Efficacy of Anakinra for Various Types of Crystal-Induced Arthritis in Complex Hospitalized Patients: A Case Series and Review of the Literature

Background. There are few data on anakinra use after failure of conventional medications for crystal-induced peripheral arthritis and/or crowned dens syndrome among complex hospitalized patients. Methods. We retrospectively analyzed the outcome of six patients affected with subacute crystal-induced arthritis who had received anakinra in second or third line therapy, including three patients with crowned dens syndrome and three others with gouty arthritis. Patients' comorbidities, reasons for anakinra use and associated drugs, and outcomes were recorded. Results. All patients presented with elevated inflammatory syndrome, systemic symptoms with poly/oligoarthritis. Except for absolute contraindications, all patients were previously treated with full or decreased dose of NSAID, colchicine, and/or glucocorticoids, with unsatisfactory response. All three gouty patients exhibited complete responses in all acute involvements under anakinra within 3 to 5 days, including one of them who needed the reintroduction of colchicine treatment that was previously unsuccessful. Crowned dens syndrome patients, including two with pseudogout and one with subacute hydroxyapatite deposition disease, needed 9 to 11 days to achieve complete response. Tolerance to anakinra was good. Conclusion. In case series of complex hospitalized patients, anakinra showed good activity in crowned dens syndrome and associated crystal-induced peripheral arthritis, with longer treatment duration than in gouty arthritis.


Introduction
The number of therapeutic drugs used to treat crystalinduced arthritis (CRIA) is relatively limited, relying on colchicine, nonsteroidal anti-inflammatory drugs, and corticosteroids. These drugs may have significant and severe side-effects or may be contraindicated and therefore cannot always be used at an optimal dosage especially among elderly patients. Thus, other therapeutic options are often required in clinical practice.
Recently, several publications have focused on the rationale and efficacy of the interleukin-1 blockade mainly in gout [1][2][3][4] and sparsely in pseudogout [5][6][7]. No clear outcome parameters concerning notably the treatment duration of anakinra are defined. Otherwise, there is no report on the use of anakinra in the acute inflammatory manifestations of hydroxyapatite deposition disease (HADD) and notably in crowned dens syndrome, which is a clinical form that this latter shares with pseudogout. Both diagnosis and management of crowned dens syndrome can be challenging, and the control of the disease requires sometimes the combination of several drugs, with increased risk of toxicity especially in older patients.
Herein, we report on the successful use of anakinra in the context of resistance or contraindications to conventional drugs, in six complex hospitalized elderly patients with crowned dens syndrome and associated crystal-induced peripheral arthritis ( = 3) and gouty polyarthritis on the other hand ( = 3).

