- Email: [email protected]
Systematic Review and the External Validity of Randomized Controlled Trials in Lupus Nephritis
CLINICAL RESEARCH
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Systematic Review and the External Validity of Randomized Controlled Trials in Lupus Nephritis Angela Pakozdi1, Ravindra Rajakariar2, Debasish Pyne1, Andrea Cove-Smith2 and Muhammad Magdi Yaqoob2
Q12
1
Department of Rheumatology, Barts Health NHS Trust, London, United Kingdom; and 2Department of Nephrology, Barts Health NHS Trust, London, United Kingdom
Introduction: Randomized controlled trials (RCTs) are considered the gold standard for assessing treatment efficacy. However, sampling bias can affect the generalization of results to routine clinical practice. Here we assessed whether patients with lupus nephritis (LN) seen in routine clinical practice would have satisfied entry criteria to the major published RCTs in LN.
Q1
Methods: A systematic literature search from January 1974 to May 2015 was carried out, identifying all RCTs investigating LN induction treatment. Patients diagnosed with proliferative or membranous LN between 1995 and 2013 were identified from the Barts Lupus Centre database; baseline characteristics were compared with each RCT’s entry criteria to assess hypothetical inclusion or exclusion. Results: Of 363 articles, 33 RCTs met inclusion criteria. Of 137 patients newly diagnosed with LN (111 with proliferative/mixed proliferative and 26 with pure membranous LN), 32% would have been excluded from RCT entry (range 8%–73%). The main reasons for exclusion would have been too severe disease, too mild disease, or prior immunosuppressant use, which were exclusion criteria in 26, 20, and 22 RCTs, respectively. A total of 27 patients with LN (20%) were re-biopsied due to flare; 68% of these would have been ineligible to enter RCTs. Conclusion: Published RCTs do not truly reflect the heterogeneity of patients with LN in routine practice at our lupus center. The external validity of RCTs could be improved by including more representative patient cohorts. RCTs should be used as a guide but consideration should be given to similarities between individual patients and the characteristics of the trial cohorts before treatment decisions being made. Kidney Int Rep (2018) -, -–-; https://doi.org/10.1016/j.ekir.2017.11.005 KEYWORDS: glomerulonephritis; lupus ª 2017 International Society of Nephrology. Published by Elsevier Inc. This is an open access article under the CC BYNC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Q2
remendous advances in immunosuppressive treatment of lupus nephritis (LN) have been made over the past 4 decades. Fortunately, most treatment decisions made in routine clinical practice are now based on results from well-designed clinical trials rather than anecdotal evidence. Randomized controlled trials (RCTs) are considered the “gold standard” to assess treatment efficacy,1,2 as they have the best potential to minimize bias through allocating participants to study arms in a random fashion. Despite their rigorous design, the applicability of results to individual patients has been debated thanks to other types of
T
Correspondence: Muhammad Magdi Yaqoob, Barts Health NHS Trust, Department of Nephrology, Royal London Hospital, Whitechapel Road, London E1 1BB, United Kingdom. E-mail: m.m. [email protected] Received 11 October 2017; accepted 6 November 2017; published online
bias, some of which lead to RCTs not adequately reflecting the heterogeneity of patients in routine clinical practice.3 Examples include “selection bias” due to poor allocation concealment, “sampling bias” when physicians consider only a certain type of patient for trials and exclude patients on the basis of the lack of personal equipoise, or “literacy bias” for those who fail to understand the consent form.4 In LN RCTs, a particular problem may be “severity of illness bias.”4 To achieve a more homogeneous study population in whom the study drug is most likely to work and less likely to harm, patients must meet a range of prespecified entry requirements.5 Inclusion criteria to LN RCTs often specify the target range of biochemical variables and serological activity markers, and the histological evidence of LN with restriction to 1 or 2 particular classes of glomerulonephritis. Moreover, exclusion criteria intentionally eliminate patients with mild or, more commonly, severe disease; those with
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
1
52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
CLINICAL RESEARCH
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
A Pakozdi et al.: External validity of LN RCTs
recent major infections or with preexisting comorbidities; and tend to prohibit certain immunosuppressive drugs before enrollment. As a consequence, patients recruited to most LN RCTs have a relatively moderate disease activity excluding those with too mild or severe disease and those developing active LN while on immunosuppressive therapies. In this study, we aimed to evaluate the external validity of published LN RCTs. We assessed whether a multiethnic cohort of patients with lupus seen in routine clinical practice at a single center would have been represented adequately by assessing RCT inclusion and exclusion criteria. METHODS Patients and Study Design This study was conducted at Barts Health NHS Trust, the largest tertiary teaching hospital in the United Kingdom. Patients with biopsy-proven proliferative (class III or IV) or membranous (class V) LN diagnosed between 1995 and 2013 were included. LN classes based on glomerular pathology were defined according to the International Society of Nephrology/Renal Pathology Society 2003 classification.6,7 All patients fulfilled at least 4 of the American College of Rheumatology revised classification criteria for systemic lupus erythematosus.8 Clinical data were collected retrospectively from electronic patient records; baseline variables included serum creatinine, glomerular filtration rate (GFR), serum albumin, proteinuria, doublestranded DNA, and complement C3 and C4 levels. Baseline immunosuppressive drugs at time of LN onset were recorded. Identification of Eligible Clinical Studies We selected RCTs that aimed to measure the efficacy of immunosuppressive agents in proliferative (class III, IV) and/or membranous (class V) LN induction treatment. Literature searches were performed using PubMed, Medline, and EMBASE databases (from 1974 to May 2015) and the Cochrane Controlled Trial Register (up to May 2015) using the OVID search engine, to identify relevant studies. Key search words included “lupus nephritis” and “randomized” and “lupus nephritis” and “clinical trial.” Additional studies were sought through bibliographic notations. Studies published in full-text English literature were included if they met the following criteria: (i) prospective controlled trial with treatment allocation by random assignment; (ii) compared at least 2 arms for LN induction treatment; (iii) involved proliferative and/or membranous LN patient population; and (iv) included adult patients. We excluded trials that published data 2
on extended follow-up of already published RCTs, studies of maintenance therapies, or refractory LN. Evaluation of Hypothetical Noninclusion Rate to RCTs Baseline characteristics of patients in the Barts LN cohort were compared with patients participating in some of the most important RCTs of LN induction treatment from the past 2 decades. For the purpose of this study, factors leading to exclusion included severe renal impairment with predefined GFR or serum creatinine in the selection criteria of the RCT; mild disease with predefined levels of proteinuria, serum albumin, GFR, or serum creatinine; or prohibited baseline immunosuppressive drugs used at time of LN onset. We did not analyze any other exclusion criteria, including confounding factors not related to their renal disease. When trials involved patients with proliferative glomerulonephritis (class III or IV V), we evaluated only the patients with proliferative LN from our cohort (n ¼ 111); where trials recruited patients with membranous LN, we evaluated only our membranous LN cohort (n ¼ 26), whereas our entire cohort was compared (n ¼ 137) when both classes were represented in the RCT. RESULTS Characteristics of RCTs Included in the Analysis Our systematic literature search yielded 363 articles, of which 309 articles were deemed unsuitable after title or abstract review (Figure 1). Of the remaining, 5 trials on LN induction treatment had to be excluded, as the study design was insufficiently described and no comment was made on inclusion and exclusion criteria used. The final analysis composed of 33 RCTs is shown in Table 1.9–41 Of the 33 RCTs, 19 journal articles (58%) failed to report the methods of randomization and/or the protocols used to assign participants to comparison groups. Furthermore, only 4 RCTs27,31,39,40 published the rate of screening failure, ranging from 19.6% to 39.4%. There were only 2 of 33 RCTs elaborating the reasons for ineligibility.27,31 The most frequently prespecified inclusion and exclusion criteria encountered in the selected RCTs are detailed in Table 2. Nearly all RCTs (97%) required fulfillment of systemic lupus erythematosus classification criteria, but a positive antinuclear antibody/lupus erythematosus test was rarely desired (18%). Three trials (9%) did not require biopsy evidence of LN; 75% of RCTs required patients to have a renal function that met predefined values, and 67% of studies demanded a Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
CLINICAL RESEARCH
Records idenfied through database searching (n = 352)
Addional records idenfied through other sources (n = 11)
Screening
Records screened (n = 106)
Eligibility
Full-text arcles assessed for eligibility (n = 54)
Included
Records aer duplicates removed (n = 106)
web 4C=FPO
211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
Idenficaon
A Pakozdi et al.: External validity of LN RCTs
Studies included in qualitave synthesis (n = 33)
Records excluded (n = 52)
Full-text arcles excluded, with reasons (n = 21) Not on inducon treatment (n = 7) No randomizaon (n = 2) No explanaon of methodology (n = 5) Extended follow-up of published trial (n = 4) Terminated trial without efficacy analysis (n = 1) Other language than English (n = 2)
Figure 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram.
minimal level proteinuria. Prohibited concomitant or prior immunosuppressive treatment was defined in 76% of RCTs. Characteristics of Barts LN Patient Cohort From our database, we identified 137 patients with a new diagnosis of biopsy-proven active LN; their baseline characteristics are shown in Table 3. Eightyone percent of our LN cohort had proliferative (n ¼ 111, 30% class III, 51% class IV, with or without membranous component), and 19% had pure membranous LN (n ¼ 26). Overall, patients with LN in our cohort tended to have more severe renal disease compared with patients in landmark RCTs of LN induction (Table 4)42; they had a mean proteinuria of 5.5 g/24 hours ( 5.1 SD), mean serum albumin of 27.3 g/L (7.4 SD), and mean serum creatinine of 159 mmol/L ( 166 SD). Although 33% (n ¼ 45) had a normal GFR (>90 ml/min per 1.73 m2), 18% (n ¼ 25) had a GFR <30 ml/min per 1.73 m2. Serological evidence of lupus activity was present in half of the cohort with strongly positive double-stranded DNA and low complements. Twenty-three patients with systemic lupus
erythematosus (16.7%) were taking immunosuppressants other than hydroxychloroquine or oral prednisolone (either mycophenolate mofetil or azathioprine) at time of LN diagnosis. Hypothetical Noninclusion Rate of Newly Diagnosed Patients With LN to LN Induction RCTs On average, one-third (31.8%) of our newly diagnosed biopsy-proven active LN cohort would have been Q3 ineligible to enter the RCTs (range 8.1%–72.9%) (Figure 1). These patients would have been excluded because of factors directly related to their renal disease or the concomitant use of immunosuppressive treatment. Severe renal impairment was a prespecified exclusion criterion in 26 RCTs (79%) leading to ineligibility in up to 61.4% of our LN cohort. Twenty RCTs (63%) would have left out patients with milder disease based on mildly impaired renal function or lower level of proteinuria, resulting in a hypothetical exclusion rate reaching up to 44.1%. Additionally, 22 RCTs (67%) prohibited the use of various immunosuppressive drugs resulting in exclusion rates of 16.1% of our LN cohort.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
3
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
CLINICAL RESEARCH
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372
A Pakozdi et al.: External validity of LN RCTs
Figure 2. Noninclusion rates to landmark LN RCTs. The proportion of our patients with LN that would not have been included in RCTs is shown. Q11 Overall on average 32% of our patients with LN would have not been eligible to participate in the RCTs listed (range 8%–73%).
