SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background  Crohn’s disease patients who have lost response to 5 mg/kg of infliximab may regain response by increasing the dose of infliximab to 10 mg/kg. Alternatively, adalimumab can be used as a rescue therapy.

Aim  To determine whether dose escalation of infliximab was a cost-effective strategy compared with adalimumab initiation after loss of response to 5 mg/kg of infliximab.

Methods  A decision-analysis model simulated two cohorts of Crohn’s patients: (i) infliximab dose was escalated to 10 mg/kg and (ii) infliximab was discontinued and patients were started on adalimumab. The time horizon was 1 year. One- and two-way sensitivity analyses were performed.

Results  The infliximab dose escalation strategy yielded more quality-adjusted life years (0.79) compared with the adalimumab strategy (0.76). The incremental cost-effectiveness ratio was $332 032/quality-adjusted life year. Sensitivity analysis demonstrated that the model findings were robust. The most significant variables were the cost of infliximab and that of adalimumab, such that a reduction in the cost of infliximab by 1/3 resulted in an incremental cost-effectiveness ratio below $80 000/quality-adjusted life year.

Conclusion  After a Crohn’s patient has lost response to 5 mg/kg of infliximab, dose escalation will yield more quality-adjusted life-year compared with switching to adaliumamb; however, the cost was considerable.


Background

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Infliximab is a chimaeric monoclonal antibody (25% murine, 75% human protein sequence homology) directed at tumour necrosis factor-alpha (TNFα). Infliximab effectively induces and maintains remission in inflammatory and fistulizing Crohn’s disease (CD).1, 2 At present, infliximab is usually prescribed to CD patients who are dependent or refractory to steroids and lack response or are unable to tolerate immunomodulators such as mercaptopurine (6-mercaptopurine), azathioprine or methotrexate. The annual cost for maintenance infliximab therapy is considerable, but the direct drug costs are offset by infliximab’s capacity to reduce hospitalizations and surgeries, as well as improve employment status.3, 4 Thus, private and governmental payers view treatment with infliximab as cost-effective.

The long-term efficacy of infliximab is limited by loss of response over time. This may be, in part, the consequence of the formation of antichimaeric antibodies, which are associated with infusion reactions as well as decreased serum infliximab levels.5 Patients who lose response to the standard dose of 5 mg/kg every 8 weeks are typically treated with doses as high as 10 mg/kg. Alternatively, the dosing interval may be reduced or a combination of both. Although dose escalation comes at a substantial drug cost, the cost is considered acceptable because medical options for patients who fail or do not tolerate infliximab have been limited.

Recently, clinical trials have demonstrated that adalimumab is effective and safe to use in patients who lose response to infliximab.6–11 Adalimumab is a subcutaneously administered human anti-TNFα monoclonal antibody. The GAIN study demonstrated that adalimumab induced clinical remission (21%) and response (52%) in CD patients who had either lost their response or developed an adverse reaction to infliximab.12 In addition, a subset of patients in the CHARM study consisted of CD patients who either lost response to or failed to tolerate infliximab. Among the patients who initially responded to adalimumab induction, one-third maintained their response at week 56 irrespective of weekly or every other week dosing. Thus, adalimumab provides a medical option for patients who lose response to infliximab therapy. However, not all patients who respond to one TNF agent will respond to a different agent in this class of compounds upon switching.

Clinicians faced with CD patients who have lost their response to standard dose infliximab may choose to escalate dose and then to use adalimumab, if dose escalation is unsuccessful. Alternatively, adalimumab could be initiated instead of dose escalation. Randomized-controlled trials comparing clinical efficacy and cost-effectiveness of these strategies have not been undertaken. Accordingly, we developed and analysed a decision model that compared the strategies of inflximab dose escalation to immediate switch to adalimumab for CD patients who have lost response to standard dose infliximab therapy.

