The Efficacy, Safety and Tolerability of Canakinumab in the Treatment of Familial Mediterranean Fever: A Systematic Review of the Literature.

Familial Mediterranean Fever (FMF) is the most prevalent genetic autoinflammatory disorder. In most patients, treatment with colchicine can prevent attacks of fever and inflammation. However, 5%-10% of patients are resistant to colchicine treatment, while a similar percentage cannot tolerate colchicine in doses needed to prevent attacks. For these patients, Canakinumab, a full human antibody against IL-1β, has been approved recently by the FDA and EMA. In this article, we present a systematic review of the long-term efficacy, safety, and tolerability of Canakinumab in FMF patients who cannot tolerate colchicine or who are resistant to colchicine treatment.

Introduction

Familial Mediterranean Fever (FMF) is the most prevalent of the genetic autoinflammatory syndromes. It is characterized by recurrent attacks of fever accompanied by signs of serosal inflammations such as peritonitis, arthritis, pericarditis and pleuritis.1 There are ~100.000 patients worldwide, with the highest prevalence of FMF found in people living in the countries bordering the Eastern and Southern Mediterranean and Armenia and in immigrant populations from these areas in Western Europe and the United States.2

A typical FMF patient suffers from short episodes of fever and inflammation lasting 12 h to 3 days, with symptom-free intervals between attacks. A subgroup of patients has a more severe phenotype with continuous inflammation and episodes of aggravation.3

Colchicine has been shown effective in preventing inflammatory attacks in most patients. However, an estimated 5–10% continue to experience inflammatory attacks despite maximal dose of colchicine.4,5 Patients with an attack frequency of >1 typical episode per 3 months and presence of amyloidosis or elevated acute phase reactants are considered to have colchicine-resistant FMF (crFMF).6 Furthermore, another 5% to 10% of patients experience serious side effects including severe diarrhea, neuropathy, rhabdomyolysis, and bone marrow suppression.7

Since inappropriate production of the pro-inflammatory cytokine interleukin 1β by cells of the innate immune system plays a central role in the pathogenesis of FMF, blocking of IL1 by biological drugs has been tried in colchicine-resistant FMF (crFMF) patients since the beginning of this century.8 Anakinra, a IL-1 receptor antagonist administered daily by subcutaneous injection, showed promising results in case reports and case series.9 In addition, Canakinumab, a full human antibody against IL-1β administered subcutaneously every 4–8 weeks, also appeared very effective in crFMF.10 It was approved by the FDA and EMA in 2016 for the treatment of colchicine-resistant FMF.

The long-term effectiveness, the rate and type of rare side effect, and adherence to therapy are hitherto unclear. Here we present a systematic review of the literature in order to fill the gap in knowledge about the safety, efficacy, and tolerability and of Canakinumab in crFMF patients.

Methods

We followed a protocol using the methodological approaches outlined in the Agency for Healthcare Research and Quality Methods Guide for Effectiveness and Comparative Effectiveness Reviews and applied the PRISMA Guidelines.11 The study was registered in the PROSPERO registry for systematic reviews (number CRD42019158100). The systematic literature review aimed to include all studies published until December 1, 2019 reporting on Canakinumab therapy in FMF patients. We searched Medscape, Web-of-Science, and SCOPUS using the following syntax: (ILARIS OR canakinumab) AND (familial mediterranean fever OR FMF) in all fields.

Study Selection

The selection was performed by two investigators (MK, JH). We included randomized controlled trials, non-randomized trials, retrospective analysis of these trials, cohort studies or cross-sectional studies, case reports, and case series. There were no language restrictions. We excluded in vitro and animal studies, review articles, and congress abstracts. We assessed all titles and abstracts identified by our search. Publications were considered eligible for the analysis if they contained data on Canakinumab in patients with FMF as defined by the Tel Hashomer criteria.12 Full-length articles were retrieved from all published papers.

Data Extraction

Using a standardized data extraction sheet, the following data were collected from the articles: lead author, publication year, study design, sample size, reason for Canakinumab use, MEFV-mutation, presence of type AA amyloidosis, the response to therapy, dosage, duration of therapy, previous Anakinra use, reasons for stopping Anakinra treatment, reported side effects. In addition, the following baseline characteristics were extracted: age of the patients, age at diagnosis and the number and proportion of male patients. Patients were considered to have a complete response if there was not a single attack whilst on Canakinumab.