Hypothetical Noninclusion Rate of Relapsed Patients With LN to LN Induction RCTs Next, we analyzed hypothetical exclusion rates of relapsed patients with LN with active LN evidenced by a repeat kidney biopsy. Twenty-nine of 137 patients with LN (21.2%) had a repeat kidney biopsy. The median time to repeat biopsy was 28 months (interquartile range 13–47 months). Of those, 27 had active LN, 24 proliferative (class III or IV with or without class V, 88.9%), and 3 pure membranous (class V, 11.1%). Two patients had end-stage LN (class VI, 4
7.4%) and because indication for repeat biopsy was persistent disease activity rather than flare, they therefore were excluded from the following analysis. We found that due to strict trial designs, on average, 67.7% of patients with LN with renal flares would have been ineligible to participate in these RCTs (range 7.4%–100%). Severe disease excluded up to 63.2% of the repeat biopsy cohort (average 19.2%), and mild disease caused ineligibility rates up to 51.9% (average 12.9%). Overall, 57.5% of patients would have been prevented from entering the 33 RCTs because of prior Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480
Table 1. An overview of 33 randomized controlled trials of lupus nephritis induction treatments published since 1970 Study design
Follow-up
n
LN classes
Treatment arms
Steinberg AD et al.9 1971
Double-blind single center (USA)
10 wk
13
Biopsy not required
G1: p.o. CYC G2: placebo
Steinberg AD et al.10 1974
Double-blind single center (USA)
10 wk
38
Diffuse proliferative (IV)
G1: p.o. CYC G2: AZA G3: placebo
Ginzler E et al.11 1976
Double-blind single center (USA)
12 mo
14
Diffuse proliferative (IV) or membranous (V)
G1: AZA G2: p.o. CYC þ AZA
Donadio JV et al.12 1978
Open label single center (USA)
Mean 43 mo
50
Diffuse proliferative (IV)
G1: p.o. CYC þ p.o. corticosteroid G2: p.o. corticosteroid
Dinant H et al.13 1982
Open label single center (USA)
Mean 42 mo
46
Biopsy not required
G1: p.o. corticosteroid G2: p.o. CYC þ AZA G3: i.v. CYC
Boumpas DT et al.14 1992
Open label single center (USA)
30 mo
65
Proliferative (III or IV) or membranous (V)
G1: i.v. corticosteroid G2: short course i.v. CYC G3: long course i.v. CYC
Lewis EJ et al.15 1992
Open label multicenter (USA)
Mean 136 wk
86
Proliferative (III or IV)
G1: p.o. corticosteroid þ p.o. CYC G2: p.o. corticosteroid þ p.o. CYC þ PE
Doria A et al.16 1994
Open label single center (Italy)
Mean 23 mo
18
Diffuse proliferative (IV)
G1: AZA G2: AZA þ PE
Sesso R et al.17 1994
Open label single center (Brazil)
Mean 15 mo
29
Biopsy not required
G1: i.v. CYC G2: i.v. corticosteroid
Gouerly MF et al.18 1996
Open label single center (USA)
Minimum 5 years
82
Proliferative (III or IV)
G1: i.v. CYC G2: i.v. corticosteroid G3: i.v. CYC þ i.v. corticosteroid
Wallace DJ et al.19 1998
Open label international
24 mo
19
Proliferative (III or IV)
G1: i.v. CYC G2: i.v. CYC þ PE
Open label single center (Hong Kong)
12 mo
42
Diffuse proliferative (IV)
G1: MMF G2: p.o. CYC
Houssiau FA et al.21 2002
Open label international (Europe)
Median 41 mo
90
Proliferative (III or IV)
G1: high dose i.v. CYC G2: low dose i.v. CYC
Yee CS et al.22 2004
Open label international (Europe)
Mean 3.5 years
32
Proliferative (III or IV)
G1: i.v. CYC G2: p.o. CYC
Ong LM et al.23 2005
Open label national (Malaysia)
6 mo
54
Proliferative (III or IV)
G1: MMF G2: i.v. CYC
Open label national (USA)
24 wk
140
Proliferative (III or IV) or membranous (V)
G1: MMF G2: i.v. CYC
Grootscholten C et al.25 2006
Open label national (Netherlands)
Median 5.7 years
87
Proliferative (III or IV)
G1: i.v. CYC G2: AZA þ i.v. corticosteroid
Bao H et al.26 2008
Open label single center (China)
9 mo
40
Diffuse proliferative and membranous (IV þ V)
G1: MMF þ Tac G2: i.v. CYC
Appel GB et al.27 2009
Open label international
24 wk
370
Proliferative (III or IV) or membranous (V)
G1: MMF G2: i.v. CYC
Austin HA et al.28 2009
Open label single center (USA)
12 mo
42
Membranous (V)
G1: p.o. corticosteroid G2: i.v. CYC G3: CsA
Open label single center (Hong Kong)
48 wk
19
Proliferative (III or IV)
G1: RTX G2: RTX þ i.v. CYC
Sabry A et al.30 2009
Open label single center (Egypt)
12 mo
46
Diffuse proliferative (IV)
G1: high dose i.v. CYC G2: low dose i.v. CYC
El-Shafey EM et al.31 2010
Open label single center (Egypt)
24 wk
47
Proliferative (III or IV)
G1: MMF G2: i.v. CYC
Zavada J et al.32 2010
Open label, international (Czech Republic, Slovakia)
18 mo
40
Proliferative (III or IV)
G1: i.v. CYC G2: CsA
Chen W et al.33 2011
Open label national (China)
6 mo
81
Proliferative (III or IV) or membranous (V)
G1: Tac G2: i.v. CYC
Open label single center (China)
24 wk
62
Proliferative (III or IV) or membranous (V)
G1: MMF G2: Tac G3: i.v. CYC
Double-blind, international (USA, Latin-America)
52 wk
144
Proliferative (III or IV)
G1: MMF G2: MMF þ RTX
Open label national (China)
24 mo
16
Membranous (V)
G1: MMF G2: Tac
Double-blind international
48 wk
381
Proliferative (III or IV)
G1: MMF or i.v. CYC G2: low dose OCR þ MMF or i.v. CYC G3: high dose OCR þ MMF or i.v. CYC
Author, Year
Chan TM et al.20 2000
Ginzler EM et al.24 2005
Li EK et al.29 2009
Li X et al.34 2011
Rovin BH et al.35 2012 Yap DY et al.36 2012 Mysler EF et al.37 2013
(Continued on next page)
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
5
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
CLINICAL RESEARCH
535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588
A Pakozdi et al.: External validity of LN RCTs
Table 1. (Continued) Study design
Follow-up
n
LN classes
Treatment arms
Double-blind international (USA, Mexico)
52 wk
137
Proliferative (III or IV)
G1: ABT þ i.v. CYC G2: i.v. CYC
Double-blind, international
52 wk
300
Proliferative (III or IV)
G1: MMF G2: low dose ABT þ MMF G3: high dose ABT þ MMF
Open label national (China)
24 wk
544
Proliferative (III or IV) or membranous (V)
G1: Tac þ MMF G2: i.v. CYC
Open label national (Hong Kong)
6 mo
150
Proliferative (III or IV) or membranous (V)
G1: MMF G2: Tac
Author, Year The ACCESS Trial group38 2014 Furie R et al.39 2014
Liu Z et al.40 2014 Mok CC et al.41 2014
ABT, abatacept; AZA, azathioprine; CsA, cyclosporin; CYC, cyclophosphamide; G, group; LN, lupus nephritis; MMF, mycophenolate mofetil; n, number of patients; OCR, ocrelizumab; PE, plasmapheresis; RTX, rituximab; Tac, Tacrolimus.
