The overall life expectancy for multiple sclerosis patients is 7 years less than normal, or 75% to 85% of expected survival (Weinshenker and Ebers 1987; Sadovnick et al 1992). Mortality increases with disability. Case fatality ratios are 1:5 for patients with Kurtzke disability scores of less than or equal to 7, but 4:4 for those with scores greater than 7 (Sadovnick et al 1992). The suicide rate is increased 7.5-fold (Sadovnick et al 1991). The lifetime risk of suicide is 2%. Less-disabled young patients within 5 years of diagnosis are the most likely actors (Sadovnick et al 1992; Stenager et al 1992).
The cause of death in 50% of a clinic population and in approximately 75% of all multiple sclerosis patients is from complications of multiple sclerosis, usually pneumonia (Sadovnick et al 1991; Reder et al 2012). Patients most commonly die when disability scores approach 8.0. Brainstem lesions occasionally cause loss of inspiratory drive, causing the patient to stop breathing; this is most common at night. Deaths from malignancy are less common than in age-matched controls (Sadovnick et al 1991). The cumulative lifetime dose of disease-modifying therapy improves prognosis. Interferon-beta-1b extends survival by at least 6 years.
Children with isolated symptoms at onset, multiple MRI lesions, elevated IgG index, and oligoclonal bands in CSF are more likely to develop multiple sclerosis than those with polyclonal onset and negative CSF tests.
The future course of multiple sclerosis is poor if there are cerebellar or pyramidal symptoms, slow timed walk test at baseline, early sphincter symptoms, multi-site onset, frequent early attacks, development of progression or a primary progressive course, and age over 40 years at onset. Good prognostic signs include optic neuritis, sensory symptoms, and an exacerbating-remitting course (Weinshenker and Ebers 1987; Phadke 1990). The course is a more important predictor than age of onset. Development of a progressive course is the strongest predictor of poor outcome. The second strongest predictor is the number of relapses in the first 2 years. After a first demyelinating episode, a second attack is more likely in younger patients with abnormal CSF, with more than 8 T2 MRI lesions (dissemination) or 1 Gd enhancing lesion (activity). Long-duration, low-disability multiple sclerosis is likely to remain stable. Complete recovery is more likely with mild severity and mono-lesional exacerbations. Clinic-based cohorts have more severe multiple sclerosis than population-based groups, where many patients remain stable or progress minimally over 10 years (Pittock et al 2004). Multiple sclerosis reduces quality of life throughout its course.
Patients with relapsing-remitting multiple sclerosis take 15 years from onset to reach an Expanded Disability Status Scale of 6.0 (using a cane to walk 100 meters), based on longitudinal studies in Ontario, Canada (Ebers 2000). Those with primary progressive multiple sclerosis take 8 years, and early progression and multi-system symptoms hasten the rate of progression. Later onset multiple sclerosis is more common in men and is often primary progressive. Even when beginning as relapsing-remitting disease, the transition to primary progressive multiple sclerosis is earlier in men than in women. Patients with 1 attack in the first 2 years do not need a cane for 20 years; those with 5 or more attacks need a cane within 7 years. Attacks after the first 2 years correlate with better prognosis, perhaps indicating that the patient has relapsing and not progressive multiple sclerosis. Once a patient becomes unable to walk 500 m (EDSS =4; typically after 11 years), progression is no longer affected by relapses. The average rate of decline is similar in all groups once multiple sclerosis becomes progressive (including primary progressive at onset, “bout onset progression,” or at the transition from relapsing-remitting to secondary progression) (Rice 1997; Ebers 2000; Kremenchutzky et al 2006). This happens on average at the age of 40, preferentially targets the corticospinal tract, and is not obviously influenced by prior relapses. Within the progressive group, however, rates vary (“sooner to cane, sooner to wheelchair”).
On MRI, bad prognostic signs include a large number of T2 lesions or high T2 volume, T1 hypointensities, many enhancing lesions, low magnetization transfer, and lesions in juxtacortical, infratentorial, and periventricular sites, and brain atrophy, especially early atrophy. Good signs are little tissue damage and sparing of important regions. An MRI with a few large lesions gives a better prognosis than one with the same volume of many, smaller lesions (Kepes 1993; Zivadinov personal communication 2005). In a 20-year follow-up of 107 relapse-onset patients, lesion growth was 0.80 cc/year in those who were relapsing-remitting but was 2.89 cc/year in secondary progression (Fisniku et al 2008).
