total = 73 million residents) retain electronic records of sociodemographic data and health service use for nearly all residents (~99%). Information on dates of birth and death, sex, and place of residence is available, as well as on hospitalizations, prescriptions filled, and physician visits.²⁵

In both regions, each hospitalization is associated with an admission and discharge date, one principal discharge diagnosis code, and multiple associated diagnosis codes from the International Classification of Diseases (ICD; France = ICD-10; BC = ICD-9 or 10). Each physician visit is associated with a visit date and the physician’s specialty. Only in BC, physician visits are associated with up to five ICD-9 diagnoses. In addition, in France, all medical procedures (e.g. chemotherapy or radiotherapy) are recorded. Moreover, patients with a chronic disease that requires either long or expensive treatments can obtain the long-term disease (LTD) status (from a specific list, including MS and cancer).

MS participants

PwMS were identified through a healthcare use-based algorithm in both regions. In BC, MS was defined as ⩾3 MS ICD codes (ICD-10:G35/ICD-9:340 in hospitalizations or physician visits) or ⩾1 filled prescription of an MS-specific DMT (Supplemental Table 1). After identification, the MS diagnosis date was defined as the date of the first MS or demyelinating-related diagnostic code or the first MS-specific DMT prescription filled (Supplemental Tables 1 and 2).

In France, as records do not include diagnosis codes of physicians’ visits, pwMS were identified using an algorithm developed in collaboration with neurologists and medico-administrative database experts, that had been extensively used in previous studies.^4,26,27 MS was identified through MS hospitalizations, MS-specific drug reimbursements, or MS-LTD status.⁴

Study design and population

This multinational cohort study investigated the impact of cancer on MS-related healthcare use. The study population comprised two population-based cohorts of pwMS. Data were available: 1 January 1991 to 31 March 2020 in BC and 1 January 2009 to 31 December 2021 in France.

Cases were defined as ⩾18 year-old pwMS with an incident cancer. The index date was defined as the first cancer-related event (Supplemental Tables 3 and 4). A cancer was considered incident if there was no evidence of cancer in the 3 years pre-index date. Each individual’s follow-up period included the 2 years before and after the first cancer-related event (index date) (Figure 1). Therefore, cancer cases were included from 1 January 1994 to 31 March 2018 (BC) and 1 January 2012 to 31 December 2019 (France). For both cases and controls, MS onset had to be at ⩾2 years pre-index. Each case was matched to two randomly selected controls from the MS population without replacement. Controls were assigned the same index date and study period as their matched cancer case and were without evidence of cancer pre-index date of their case and in the 2 years following. Matching criteria were calendar year, sex, birth year (±3 years), place of residence (France = department/ecological deprivation quintile;²⁸ BC = first three digits of the postal code), MS duration (defined as the time between the first recorded MS event and the index date), and DMT use in the 12–24 months pre-index (France = none; ⩾ 1 moderate efficacy, no high efficacy; ⩾ 1 high efficacy; BC = none, ⩾ 1 moderate or high efficacy). In addition, DMT use was studied according to each region’s prescription practices. During the study period, off-label drugs (i.e. DMT without formal regulatory approval to treat MS) could be identified as part of MS management in France but not in BC. Therefore, off-label drugs were only included in the French matching/subgroup analysis. In the absence of an exact match, residence criteria were gradually relaxed to become less stringent (Supplemental Figure 1).

Figure 1.

Study design.

Cancer

In BC, cancer cases were identified using a validated definition (⩾1 cancer-related hospitalization or physician visit claims).²⁹ Although this previous study focused on the most common cancer types (breast, lung, prostate, and colorectal), this definition was applied to all cancer types. In France, cancer cases were identified using a previously used definition developed with the French Cancer Institute and based on hospital records, chemotherapy/radiotherapy use, and LTD status.^12,30 Information on cancer treatment was only available in France.

Outcomes

MS-related healthcare use was determined through three outcomes, using administrative data: (1) neurologist visits, (2) MS hospitalizations, and (3) DMT use. Overnight MS hospitalizations had a principal diagnosis code of MS (ICD-10:G35/ICD-9:340). DMT use was defined as filled prescriptions of (1) MS-specific DMTs (Supplemental Table 1) and (2) non-specific/off-label drugs used to treat MS (Supplemental Table 2).

Current treatment for MS was based on each DMT and off-label drug used during the study period grouped as: high efficacy (alemtuzumab, fingolimod, natalizumab, ocrelizumab, cladribine, siponimod), moderate efficacy (daclizumab, dimethyl fumarate, glatiramer acetate, interferon beta, teriflunomide) or off-label/non-specific to MS. For all treatments, except anti-CD20 drugs (i.e. ocrelizumab, rituximab), participants were considered treated for 90 days, and discontinuation occurred if there was no dispensation for ⩾ 90 consecutive days, plus a 30-day grace period. For anti-CD20, participants were considered treated for 12 months and for 6 months, respectively. Similarly, there was a 30-day grace period. Each of these DMT and off-label drugs was considered discontinued if no further prescription was filled. There were no cladribine or alemtuzumab prescription fills in the study population, so we did not include additional specific treatment window considerations.