Results

Study and Patient Characteristics

Among the 413 results provided by the database search, 192 unique studies were identified, of which 19 were eligible studies reporting on the effect of Canakinumab in FMF patients (Figure 1). These included 2 case reports, 6 case series, 7 cohort studies, 3 open-label prospective trials, and one randomised, placebo-controlled trial (Table 1).

Figure 1

Search strategy and article selection.

Table 1

Characteristics of Reviewed Studies and Participants

Author	Year	Study Design	Total	Female	Paediatric	Age at Treatment Start [Year]	Colchicine Resistance	Colchicine Intolerance	Anakinra Tried	Anakinra Resistance	Anakinra Intolerance	Renal Amyloidosis	M694V Homozygous [%]	Attack Frequency [/Month]
Alpa18	2015	Case report	1	1	0	30	1	0	1	0	1	NA	0.0%	12.0
Babaoglu14	2018	Retrospective analysis	23	15	0	Median 32 (24–58)	12	2	23	7	16	7	73.9%	Mean 2.0 ± 0.9
Basaran40	2015	Case series	4	1	3	Median 12.5 (10.6–19)	4	0	4	0	4	NA	100.0%	1, 4+, 4+, 4+, 4+‡
Berdeli30	2019	Open-label single-arm study	22	9	22	Mean 13.4 ± 4.0 (6–18)	22	0	NA	NA	NA	NA	59.1%	NA
Brik31	2014	Open-label single-arm study	7	2	7	Median 9.5 (6.8–14.9)	7	0	NA	NA	NA	NA	0.0%	Mean 2.1 ± 0.8
De Benedetti10	2018	Prospective randomized double-blinded study	63	14	29	Mean 22.5±15(can) 21.8±13.8(plac)	63	2	NA	NA	NA	0	68.3%	Mean 2.3 ± 2.5
Eren akarcan23	2019	Case series	9	6	9	Mean 14.3 (7–19)	7	2	NA	NA	NA	NA	55.6%	NA
Gül19	2015	Open-label single-arm study	9	7	NA	Median 22(12–34)	9	0	NA	NA	NA	NA	66.7%	Median 3.29/3mo
Gülez15	2018	Retrospective analysis	15	8	15	Median 16.5(8–19)	15	0	3	2	1	1	86.7%	NA
Kisla Ekinci16	2019	Case series	14	10	12	Median 11(4–19)	11	2	1	NA	NA	1	78.6%	Mean 1.8 ± 0.5
Laskari20	2017	Retrospective analysis	14	7	3	Median 38.5 (13–70)	14	0	5	5	0	NA	21.4%	NA
Meinzer21	2011	Case series	2	1	2	7, 7	2	0	1	1	0	NA	100.0%	NA
Sendogan17	2019	Case series	4	2	0	Median 34(24–44)	2	2	4	1	2	4	50.0%	NA
Trabulus28	2018	Retrospective analysis	9	4	0	Median 33 (27–62)	8	0	2	0	2	4	44.4%	Mean 1.1 ± 1.5
Yasuda22	2019	Case report	1	1	1	12	1	0	0	0	0	0	0.0%	5.0
Yazilitas27	2018	Retrospective analysis	11	5	11	Median 14 (6–17)	8	0	NA	NA	NA	3	100.0%	Mean 1.0 ± 0.4
Yildirim26	2018	Case series	3	3	0	Median 57 (46–61)	2	1	0	0	0	3	66.7%	NA
Kucuksahin25	2017	Retrospective analysis	2	2	0	34, 57	1	1	0	0	0	0	50.0%	Mean 0.7 ± 0.3
Cetin24	2015	Retrospective analysis	8	4	2	Median 19(14–22)	8	0	0	NA	0	1	100.0%	Median 1

Notes: Data are provided as mean ± standard deviation or median and range. ‡Attack frequency exceeding 4/month reported as “4+”.

Abbreviations: NA, not available; can, canakinumab arm; plac, placebo arm.

The 19 studies included data on 221 patients from 16 countries, 102 (46%) of whom were female.

The majority of patients (65.6%,) were homozygous for the M694V mutation. The pre-treatment attack frequency was reported in 12 studies and varied from 0 to 12 attacks per month. Seventy-five percent of patients with reported attack frequencies had 6 or more attacks per year before initiation of anti-IL1 therapy. The cumulative exposure to Canakinumab treatment was 265.4 patient-years.

Indications for Canakinumab

Colchicine resistance was reported in 89% of patients, with 12% having dose-limiting intolerance to the drug.