Q6
use of immunosuppressive agents, with 11 RCTs (33.3%) excluding more than 80% of our cohort.
patients with LN seen in everyday practice for inclusion into published RCTs. Our results indicate that LN RCT populations may only partially reflect real-world
DISCUSSION Current standard of care in LN induction remains suboptimal and kidney damage occurs despite using the best available therapies. The rationale behind choosing one immunosuppressive drug over another or in combination is based on knowledge gained from trials conducted in the past 5 decades. Among those, explanatory RCTs are most valuable, but can be susceptible to bias, using narrow inclusion and exclusion criteria to recruit those most likely to benefit from a drug and least likely to experience harm.4,5 RCTs must not only be internally valid but clinically useful and relevant to lupus patient populations. The present study is the first to evaluate eligibility of
Table 3. Baseline characteristics of our LN cohort
Table 2. Commonly used eligibility and exclusion criteria of the 33 RCTs
Q7
Barts LN cohort, n [ 137 Gender, n (%) Male
20 (15.6)
Female
117 (85.4)
Ethnicity, n (%) White Caucasian
22 (16.1)
Black
69 (50.4)
Asiana
46 (33.5)
Age, y, mean (SD) at SLE diagnosis
30.8 (11.3)
at LN diagnosis
33.6 (11.5)
Serology, n (%) ANA positive
131 (97.0)
ENA antibody positive
97 (75.8)
RNP antibody positive
57 (44.5)
Sm antibody positive
40 (31.3)
Ro antibody positive
54 (42.2)
aPL antibodies
30 (25.4)
Renal histology class, n (%)
n (%)
Class III LN ( class V)
41 (30.0)
Fulfillment of systemic lupus erythematosus criteria according to ARA or American College of Rheumatology
32 (97.0)
Class IV LN ( class V)
70 (51.1)
Class V LN
26 (19.0)
Predefined serum creatinine or glomerular filtration rate
25 (75.8)
Eligibility criteria
Biochemical variables, mean SD
Predefined minimal proteinuria
22 (66.7)
Proteinuria (g/24 h)
5.5 5.1
Age limitation
19 (57.6)
Serum albumin (g/L)
27.3 7.4
Active urinary sediment
12 (36.4)
Serum creatinine (mmol/L)
Predefined histopathological changes (other than class)
10 (30.3)
Range of GFR (ml/min per 1.73 m2), n (%)
159.2 166.1
Positive antinuclear antibody/lupus erythematosus test
6 (18.2)
$90
45 (32.8)
Abnormal lupus activity markers, double-stranded DNA, complements
4 (12.1)
$60 to <90
31 (22.6)
n (%)
$30 to <60
36 (26.3)
Prohibited immunosuppressive medication
25 (75.8)
$15 to <30
11 (8.0)
Pregnancy
24 (72.7)
<15
14 (10.2)
Major infection
22 (66.7)
Exclusion criteria
Lupus activity markers, n (%) Range of anti-dsDNA titer, IU/ml
Malignancy
11 (33.3)
Cytopenia
11 (33.3)
<30 (negative)
37 (30.1)
Hypersensitivity to a study drug
10 (30.3)
30 to 60 (low positive)
14 (11.4)
Diabetes
10 (30.3)
60 to 200 (positive)
15 (12.2)
>200 (strong positive)
57 (46.3)
Gastrointestinal hemorrhage, peptic ulcer disease
8 (24.2)
Liver disease
7 (21.2)
Low complement C3
78 (68.4)
Cerebral lupus
6 (18.2)
Low complement C4
82 (69.5)
Non-lupus renal disease
5 (15.2)
Unwilling to use contraception
5 (15.2)
Anticipation of poor compliance
4 (12.1)
6
ANA, antinuclear antibody; aPL, antiphospholipid; dsDNA, double-stranded DNA; ENA, extractable nuclear antigens; GFR, glomerular filtration rate; LN, lupus nephritis; n, number of patients; SLE, systemic lupus erythematosus. a Q8 South Asians (n ¼ 39) and Oriental Asians (n ¼ 7). Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
Table 4. Overview of patient cohorts in important RCTs of LN induction Barts cohort
Male
20 (15.6)
14 (16.7)
3 (7.1)
6 (7.0)
14 (10.0)
57 (15.4)
14 (9.7)
47 (15.8)
Female
117 (85.4)
68 (82.9)
39 (92.9)
84 (93.0)
126 (90.0)
313 (84.6)
130 (90.3)
251 (84.2) 111 (37.2)
Q10
Chan TM et al.20 2000
Q9
Gouerly MF et al.18 1996
Houssiau FA et al.21 2002
Ginzler EM et al.24 2005
Appel GB et al.27 2009
Rovin BH et al.35 2012
Furie et al.39 2014
Gender, n (%)
Ethnicity, n (%) White Caucasian
22 (16.1)
57 (69.5)
0
76 (84.4)
24 (17.1)
147 (39.7)
45 (31.3)
Black
69 (50.4)
18 (22.0)
0
8 (8.9)
79 (56.4)
46 (12.4)
40 (27.8)
14 (4.7)
Asian
46 (33.5)
2 (2.4)
42 (100)
6 (7)
8 (5.7)
123 (33.2)
7 (4.9)
164 (55.0)
Renal histology class, n (%) Class III LN ( V)
41 (30.0)
62 (78.5)
0
21 (23.3)
NK
58 (15.7)
49 (34.0)
78 (26.2)
Class IV LN ( V)
70 (51.1)
17 (21.5)
42 (100)
62 (68.9)
NK
252 (68.1)
95 (66.0)
220 (73.8)
Class V LN
26 (19.0)
0
0
7 (7.8)
39 (27.9)
60 (16.2)
0
0
Proteinuria (g/24 h)
5.5 5.1
4.0 0.7a
4.8 3.2b
3.04 2.4
4.2 2.4b
4.1 3.7
4.0 2.9
Serum albumin (g/L)
27.3 7.4
NK
28.0 5.5
30.3 6.1
27.5 7.6
NK
26.5 7.5
NK
159.2 166.1
102.0 9.9a
106.1 39.8b
101.7 58.3
94.6 45.1b
101.0 1.1
88.4 44.2
73.6 (35–239)c
Biochemical variables, mean SD
Serum creatinine (mmol/L)
b
b
4.4 5.5b b
LN, lupus nephritis; n, number of patients. a Combined mean standard error. b Combined mean standard deviation. c Combined median (range).
LN cohorts and caution should be exercised in extrapolating trial data to patient subgroups that have not been adequately represented in studies. Overall, 32% of our LN cohort would have been excluded from the reviewed 33 RCTs because of the severity of their renal disease and/or the use of an immunosuppressive drug at the time of diagnosis. Eligibility criteria and hence exclusion rates showed a wide range of variability among the examined trials, and reached up to 61% due to severe disease (low GFR or high serum creatinine), up to 44% due to mild disease (low-level proteinuria or preserved renal function), and up to 16% due to concomitant immunosuppressive drugs at time of LN onset. These findings suggest that RCT results are less likely to be relevant to patients with severe LN who would have never made it to the trial. Exclusion rates were even higher, on average 68%, when examining our existing LN population, with new renal flares having had a repeat kidney biopsy proving active LN. This further queries the application of RCT results to patients with severe LN and those who are nonresponders to other immunosuppressive treatments. In this study, we focused only on exclusion criteria directly related to LN: disease severity and immunosuppressive status at time of LN diagnosis. There are other factors listed in trial designs that would have further increased exclusion rates if examined. The most important ones are previous immunosuppression, pregnancy, and medical conditions associated with safety of study subjects: infections, malignancy, and comorbidities, such as cytopenia, diabetes, peptic ulcer disease, and cerebral lupus. Some of these criteria further weaken the representation of a real-life LN
population; nevertheless, there is no solution that could address the problem without jeopardizing patient safety. However, implementing less conservative exclusion criteria, particularly with regard to comorbidities, in late-stage clinical trials might improve the generalizability. Stringent eligibility criteria are likely to limit the external validity of RCTs; nevertheless, clinicians should be able to use their judgment when interpreting the results of published trials. Having said that, inadequate reporting of trials can often mislead decision making in prescribing practice. We found a large variation in the quality of reporting of LN RCTs. None of the RCTs published the “screening rate,” the number of patients who were diagnosed with LN and assessed for eligibility using inclusion/exclusion criteria. What is more disappointing is that only 4 of the 33 RCTs published “ineligibility rates,” and only 2 published the factors leading to nonparticipation. In one of the most cited and influential LN articles, in the Aspreva Lupus Management Study (ALMS) cohort, Q4 13% of screened patients with LN did not meet entry criteria due to mild clinical activity or prohibited concurrent medications.30 In our cohort, the hypothetical rate of exclusion from the ALMS trial for these reasons would have reached 29%, despite that our LN cohort seemed to have a more severe disease compared with ALMS study subjects. An explanation for the discrepancy could be the ethnic background of the participating subjects, with a higher proportion of Caucasian patients (40%) in the ALMS versus our Q5 cohort (16%), and the low proportion of black patients (12% vs. 50%, ALMS vs. our LN cohort, respectively) who generally tend to present with more severe LN.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
7
697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
CLINICAL RESEARCH
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804
A Pakozdi et al.: External validity of LN RCTs
Both gender and ethnicity are factors affecting the incidence and severity of LN, as well as treatment outcomes. There is evidence that Hispanic and AfricanAmerican individuals have a higher incidence and prevalence rates of LN and they may also show different response to treatment.43,44 Clinical studies should account for these ethnic disparities, as indicated by the National Institutes of Health Revitalization Act of 1993.45 Some issues could certainly be improved if demographic profiles would be incorporated in trial designs, and medical systems were able to address poverty, lack of education, impaired access to trials, and language barrier. Ultimately, more diverse recruitment strategies should be used when enrolling minority groups to achieve a more representative patient population, which would lead to a better generalizability of the results. The lack of transparency due to poor reporting rate can eventually cause serious damage; by preventing readers from evaluation of the validity and reliability of these trials, which may result in biased estimates of treatment effects. The CONSORT 2010 statement (Consolidated Standards of Reporting Trials) gives a clear guidance on the reporting of RCTs using an essential checklist.46 The checklist includes the reporting of exclusion rates. Since this CONSORT was published, sadly, only 1 journal article included in this analysis followed these recommendations. We recommend that editors improve the reporting quality of clinical trials for transparency by giving authors clear guidance on items to be reported. In summary, the present study was the first to evaluate how representative participants in published LN RCTs were in comparison with patients seen in usual clinical practice from a large tertiary lupus center. A third of newly diagnosed real-world patients with LN would have been excluded from LN RCTs due to strict eligibility criteria solely related to the severity of their renal disease. Although some trials have excellent validity, many do not; and factors determining their validity are rarely reported, preventing the clinician from using thoughtful judgment. Specifically, we highlight the need for more pragmatic trials designed for those with more severe LN.