Statistical analysis

For each follow-up year, neurologist visits and hospitalizations were assessed as yearly rates and as percentages of individuals with at least one encounter during the year. Periods of hospitalization were discounted when computing individual neurologist visit rates because physician visits were not consistently recorded during hospitalizations. DMT use was described as a percentage of individuals on treatment at each time point.

Mixed-effect models were used to assess the impact of cancer on MS management in cases compared with controls, and time was modeled using a piecewise linear spline with a knot at the time of cancer diagnosis. The inclusion of the knot allowed fitting two slopes (pre- and post-cancer) that could then be compared through a ratio of the estimates, a “slope ratio” (SR). An SR < 1 indicates that the pre-cancer estimate was higher than the post-cancer estimate. The type of mixed-effect model used differed depending on the outcome structure: neurologist visits and MS hospitalizations were modeled using a Poisson regression yielding incidence rate ratios (IRRs) and DMT use using a logistic regression yielding odds ratios (ORs).

The three following additional analyses were carried out in the French cohort only. DMT use models were stratified by cancer treatment type (≥1 chemotherapy sessions, ≥1 radiotherapy sessions without any chemotherapy, cancer surgery only, other). Neurologist visits and MS-related hospitalization models were stratified by type of DMT used in the 24–12 months pre-index. To explore a potential period effect, all three outcomes were stratified into cancer diagnosis years (2012–2017 vs. 2018–2019).

As per data access and privacy agreements, analyses were performed separately in each region, with aggregated results pooled. The I² statistic was computed; if ⩽ 50%, this suggested no evidence of regional heterogeneity, and estimates were pooled by meta-analysis. Otherwise, more conservative pooling estimates were derived from mixed models. Analyses were performed using SAS (v.8.3.7) and R (v.4.1.2).

Results

A total of 20,706 pwMS were included 6902 MS cases (France = 4555; BC = 2347) matched 1:2 to 13,804 MS controls (France = 9110; BC = 4694) (Figure 2). Their characteristics at the index date are described in Table 1. Most were female (France 71.6%; BC 75.9%) and mean age at index = 58.6 years (SD = 12.5) in France and 55.9 years (SD = 11.6) in BC. The most common cancer types were breast (females only, France = 36.4%; BC = 23.9%), prostate (males only, France = 26.1%; BC = 19.8%), and colorectal cancer (France = 10.3%; BC = 10.9%).

Figure 2.

Population flowchart.

Table 1.

Characteristics of patients living with multiple sclerosis with (cases) and without (controls) a cancer diagnosis at the index date (n = 20,706).

	France		British Columbia
	Cases	Controls	Cases	Controls
Characteristic, n (%)	n = 4555	n = 9110	n = 2347	n = 4694
Sex
Male	1292 (28.4)	2584 (28.4)	566 (24.1)	1132 (24.1)
Female	3263 (71.6)	6526 (71.6)	1781 (75.9)	3562 (75.9)
Age (in years) at the index date
Mean (SD)	58.6 (12.5)	58.5 (12.4)	55.9 (11.6)	55.6 (11.5)
18-40	382 (8.4)	767 (8.4)	237 (10.1)	473 (10.1)
41-50	771 (16.9)	1524 (16.7)	509 (21.7)	1030 (21.9)
51-60	1225 (26.9)	2526 (27.7)	734 (31.3)	1544 (32.9)
61-70	1440 (31.6)	2890 (31.7)	637 (27.1)	1219 (26.0)
71-80	598 (13.1)	1140 (12.5)	206 (8.8)	378 (8.1)
⩾81	139 (3.1)	263 (2.9)	24 (1.0)	50 (1.1)
Cancer diagnosis year
1994–2011	N/A	N/A	1343 (57.2)	2686 (57.2)
2012–2018	3846 (84.4)	7692 (84.4)	1004 (42.8)	2008 (42.8)
2019	709 (15.5)	1418 (15.5)	N/A	N/A
MS duration, in years
2–5	673 (14.8)	1346 (14.8)	555 (23.6)	1110 (23.6)
6–10	1048 (23.0)	2096 (23.0)	647 (27.6)	1294 (27.6)
11–15	925 (20.3)	1850 (20.3)	564 (24.0)	1128 (24.0)
16–20	817 (17.9)	1634 (17.9)	362 (15.4)	724 (15.4)
21+	1092 (24.0)	2184 (24.0)	219 (9.3)	438 (9.3)
MS-related care consumption
DMT receipt (12–24 months before the index date)
No	2809 (61.7)	5618 (61.7)	2121 (90.4)	4242 (90.4)
Yes	1746 (38.3)	3492 (38.3)	226 (9.6)	452 (9.6)
Cancer characteristics
Tumor type
Breast^a	1187 (36.4)		426 (23.9)
Prostate^b	337 (26.1)		112 (19.8)
Colorectal	468 (10.3)		256 (10.9)
Cervical^a	270 (7.5)		43 (2.4)
Skin	154 (3.4)		227 (9.7)
Bladder	246 (5.4)		135 (5.8)
Lung	197 (4.3)		103 (4.4)
Kidney	152 (3.3)		24 (1.0)
Brain and CNS	88 (1.9)		47 (2.0)
Other	1456 (32.0)		974 (41.5)
Chemotherapy or radiotherapy	2790 (61.3)		N/A
Chemotherapy	1484 (32.6)
Radiotherapy	2114 (46.4)
Cancer surgery	2490 (54.6)
Palliative care	62 (1.4)
Immune checkpoint inhibitors	32 (0.7)