Methods

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Decision model

A decision-analytic model was constructed to analyse different management strategies in CD patients who have lost response to standard dose infliximab therapy. The model’s base-case was a hypothetical cohort of 35-year-old patients with moderate to severely active CD who achieved remission following induction by 5 mg/kg of infliximab at weeks 0, 2 and 6, but lost their response during maintenance therapy dosed every 8 weeks. The time horizon for the model was 1 year.

The two strategies compared were: (i) infliximab dose was escalated to 10 mg/kg and (ii) infliximab was discontinued and patients were started on adalimumab. In the adalimumab strategy, adalimumab was initiated with a 160 mg injection followed by an 80 mg dose 2 weeks later with subsequent maintenance of 40 mg every other week. If the hypothetical patients lost response, the dose interval was reduced to 40 mg weekly. Adalimumab failures (no initial response, intolerance to drug or no response to weekly dosing) discontinued the drug and received non-anti-TNFα-based management. Non-anti-TNFα therapies resulted in the transition to one of the following states: surgery; remission on a non-anti-TNFα medication (e.g. azathioprine, mercaptopurine, methotrexate and/or corticosteroids); improve to a mild disease state on non-anti-TNFα therapies; no response to non-anti-TNFα therapies and remain in severe disease state; or death. In the dose escalation strategy, infliximab was increased to 10 mg/kg every 8 weeks. Patients who did not tolerate dose escalation of infliximab or lost response to this strategy were subsequently treated with adalimumab as described in the adalimumab strategy. Similarly, adalimumab failures received non-anti-TNFα management (Figure 1).

image

Figure 1.  Schematic representation of decision-analysis model comparing infliximab dose escalation strategy to switch immediately to adalimumab strategy.

Download figure to PowerPoint

The model was constructed and all simulations were performed using treeage pro 2006 (TreeAge Software Inc., Williamstown, MA, USA).

Model inputs

The initial response rate to adalimumab was defined as a decrease in the Crohn’s Disease Activity Index (CDAI)13 of 70 points at week 4 following induction of 160 mg at baseline and 80 mg 2 weeks later.12 The model inputs were based on the GAIN study12 that evaluated adalimumab induction following infliximab failure. Although a CDAI decrease of 100 has been proposed as a conservative definition of response,14 we chose to use 70 assuming that even a mild initial response would lead to continuation of the drug. In addition, the 70-point decrease in CDAI was consistent with studies of infliximab that previously used this value as a definition of clinical response.2 Remission on adalimumab at 1 year was defined as the proportion of patients who had a CDAI score <150 after 1 year of maintenance therapy. The base-case estimate was derived from a subset of the CHARM study, which randomized infliximab failures (loss of response or intolerance) who had an initial response to adalimumab, to either adalimumab or placebo maintenance for 1 year. Although the CHARM study published the clinical response rate (CDAI decrease of 70) at 1 year for the entire study population, these data were not available for the subset of patients who were infliximab failures. We extrapolated the response rate (CDAI decrease of 70) of the entire study population to the infliximab failure subset for the purpose of the model (Table 1).

Table 1.   Model inputs and utility scores
Parameter estimateBaseRangeReference
  1. * Responders include both patients in remission and patients who had a clinical response (CDAI score: <70); the proportion of patients only in clinical response is the difference between remitters and responders.