3. Kennedy-Martin T, Curtis S, Faries D, et al. A literature review on the representativeness of randomized controlled trial samples and implications for the external validity of trial results. Trials. 2015;16:495. 4. Jadad AR, Enkin M, Jadad AR. Randomized Controlled Trials: Questions, Answers, and Musings. 2nd ed. Malden, MA: Blackwell Pub; 2007. 5. Roland M, Torgerson DJ. What are pragmatic trials? BMJ. 1998;316:285. 6. Churg J, Sobin LH. Renal Disease: Classification and Atlas of Glomerular Disease. Tokyo, Japan: Igaku-Shoin; 1982. 7. Weening JJ, D’Agati VD, Schwartz MM, et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int. 2004;65:521–530. 8. Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1982;25:1271–1277. 9. Steinberg AD, Kaltreider HB, Staples PJ, et al. Cyclophosphamide in lupus nephritis:a controlled trial. Ann Intern Med. 1971;75:165–171. 10. Steinberg AD, Decker JL. A double-blind controlled trial comparing cyclophosphamide, azathioprine and placebo in the treatment of lupus glomerulonephritis. Arthritis Rheum. 1974;17:923–937. 11. Ginzler E, Diamond H, Guttadauria M, et al. Prednisone and azathioprine compared to prednisone plus low-dose azathioprine and cyclophosphamide in the treatment of diffuse lupus nephritis. Arthritis Rheum. 1976;19:693–699. 12. Donadio JV Jr., Holley KE, Ferguson RH, et al. Treatment of diffuse proliferative lupus nephritis with prednisone and combined prednisone and cyclophosphamide. N Engl J Med. 1978;299:1151–1155. 13. Dinant HJ, Decker JL, Klippel JH, et al. Alternative modes of cyclophosphamide and azathioprine therapy in lupus nephritis. Ann Intern Med. 1982;96:728–736. 14. Boumpas DT, Austin HA 3rd, Vaughn EM, et al. Controlled trial of pulse methylprednisolone versus two regimens of pulse cyclophosphamide in severe lupus nephritis. Lancet. 1992;340:741–745. 15. Lewis EJ, Hunsicker LG, Lan SP, et al. A controlled trial of plasmapheresis therapy in severe lupus nephritis. The Lupus Nephritis Collaborative Study Group. N Engl J Med. 1992;326: 1373–1379. 16. Doria A, Piccoli A, Vesco P, et al. Therapy of lupus nephritis. A two-year prospective study. Ann Med Intern (Paris). 1994;145: 307–311. 17. Sesso R, Monteiro M, Sato E, et al. A controlled trial of pulse cyclophosphamide versus pulse methylprednisolone in severe lupus nephritis. Lupus. 1994;3:107–112.
DISCLOSURES All the authors declared no competing interests.
REFERENCES 1. Riegelman RK. Studying a Study and Testing a Test: How to Read the Medical Evidence. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2000. 2. Nallamothu BK, Hayward RA, Bates ER. Beyond the randomized clinical trial: the role of effectiveness studies in 8
evaluating cardiovascular therapies. Circulation. 2008;118: 1294–1303.
18. Gourley MF, Austin HA 3rd, Scott D, et al. Methylprednisolone and cyclophosphamide, alone or in combination, in patients with lupus nephritis. A randomized, controlled trial. Ann Intern Med. 1996;125:549–557. 19. Wallace DJ, Goldfinger D, Pepkowitz SH, et al. Randomized controlled trial of pulse/synchronization cyclophosphamide/ apheresis for proliferative lupus nephritis. J Clin Apher. 1998;13:163–166. Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
20. Chan TM, Li FK, Tang CS, et al. Efficacy of mycophenolate mofetil in patients with diffuse proliferative lupus nephritis. Hong Kong-Guangzhou Nephrology Study Group. N Engl J Med. 2000;343:1156–1162. 21. Houssiau FA, Vasconcelos C, D’Cruz D, et al. Immunosuppressive therapy in lupus nephritis: the Euro-Lupus Nephritis Trial, a randomized trial of low-dose versus high-dose intravenous cyclophosphamide. Arthritis Rheum. 2002;46:2121–2131. 22. Yee CS, Gordon C, Dostal C, et al. EULAR randomised controlled trial of pulse cyclophosphamide and methylprednisolone versus continuous cyclophosphamide and prednisolone followed by azathioprine and prednisolone in lupus nephritis. Ann Rheum Dis. 2004;63:525–529. 23. Ong LM, Hooi LS, Lim TO, et al. Randomized controlled trial of pulse intravenous cyclophosphamide versus mycophenolate mofetil in the induction therapy of proliferative lupus nephritis. Nephrology. 2005;10:504–510. 24. Ginzler EM, Dooley MA, Aranow C, et al. Mycophenolate mofetil or intravenous cyclophosphamide for lupus nephritis. N Engl J Med. 2005;353:2219–2228. 25. Grootscholten C, Ligtenberg G, Hagen EC, et al. Azathioprine/ methylprednisolone versus cyclophosphamide in proliferative lupus nephritis. A randomized controlled trial. Kidney Int. 2006;70:732–742. 26. Bao H, Liu ZH, Xie HL, et al. Successful treatment of class VþIV lupus nephritis with multitarget therapy. J Am Soc Nephrol. 2008;19:2001–2010. 27. Appel GB, Contreras G, Dooley MA, et al. Mycophenolate mofetil versus cyclophosphamide for induction treatment of lupus nephritis. J Am Soc Nephrol. 2009;20:1103–1112. 28. Austin HA 3rd, Illei GG, Braun MJ, et al. Randomized, controlled trial of prednisone, cyclophosphamide, and cyclosporine in lupus membranous nephropathy. J Am Soc Nephrol. 2009;20:901–911. 29. Li EK, Tam LS, Zhu TY, et al. Is combination rituximab with cyclophosphamide better than rituximab alone in the treatment of lupus nephritis? Rheumatology. 2009;48:892–898. 30. Sabry A, Abo-Zenah H, Medhat T, et al. A comparative study of two intensified pulse cyclophosphamide remissioninducing regimens for diffuse proliferative lupus nephritis: an Egyptian experience. Int Urol Nephrol. 2009;41:153–161. 31. El-Shafey EM, Abdou SH, Shareef MM. Is mycophenolate mofetil superior to pulse intravenous cyclophosphamide for induction therapy of proliferative lupus nephritis in Egyptian patients? Clin Exp Nephrol. 2010;14:214–221. 32. Zavada J, Pesickova S, Rysava R, et al. Cyclosporine A or intravenous cyclophosphamide for lupus nephritis: the Cyclofa-Lune study. Lupus. 2010;19:1281–1289. 33. Chen W, Tang X, Liu Q, et al. Short-term outcomes of induction therapy with tacrolimus versus cyclophosphamide
for active lupus nephritis: a multicenter randomized clinical trial. Am J Kidney Dis. 2011;57:235–244. 34. Li X, Ren H, Zhang Q, et al. Mycophenolate mofetil or tacrolimus compared with intravenous cyclophosphamide in the induction treatment for active lupus nephritis. Nephrol Dial Transplant. 2012;27:1467–1472. 35. Rovin BH, Furie R, Latinis K, et al. Efficacy and safety of rituximab in patients with active proliferative lupus nephritis: the Lupus Nephritis Assessment with Rituximab study. Arthritis Rheum. 2012;64:1215–1226. 36. Yap DY, Yu X, Chen XM, et al. Pilot 24 month study to compare mycophenolate mofetil and tacrolimus in the treatment of membranous lupus nephritis with nephrotic syndrome. Nephrology. 2012;17:352–357. 37. Mysler EF, Spindler AJ, Guzman R, et al. Efficacy and safety of ocrelizumab in active proliferative lupus nephritis:results from a randomized, double-blind, phase III study. Arthritis Rheum. 2013;65:2368–2379. 38. The ACCESS Trial group. Treatment of lupus nephritis with abatacept: the Abatacept and Cyclophosphamide Combination Efficacy and Safety Study. Arthritis Rheum. 2014;66: 3096–3104. 39. Furie R, Nicholls K, Cheng TT, et al. Efficacy and safety of abatacept in lupus nephritis:a twelve-month, randomized, double-blind study. Arthritis Rheum. 2014;66:379–389. 40. Liu Z, Zhang H, Liu Z, et al. Multitarget therapy for induction treatment of lupus nephritis:a randomized trial. Ann Intern Med. 2015;162:18–26. 41. Mok CC, Ying KY, Yim CW, et al. Tacrolimus versus mycophenolate mofetil for induction therapy of lupus nephritis: a randomised controlled trial and long-term follow-up. Ann Rheum Dis. 2016;75:30–36. 42. Corapi KM, Dooley MA, Pendergraft WF 3rd. Comparison and evaluation of lupus nephritis response criteria in lupus activity indices and clinical trials. Arthritis Res Ther. 2015;17:110. 43. Contreras G, Lenz O, Pardo V, et al. Outcomes in African Americans and Hispanics with lupus nephritis. Kidney Int. 2006;69:1846–1851. 44. Isenberg D, Appel GB, Contreras G, et al. Influence of race/ ethnicity on response to lupus nephritis treatment:the ALMS study. Rheumatology. 2010;49:128–140. 45. Freedman LS, Simon R, Foulkes MA, et al. Inclusion of women and minorities in clinical trials and the NIH Revitalization Act of 1993–the perspective of NIH clinical trialists. Control Clin Trials. 1995;16:277–285. discussion 286–279, 293–309. 46. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration:updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c869.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
9
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Systematic Review and the External Validity of Randomized Controlled Trials in Lupus Nephritis Angela Pakozdi1, Ravindra Rajakariar2, Debasish Pyne1, Andrea Cove-Smith2 and Muhammad Magdi Yaqoob2
Q12
1
Department of Rheumatology, Barts Health NHS Trust, London, United Kingdom; and 2Department of Nephrology, Barts Health NHS Trust, London, United Kingdom
Introduction: Randomized controlled trials (RCTs) are considered the gold standard for assessing treatment efficacy. However, sampling bias can affect the generalization of results to routine clinical practice. Here we assessed whether patients with lupus nephritis (LN) seen in routine clinical practice would have satisfied entry criteria to the major published RCTs in LN.