In clinically isolated syndromes (CIS; lesions in optic nerve, brainstem, or spinal cord), total T2 lesion volume on MRI at onset correlates with disability at 10 years (r = 0.45), and all patients with a total lesion volume greater than 3cc progress to definite multiple sclerosis by 4 years (Sailer et al 1999). Ventricular enlargement or new T2 MRI lesions 3 months after the first symptoms strongly predict clinically definite multiple sclerosis – 88% after a positive MRI versus 19% after a negative MRI (Brex et al 2002). In 532 patients with clinically isolated syndromes followed for up to 9 years, definite multiple sclerosis developed in only 35% of those with no asymptomatic baseline lesions but in 74% of those with 3 of 4 of the following lesions: a) 1 enhancing or 9 T2, b) 3 periventricular, c) 1 juxtacortical, or d) 1 infratentorial (revised McDonald criteria) (Korteweg et al 2006). In the BENEFIT CIS study, 9 baseline T2 lesions (hazard ratio = 1.6) or 3 periventricular lesions (hazard ratio = 1.7), or when one lesion changed to more than one, predicted conversion to multiple sclerosis. Corpus callosum lesions have a hazard ratio of 2.7.
After starting thrice-weekly interferon beta-1a therapy, first-year MRI lesions increase the chance of progression at 2 years by 4%. Clinical relapses after the first year increase the chance of progression by 49% (Sormani et al 2011).
Brain atrophy is often present in mild-to-moderate multiple sclerosis. Atrophy is most likely to progress when there are Gd-enhancing lesions at baseline (Simon et al 2000). Ventricles in relapsing-remitting multiple sclerosis increase in size by 5% per year, compared to 1% to 2% in normal controls. In most cases of multiple sclerosis that come to medical attention, disease activity never sleeps and atrophy progresses relentlessly in all multiple sclerosis subtypes. However, there are a large number of subclinical cases with benign courses and presumably much milder inflammation.
Magnetic (motor) and electric evoked potentials do not correlate with disability at first presentation. However, abnormal evoked potentials do predict disability at 2 years (r = 0.6 with motor and visual potentials) (Fuhr et al 2001) and at 5 years (r = 0.5 with motor and sensory) (Kallmann et al 2006).
CSF with a high white cell count predicts Gd+ MRI lesions (Rudick et al 1999). High levels of IgG predict faster progression. CSF with high numbers of natural killer cells and monocytes augurs slower progression (Cepok et al 2001), although subsets of both of these cell types can damage oligodendroglia. B cells and plasma cells and a high B cell to monocyte ratio in CSF predict faster progression, as do increased myelin basic protein, increased IgM oligoclonal bands and homozygous HLA-DRB1*1501, and polymorphisms in multiple genes (Frohman et al 2005). Antibodies to myelin basic protein in clinically isolated syndromes were highly predictive for development of multiple sclerosis in one study, but other labs have not been able to replicate these findings. Conversely, the 10% of multiple sclerosis patients with negative oligoclonal bands in CSF are more likely to have progressive forms of multiple sclerosis, non-specific “supratentorial” symptoms, lower cells and IgG in CSF, and less well-defined MRI lesions (Siritho and Freedman 2009).
Other markers may be helpful in the future. Some of these include molecular indicators of interferon efficacy, B7 costimulatory molecules (CD8, CD86, PD-1, PD-L1) and adhesion molecule expression on mononuclear cells, serum cytokines or cytokine receptors, kallikrein proteases and matrix metalloproteases, IgM antibodies to glycans (anti-GAGA4), antibodies to CNS proteins (galactocerebroside, myelin basic protein, myelin-oligodendrocyte glycoprotein, neurofilaments, proteolipid protein), heat shock proteins, newer MRI techniques, increased CSF 14-3-3 protein (neuronal loss) and myelin basic protein (oligodendrocyte damage), and CSF or serum soluble intercellular adhesion molecule (ICAM-1) and vascular cell adhesion molecule (VCAM) (correlate with MRI activity).
The lifetime cost of multiple sclerosis in the United States is high compared to other neurologic diseases. Multiple sclerosis has a total lifetime cost of $2,200,000 and an annual cost of $34,000 to $47,000 and is more expensive than ischemic stroke ($100,000 lifetime) and Alzheimer disease ($49,000 to $490,000 lifetime) (Whetten-Goldstein et al 1998). Expense in diseases of long duration accrues from earnings lost, rehabilitation, drug therapy, medical equipment, and formal and informal care. Fifty-three percent is direct (40% drugs, 3% inpatient hospital care), 10% is informal care, and 37% is productivity loss (reduced work time and retirement) (Kobelt et al 2006). Costs increase dramatically with more severe disease. A British cost-effectiveness analysis suggests that each relapse avoided during interferon therapy costs 28,700 British pounds (Parkin et al 2000). Note that relapses are easy to count, but have weak correlation with progression--a better indicator of disability. Longer survival with interferon beta-1b has not yet been factored into these analyses.