Transition probabilities
Dose escalation with infliximab
 Initial response to infliximab and tolerate the medication0.830.31–0.99 
  Initial response to dose escalation0.9 15
  Discontinue to infliximab because of serious adverse event0.08 15
 Initial responders maintain remission at 1 year0.390.13–0.862, 24
 Initial responders maintain response at 1 year*0.53 2, 24
 Mortality associated with infliximab0.0040.0–0.0516
Adalimumab
 Initial response to adalimumab and tolerate the medication0.480.0–0.99 
  Initial response to adalimumab0.52 12
  Discontinue adalimumab because of a serious adverse event0.07 6
 Initial responders maintain remission at 1 year with every other week dosing0.300.0–0.936
 Initial responders maintain response at 1 year with every other week dosing*0.37 6
 Loss of response to adalimumab0.63  
  Remission at 1 year after escalating to weekly dosing0.17 32
  Response at 1 year after escalating to weekly dosing0.31 32
 Mortality associated with adalimumab0.0010.0–0.286
Non-anti-TNFα management
 Remission0.210.0–0.5917
 Improve to mild disease0.21 17
 Remain in state of moderate–severe activity0.39 17
 Require surgery0.180.0–0.5717
 Death from flare0.008 17
 Death from surgery0.0015 17
Quality of life utilities
 Medical remission0.89 18
 Surgical remission0.86 18
 Mild flare0.77 18
 Severe flare0.62 18
 Death0  

The initial response rate to dose escalation of infliximab was based on the proportion of patients in the ACCENT 1 study who lost their response to 5 mg/kg, but regained it after they crossed-over to 10 mg/kg.15 Despite a systematic review of the literature, we were unable to find a study that provided the probability of remission (CDAI < 150) and clinical response (CDAI decrease of 70) 1 year following dose escalation. Consequently, we used the results from the cohort in ACCENT 1 that was randomized to 10 mg/kg to define response (14%) and remission rates (39%) at 1 year (Table 1). In ACCENT 1, 59% of patients who were dose escalated from 5 to 10 mg/kg continued the drug to the end of the study period (54 weeks). However, the exact proportion in remission and that in response were not reported and over half these patients received dose escalation for <6 months before the study period ended; thus, the true estimate of overall response is likely to be <59%.

The initial response rate to dose escalation of infliximab and adalimumab also included the possibility that a patient would discontinue the anti-TNFα agent because of a serious adverse event. The probability that a hypothetical individual in the model would discontinue infliximab because of an adverse event following dose escalation was derived from data on the group of patients who crossed-over from 5 to 10 mg/kg in the ACCENT 1 study.2, 15 For adalimumab, the adverse event rate was estimated from the entire cohort in the CHARM study because this probability was not provided for the subset of patients who were infliximab failures. The mortality rate associated with infliximab was based on a systematic literature review as previously reported.16 However, the mortality rate associated with adalimumab was derived solely from the CHARM study because of the paucity of adalimumab studies in CD (Table 1).

The base-case estimates for non-anti-TNFα management were derived from a previously published population-based longitudinal cohort Markov model of the natural history of CD before the use of infliximab and adalimumab.17

Utilities

Quality of life utilities were estimated for the following outcomes: medically induced remission; surgically induced remission; mild flare; severe flare; and death (Table 1). The utilities for the CD states were derived from Gregor et al. who used a standard gamble approach to define utility scores and correlate them with the CDAI.18 Hypothetical patients could transition between various health states of the model in 4-week intervals. The utilities assigned per 4-week interval for each branch of the decision tree were based on the following model assumptions.