Q1
Methods: A systematic literature search from January 1974 to May 2015 was carried out, identifying all RCTs investigating LN induction treatment. Patients diagnosed with proliferative or membranous LN between 1995 and 2013 were identified from the Barts Lupus Centre database; baseline characteristics were compared with each RCT’s entry criteria to assess hypothetical inclusion or exclusion. Results: Of 363 articles, 33 RCTs met inclusion criteria. Of 137 patients newly diagnosed with LN (111 with proliferative/mixed proliferative and 26 with pure membranous LN), 32% would have been excluded from RCT entry (range 8%–73%). The main reasons for exclusion would have been too severe disease, too mild disease, or prior immunosuppressant use, which were exclusion criteria in 26, 20, and 22 RCTs, respectively. A total of 27 patients with LN (20%) were re-biopsied due to flare; 68% of these would have been ineligible to enter RCTs. Conclusion: Published RCTs do not truly reflect the heterogeneity of patients with LN in routine practice at our lupus center. The external validity of RCTs could be improved by including more representative patient cohorts. RCTs should be used as a guide but consideration should be given to similarities between individual patients and the characteristics of the trial cohorts before treatment decisions being made. Kidney Int Rep (2018) -, -–-; https://doi.org/10.1016/j.ekir.2017.11.005 KEYWORDS: glomerulonephritis; lupus ª 2017 International Society of Nephrology. Published by Elsevier Inc. This is an open access article under the CC BYNC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Q2
remendous advances in immunosuppressive treatment of lupus nephritis (LN) have been made over the past 4 decades. Fortunately, most treatment decisions made in routine clinical practice are now based on results from well-designed clinical trials rather than anecdotal evidence. Randomized controlled trials (RCTs) are considered the “gold standard” to assess treatment efficacy,1,2 as they have the best potential to minimize bias through allocating participants to study arms in a random fashion. Despite their rigorous design, the applicability of results to individual patients has been debated thanks to other types of
T
Correspondence: Muhammad Magdi Yaqoob, Barts Health NHS Trust, Department of Nephrology, Royal London Hospital, Whitechapel Road, London E1 1BB, United Kingdom. E-mail: m.m. [email protected] Received 11 October 2017; accepted 6 November 2017; published online
bias, some of which lead to RCTs not adequately reflecting the heterogeneity of patients in routine clinical practice.3 Examples include “selection bias” due to poor allocation concealment, “sampling bias” when physicians consider only a certain type of patient for trials and exclude patients on the basis of the lack of personal equipoise, or “literacy bias” for those who fail to understand the consent form.4 In LN RCTs, a particular problem may be “severity of illness bias.”4 To achieve a more homogeneous study population in whom the study drug is most likely to work and less likely to harm, patients must meet a range of prespecified entry requirements.5 Inclusion criteria to LN RCTs often specify the target range of biochemical variables and serological activity markers, and the histological evidence of LN with restriction to 1 or 2 particular classes of glomerulonephritis. Moreover, exclusion criteria intentionally eliminate patients with mild or, more commonly, severe disease; those with
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
1
52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
CLINICAL RESEARCH
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
A Pakozdi et al.: External validity of LN RCTs
recent major infections or with preexisting comorbidities; and tend to prohibit certain immunosuppressive drugs before enrollment. As a consequence, patients recruited to most LN RCTs have a relatively moderate disease activity excluding those with too mild or severe disease and those developing active LN while on immunosuppressive therapies. In this study, we aimed to evaluate the external validity of published LN RCTs. We assessed whether a multiethnic cohort of patients with lupus seen in routine clinical practice at a single center would have been represented adequately by assessing RCT inclusion and exclusion criteria. METHODS Patients and Study Design This study was conducted at Barts Health NHS Trust, the largest tertiary teaching hospital in the United Kingdom. Patients with biopsy-proven proliferative (class III or IV) or membranous (class V) LN diagnosed between 1995 and 2013 were included. LN classes based on glomerular pathology were defined according to the International Society of Nephrology/Renal Pathology Society 2003 classification.6,7 All patients fulfilled at least 4 of the American College of Rheumatology revised classification criteria for systemic lupus erythematosus.8 Clinical data were collected retrospectively from electronic patient records; baseline variables included serum creatinine, glomerular filtration rate (GFR), serum albumin, proteinuria, doublestranded DNA, and complement C3 and C4 levels. Baseline immunosuppressive drugs at time of LN onset were recorded. Identification of Eligible Clinical Studies We selected RCTs that aimed to measure the efficacy of immunosuppressive agents in proliferative (class III, IV) and/or membranous (class V) LN induction treatment. Literature searches were performed using PubMed, Medline, and EMBASE databases (from 1974 to May 2015) and the Cochrane Controlled Trial Register (up to May 2015) using the OVID search engine, to identify relevant studies. Key search words included “lupus nephritis” and “randomized” and “lupus nephritis” and “clinical trial.” Additional studies were sought through bibliographic notations. Studies published in full-text English literature were included if they met the following criteria: (i) prospective controlled trial with treatment allocation by random assignment; (ii) compared at least 2 arms for LN induction treatment; (iii) involved proliferative and/or membranous LN patient population; and (iv) included adult patients. We excluded trials that published data 2
on extended follow-up of already published RCTs, studies of maintenance therapies, or refractory LN. Evaluation of Hypothetical Noninclusion Rate to RCTs Baseline characteristics of patients in the Barts LN cohort were compared with patients participating in some of the most important RCTs of LN induction treatment from the past 2 decades. For the purpose of this study, factors leading to exclusion included severe renal impairment with predefined GFR or serum creatinine in the selection criteria of the RCT; mild disease with predefined levels of proteinuria, serum albumin, GFR, or serum creatinine; or prohibited baseline immunosuppressive drugs used at time of LN onset. We did not analyze any other exclusion criteria, including confounding factors not related to their renal disease. When trials involved patients with proliferative glomerulonephritis (class III or IV V), we evaluated only the patients with proliferative LN from our cohort (n ¼ 111); where trials recruited patients with membranous LN, we evaluated only our membranous LN cohort (n ¼ 26), whereas our entire cohort was compared (n ¼ 137) when both classes were represented in the RCT. RESULTS Characteristics of RCTs Included in the Analysis Our systematic literature search yielded 363 articles, of which 309 articles were deemed unsuitable after title or abstract review (Figure 1). Of the remaining, 5 trials on LN induction treatment had to be excluded, as the study design was insufficiently described and no comment was made on inclusion and exclusion criteria used. The final analysis composed of 33 RCTs is shown in Table 1.9–41 Of the 33 RCTs, 19 journal articles (58%) failed to report the methods of randomization and/or the protocols used to assign participants to comparison groups. Furthermore, only 4 RCTs27,31,39,40 published the rate of screening failure, ranging from 19.6% to 39.4%. There were only 2 of 33 RCTs elaborating the reasons for ineligibility.27,31 The most frequently prespecified inclusion and exclusion criteria encountered in the selected RCTs are detailed in Table 2. Nearly all RCTs (97%) required fulfillment of systemic lupus erythematosus classification criteria, but a positive antinuclear antibody/lupus erythematosus test was rarely desired (18%). Three trials (9%) did not require biopsy evidence of LN; 75% of RCTs required patients to have a renal function that met predefined values, and 67% of studies demanded a Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
CLINICAL RESEARCH
Records idenfied through database searching (n = 352)
Addional records idenfied through other sources (n = 11)
Screening
Records screened (n = 106)
Eligibility
Full-text arcles assessed for eligibility (n = 54)
Included
Records aer duplicates removed (n = 106)
web 4C=FPO
211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
Idenficaon
A Pakozdi et al.: External validity of LN RCTs
Studies included in qualitave synthesis (n = 33)
Records excluded (n = 52)
Full-text arcles excluded, with reasons (n = 21) Not on inducon treatment (n = 7) No randomizaon (n = 2) No explanaon of methodology (n = 5) Extended follow-up of published trial (n = 4) Terminated trial without efficacy analysis (n = 1) Other language than English (n = 2)
Figure 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram.
minimal level proteinuria. Prohibited concomitant or prior immunosuppressive treatment was defined in 76% of RCTs. Characteristics of Barts LN Patient Cohort From our database, we identified 137 patients with a new diagnosis of biopsy-proven active LN; their baseline characteristics are shown in Table 3. Eightyone percent of our LN cohort had proliferative (n ¼ 111, 30% class III, 51% class IV, with or without membranous component), and 19% had pure membranous LN (n ¼ 26). Overall, patients with LN in our cohort tended to have more severe renal disease compared with patients in landmark RCTs of LN induction (Table 4)42; they had a mean proteinuria of 5.5 g/24 hours ( 5.1 SD), mean serum albumin of 27.3 g/L (7.4 SD), and mean serum creatinine of 159 mmol/L ( 166 SD). Although 33% (n ¼ 45) had a normal GFR (>90 ml/min per 1.73 m2), 18% (n ¼ 25) had a GFR <30 ml/min per 1.73 m2. Serological evidence of lupus activity was present in half of the cohort with strongly positive double-stranded DNA and low complements. Twenty-three patients with systemic lupus
erythematosus (16.7%) were taking immunosuppressants other than hydroxychloroquine or oral prednisolone (either mycophenolate mofetil or azathioprine) at time of LN diagnosis. Hypothetical Noninclusion Rate of Newly Diagnosed Patients With LN to LN Induction RCTs On average, one-third (31.8%) of our newly diagnosed biopsy-proven active LN cohort would have been Q3 ineligible to enter the RCTs (range 8.1%–72.9%) (Figure 1). These patients would have been excluded because of factors directly related to their renal disease or the concomitant use of immunosuppressive treatment. Severe renal impairment was a prespecified exclusion criterion in 26 RCTs (79%) leading to ineligibility in up to 61.4% of our LN cohort. Twenty RCTs (63%) would have left out patients with milder disease based on mildly impaired renal function or lower level of proteinuria, resulting in a hypothetical exclusion rate reaching up to 44.1%. Additionally, 22 RCTs (67%) prohibited the use of various immunosuppressive drugs resulting in exclusion rates of 16.1% of our LN cohort.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
3
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
CLINICAL RESEARCH
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372
A Pakozdi et al.: External validity of LN RCTs
Figure 2. Noninclusion rates to landmark LN RCTs. The proportion of our patients with LN that would not have been included in RCTs is shown. Q11 Overall on average 32% of our patients with LN would have not been eligible to participate in the RCTs listed (range 8%–73%).