Data on previous immunosuppressive therapy were available for 100 patients.13–21 Forty-four patients used Anakinra prior to initiation of Canakinumab therapy. Of them, 26 had an intolerance to Anakinra and in 16 patients Anakinra was ineffective in preventing attacks. One patient had a good response to anakinra but was switched to Canakinumab due to ease of use,17 whereas the reason for the switch was not given in one.16

One hundred and seventy-four patients continued to use colchicine in association with Canakinumab. In 2 patients colchicine was stopped during treatment.14,22

Dosing of Canakinumab

The most common dose of Canakinumab was 2mg/kg in patients <40kg, or 150mg in those above.13,17,18,20–28 In case of inadequate disease control, an increase in dose to 3–4mg/kg in those below 40kg or 300mg in those over was used in 8 studies.10,14,15,19,23,29–31

In 8 studies,10,19,22,25–28,31 Canakinumab was given at 4-weekly intervals, in 3 studies18,21,24 it was given at 8-weekly intervals, whereas the remaining studies had dosing frequencies tailored to clinical needs and/or protocols, with the medication being given every 4, 6 or 8 weeks.

The study by Eren Akarcan et al23 included a reduction in frequency at 6 months with treatment discontinuation at 12 months. The patients experiencing flares (n=4) during follow-up were given Canakinumab at 12-weekly intervals.

Efficacy of Canakinumab

Response data were available for 189 patients, 2 of which did not meet the Tel-Hashomer criteria and will be reported separately. At the initial Canakinumab dose of 150mg or 2mg/kg in <40kg, 78.1% had a complete response, defined as the complete absence of attacks during Canakinumab treatment (Table 2). An additional 19.8% had a partial response with fewer attacks and/or reduced severity, with only 2.1% failing to attain any response at all.

Table 2

Treatment Regimens, Responses and Safety Data

Author	Total	Females	Paediatric	Dosage†	Dosing Intervals	CR‡ [%]	PR‡ [%]	NR‡ [%]	Concurrent Colchicine [%]	Duration of Therapy (Months)	Cumulative Duration (Months)	Number Discontinued/Reason	SAE	Side Effects
Alpa18	1	1	0	150mg	q8w	0.0	100.0	0.0	0.0	6	6	0	0	None
Babaoglu14	23	15	0	150mg OR 300mg	q4w/q6w/q8w	60.9	39.1	0.0	60.9	Median 7 (1–44)	257	7 (2 pregnancy, 2 Ax SpA, 2 DC, 1 IBD)	0	3 weight gain (1 DC), 5 URTI, 1 leukopenia
Basaran40	4	1	3	2mg/kg OR 150mg	q4w/q6w/q8w	50.0	25.0	25.0	50.0	Median 25.5 (24–28)	103	1 (exacerbation)	0	1 flare of FMF
Berdeli30	22	9	22	4mg/kg	q4w/q82	100.0	0.0	0.0	100.0	Mean 15.9 ±8.8	349.8	0	0	None
Brik31	7	2	7	2mg/kg OR 150mg (up to 4mg/kg OR 300mg)	q4w	42.9	42.9	14.3	42.9	3	21	0	1 (infection)	11 AE, 1 mild infection and 1 moderate throat infection)
De Benedetti10	31	14	14	4mg/kg OR 150mg (up to 300mg)	q4w	71.0	29.0	0.0	71.0	10	547.2	0	14 (3 infections)	332 AE including flare (79 infections)
Eren Akarcan23	9	6	9	4mg/kg OR 150mg	q4w/q8w/q12w	100.0	0.0	0.0	100.0	12*	108	9* (5 relapses)	0	NA
Gül19	9	7	NA	150mg (up to 300mg)	q4w	88.9	11.1	0.0	88.9	3	27	0	1 (headache)	4 headache, 2 URTI
Gülez15	15	8	15	2mg/kg OR 150mg (up to 4mg/kg OR 300mg)	q4w/q8w	93.3	6.7	0.0	93.3	Median 26 (12–58)	359	0	1 (bronchopneumonia)	2 UTI, 1 bronchopneumonia, 2 dental abscess
Kisla Ekinci16	14	10	12	2–4mg/kg	q4w/q8w	71.4	28.6	0.0	71.4	Median 22 (4–62)	416	1 (breakthrough)	0	None
Laskari20	14	7	3	150mg	q4w/q6w/q8w	92.9	7.1	0.0	92.9	Median 18 (13–53)	310	0	0	1 UTI, 1 viral gastroenteritis
Meinzer21	2	1	2	2mg/kg	q8w	100.0	0.0	0.0	100.0	Median 4.5 (4–5)	9	0	0	NA
Sendogan17	4	2	0	150mg	q4w/q8w	100.0	0.0	0.0	100.0	Median 36 (30–36)	138	0	0	1 mild UTI
Trabulus28	8	4	0	150mg	q4w	100.0	0.0	0.0	100.0	Median 9 (3–14)	83	0	1°° (CMV pneumonia)	1 CMV pneumonia, 1 injection reaction
Yasuda22	1	1	1	150mg	q4w	100.0	0.0	0.0	100.0	18	18	0	0	None
Yazilitas27	11	5	11	2mg/kg	q4w	90.9	9.1	0.0	90.9	Median 21 (5–49)	266	1 (breakthrough)”	1 (pneumonia)	1 pneumonia
Yildirim26	3	3	0	150mg	q4w	100.0	0.0	0.0	100.0	Median 6 (3–12)	21	0	1 (pyelonephritis)	1 pyelonephritis, 2 injection reactions
Kucuksahin25	1	1	0	150mg	q4w	0.0	100.0	0.0	0.0	Median 4.5 (3–6)	9	1 (progression of arthritis)	0	None
Cetin24	8	4	2	150mg	q8w	50.0	50.0	0.0	50.0	Median 18 (4–25)	137	0	0	None