Adalimumab strategy
  • 1
    In patients who tolerated and responded to adalimumab and were in remission at the end of year, they were considered to be in remission for the entire year. Similarly, if patients were in clinical response (mild flare) at the end of 1 year, they were maintained in this state for the entire year.
  • 2
    In patients who tolerated and responded to adalimumab initially, but lost response to the drug, we assumed that response was lost at interval 6 because in the CHARM study, more than half the patients lost response by week 26.6 In interval 7, patients were dose escalated to 40 mg weekly. During block 6 and 7, patients were in a severe flare state. Patients who responded to dose escalation entered remission or mild flare at interval 8 and maintained this clinical state to interval 13. Patients who did not respond to dose escalation of adalimumab were treated with a non-anti-TNFα approach and entered one of the following five states: surgery; remission on non-anti-TNFα therapies; improve to a mild disease state; remain in severe disease state; or death.
  • 3
    Patients who did not respond or did not tolerate initial induction with adalimumab experienced a severe flare state in interval 1. In interval 2, they were managed with non-anti-TNFα therapies. Patients who achieved remission or improved to a mild flare following non-anti-TNFα agents achieved these states after 2 months of therapy.
Infliximab dose escalation strategy
  • 1
    Patients who after infliximab dose escalation were in remission or in a mild flare state at the end of 1 year were considered to have maintained their response throughout the entire year.
  • 2
    In patients who tolerated and responded to infliximab initially, but lost response to the drug, we assumed that response was lost at interval 6, because 55% of patients who crossed-over from 5 to 10 mg/kg did so by week 22 in ACCENT 1.15 In interval 7, infliximab was discontinued and adalimumab was induced with 160 mg initially and 80 mg at week 2. Patients who responded to adalimumab maintained their response (remission or mild flare) to interval 13. Among patients who initially responded to adalimumab, but then lost their response, we assumed that response was lost at interval 13. Patients who did not initially respond to or tolerate adalimumab therapy were in a state of severe flare during interval 7. In interval 8, they received non-anti-TNFα management.
  • 3
    Patients who did not tolerate or respond to infliximab dose escalation initially remained in a state of severe flare during interval 1. These patients were started on adalimumab in interval 2. Patients who tolerated and responded to adalimumab did so immediately in interval 2 and maintained their response (remission or mild flare) to interval 13. Among patients who initially responded to adalimumab, but then lost their response, we assumed that response was lost at interval 7. In interval 8, patients were dose escalated to 40 mg weekly. Patients who responded to dose escalation maintained that response (remission or mild flare) to interval 13. Patients who did not respond to dose escalation were in state of severe flare in interval 8 and received non-anti-TNFα management thereafter.
  • 4
    Patients who failed induction to infliximab dose escalation and then to adalimumab were assumed to be in a state of severe flare for intervals 1 and 2. In interval 3, they received non-anti-TNFα management.

Cost

The total costs for each strategy that could occur over 1 year were estimated (Table 2). Costs were reported using 2006 US dollars. For each 4-week interval in the decision tree, we determined the amount of adalimumab and infliximab used to treat a 70 kg patient. The average wholesale prices for adalimumab and infliximab were obtained from the 2006 Drug Topics Red Book.19 Additional costs of an infusion were estimated from the American Medical Association Current Procedural Terminology code.20

Table 2.   Costs of drug and infusions in 2006 US dollars
Parameter estimateBase ($)Range ($)Reference
  1. AWP, average wholesale price.

Cost of a 70 kg patient to receive 10 mg/kg (700 mg) based on the AWP46392498–774919
Cost of administrating i.v. infusion193 20
Cost of 40 mg vial of adalimumab based on the AWP9440–160219
Monthly cost of non-anti-TNFα management in a patient who responds to therapy817 17
Monthly cost of standard drug therapy in a patient who is refractory to therapy786 17
Cost of undergoing surgery11 341 17
Monthly cost of post-operative remission114 17

The monthly costs for non-anti-TNFα management were derived from a Markov model of the natural history of CD before the use of anti-TNFα therapies.17 The paper published the median charges per month to manage a CD patient in a different state of health. The costs derived from this study were adjusted to 2006 dollars using the Consumer Price Index.21

Sensitivity analysis

Extensive sensitivity analysis was performed to evaluate the effect of varying model parameter estimates on model outputs (Tables 1 and 2). The parameter estimates evaluated included: initial response and tolerability of adalimumab and infliximab; remission at 1 year for adalimumab and infliximab; mortality associated with adalimumab and infliximab; remission on non-anti-TNFα therapy; surgery; and cost of infliximab and adalimumab. The parameter estimates were varied over a wide range to determine: (i) the threshold value where both strategies would result in equivalent quality-adjusted life year (QALY) and (ii) the threshold value where the incremental cost-effectiveness ratio (ICER) was <$80 000/QALY.22 In this paper, the ICER represents the additional cost necessary to achieve one QALY when the strategy of dose escalation of infliximab is compared to initiating adalimumab immediately.

