Methotrexate in chronic-recurrent calcium pyrophosphate deposition disease: no significant effect in a randomized crossover trial

Calcium pyrophosphate deposition (CPPD) disease is a very common and occasionally severe arthropathy associated with chondrocalcinosis. Chondrocalcinosis is itself a metabolic condition due to CPPD, particularly frequent in the second half of life. It can be asymptomatic, at least at the beginning, but various forms of arthropathies develop in affected individuals. Symptomatic presentations include acute CPP arthritis, sometimes called pseudogout, and a chronic arthropathy, generally recurrent, associating osteoarthritis (OA) and CPPD [1]. Recurrent acute CPP arthritis and chronic recurrent CPP arthropathy can lead to severe joint destruction, sometimes very rapidly.

Currently, no specific treatment has demonstrated efficacy to prevent or slow CPP deposition [2]. Acute inflammation induced by CPPD crystals can be nonspecifically treated by nonsteroidal anti-inflammatory agents (NSAIDs), glucocorticoids or colchicine, but these symptomatic treatments are very often ineffective, particularly in recurrent or chronic cases [2]. Furthermore, these agents are frequently not tolerated or contraindicated in older patients, the population most frequently affected by chondrocalcinosis. More importantly, none of the available treatments proved able to stop gradual joint damage due to chondrocalcinosis. More recently, promising results have been reported with biologic anti-interleukin 1 (IL-1) agents [3],[4], but these agents are extremely expensive and not reimbursed by most health systems or insurances for this indication.

At low doses, methotrexate (MTX) is a potent anti-inflammatory agent, considered the gold standard therapy for many inflammatory arthritides. MTX could conceptually be effective in CPP arthropathies since it interferes with glutathione metabolism, decreasing cell recruitment to the inflamed joint and promoting the release of the endogenous anti-inflammatory mediator adenosine and IL-1Ra [5]. Given its good safety profile and low cost, MTX could represent an interesting therapeutic option, particularly for patients resistant to classic anti-inflammatory treatments. Two case series suggested that MTX is a safe and effective therapeutic alternative for CPPD arthropathy refractory to standard therapy [6],[7], but were contradicted later by a small French case series [8]. To date, no controlled study has examined the impact of MTX in refractory CPP arthropathy. We therefore conducted a randomized controlled trial to test the efficacy of MTX on symptoms and signs of chronic or recurrent CPP arthropathy.

Forty-nine patients were screened for participation in the trial, and after a medical assessment, 26 patients were randomized (Figure 2). Seven patients did not complete the whole trial; the most common reason for discontinuation was a transaminase elevation above three times the normal values. The baseline characteristics at the initiation of the trial treatment were balanced (Table 1). Subjects ranged from 46 to 84 years (median = 71, IQR (54.5 to 76.0), mean = 66.4 ± 12.8) of age and were all Caucasian.

Disease activity, as measured by the DAS44 improved over time (P = 0.005) but was not significantly improved by treatment with MTX. As for the secondary outcomes, none of the other secondary outcomes improved significantly by treatment with MTX (Table 2, Figure 3).

Outcomes	Treatment course: MTX	Treatment course: PBO	Pvalue
	(N = 25)	(N = 21)
DAS44	−0.08 [−0.6, 0.1]	−0.13 [−0.6, 0.1]	0.44
Number of tender joints [0 – 44]	0 [−1.0, 1.3]	−1 [−3.5, 0.5]	0.17
Number of swollen joints [0 – 44]	−1 [−3.5, 0]	0 [−1.5, 2.0]	0.56
CRP level	0.2 [−0.8, 1.9]	−0.3 [−1.6, 1.6]	0.33
Number of analgesic pills/week, median	0 [−7, 4]	0 [−2.0, 1.5]	0.36
Number of flares/3 months	0 [−1, 0]	0 [−1, 0]	0.10
VAS Pain	−1 [−2.8, 0]	0 [−2, 1]	0.43

All results are displayed in medians [interquartile ranges]. MTX, methotrexate; PBO, placebo; DAS44, disease activity score based on 44 joints; CRP, C-reactive protein; VAS Pain, visual analog scale for pain.

We also examined the possibility of an effect modification by the type of CPPD arthropathy (monoarticular presentations vs. polyarticular presentations) and found no evidence that MTX was more effective in patients with polyarticular presentation (test for effect modification, P = 0.27).

During the study follow-up, 35 adverse events (AEs) were reported, 16 on MTX and 18 on PBO (P = 0.81) and 1 during the wash-out period. However, more patients dropped out on MTX than on PBO (5 vs. 0), mostly for common AE related to MTX. One serious AE was observed on a patient on methotrexate: an acute bicytopenia, which imposed hospitalization, fortunately without permanent damage.

This randomized controlled, crossover trial was conducted to assess the efficacy of MTX in chronic CPPD arthropathy. We found that after three months of treatment, 15 mg of sc MTX, did not significantly improve the symptoms of CPPD disease, while MTX was associated with increased AEs typically related to this therapy. To our knowledge, this is the first PBO-controlled trial to investigate the efficacy of MTX, an inexpensive and well-tolerated anti-rheumatic drug, in patients with chronic CPPD arthropathy.