CNS, central nervous system; DMT, disease-modifying therapy; MS, multiple sclerosis; N/A, not available; SD, standard deviation.

Denominator is the female population.

Denominator is the male population.

Neurologist visits

Overall, the percentage of pwMS who saw a neurologist was higher among cases than controls (France 75.1% vs. 72.7%; BC 72.0% vs. 68.7%) with an overall decrease in both groups over time (Table 2). Before and after cancer, the neurologist visit rate was slightly higher for cases than controls, but the post-cancer IRR was only modestly significant (IRR; pre-cancer = 1.08 [1.04–1.13], p = 0.001; post-cancer = 1.04 [1.000–1.09], p = 0.05) (Table 3). When comparing pre- and post-cancer IRRs, they did not differ (SR = 0.96 [0.91–1.005], p = 0.08). Results were similar across regions and age groups (Figure 3 and Supplemental Figure 2). In addition, post- vs. pre-SR were comparable across time periods (2018–2019 = 0.97 [0.88–1.07]; 2012–2017 = 0.94 [0.89–1.00], Figure 4, Supplemental Table 5).

Table 2.

Yearly multiple sclerosis-related care consumption.

	(Year − 2)		(Year − 1)		(Year + 1)		(Year + 2)		Whole period
	24–12 months before the index date		12–0 months before the index date		0–12 months after the index date		12–24 months after the index date		Whole period
	Cases	Controls	Cases	Controls	Cases	Controls	Cases	Controls	Cases	Controls
France
Neurologist visit
At least one, n (%)	2510 (55.1)	5058 (55.5)	2650 (58.2)	5032 (55.2)	2498 (54.8)	5043 (55.4)	2441 (53.6)	4945 (54.3)	3422 (75.1)	6622 (72.7)
Yearly rate, mean [SD]	1.32 [1.93]	1.26 [1.80]	1.38 [1.99]	1.21 [1.64]	1.22 [1.75]	1.16 [1.58]	1.16 [1.76]	1.13 [1.52]	1.27 [1.63]	1.19 [1.41]
Hospitalization specific to MS
At least one, n (%)	152 (3.3)	300 (3.3)	185 (4.1)	248 (2.7)	119 (2.6)	244 (2.7)	98 (2.2)	210 (2.3)	434 (9.5)	797 (8.7)
Yearly rate, mean [SD]	0.04 [0.26]	0.04 [0.25]	0.05 [0.31]	0.03 [0.23]	0.03 [0.23]	0.03 [0.21]	0.03 [0.22]	0.03 [0.18]	0.04 [0.17]	0.03 [0.14]
Length of stay, median [IQR]	5 [3-8]	4 [3-8]	6 [3-12]	4 [3-8]	6 [3-13]	5 [3-10]	5 [3-11]	5 [3-9]	6 [3-14]	6 [3-11]
MS-specific DMT use
At least one, n (%)	1391 (30.5)	2896 (31.8)	1399 (30.7)	2880 (31.6)	1208 (26.5)	2854 (31.3)	1169 (25.7)	2820 (31.0)	1595 (35.0)	3323 (36.5)
High efficacy	328 (7.2)	618 (6.8)	403 (8.8)	707 (7.8)	357 (7.8)	776 (8.5)	358 (7.9)	849 (9.3)	522 (11.5)	987 (10.8)
Moderate efficacy	1104 (24.2)	2339 (25.7)	1037 (22.8)	2252 (24.7)	884 (19.4)	2138 (23.5)	835 (18.3)	2032 (22.3)	1265 (27.8)	2670 (29.3)
British Columbia
Neurologist visit
At least one, n (%)	1058 (45.1)	2047 (43.6)	1053 (44.9)	1971 (42.0)	1001 (42.7)	1986 (42.3)	969 (41.3)	1833 (39.0)	1690 (72.0)	3224 (68.7)
Yearly rate, mean [SD]	0.78 [1.2]	0.75 [1.2]	0.76 [1.2]	0.66 [1.0]	0.71 [1.1]	0.66 [1.0]	0.67 [1.1]	0.63 [1.1]	0.73 [1.0]	0.68 [0.9]
Hospitalization specific to MS
At least one, n (%)	32 (1.4)	64 (1.4)	40 (1.7)	69 (1.5)	27 (1.2)	62 (1.3)	36 (1.5)	69 (1.5)	143 (6.1)	245 (5.2)
Yearly rate, mean [SD]	0.02 [0.2]	0.02 [0.1]	0.02 [0.1]	0.02 [0.2]	0.01 [0.1]	0.02 [0.2]	0.02 [0.1]	0.02 [0.1]	0.02 [0.09]	0.02 [0.08]
Length of stay, median [IQR]	11 [6-26]	12 [4-38.5]	28.5 [5-118]	12 [4.75-30]	8.5 [4-21.75]	9 [3-30]	8 [2-20]	6 [3-31]	11 [4-42]	9 [3-32.5]
MS-specific DMT use
At least one, n (%)	232 (9.9)	443 (9.4)	234 (10.0)	454 (9.7)	213 (9.1)	442 (9.4)	216 (9.2)	445 (9.5)	322 (13.7)	624 (13.3)
High efficacy	17 (0.7)	31 (0.7)	19 (0.8)	44 (0.9)	20 (0.9)	48 (1.0)	25 (1.1)	60 (1.3)	37 (1.6)	82 (1.8)
Moderate efficacy	219 (9.3)	414 (8.8)	218 (9.3)	413 (8.8)	197 (8.4)	400 (8.5)	193 (8.2)	390 (8.3)	311 (13.3)	583 (12.4)