Additionally, a two-way sensitivity analysis was performed on the cost of adalimumab and the cost of infliximab. The values were varied from $0 to the 2006 average wholesale price.

Results

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Cost-effectiveness

For ease of comparison, all results assumed that the model simulated 100 000 hypothetical patients in each strategy. The model demonstrated (Table 3) that at the end of 1 year, when compared with the adalimumab strategy, the dose escalation with infliximab strategy would result in 13 989 more patients in remission and 6428 fewer surgeries. However, 5334 fewer patients would have improved without remission and 250 more patients would have died. The infliximab dose escalation strategy resulted in greater QALYs per patient than the adalimumab strategy (0.79 QALYs/patient vs. 0.76 QALYs/patient); primarily because only 24% of the inflixmab dose-escalated patients would remain in a severe state of CD flare at the end of the year compared to 27% of the adalimumab-treated patients (Table 3).

Table 3.   Results of base-case analysis*
Parameter estimateDose escalation of infliximabAdalimumab therapyDifference
  1. (+) signifies more in the dose escalation infliximab group; (−) signifies more in the adalimumab therapy group. QALYs per patient range from 0 (death) to 1 (perfect health). * Patient outcomes and costs at the end of 1 year in a hypothetical cohort of 100 000 Crohn’s disease patients who have lost response to 5 mg/kg of infliximab.

  2. QALY, quality-adjusted life year; ICER, incremental cost-effectiveness ratio.

Remission without surgery47 83933 850+13 989
Improvement without remission21 65226 986−5334
No change, remain in severe state24 36326 840−2477
Surgical remission559812 026−6428
Death548298+250
QALYs per patient0.790.76+0.03
Cost$28 367$18 074+$10 293
ICER$332 032/QALY

The infliximab dose escalation strategy cost $28 367 per patient vs. $18 074 per patient for a net difference of $10 293. The resulting incremental cost-effectiveness ratio of the infliximab dose escalation strategy vs. the switch to adalimumab therapy was $332,032/QALY (Table 3).

Sensitivity analysis

In one-way sensitivity analysis, the infliximab dose escalating strategy yielded more QALYs unless: (i) the initial response rate for infliximab was below 31%; (ii) the rate of remission for infliximab dose escalation at 1 year was <13%; (iii) the mortality rate of infliximab was >4.8%; (iv) the remission at 1 year for patients receiving non-anti-TNFα therapy exceeded 59%; or (v) the proportion of patients receiving non-anti-TNFα therapy who required surgery was more than 57%. At these threshold values, the adalimumab strategy dominated the infliximab dose escalation strategy because the adalimumab strategy generated more QALYs and cost less money (Table 4).

Table 4.   Sensitivity analysis
ParameterEstimate*Dose escalation infliximabAdalimumabICER ($)
QALYCost ($)QALYCost ($)
  1. * Indicates base-case.

  2. † As the estimate changes in the sensitivity analysis, loss of response changes accordingly.

  3. ‡ Denotes the threshold estimate when the QALY of the adalimumab strategy equals the infliximab dose escalation strategy.

  4. § Denotes the threshold estimate when the ICER of the adalimumab strategy equals the infliximab dose escalation strategy.