Hypothetical Noninclusion Rate of Relapsed Patients With LN to LN Induction RCTs Next, we analyzed hypothetical exclusion rates of relapsed patients with LN with active LN evidenced by a repeat kidney biopsy. Twenty-nine of 137 patients with LN (21.2%) had a repeat kidney biopsy. The median time to repeat biopsy was 28 months (interquartile range 13–47 months). Of those, 27 had active LN, 24 proliferative (class III or IV with or without class V, 88.9%), and 3 pure membranous (class V, 11.1%). Two patients had end-stage LN (class VI, 4
7.4%) and because indication for repeat biopsy was persistent disease activity rather than flare, they therefore were excluded from the following analysis. We found that due to strict trial designs, on average, 67.7% of patients with LN with renal flares would have been ineligible to participate in these RCTs (range 7.4%–100%). Severe disease excluded up to 63.2% of the repeat biopsy cohort (average 19.2%), and mild disease caused ineligibility rates up to 51.9% (average 12.9%). Overall, 57.5% of patients would have been prevented from entering the 33 RCTs because of prior Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480
Table 1. An overview of 33 randomized controlled trials of lupus nephritis induction treatments published since 1970 Study design
Follow-up
n
LN classes
Treatment arms
Steinberg AD et al.9 1971
Double-blind single center (USA)
10 wk
13
Biopsy not required
G1: p.o. CYC G2: placebo
Steinberg AD et al.10 1974
Double-blind single center (USA)
10 wk
38
Diffuse proliferative (IV)
G1: p.o. CYC G2: AZA G3: placebo
Ginzler E et al.11 1976
Double-blind single center (USA)
12 mo
14
Diffuse proliferative (IV) or membranous (V)
G1: AZA G2: p.o. CYC þ AZA
Donadio JV et al.12 1978
Open label single center (USA)
Mean 43 mo
50
Diffuse proliferative (IV)
G1: p.o. CYC þ p.o. corticosteroid G2: p.o. corticosteroid
Dinant H et al.13 1982
Open label single center (USA)
Mean 42 mo
46
Biopsy not required
G1: p.o. corticosteroid G2: p.o. CYC þ AZA G3: i.v. CYC
Boumpas DT et al.14 1992
Open label single center (USA)
30 mo
65
Proliferative (III or IV) or membranous (V)
G1: i.v. corticosteroid G2: short course i.v. CYC G3: long course i.v. CYC
Lewis EJ et al.15 1992
Open label multicenter (USA)
Mean 136 wk
86
Proliferative (III or IV)
G1: p.o. corticosteroid þ p.o. CYC G2: p.o. corticosteroid þ p.o. CYC þ PE
Doria A et al.16 1994
Open label single center (Italy)
Mean 23 mo
18
Diffuse proliferative (IV)
G1: AZA G2: AZA þ PE
Sesso R et al.17 1994
Open label single center (Brazil)
Mean 15 mo
29
Biopsy not required
G1: i.v. CYC G2: i.v. corticosteroid
Gouerly MF et al.18 1996
Open label single center (USA)
Minimum 5 years
82
Proliferative (III or IV)
G1: i.v. CYC G2: i.v. corticosteroid G3: i.v. CYC þ i.v. corticosteroid
Wallace DJ et al.19 1998
Open label international
24 mo
19
Proliferative (III or IV)
G1: i.v. CYC G2: i.v. CYC þ PE
Open label single center (Hong Kong)
12 mo
42
Diffuse proliferative (IV)
G1: MMF G2: p.o. CYC
Houssiau FA et al.21 2002
Open label international (Europe)
Median 41 mo
90
Proliferative (III or IV)
G1: high dose i.v. CYC G2: low dose i.v. CYC
Yee CS et al.22 2004
Open label international (Europe)
Mean 3.5 years
32
Proliferative (III or IV)
G1: i.v. CYC G2: p.o. CYC
Ong LM et al.23 2005
Open label national (Malaysia)
6 mo
54
Proliferative (III or IV)
G1: MMF G2: i.v. CYC
Open label national (USA)
24 wk
140
Proliferative (III or IV) or membranous (V)
G1: MMF G2: i.v. CYC
Grootscholten C et al.25 2006
Open label national (Netherlands)
Median 5.7 years
87
Proliferative (III or IV)
G1: i.v. CYC G2: AZA þ i.v. corticosteroid
Bao H et al.26 2008
Open label single center (China)
9 mo
40
Diffuse proliferative and membranous (IV þ V)
G1: MMF þ Tac G2: i.v. CYC
Appel GB et al.27 2009
Open label international
24 wk
370
Proliferative (III or IV) or membranous (V)
G1: MMF G2: i.v. CYC
Austin HA et al.28 2009
Open label single center (USA)
12 mo
42
Membranous (V)
G1: p.o. corticosteroid G2: i.v. CYC G3: CsA
Open label single center (Hong Kong)
48 wk
19
Proliferative (III or IV)
G1: RTX G2: RTX þ i.v. CYC
Sabry A et al.30 2009
Open label single center (Egypt)
12 mo
46
Diffuse proliferative (IV)
G1: high dose i.v. CYC G2: low dose i.v. CYC
El-Shafey EM et al.31 2010
Open label single center (Egypt)
24 wk
47
Proliferative (III or IV)
G1: MMF G2: i.v. CYC
Zavada J et al.32 2010
Open label, international (Czech Republic, Slovakia)
18 mo
40
Proliferative (III or IV)
G1: i.v. CYC G2: CsA
Chen W et al.33 2011
Open label national (China)
6 mo
81
Proliferative (III or IV) or membranous (V)
G1: Tac G2: i.v. CYC
Open label single center (China)
24 wk
62
Proliferative (III or IV) or membranous (V)
G1: MMF G2: Tac G3: i.v. CYC
Double-blind, international (USA, Latin-America)
52 wk
144
Proliferative (III or IV)
G1: MMF G2: MMF þ RTX
Open label national (China)
24 mo
16
Membranous (V)
G1: MMF G2: Tac
Double-blind international
48 wk
381
Proliferative (III or IV)
G1: MMF or i.v. CYC G2: low dose OCR þ MMF or i.v. CYC G3: high dose OCR þ MMF or i.v. CYC
Author, Year
Chan TM et al.20 2000
Ginzler EM et al.24 2005
Li EK et al.29 2009
Li X et al.34 2011
Rovin BH et al.35 2012 Yap DY et al.36 2012 Mysler EF et al.37 2013
(Continued on next page)
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
5
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
CLINICAL RESEARCH
535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588
A Pakozdi et al.: External validity of LN RCTs
Table 1. (Continued) Study design
Follow-up
n
LN classes
Treatment arms
Double-blind international (USA, Mexico)
52 wk
137
Proliferative (III or IV)
G1: ABT þ i.v. CYC G2: i.v. CYC
Double-blind, international
52 wk
300
Proliferative (III or IV)
G1: MMF G2: low dose ABT þ MMF G3: high dose ABT þ MMF
Open label national (China)
24 wk
544
Proliferative (III or IV) or membranous (V)
G1: Tac þ MMF G2: i.v. CYC
Open label national (Hong Kong)
6 mo
150
Proliferative (III or IV) or membranous (V)
G1: MMF G2: Tac
Author, Year The ACCESS Trial group38 2014 Furie R et al.39 2014
Liu Z et al.40 2014 Mok CC et al.41 2014
ABT, abatacept; AZA, azathioprine; CsA, cyclosporin; CYC, cyclophosphamide; G, group; LN, lupus nephritis; MMF, mycophenolate mofetil; n, number of patients; OCR, ocrelizumab; PE, plasmapheresis; RTX, rituximab; Tac, Tacrolimus.