All time periods are left-close, right-open intervals. IQR, interquartile range; DMT, disease-modifying therapy; MS, multiple sclerosis; SD, standard deviation.

Table 3.

Mixed-effect model estimates to assess the impact of the cancer diagnosis on multiple sclerosis-related care by region and pooled.

	Cases vs. controls	British Columbia			France			I ²	Pooled
		IRR	95% CI	p-value	IRR	95% CI	p-value		IRR	95% CI	p-value
Neurologist visits	Pre-cancer	1.09	1.02-1.18	0.02	1.08	1.03-1.14	0.002	0%	1.08	1.04-1.13	< 0.001
	Post-cancer	1.07	0.99-1.15	0.08	1.03	0.98-1.08	n.s.	0%	1.04	1.000-1.09	0.05
	Post- vs. pre-ratio	0.98	0.91-1.06	n.s.	0.95	0.90-1.001	0.06	0%	0.96	0.91-1.005	0.08
MS hospitalizations	Pre-cancer	1.24	0.65-2.35	n.s.	1.12	0.87-1.43	n.s.	0%	1.13	0.90-1.43	n.s.
	Post-cancer	1.49	0.80-2.76	n.s.	1.10	0.86-1.42	n.s.	0%	1.15	0.91-1.45	n.s.
	Post- vs. pre-ratio	1.20	0.74-1.58	n.s.	0.99	0.74-1.31	n.s.	0%	1.06	0.84-1.33	n.s.
		OR	95% CI	p-value	OR	95% CI	p-value		OR	95% CI	p-value
MS-specific DMT use	Pre-cancer	1.30	0.79-2.16	n.s.	1.29	0.97-1.72	0.08	0%	1.29	1.01-1.66	0.04
	Post-cancer	0.70	0.42-1.18	n.s.	0.35	0.26-0.47	< 0.001	80%	0.48	0.25-0.94	0.03
	Post- vs. pre-ratio	0.54	0.28-1.03	0.06	0.27	0.20-0.38	< 0.001	71%	0.36	0.19-0.70	0.002
MS-specific DMT use and off-label drugs	Pre-cancer	1.38	0.88-2.16	n.s.	1.34	1.05-1.72	0.020	0%	1.35	1.09-1.67	0.007
MS-specific DMT use and off-label drugs	Post-cancer	0.94	0.60-1.49	n.s.	0.44	0.35-0.57	< 0.001	88%	0.63	0.30-1.31	n.s.
	Post- vs. pre-ratio	0.69	0.39-1.20	n.s.	0.33	0.25-0.44	< 0.001	81%	0.46	0.22-0.94	0.03

CI, confidence interval; MS, multiple sclerosis; IRR, incidence rate ratio; OR, odds ratio; DMT, disease-modifying therapy.

All p-values > 0.1 were marked as non-significant (n.s).

Figure 3.