Notes: †Weight-adjusted dosage was uniformly used in patients weighing less than 40kg, with standardized dosing at 150 or 150mg thereafter; q4w/q6w/q8w/q12w – 4-/6-/8- or 12-weekly; ‡Response rates to optimum Canakinumab therapy; *Treatment regimen included discontinuation at 12 months, with relapsing patients offered Canakinumab q12w; ”Patient on peritoneal dialysis discontinued due to breakthroughs, developed bacterial peritonitis 1 year after discontinuation, and died; °°The reported serious adverse event occurred in the patient not meeting Tel-Hashomer criteria.

Abbreviations: CR, complete response; PR, partial response; NR, no response; (S) AE, (serious) adverse event; Ax SpA, axial spondyloarthritis; DC, patient decision to discontinue; IBD, inflammatory bowel disease; CMV, cytomegalovirus; URTI, upper respiratory tract infection; UTI, urinary tract infection.

The dose was increased to 300mg in 11 of the 37 partial responders and 3 of the 4 non-responders. In total, treatment with the optimal dosage of Canakinumab led to CR in 79.9% and PR in 19.3%, with only 1.1% of patients remaining non-responders.

The two patients who did not fulfil Tel-Hashomer criteria were both heterozygote for a disease-causing mutation.25,28 Complete response to Canakinumab could not be achieved in either case.

Safety and Tolerability of Canakinumab

Reporting thresholds for adverse events (AE) varied by study, with only serious AE being reported consistently. Safety data could be assessed for 265.4 patient-years. During this period, there were 8 cases of moderate to severe infections (three cases of pneumonia, two cases of pharyngotonsillitis, one case of pyelonephritis, one case of pelvic abscess and one of cellulitis). No deaths, mycobacterial or opportunistic infections were reported. For purposes of this review, non-SAE were assessed only in prospective studies (n=4) to avoid recall bias. Of those, only the prospective randomized, placebo-controlled trial by De Benedetti et al provided sufficient information about AE. The most common among the 332 reported AE were infections (23.8% of all AE), injection site reactions (6.0%), headache (3.9%) and abdominal pain (3.6%).10

None of the eleven reported unscheduled treatment discontinuations were due to SAE. The discontinuations occurred for the following reasons: progression of arthritis (n = 3), lack of effect (n = 3), pregnancy (n = 2), weight gain (n = 1), progression of inflammatory bowel disease necessitating alternative biological therapy (n =1) or patient decision not otherwise stated (n = 1).

Pregnancy

The articles by Gul et al and Babaoglu et al reported on one Canakinumab-exposed pregnancy each.14,19 The medication was discontinued in the peri-conception period and at 8 weeks’ gestation, respectively. There were no signs of congenital malformations or intra-uterine growth retardation, both infants were carried to term and no developmental abnormalities were reported.