Patients and Methods
We retrospectively collected data from six adult patients who had received off-label anakinra (Kineret 100 mg/day; subcutaneous injection) for CRIA (gout, pseudogout, or HADD) between January 2012 and March 2013. All patients included in this study were hospitalized or referred to the department of internal medicine. Approvals from institutional medicalpharmaceutical committee and review board were obtained for this study, and the patients' consent was obtained after providing them with information about anakinra.
The diagnosis of gout was based on the presence of monosodium urate (MSU) crystals in synovial fluid, the presence of tophus, or the presence of at least six of the twelve other crystal-induced types of arthritis referred to in the criteria of the American College of Rheumatology [8]. The diagnosis of other types of CRIA was based on the same scheme, except for the characteristic signs of gout, including hydroxyapatite crystals for HADD or calcium pyrophosphate (CPP) crystals for pseudogout.
In the setting of complex hospitalized patients, diagnosis of an attack of CRIA was based on the presence of compatible clinical signs, significant increase in C-reactive protein levels (CRP; normal < 5 mg/L) to at least 15 mg/L, and associated typical imaging. Some differential diagnoses were ruled out by performing minimal autoimmune (antinuclear factor, anti-DNA, rheumatoid factor, anti-CCP, and ANCA antibodies) and routine microbiological exams. Serum uric acid levels were assessed (>6.0 mg/dL in women) also taking account of creatinine level. None of our six patients had a medical history or other clinical manifestations that evoked an inherited autoinflammatory syndrome.
For an attack of CRIA, colchicine was given to patients 1 and 6, at 1 mg/three times a day, on days 1 and 2; then 1 mg twice daily, on days 3 and 4; and then 1 mg a day on days 5 and 6. Reduced doses, for best tolerance with similar efficacy [2], were given to patients 2 and 4, whereas this drug was contraindicated in patients 3 and 5 (Table 1). Longterm prophylaxis was given to patients who had exhibited previous recurrent attacks of polyarthritis with respect to renal function [2]. The nonsteroidal anti-inflammatory drugs that were routinely used were indomethacin (orally, 25 mg, 3-4 times daily) or ketoprofen (100 mg twice daily, orally and/or intravenously) given for 10-14 days; doses were reduced by 30% for elderly patients and those at risk of renal damage with close creatinine monitoring. Prednisone was given at 0.5 mg/kg/day (max. 40 mg/day) for 15 days. Anakinra therapy was maintained until there was both a complete clinical response and normalization of CRP level. Treatment with anakinra was stopped by day 6 in the absence of any clinical improvement and no reduction in CRP level, which was monitored every 48-72 h.
Patients with crowned dens syndrome were more likely aged females (three women with mean age of 79 years) compared to gouty patients (two men, one woman with mean age of 69.6 years). For the three patients with crowned dens syndrome, the diagnosis was retained on clinical signs, namely, fever, multidirectional neck stiffness and pain, exclusion of meningitis with lumbar picture in two, and compatible features of neck CT-scan without any criteria for vertebral or soft-tissue infectious or tumoral process ( Figure 1). The diagnosis of crowned dens syndrome was moreover based on concomitant peripheral joint involvements criteria (Table  1). Two patients had, respectively, a history of pseudogout and HADD diagnosed on typical crystal finding on joint puncture and compatible X-ray features. The latter exhibited calcium pyrophosphate crystals on knee joint puncture and iconographic pictures of pseudogout location, but also thick arciform calcification of the retrodens ligament on cervical CT-scan and HADD features on scapular MRI (Figures 1(c)(A) and 1(c)(B)).
Among the three gouty patients, two were treated, respectively, with hydroxyurea and allopurinol for polycythemia vera and with pentostatin for a prolymphocytic leukemia without exhibiting criteria for tumor lysis syndrome; the latter exhibited severe chronic renal insufficiency related to gouty nephropathy.
The localization of arthritis was different in the two groups: gouty patients presented symmetrical large and small joints arthritis whereas patients with crowned dens syndrome displayed by definition a periodontoid arthritis and often asymmetrical shoulder and/or wrist involvements.

Anakinra Efficacy among Patients with Crowned Dens
Syndrome and Gout. Indication for anakinra was mostly similar for both groups, including failure or contraindications to colchicine and/or nonsteroidal anti-inflammatory drugs or corticosteroids, especially for the associated comorbidities.
Complete remission, defined both clinically and biologically with the normalization of CRP level (Figure 2), was obtained in all three gouty patients, whereas for the crowned   dens syndrome group only two of three patients underwent complete remission with only anakinra. In the crowned dens syndrome group (patients 1, 2, and 3; Table 1), 9 and 11 days were, respectively, needed to induce complete remission for the two responders, whereas in gout only a median of 5 days (3, 5, and 5 days, resp.) was needed. For the third patient with crowned dens syndrome, anakinra withdrawal on the 7th day in the context of marked reduction but persistent neck pains was followed by a rapid disease flare-up 48 h later; the reintroduction of a second course of anakinra led to rapid, complete, and persistent remission of all acute clinical signs associated with CRP level normalization at the 11th day, allowing for withdrawing the treatment this same day. Finally, a median of 11 days (9, 11, and 11 days, resp.) was required for complete remission in crowned dens syndrome patients. Finally, a gouty patient (patient 6) exhibited a relapse six weeks after anakinra withdrawal. While the reintroduction of anakinra did not allow disease control at the 5th day, adjunction of colchicine led to persistent complete remission at the 3rd day of the combined treatment. For all six patients, after anakinra withdrawal, dietary recommendations associated with allopurinol, febuxostat, and/or colchicine were continued as prophylactic measures; no further relapse occurred within a follow-up of 6 to 16 months.

Treatment Safety.
Anakinra showed good or excellent tolerance profiles among all patients. Two patients (patients 1 and 4) had mild injection-site reactions and transient diffuse pruritus. No episodes of neutropenia or infection were observed. Despite initial decrease in renal function (creatinine clearance: 40-20 mL/min) in some patients, daily injections of anakinra were not harmful and even improved the renal parameters of patient 4 who probably had chronic gouty nephropathy. Colchicine tolerance was also good, although patients 2 and 4 presented episodic diarrhea. Finally, for patient 3 who had an ongoing staphylococcal soft-tissue infection, the use of anakinra rather than colchicine was harmless.