MS-related outcomes over time, cases vs. controls. (a) Yearly neurologist visit rates. (b) Yearly MS hospitalization rates. (c) Percentages of patients on MS-specific DMT.

Figure 4.

MS-related outcomes over time, cases vs. controls, according to time period in France (n = 13,665). (a) Yearly neurologist visit rates. (b) Yearly MS hospitalization rates. (c) Percentages of patients on MS-specific DMT.

By DMT use

In the French cohort, those receiving DMT had a more pronounced neurological follow-up overall. Over the study period, >90% of PwMS in the high- and moderate-efficacy DMT groups saw a neurologist but only 62% in the untreated group (Supplemental Table 6). Those using a high-efficacy DMT were younger, had the highest mean annual neurologist visit rate (Supplemental Table 6), and were more frequently on chemotherapy (Supplemental Table 7). Among those receiving high- and moderate-efficacy DMT, cancer did not have any impact on neurologist visit rates (SR = 0.98 [0.83–1.16] and 0.98 [0.93–1.03], respectively). Conversely, among those not receiving DMT, the neurologist visit rate was higher among cases than controls only before cancer diagnosis (IRR, pre-cancer = 1.13 [1.04–1.22]; post-cancer = 1.03 [0.95–1.11]) which resulted in a significant post- vs. pre-cancer SR = 0.91 [0.84–0.99].

MS hospitalizations

During the study period, there were few MS hospitalizations in France (cases = 434, 9.5%; controls = 797, 8.7%) and BC (cases = 143, 5.9%; controls = 245, 5.1%, Table 2). In both regions, the pre- and post-cancer MS hospitalization rates did not differ between cases and controls (IRR, pre-cancer = 1.13 [0.90–1.43]; post-cancer = 1.15 [0.91–1.45]) (Table 3). The pre- and post-cancer IRRs also did not differ, showing no impact of cancer on MS hospitalization rates (SR = 1.06 [0.84–1.33]). In the French cohort, similar results showing no difference were found when stratifying by DMT use (Supplemental Figure 3). Those findings were consistent across time periods, showing no difference (2018–2019 = 1.21 [0.62–2.33]; 2012–2017 = 0.92 [0.67–1.28], Figure 4, Supplemental Table 5).

DMT use

Over the study period, fewer individuals received MS-specific DMT in BC than in France (13.4% vs. 36.0%), especially high-efficacy DMT (1.7% vs. 11.0%). After cancer, fewer cases were on DMT than controls (post-cancer OR = 0.48 [0.25–0.94]), specifically in the French cohort (post-cancer OR = 0.35 [0.26–0.47]) (Table 3). The comparison of the pre- and post-cancer estimates showed that cancer had a negative effect on DMT use (SR = 0.36 [0.19–0.70]). This effect, albeit less strong, remained when including off-label drugs (SR = 0.46 [0.22–0.94]). This decrease in DMT use was observed across all age groups (Supplemental Figure 4) as well as across all time periods (2018–2019 = 0.16 [0.08–0.34]; 2012–2017 = 0.33 [0.22–0.47], Figure 4, Supplemental Table 5). In the French cohort, post hoc analyses by selected individual DMTs, including S1P-modulators and anti-CD20, are shown in Supplemental Figure 5.^31,32

By cancer treatment

In the French cohort, cases who received chemotherapy were on average younger and more likely to be on DMT (Supplemental Table 8). In all cancer treatment groups, the percentage of DMT users decreased after the index date; however, the extent and statistical significance varied across groups (Figure 5). The chemotherapy group experienced the most significant decrease in DMT use (SR = 0.24 [0.14–0.40]). DMT use also decreased in the radiotherapy alone (SR = 0.36 [0.19–0.67]) and surgical intervention-only groups (SR = 0.38 [0.18–0.78]). Findings remain virtually unchanged in the post hoc descriptive analysis excluding off-label/non-specific DMTs (including rituximab) from the analysis (Supplemental Figure 6).

Figure 5.

Percentage of patients using MS-specific DMT and off-label drugs: mixed-effect model results according to the cancer treatment type, cases vs. controls, France.

Discussion

We assessed the impact of cancer on MS-related care and DMT use in a large multinational population-based cohort of matched pwMS with and without cancer. Overall, cancer did not have a major effect on neurological care and MS hospitalizations in the 2 years after the diagnosis, independently of age, sex, residence, and MS duration. These findings remained consistent in pwMS across all MS treatment groups as well as across time periods.