Amyloidosis

Information about the presence of renal or systemic amyloidosis could be obtained from 121 patients in 11 of the assessed studies.14,17,24,26–28 Twenty-four patients had type AA amyloidosis at the start of treatment. Of those, six had proteinuria and reduced renal function which improved in 5 and remained stable in one patient after initiation of Canakinumab. Four patients had end-stage renal disease requiring dialysis when Canakinumab was started, of whom 2 experienced improvement in renal function, one discontinued Canakinumab due to disease progression, and one of them continued to renal transplantation without evidence of recurrence under Canakinumab treatment.17 A further 14 patients were renal allograft recipients following amyloid-induced end-stage renal failure.17,26,28 Of these, 7 had evidence of amyloid deposition in the allograft, leading to the introduction of Canakinumab,28 whilst poorly controlled inflammatory attacks were the indication for implementation of Canakinumab in the remaining patients. Renal function remained stable in all cases, with no improvement or exacerbation of proteinuria or significant changes to serum creatinine levels reported.

No patients developed new-onset systemic or renal amyloidosis whilst receiving Canakinumab.

Discussion

In this systematic review, we found that Canakinumab is highly effective and safe in patients with colchicine-resistant or colchicine-intolerant FMF. Nearly 80% of patients did not experience a single attack after the initiation of Canakinumab. Almost all the remaining patients had significantly reduced disease activity. These results are remarkable taking into consideration that most patients included in the studies had a particularly severe phenotype. The majority were homozygote for M694V, which is known to cause more severe presentation of FMF with higher rates of colchicine resistance.32 Where pre-Canakinumab attack frequencies could be assessed (n=72) 75.0% of patients had at least 6 attack per year, with the mean frequency exceeding 2.1 attacks per month and the maximum as high as 12 attacks per month. Since FMF severely affects the quality of life,33 efficacious therapy is crucial. Moreover, recurrent inflammation is directly related to the risk of type AA amyloidosis.34,35 If inflammation cannot be controlled, type AA amyloidosis develops in up to three quarters of patients during life.36 In the 97 patients without amyloidosis before initiation of therapy, none developed it during Canakinumab treatment. In addition, in patients with already established nephrotic syndrome caused by type AA amyloidosis, a significant reduction in proteinuria was consistently observed. Furthermore, in patients with renal transplantation, no recurrences of amyloidosis were seen when patients were on Canakinumab treatment.

Colchicine remains the mainstay of therapy in FMF. It has been shown to prevent type AA amyloidosis.37 Two patients stopped colchicine therapy whilst on anti-IL1 treatment. An additional 11 patients were unable to tolerate colchicine. Since type AA amyloidosis only develops in the presence of the acute phase reactant serum amyloid A (SAA), it is unclear if patients under anti-IL1 treatment with a complete response clinical and biochemical response benefit from additional colchicine.34 The current guideline still advice to continue colchicine during anti-IL1 treatment.38

Canakinumab was not only highly efficacious, it was also safe and well-tolerated. Based on the 265.4 patient-years of cumulative exposure described in the reviewed studies, only 8 severe infections were reported in 7 patients. There were no reports of opportunistic infections.

Of the 221 patients treated with Canakinumab, only three stopped due to lack of efficacy, two of which did not fulfil the Tel-Hashomer criteria. No patients discontinued treatment due to side effects, a marked contrast to the 59% of Anakinra discontinuations due to side effects reported in the reviewed patients who had received anakinra before. Canakinumab seems to be better tolerated than Anakinra, the other anti-IL1 that is used for the treatment of crFMF. In general, between 6.5% and 30% of patients using Anakinra therapy withdraw due to adverse events.39,40 The main reason for stopping Anakinra therapy is severe injection site reaction. This was not reported in any of the patients on Canakinumab. Although the efficacy of Anakinra seems similar to Canakinumab,41 Canakinumab was initiated in 16 patients because of insufficient response to Anakinra. Response to Canakinumab among that subset could be assessed in 14 patients; CR was achieved in 64.3%. However, the lack of high-quality evidence precludes judgment about the superiority of one the anti-IL1 treatments over another.

The current literature does not allow us to make a statement about optimal dosing. Although the majority of patients were treated with 150 mg every 4 weeks, there was considerable variation in the dosing interval between studies. A lower dose interval did not seem to give high numbers of treatment failures. This suggests that at least in some patients receiving 150mg, an interval length exceeding 4 weeks may be feasible. This would be desirable considering the high cost of therapy and potential side effects.

Conclusion

Cumulative evidence shows that Canakinumab is very effective in preventing attacks in almost all patients with colchicine resistant or colchicine-intolerant FMF. It may prevent the development of type AA amyloidosis and it appears safe and well-tolerated. Future research is needed to asses the optimal dosing strategy.