Discussion
In the context of failure or contraindications to conventional therapies, anakinra exhibited good efficacy in all three types of CRIA with a longer treatment duration in crowned dens syndrome and associated peripheral arthritis related to pseudogout and/or HADD (9, 11, and 11 days, resp.) than in gouty arthritis (3,5, and 5 days, resp.). Moreover, adjunction of colchicine to anakinra seems to allow a best outcome in CRIA refractory to each drug alone. To the best of our knowledge, this is the first report of anakinra use and efficacy for crowned dens syndrome related to pseudogout and HADD and also for HADD peripheral subacute arthritis. The tolerance in elderly patients was good in our series. Indeed, several articles showed that anti-IL-1 agents, such as anakinra, are safe in comparison with other biotherapies, such as anti-TNF-alpha, considering the risk of tuberculosis reactivation [9,10].
Only few small series have been reported on the use of anakinra in gout and pseudogout as curative and/or prophylactic treatment. Moltó et al. showed that, in four out of five patients suffering from refractory subacute oligo/polyarticular flare-ups of pseudogout, anakinra use for three days allowed dramatic clinical control and normalization of CRP as soon as the third day [6]. Announ et al. and McGonagle et al. showed similar results for anakinra, used as a curative or prophylactic treatment for recurrent subacute or chronic polyarticular pseudogout in two patients for at least 6 months [5,11]. Although Couderc et al. reported three patients with a satisfactory result with anakinra to prevent acute flare-ups, this drug was unable to control chronic synovitis symptoms in two of them [7].
Whereas gouty acute/subacute arthritis usually needed only three days of anakinra treatment, a longer treatment duration was necessary to control acute pseudogout flareups (14 days for the patient reported by McGonagle et al.) or chronic synovitis (i.e., at least three months), as observed for our patients and those in the literature [5,7]. For two pseudogout polyarthritis patients, reported by Verhoeven et al., anakinra prescribed for five days showed no or only slight response [12]. These authors concluded that this poor efficacy of anakinra might be related to the systemic inflammation and/or to the long duration of this disease (2-6 months); these data also correlate with those of Couderc et al. [7].
Considering our pseudogout patients and those reported by others, this hypothesis appears only partially consistent. Treatment duration of five days leading to poor efficacy of anakinra in the series of Verhoeven et al. [12] was much shorter than the minimum of nine days in our series and for the patient reported by McGonagle et al. [5]. Thus, we hypothesize that treatment duration is the major factor in anakinra success to treat pseudogout and probably also HADD attacks. Indeed, in the pathophysiology of CRIA, the key role of IL-1 is now well documented: induction of IL-6 and TNF-alpha upregulation, complex retroinduction of other cytokines, leading to self-sustaining multicytokine vicious circle notably in cases of prolonged attacks [13][14][15].
Deposition of hydroxyapatite crystals in joints and periarticular tissues can cause acute attacks, qualified as pseudopodagra, which mimic gout and/or pseudogout attacks [13,16]. Pazár et al. have shown that, similar to MSU and CPP, hydroxyapatite crystals can also act as alarmins and trigger NLRP3 inflammasome activation [13]. The consequent overproduction of IL-1 seems indirectly mediated by increased levels of other proinflammatory stimuli, including TNF-or uric acid [17,18]. Pazár et al. thus suggested that the IL-1 blockade may be clinically useful in acute HADD manifestations; this hypothesis was clinically confirmed for two of our patients [13]. Adding anakinra to previous ongoing treatments and, notably, colchicine should lead to the best response, as seen in the report of McGonagle et al. and for one of our polyarticular gouty patients [5].
To conclude, IL-1 blockade seems to be an interesting therapeutic option for all three types of CRIA whatever the type of joint involvement, including the debilitating and often misdiagnosed crowned dens syndrome. As observed in this short series and in sparse literature data, acute/subacute pseudogout or acute hydroxyapatite deposition arthritis should require extended duration of treatment with anakinra than in gouty arthritis [15]. Moreover, adding anakinra to conventional drugs could show better efficacy than each drug alone. Therefore, larger studies are needed to best assess the treatment modalities of anakinra in acute/subacute manifestations of the various types of CRIA, including its association with conventional drugs in curative and prophylactic purposes.