Conversely, DMT use decreased sharply after the cancer diagnosis, overall and across all cancer treatment groups, but especially in individuals receiving chemotherapy. This decrease was also observed in those diagnosed with cancer after 2018, reflecting current MS practices. Cancer did not affect the rates of neurologist visits and MS hospitalizations, possibly suggesting a decision to withhold DMT and not patient-led care withdrawal. This decision could be due to multiple factors: oncology-driven treatment prioritization, contraindications imposed by oncologists, or temporary clinical instability. It could also be neurologist or patient-driven due to concerns about the role of DMT in cancer incidence or severity. The fact that chemotherapy can have a beneficial and long-lasting immunomodulatory effect on the course of MS could justify temporary DMT cessation.^15,22,24 Furthermore, some DMT–chemotherapy combinations may be contraindicated due to cumulative risks, most notably infectious ones. However, radiotherapy and surgery do not act as MS treatment; therefore, the decrease in DMT use in these treatment groups is more challenging to interpret. In addition, the adverse effect of cancers on DMT use seemed larger in the French cohort.

A key challenge was ensuring comparability between pwMS with cancer (cases) and matched cancer-free pwMS (controls), both alive during the study period. The case population was highly specific: ill enough to develop cancer yet healthy enough to survive ⩾2 years post-diagnosis. To maximize compatibility and ensure similar levels of health and health-related behaviors, individuals were matched on criteria as stringent as possible. In addition, pre-cancer MS-related healthcare utilization indicators were broadly similar, suggesting a high degree of comparability between cases and controls.

Several differences emerged between the French and BC cohorts. First, at cancer diagnosis, BC patients were younger (55.9 vs. 58.6 years), likely reflecting differences in the cancer identification algorithm (cancer identification through physicians’ visits in BC which may identify earlier or less severe tumors), cancer type distribution, and study periods (France: 2011–2021; BC: 1992–2020). Indeed, the study period in BC included a time of elevated MS mortality, resulting in a younger overall population. Second, for all three outcomes, MS-related healthcare consumption was higher in France than in BC, especially concerning DMT use, for both cases and controls. This could be explained by differences in care practice and use of healthcare services between these regions. Regardless of these differences, the impact of cancer on MS-related care consumption was similar in both cohorts.

Although population-based medico-administrative databases allow studying large cohorts and diminishing the selection bias, they can be limited in terms of available information, such as lifestyle-related factors and clinical data (e.g. disease progression or activity [relapses]). In addition, cancer severity or stage at diagnosis was not available and could be an important factor regarding treatment choices, specifically regarding discontinuation, as some DMTs are contraindicated in the case of progressive cancer.^33,34 Finally, DMT exposure windows, based on filled prescriptions, may introduce minor misclassification. Regional differences also existed; for example, information on cancer treatment types was available only in France, whereas diagnoses (ICD codes) related to physician visits were available only in BC. The modest use of high-efficacy DMTs likely reflects an older, less inflammatory population, the study covering multiple treatment paradigms, and estimates align with recent population-based studies.^4,35

To conclude, this study offers valuable insights into how the diagnosis of cancer affects MS neurological care and treatment decisions, yielding reassuring and consistent results across two regions on separate continents. We used population-based data that allowed the inclusion of virtually all pwMS with a cancer diagnosis in France and BC between 1994 and 2019, resulting in a large, population-based cohort. It also allowed addressing the same questions in two regions with different MS care practices over a long time period. Our time-period–stratified analysis, particularly for the 2018–2019 cancer diagnosis group followed up until 2021 in France, demonstrated that our observations remained consistent over time, thereby reinforcing the relevance of these findings within the context of contemporary MS management practices. As the MS population grows and ages, this issue is becoming increasingly relevant. Further research is needed on care coordination to develop unified, evidence-based guidelines for managing MS in the context of cancer.²³

Supplemental Material

sj-docx-1-msj-10.1177_13524585261437966 – Supplemental material for Impact of cancer on multiple sclerosis–related healthcare and disease-modifying drug use: A multinational cohort study

Supplemental material, sj-docx-1-msj-10.1177_13524585261437966 for Impact of cancer on multiple sclerosis–related healthcare and disease-modifying drug use: A multinational cohort study by Chloe Pierret, Geraldine Androdias, Christine Lebrun-Frenay, Sandra Vukusic, Myriam Maumy, Huah Shin Ng, Feng Zhu, Helen Tremlett and Emmanuelle Leray in Multiple Sclerosis Journal

Footnotes

Acknowledgements

We would like to acknowledge the members of the team who facilitated gaining funding for the wider program of work related to MS (Dr Karim, Dr Wijnands, Dr Marrie, Dr Maxwell, Dr Zhao, and Dr Yusuf).