Q6
use of immunosuppressive agents, with 11 RCTs (33.3%) excluding more than 80% of our cohort.
patients with LN seen in everyday practice for inclusion into published RCTs. Our results indicate that LN RCT populations may only partially reflect real-world
DISCUSSION Current standard of care in LN induction remains suboptimal and kidney damage occurs despite using the best available therapies. The rationale behind choosing one immunosuppressive drug over another or in combination is based on knowledge gained from trials conducted in the past 5 decades. Among those, explanatory RCTs are most valuable, but can be susceptible to bias, using narrow inclusion and exclusion criteria to recruit those most likely to benefit from a drug and least likely to experience harm.4,5 RCTs must not only be internally valid but clinically useful and relevant to lupus patient populations. The present study is the first to evaluate eligibility of
Table 3. Baseline characteristics of our LN cohort
Table 2. Commonly used eligibility and exclusion criteria of the 33 RCTs
Q7
Barts LN cohort, n [ 137 Gender, n (%) Male
20 (15.6)
Female
117 (85.4)
Ethnicity, n (%) White Caucasian
22 (16.1)
Black
69 (50.4)
Asiana
46 (33.5)
Age, y, mean (SD) at SLE diagnosis
30.8 (11.3)
at LN diagnosis
33.6 (11.5)
Serology, n (%) ANA positive
131 (97.0)
ENA antibody positive
97 (75.8)
RNP antibody positive
57 (44.5)
Sm antibody positive
40 (31.3)
Ro antibody positive
54 (42.2)
aPL antibodies
30 (25.4)
Renal histology class, n (%)
n (%)
Class III LN ( class V)
41 (30.0)
Fulfillment of systemic lupus erythematosus criteria according to ARA or American College of Rheumatology
32 (97.0)
Class IV LN ( class V)
70 (51.1)
Class V LN
26 (19.0)
Predefined serum creatinine or glomerular filtration rate
25 (75.8)
Eligibility criteria
Biochemical variables, mean SD
Predefined minimal proteinuria
22 (66.7)
Proteinuria (g/24 h)
5.5 5.1
Age limitation
19 (57.6)
Serum albumin (g/L)
27.3 7.4
Active urinary sediment
12 (36.4)
Serum creatinine (mmol/L)
Predefined histopathological changes (other than class)
10 (30.3)
Range of GFR (ml/min per 1.73 m2), n (%)
159.2 166.1
Positive antinuclear antibody/lupus erythematosus test
6 (18.2)
$90
45 (32.8)
Abnormal lupus activity markers, double-stranded DNA, complements
4 (12.1)
$60 to <90
31 (22.6)
n (%)
$30 to <60
36 (26.3)
Prohibited immunosuppressive medication
25 (75.8)
$15 to <30
11 (8.0)
Pregnancy
24 (72.7)
<15
14 (10.2)
Major infection
22 (66.7)
Exclusion criteria
Lupus activity markers, n (%) Range of anti-dsDNA titer, IU/ml
Malignancy
11 (33.3)
Cytopenia
11 (33.3)
<30 (negative)
37 (30.1)
Hypersensitivity to a study drug
10 (30.3)
30 to 60 (low positive)
14 (11.4)
Diabetes
10 (30.3)
60 to 200 (positive)
15 (12.2)
>200 (strong positive)
57 (46.3)
Gastrointestinal hemorrhage, peptic ulcer disease
8 (24.2)
Liver disease
7 (21.2)
Low complement C3
78 (68.4)
Cerebral lupus
6 (18.2)
Low complement C4
82 (69.5)
Non-lupus renal disease
5 (15.2)
Unwilling to use contraception
5 (15.2)
Anticipation of poor compliance
4 (12.1)
6
ANA, antinuclear antibody; aPL, antiphospholipid; dsDNA, double-stranded DNA; ENA, extractable nuclear antigens; GFR, glomerular filtration rate; LN, lupus nephritis; n, number of patients; SLE, systemic lupus erythematosus. a Q8 South Asians (n ¼ 39) and Oriental Asians (n ¼ 7). Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
Table 4. Overview of patient cohorts in important RCTs of LN induction Barts cohort
Male
20 (15.6)
14 (16.7)
3 (7.1)
6 (7.0)
14 (10.0)
57 (15.4)
14 (9.7)
47 (15.8)
Female
117 (85.4)
68 (82.9)
39 (92.9)
84 (93.0)
126 (90.0)
313 (84.6)
130 (90.3)
251 (84.2) 111 (37.2)
Q10
Chan TM et al.20 2000
Q9
Gouerly MF et al.18 1996
Houssiau FA et al.21 2002
Ginzler EM et al.24 2005
Appel GB et al.27 2009
Rovin BH et al.35 2012
Furie et al.39 2014
Gender, n (%)
Ethnicity, n (%) White Caucasian
22 (16.1)
57 (69.5)
0
76 (84.4)
24 (17.1)
147 (39.7)
45 (31.3)
Black
69 (50.4)
18 (22.0)
0
8 (8.9)
79 (56.4)
46 (12.4)
40 (27.8)
14 (4.7)
Asian
46 (33.5)
2 (2.4)
42 (100)
6 (7)
8 (5.7)
123 (33.2)
7 (4.9)
164 (55.0)
Renal histology class, n (%) Class III LN ( V)
41 (30.0)
62 (78.5)
0
21 (23.3)
NK
58 (15.7)
49 (34.0)
78 (26.2)
Class IV LN ( V)
70 (51.1)
17 (21.5)
42 (100)
62 (68.9)
NK
252 (68.1)
95 (66.0)
220 (73.8)
Class V LN
26 (19.0)
0
0
7 (7.8)
39 (27.9)
60 (16.2)
0
0
Proteinuria (g/24 h)
5.5 5.1
4.0 0.7a
4.8 3.2b
3.04 2.4
4.2 2.4b
4.1 3.7
4.0 2.9
Serum albumin (g/L)
27.3 7.4
NK
28.0 5.5
30.3 6.1
27.5 7.6
NK
26.5 7.5
NK
159.2 166.1
102.0 9.9a
106.1 39.8b
101.7 58.3
94.6 45.1b
101.0 1.1
88.4 44.2
73.6 (35–239)c
Biochemical variables, mean SD
Serum creatinine (mmol/L)
b
b
4.4 5.5b b
LN, lupus nephritis; n, number of patients. a Combined mean standard error. b Combined mean standard deviation. c Combined median (range).
LN cohorts and caution should be exercised in extrapolating trial data to patient subgroups that have not been adequately represented in studies. Overall, 32% of our LN cohort would have been excluded from the reviewed 33 RCTs because of the severity of their renal disease and/or the use of an immunosuppressive drug at the time of diagnosis. Eligibility criteria and hence exclusion rates showed a wide range of variability among the examined trials, and reached up to 61% due to severe disease (low GFR or high serum creatinine), up to 44% due to mild disease (low-level proteinuria or preserved renal function), and up to 16% due to concomitant immunosuppressive drugs at time of LN onset. These findings suggest that RCT results are less likely to be relevant to patients with severe LN who would have never made it to the trial. Exclusion rates were even higher, on average 68%, when examining our existing LN population, with new renal flares having had a repeat kidney biopsy proving active LN. This further queries the application of RCT results to patients with severe LN and those who are nonresponders to other immunosuppressive treatments. In this study, we focused only on exclusion criteria directly related to LN: disease severity and immunosuppressive status at time of LN diagnosis. There are other factors listed in trial designs that would have further increased exclusion rates if examined. The most important ones are previous immunosuppression, pregnancy, and medical conditions associated with safety of study subjects: infections, malignancy, and comorbidities, such as cytopenia, diabetes, peptic ulcer disease, and cerebral lupus. Some of these criteria further weaken the representation of a real-life LN
population; nevertheless, there is no solution that could address the problem without jeopardizing patient safety. However, implementing less conservative exclusion criteria, particularly with regard to comorbidities, in late-stage clinical trials might improve the generalizability. Stringent eligibility criteria are likely to limit the external validity of RCTs; nevertheless, clinicians should be able to use their judgment when interpreting the results of published trials. Having said that, inadequate reporting of trials can often mislead decision making in prescribing practice. We found a large variation in the quality of reporting of LN RCTs. None of the RCTs published the “screening rate,” the number of patients who were diagnosed with LN and assessed for eligibility using inclusion/exclusion criteria. What is more disappointing is that only 4 of the 33 RCTs published “ineligibility rates,” and only 2 published the factors leading to nonparticipation. In one of the most cited and influential LN articles, in the Aspreva Lupus Management Study (ALMS) cohort, Q4 13% of screened patients with LN did not meet entry criteria due to mild clinical activity or prohibited concurrent medications.30 In our cohort, the hypothetical rate of exclusion from the ALMS trial for these reasons would have reached 29%, despite that our LN cohort seemed to have a more severe disease compared with ALMS study subjects. An explanation for the discrepancy could be the ethnic background of the participating subjects, with a higher proportion of Caucasian patients (40%) in the ALMS versus our Q5 cohort (16%), and the low proportion of black patients (12% vs. 50%, ALMS vs. our LN cohort, respectively) who generally tend to present with more severe LN.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
7
697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
CLINICAL RESEARCH
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804
A Pakozdi et al.: External validity of LN RCTs
Both gender and ethnicity are factors affecting the incidence and severity of LN, as well as treatment outcomes. There is evidence that Hispanic and AfricanAmerican individuals have a higher incidence and prevalence rates of LN and they may also show different response to treatment.43,44 Clinical studies should account for these ethnic disparities, as indicated by the National Institutes of Health Revitalization Act of 1993.45 Some issues could certainly be improved if demographic profiles would be incorporated in trial designs, and medical systems were able to address poverty, lack of education, impaired access to trials, and language barrier. Ultimately, more diverse recruitment strategies should be used when enrolling minority groups to achieve a more representative patient population, which would lead to a better generalizability of the results. The lack of transparency due to poor reporting rate can eventually cause serious damage; by preventing readers from evaluation of the validity and reliability of these trials, which may result in biased estimates of treatment effects. The CONSORT 2010 statement (Consolidated Standards of Reporting Trials) gives a clear guidance on the reporting of RCTs using an essential checklist.46 The checklist includes the reporting of exclusion rates. Since this CONSORT was published, sadly, only 1 journal article included in this analysis followed these recommendations. We recommend that editors improve the reporting quality of clinical trials for transparency by giving authors clear guidance on items to be reported. In summary, the present study was the first to evaluate how representative participants in published LN RCTs were in comparison with patients seen in usual clinical practice from a large tertiary lupus center. A third of newly diagnosed real-world patients with LN would have been excluded from LN RCTs due to strict eligibility criteria solely related to the severity of their renal disease. Although some trials have excellent validity, many do not; and factors determining their validity are rarely reported, preventing the clinician from using thoughtful judgment. Specifically, we highlight the need for more pragmatic trials designed for those with more severe LN.