This controlled study therefore contradicts our preliminary observational study as well as another Spanish report [6],[7]. Several hypotheses can explain the contradiction between favorable observational findings and a negative trial. First, this trial was possibly underpowered, leading to the possibility of a false negative result. Indeed, it proved extremely difficult to set up a multicenter, multinational, investigator-based study allowing the recruitment of enough patients. However, because of the crossover design, the trial still had a predicted power of 74% to detect a moderate change in DAS44, the study’s primary outcome. While the possibility of a false negative result exists, the likelihood that MTX has a large beneficial effect on the whole spectrum of chronic CPPD arthropathies is low. Furthermore, our outcome measures might not have been adequate to capture a potential beneficial effect of MTX; however, currently no validated outcome measures are available for the different clinical presentations of chronic CPPD arthropathy. The fact that MTX did not demonstrate a significant effect on the primary outcome, nor any of the five secondary outcomes, strongly suggests that this is not a false negative result. Another possible explanation for this finding is that we included patients not necessarily during a flare, with baseline disease activity too low (DAS44 between 2 and 3, median CRP around 3) to demonstrate a large effect size. Furthermore, as the natural course of the disease is one of flares and spontaneous improvements, a phenomenon of regression to the mean, makes the demonstration of a beneficial treatment effect even more difficult. This phenomenon could be amplified by the fact that flares frequently occur at intervals longer than three months, especially in patients with early disease.

In addition, CPPD disease is notoriously heterogeneous [1] and differential responses to a given medication are likely. This study has included patients with a variety of CPPD arthropathy clinical presentations (chronic or recurrent arthropathies, mono-, oligo- or poly-articular), which may have further diluted the study’s ability to demonstrate an effect of MTX. Another major difficulty is the fact that this aged population is frequently affected by multiple coexisting pathologies, including osteoarthritis and other seronegative arthritides, which can potentially mask a beneficial therapeutic effect. In addition, induction of transient arthralgias in some patients is a well-recognized adverse effect of MTX [18], which could have further obscured a potential beneficial effect. Finally, our study design allowed patients to take auto-medication of analgesics, NSAIDs and low-dose glucocorticoids, which may further have obliterated our capacity to demonstrate a potential treatment effect of MTX.

The crossover design assumes that the effect of the medication is transitory and removed quickly after treatment discontinuation during the ‘wash-out period’, with no carry-over effect thereafter. The pharmacological half-life of MTX is typically around four to six weeks in healthy individuals, thus a two-month wash-out period appears adequate. We examined a possible carry-over effect in the analysis, but did not find any statistically significant one. We cannot exclude the possibility that a potential beneficial effect of MTX might have appeared after three months. It is possible also that a dose of 15 mg of sc MTX was insufficient for some of the patients; it was selected because this dose of MTX has proved to be effective in older RA patients [11], and in our preliminary study, but possibly a higher dosage could have been more effective. The compliance of the study medication was good, as the sc injections were either performed by a nurse, at least in the beginning, and when not, all syringes were verified after use to control proper emptying. No specific outcome measures have been validated for CPPD arthropathy, which forced us to use instruments developed and validated for other rheumatic conditions. However, the clinical presentation of complicated CPPD is often similar to other common rheumatic conditions, such as gout or seronegative inflammatory arthritis, which should make the use of outcome measures for these diseases good instruments for chondrocalcinosis as well.

Famous precedents actually exist on the sometimes wide discrepancies between observational and controlled studies, which can be explained by designs of the later ones that afterward appeared unsuitable to demonstrate an effect [19]. We therefore strongly believe that further studies involving more centers and patients, a better selection of the cases, if possible with less comorbidities, longer treatment periods and a more strictly defined intake of other medications is warranted.

Chondrocalcinosis is indeed one of the most common joint disorders, affecting up to 5% of the human population, with the prevalence rising to 15% over 60 years of age [20]; its incidence will continue to rise because of the increasing age of the population. CPPD disease is certainly under-recognized, often misdiagnosed and frequently only treated symptomatically, which can lead to accelerated joint destruction. Existing treatments are often unsatisfactory and currently no anti-rheumatic therapy has been formally tested in CPPD arthropathy. The biologic anti-IL-1 agents are promising for some patients, but certainly do not represent an option for a vast majority of patients throughout the world at least for a while, in particular because of their price. Effective, safe and cheap therapeutic options are therefore clearly needed to improve the quality of life of patients not responding to simple NSAID therapy. At this time, in view of this study and other reports [8], MTX is clearly not a treatment solution for all patients with chronic CPPD arthropathy. However, some patients continue to take MTX because of strong perceived benefits, after being informed of their treatment arm. In addition, MTX may have the potential to reduce joint damage in chronic CPPD arthropathy, given its pharmacological properties and established protective effects in RA [5], and this might be considered as an outcome in future studies. We therefore believe that MTX, despite this negative study, is still an option for selected patients with inadequate response or intolerance to the other available medications, and that further studies are warranted.