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: C.P. has nothing to disclose. G.A. has received personal compensation for consulting, serving on a scientific advisory board, speaking, or other activities with Biogen, Janssen, Merck, Novartis, Roche, and Sanofi-Genzyme. C.L.-F. has research support in the last 3 years from the EDMUS Foundation. S.V. has received non-personal consulting and lecturing fees, travel grants, and unconditional research support from Biogen, Janssen, Merck, Novartis, Roche, Sandoz, and Sanofi. M.M. has nothing to disclose. H.S.N. has received research support from the Cancer Council South Australia’s Early Career Research Fellowship. F.Z. has nothing to disclose. H.T. has received research support in the last 3 years from the Canada Research Chair Program, National MS Society, Canadian Institutes of Health Research, Canada Foundation for Innovation, MS Canada, and the EDMUS Foundation. Advisor for Canada’s Drug Agency (2024–2025, payment declined). E.L. consulting and lecture fees or travel grants from Alexion, Biogen, Merck, Novartis, Roche, and Sanofi are not related to the submitted work.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by the French Foundation for Research in Multiple Sclerosis (ARSEP) mobility grant, the French Public Health Doctoral Network (RDSP) mobility grant, the ARSEP-EDMUS Foundation, the Clinical Neuroscience Institute of Rennes (INCR), and the National MS Society and MS Canada (RG5063A4/1/RFA-2103-37392; EGID: P002/903124; PI: H.T.). None of the study funders had a role in the methods, analyses, writing of the manuscript, or the decision to submit it for publication.

Ethical Considerations

France: Data access for the study was approved according to the current French legislation (Articles R.1461–11 to R.1461-17 of the French Public Health Code and Data Protection Authority decision CNIL-2016-316), and the study was declared to the EHESP SNDS registry, equivalent to an Institutional Review Board approval. British Columbia: This study was approved by the University of British Columbia’s Clinical Research Ethics Board. Access to data provided by the Data Steward(s) is subject to approval but can be requested for research projects through the Data Steward(s) or their designated service providers.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

ORCID iDs

Chloe Pierret

Geraldine Androdias

Christine Lebrun-Frenay

Sandra Vukusic

Huah Shin Ng

Helen Tremlett

Emmanuelle Leray

Data Availability

A request for data access may be made, but it would require the previous approval of the French regulatory authorities and/or Canadian Data Stewards from PopulationData BC. France: The SNDS database is the property of the French health insurance system. British Columbia: Access to data provided by the Data Stewards is subject to approval but can be requested for research projects through the Data Stewards or their designated service providers. The following datasets were used in this study: PharmaNet, Discharge Abstract Database, Medical Services Plan, BC Vital Events, and Statistics and Census Geodata. You can find further information regarding these data sets by visiting the PopData project webpage at . All inferences, opinions, and conclusions drawn in this publication are those of the authors and do not reflect the opinions or policies of the Data Stewards.

Supplemental Material

Supplemental material for this article is available online.

References

Lunde

HMB

Assmus

Myhr

, et al. Survival and cause of death in multiple sclerosis: A 60-year longitudinal population study. J Neurol Neurosurg Psychiatry 2017; 88(8): 621–625.

Koch-Henriksen

Laursen

Stenager

, et al. Excess mortality among patients with multiple sclerosis in Denmark has dropped significantly over the past six decades: A population based study. J Neurol Neurosurg Psychiatry 2017; 88(8): 626–631.

Zhu

Kingwell

, et al. Disease-modifying drugs for multiple sclerosis and association with survival. Neurol Neuroimmunol Neuroinflamm 2022; 9(5): e200005.

Pierret

Mainguy

Leray

Prevalence of multiple sclerosis in France in 2021: Data from the French health insurance database. Rev Neurol 2024; 180(5): 429–437.

Foulon

Maura

Dalichampt

, et al. Prevalence and mortality of patients with multiple sclerosis in France in 2012: A study based on French health insurance data. J Neurol 2017; 264(6): 1185–1192.

Wallin

Culpepper

Campbell

, et al. The prevalence of MS in the United States. Neurology 2019; 92(10): e1029–e1040.

Kingwell

Zhu

Marrie

, et al. High incidence and increasing prevalence of multiple sclerosis in British Columbia, Canada: Findings from over two decades (1991–2010). J Neurol 2015; 262(10): 2352–2363.

Marrie

Reider

Cohen

, et al. A systematic review of the incidence and prevalence of cancer in multiple sclerosis. Mult Scler 2015; 21(3): 294–304.

Handel

Ramagopalan

SV.

Multiple sclerosis and risk of cancer: A meta-analysis. J Neurol Neurosurg Psychiatry 2010; 81(12): 1413–1414.

10.

Hongell

Kurki

Sumelahti

, et al. Risk of cancer among Finnish multiple sclerosis patients. Mult Scler Relat Disord 2019; 35: 221–227.

11.

Bosco-Lévy

Foch

Grelaud

, et al. Incidence and risk of cancer among multiple sclerosis patients: A matched population-based cohort study. Eur J Neurol 2022; 29(4): 1091–1099.

12.

Pierret

Mulliez

Le Bihan-Benjamin

, et al. Cancer risk among patients with multiple sclerosis: A 10-year nationwide retrospective cohort study. Neurology 2024; 103(9): e209885.

13.