3. Kennedy-Martin T, Curtis S, Faries D, et al. A literature review on the representativeness of randomized controlled trial samples and implications for the external validity of trial results. Trials. 2015;16:495. 4. Jadad AR, Enkin M, Jadad AR. Randomized Controlled Trials: Questions, Answers, and Musings. 2nd ed. Malden, MA: Blackwell Pub; 2007. 5. Roland M, Torgerson DJ. What are pragmatic trials? BMJ. 1998;316:285. 6. Churg J, Sobin LH. Renal Disease: Classification and Atlas of Glomerular Disease. Tokyo, Japan: Igaku-Shoin; 1982. 7. Weening JJ, D’Agati VD, Schwartz MM, et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int. 2004;65:521–530. 8. Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1982;25:1271–1277. 9. Steinberg AD, Kaltreider HB, Staples PJ, et al. Cyclophosphamide in lupus nephritis:a controlled trial. Ann Intern Med. 1971;75:165–171. 10. Steinberg AD, Decker JL. A double-blind controlled trial comparing cyclophosphamide, azathioprine and placebo in the treatment of lupus glomerulonephritis. Arthritis Rheum. 1974;17:923–937. 11. Ginzler E, Diamond H, Guttadauria M, et al. Prednisone and azathioprine compared to prednisone plus low-dose azathioprine and cyclophosphamide in the treatment of diffuse lupus nephritis. Arthritis Rheum. 1976;19:693–699. 12. Donadio JV Jr., Holley KE, Ferguson RH, et al. Treatment of diffuse proliferative lupus nephritis with prednisone and combined prednisone and cyclophosphamide. N Engl J Med. 1978;299:1151–1155. 13. Dinant HJ, Decker JL, Klippel JH, et al. Alternative modes of cyclophosphamide and azathioprine therapy in lupus nephritis. Ann Intern Med. 1982;96:728–736. 14. Boumpas DT, Austin HA 3rd, Vaughn EM, et al. Controlled trial of pulse methylprednisolone versus two regimens of pulse cyclophosphamide in severe lupus nephritis. Lancet. 1992;340:741–745. 15. Lewis EJ, Hunsicker LG, Lan SP, et al. A controlled trial of plasmapheresis therapy in severe lupus nephritis. The Lupus Nephritis Collaborative Study Group. N Engl J Med. 1992;326: 1373–1379. 16. Doria A, Piccoli A, Vesco P, et al. Therapy of lupus nephritis. A two-year prospective study. Ann Med Intern (Paris). 1994;145: 307–311. 17. Sesso R, Monteiro M, Sato E, et al. A controlled trial of pulse cyclophosphamide versus pulse methylprednisolone in severe lupus nephritis. Lupus. 1994;3:107–112.
DISCLOSURES All the authors declared no competing interests.
REFERENCES 1. Riegelman RK. Studying a Study and Testing a Test: How to Read the Medical Evidence. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2000. 2. Nallamothu BK, Hayward RA, Bates ER. Beyond the randomized clinical trial: the role of effectiveness studies in 8
evaluating cardiovascular therapies. Circulation. 2008;118: 1294–1303.
18. Gourley MF, Austin HA 3rd, Scott D, et al. Methylprednisolone and cyclophosphamide, alone or in combination, in patients with lupus nephritis. A randomized, controlled trial. Ann Intern Med. 1996;125:549–557. 19. Wallace DJ, Goldfinger D, Pepkowitz SH, et al. Randomized controlled trial of pulse/synchronization cyclophosphamide/ apheresis for proliferative lupus nephritis. J Clin Apher. 1998;13:163–166. Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858
CLINICAL RESEARCH
A Pakozdi et al.: External validity of LN RCTs
859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
20. Chan TM, Li FK, Tang CS, et al. Efficacy of mycophenolate mofetil in patients with diffuse proliferative lupus nephritis. Hong Kong-Guangzhou Nephrology Study Group. N Engl J Med. 2000;343:1156–1162. 21. Houssiau FA, Vasconcelos C, D’Cruz D, et al. Immunosuppressive therapy in lupus nephritis: the Euro-Lupus Nephritis Trial, a randomized trial of low-dose versus high-dose intravenous cyclophosphamide. Arthritis Rheum. 2002;46:2121–2131. 22. Yee CS, Gordon C, Dostal C, et al. EULAR randomised controlled trial of pulse cyclophosphamide and methylprednisolone versus continuous cyclophosphamide and prednisolone followed by azathioprine and prednisolone in lupus nephritis. Ann Rheum Dis. 2004;63:525–529. 23. Ong LM, Hooi LS, Lim TO, et al. Randomized controlled trial of pulse intravenous cyclophosphamide versus mycophenolate mofetil in the induction therapy of proliferative lupus nephritis. Nephrology. 2005;10:504–510. 24. Ginzler EM, Dooley MA, Aranow C, et al. Mycophenolate mofetil or intravenous cyclophosphamide for lupus nephritis. N Engl J Med. 2005;353:2219–2228. 25. Grootscholten C, Ligtenberg G, Hagen EC, et al. Azathioprine/ methylprednisolone versus cyclophosphamide in proliferative lupus nephritis. A randomized controlled trial. Kidney Int. 2006;70:732–742. 26. Bao H, Liu ZH, Xie HL, et al. Successful treatment of class VþIV lupus nephritis with multitarget therapy. J Am Soc Nephrol. 2008;19:2001–2010. 27. Appel GB, Contreras G, Dooley MA, et al. Mycophenolate mofetil versus cyclophosphamide for induction treatment of lupus nephritis. J Am Soc Nephrol. 2009;20:1103–1112. 28. Austin HA 3rd, Illei GG, Braun MJ, et al. Randomized, controlled trial of prednisone, cyclophosphamide, and cyclosporine in lupus membranous nephropathy. J Am Soc Nephrol. 2009;20:901–911. 29. Li EK, Tam LS, Zhu TY, et al. Is combination rituximab with cyclophosphamide better than rituximab alone in the treatment of lupus nephritis? Rheumatology. 2009;48:892–898. 30. Sabry A, Abo-Zenah H, Medhat T, et al. A comparative study of two intensified pulse cyclophosphamide remissioninducing regimens for diffuse proliferative lupus nephritis: an Egyptian experience. Int Urol Nephrol. 2009;41:153–161. 31. El-Shafey EM, Abdou SH, Shareef MM. Is mycophenolate mofetil superior to pulse intravenous cyclophosphamide for induction therapy of proliferative lupus nephritis in Egyptian patients? Clin Exp Nephrol. 2010;14:214–221. 32. Zavada J, Pesickova S, Rysava R, et al. Cyclosporine A or intravenous cyclophosphamide for lupus nephritis: the Cyclofa-Lune study. Lupus. 2010;19:1281–1289. 33. Chen W, Tang X, Liu Q, et al. Short-term outcomes of induction therapy with tacrolimus versus cyclophosphamide
for active lupus nephritis: a multicenter randomized clinical trial. Am J Kidney Dis. 2011;57:235–244. 34. Li X, Ren H, Zhang Q, et al. Mycophenolate mofetil or tacrolimus compared with intravenous cyclophosphamide in the induction treatment for active lupus nephritis. Nephrol Dial Transplant. 2012;27:1467–1472. 35. Rovin BH, Furie R, Latinis K, et al. Efficacy and safety of rituximab in patients with active proliferative lupus nephritis: the Lupus Nephritis Assessment with Rituximab study. Arthritis Rheum. 2012;64:1215–1226. 36. Yap DY, Yu X, Chen XM, et al. Pilot 24 month study to compare mycophenolate mofetil and tacrolimus in the treatment of membranous lupus nephritis with nephrotic syndrome. Nephrology. 2012;17:352–357. 37. Mysler EF, Spindler AJ, Guzman R, et al. Efficacy and safety of ocrelizumab in active proliferative lupus nephritis:results from a randomized, double-blind, phase III study. Arthritis Rheum. 2013;65:2368–2379. 38. The ACCESS Trial group. Treatment of lupus nephritis with abatacept: the Abatacept and Cyclophosphamide Combination Efficacy and Safety Study. Arthritis Rheum. 2014;66: 3096–3104. 39. Furie R, Nicholls K, Cheng TT, et al. Efficacy and safety of abatacept in lupus nephritis:a twelve-month, randomized, double-blind study. Arthritis Rheum. 2014;66:379–389. 40. Liu Z, Zhang H, Liu Z, et al. Multitarget therapy for induction treatment of lupus nephritis:a randomized trial. Ann Intern Med. 2015;162:18–26. 41. Mok CC, Ying KY, Yim CW, et al. Tacrolimus versus mycophenolate mofetil for induction therapy of lupus nephritis: a randomised controlled trial and long-term follow-up. Ann Rheum Dis. 2016;75:30–36. 42. Corapi KM, Dooley MA, Pendergraft WF 3rd. Comparison and evaluation of lupus nephritis response criteria in lupus activity indices and clinical trials. Arthritis Res Ther. 2015;17:110. 43. Contreras G, Lenz O, Pardo V, et al. Outcomes in African Americans and Hispanics with lupus nephritis. Kidney Int. 2006;69:1846–1851. 44. Isenberg D, Appel GB, Contreras G, et al. Influence of race/ ethnicity on response to lupus nephritis treatment:the ALMS study. Rheumatology. 2010;49:128–140. 45. Freedman LS, Simon R, Foulkes MA, et al. Inclusion of women and minorities in clinical trials and the NIH Revitalization Act of 1993–the perspective of NIH clinical trialists. Control Clin Trials. 1995;16:277–285. discussion 286–279, 293–309. 46. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration:updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c869.
Kidney International Reports (2018) -, -–-
FLA 5.5.0 DTD EKIR263_proof 24 November 2017 2:42 pm ce
9
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966