Kingwell

Bajdik

Phillips

, et al. Cancer risk in multiple sclerosis: Findings from British Columbia, Canada. Brain 2012; 135(pt 10): 2973–2979.

14.

Nylander

Singh

Poole

, et al. Clinical course of multiple sclerosis and patient experiences during breast cancer treatment. Mult Scler 2023; 29(8): 967–978.

15.

Yalachkov

Dabanli

Wenger

, et al. Concurrent CNS tumors and multiple sclerosis: Retrospective single-center cohort study and lessons for the clinical management. Neurol Sci 2022; 43(9): 5513–5522.

16.

Miller

Lachance

Lucchinetti

, et al. Multiple sclerosis, brain radiotherapy, and risk of neurotoxicity: The Mayo Clinic experience. Int J Radiat Oncol Biol Phys 2006; 66(4): 1178–1186.

17.

Milic

Rees

JH.

Acute demyelination following radiotherapy for glioma: A cautionary tale. Pract Neurol 2017; 17(1): 35–38.

18.

Zhang

Tremlett

Leung

, et al. Examining the effects of comorbidities on disease-modifying therapy use in multiple sclerosis. Neurology 2016; 86(14): 1287–1295.

19.

Salter

Lancia

Kowalec

, et al. Comorbidity and disease activity in multiple sclerosis. JAMA Neurol 2024; 81(11): 1170–1177.

20.

Al-Sakran

Marrie

Blackburn

, et al. Impact of comorbidity on hospitalizations in individuals newly diagnosed with multiple sclerosis: A longitudinal population-based study. Mult Scler Relat Disord 2020; 40: 101955.

21.

Marrie

Fisk

Fitzgerald

, et al. Etiology, effects and management of comorbidities in multiple sclerosis: Recent advances. Front Immunol 2023; 14: 1197195.

22.

Nesbitt

Van Der Walt

Butzkueven

, et al. Multiple sclerosis and cancer: Navigating a dual diagnosis. Mult Scler 2024; 30(14): 1714–1736.

23.

Castaño-Amores

Nieto-Gómez

The crossroads of multiple sclerosis and cancer: Urgent call for clinical direction. Mult Scler 2024; 30(3): 455–456.

24.

Collongues

Durand-Dubief

Lebrun-Frenay

, et al. Cancer and multiple sclerosis: 2023 recommendations from the French Multiple Sclerosis Society. Mult Scler 2024; 30(7): 899–924.

25.

Tuppin

Rudant

Constantinou

, et al. Value of a national administrative database to guide public decisions: From the système national d’information interrégimes de l’Assurance Maladie (SNIIRAM) to the système national des données de santé (SNDS) in France. Rev Epidemiol Sante Publique 2017; 65 suppl 4: S149–S167.

26.

Mainguy

Tillaut

Degremont

, et al. Assessing the risk of relapse requiring corticosteroids after in vitro fertilization in women with multiple sclerosis. Neurology 2022; 99(17): e1916–e1925.

27.

Moisset

Leray

Chenaf

, et al. Risk of relapse after COVID-19 vaccination among patients with multiple sclerosis in France. Neurology 2024; 103(5): e209662.

28.

Rey

Jougla

Fouillet

, et al. Ecological association between a deprivation index and mortality in France over the period 1997–2001: Variations with spatial scale, degree of urbanicity, age, gender and cause of death. BMC Public Health 2009; 9(1): 33.

29.

Lix

Smith

Pitz

, et al. Cancer data linkage in Manitoba: Expanding the infrastructure for research. Winnipeg, MB: Manitoba Centre for Health Policy, 2016.

30.

Bousquet

Lefeuvre

Tuppin

, et al. Cancer care and public health policy evaluations in France: Usefulness of the national cancer cohort. PLoS ONE 2018; 13(10): e0206448.

31.

Alping

Askling

Burman

, et al. Cancer risk for fingolimod, natalizumab, and rituximab in multiple sclerosis patients. Ann Neurol 2020; 87(5): 688–699.

32.

Stamatellos

Rigas

Stamoula

, et al. S1P receptor modulators in multiple sclerosis: Detecting a potential skin cancer safety signal. Mult Scler Relat Disord 2022; 59: 103681.

33.

EMD Serono. MAVENCLAD® (cladribine) [package insert] 2022, https://www.mavenclad.com/content/dam/web/healthcare/neurology/united-states/pdfs/Prescribing%20Information.pdf

34.

Genzyme Corporation. LEMTRADA® (alemtuzumab) [package insert] 2026, https://products.sanofi.us/lemtrada/lemtrada.pdf

35.

Goereci

Ellenberger

Rommer

, et al. Persons with multiple sclerosis older than 55 years: An analysis from the German MS registry. J Neurol 2024; 271(6): 3409–3416.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.09 MB

0.00 MB