  5. QALY, quality-adjusted life year; ICER, incremental cost-effectiveness ratio.

Dose escalation
 Initial response0.31‡0.7624 2060.7618 074Dominated
0.440.7725 2460.7618 074935 755
0.83*0.7928 3670.7618 074332 032
0.990.8029 6640.7618 074287 648
 Remission at 1 year†0.13‡0.7626 5440.7618 074Dominated
0.280.7827 5620.7618 074553 540
0.39*0.7928 3670.7618 074332 032
0.710.8330 6570.7618 074181 572
0.860.8531 6890.7618 074157 057
 Infliximab mortality0.00.7928 4350.7618 074309 186
0.004*0.7928 3670.7618 074332 032
0.010.7928 2320.7618 074404 264
0.030.7827 9260.7618 074785 150
0.05‡0.7627 9830.7618 074Dominated
Adalimumab
 Initial response0.00.7827 6540.7214 710228 238
0.200.7827 9460.7416 086257 375
0.48*0.7928 3670.7618 074332 032
0.790.8028 8210.7920 213606 955
0.990.8129 1120.8021 5892 120 394
 Remission at 1 year†0.00.7928 4650.7418 702238 865
0.190.7928 4040.7618 313291 879
0.30*0.7928 3670.7618 074332 032
0.740.8028 2230.7917 144706 163
0.930.8128 1630.7916 7541 214 396
 Adalimumab mortality0.00.7928 3720.7618 085339 774
0.001*0.7928 3670.7618 074332 032
0.0040.7928 5390.7618 041321 892
0.0100.7928 3290.7617 975298 482
0.2750.7327 2080.5815 07879 994
Non-anti-TNFα therapy
 Remission at 1 year†0.00.7928 3070.7417 884217 785
0.120.7928 3410.7517 990270 537
0.21*0.7928 3670.7618 074332 032
0.350.7928 4090.7818 203535 401
0.59‡0.8128 4770.8118 415Dominated
 Proportion requiring surgery†0.00.7928 0230.7417 616233 002
0.18*0.7928 3670.7618 074333 032
0.340.7928 6770.7718 486561 119
0.57‡0.8129 1120.8119 066Dominated
Cost
 40 mg vial of adalimumab$00.7923 6470.766656553 162
$3200.7926 0420.7612 449442 552
$632*0.7928 3670.7618 074333 032
$12820.7933 2270.7629 830110 603
$1602§0.7935 6230.7635 623Dominated
 Infliximab 10 mg/kg (70 kg)$2498§0.7918 0740.7618 074Dominant
$28460.7919 6080.7618 07449 962
$30550.7920 5300.7618 07479 977
$32200.7921 2560.7618 074103 587
$4832*0.7928 3670.7618 074333 032
$77490.7941 2370.7618 074754 099

The ICER was below $80 000/QALY when the mortality rate for adalimumab was above 28%; otherwise, altering the estimate parameters for adalimumab, infliximab or non-anti-TNFα therapy response did not produce an ICER lower than $80 000/QALY.

The most important factor that influenced the ICER in the sensitivity analysis was the drug cost for infliximab and adalimumab. When the average wholesale price of adalimumab was increased nearly three times its cost ($1602 per 40 mg dose), the infliximab dose escalating strategy dominated both in cost and in effectiveness. In contrast, when the average wholesale price of infliximab for a 70 kg patient dosed at 10 mg/kg was reduced nearly by half ($2498), the infliximab strategy was dominant. A decrease in cost by 33% ($3055) yielded an ICER that was below $80 000/QALY (Table 4).

Two-way sensitivity analysis was performed on the cost of adalimumab and the cost of infliximab. Figure 2 illustrates the different threshold values at which the infliximab strategy has an ICER below $80 000/QALY and when it becomes the dominant strategy.

image

Figure 2.  Two-way sensitivity analysis of the cost of adalimumab and the cost of infliximab. The cost of adalimumab was based on the cost of a 40 mg dose, while the cost of infliximab was based on a 70 kg patient dosed at 10 mg/kg and the infusion costs.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Although infliximab has proven to be an effective medication for patients with moderate-to-severe CD, loss of response to the 5 mg/kg dose is a common occurrence.15, 23, 24 After drug response is lost, gastroenterologists typically escalate the dose of infliximab to 10 mg/kg and/or reduce the interval of administration. With evidence from a large randomized-controlled trial6 and recent FDA approval, adalimumab offers an effective and safe option for infliximab failures. Our decision analysis found that dose escalating infliximab prior to starting adalimumab yielded an excess of 0.03 QALY/patient. This gain in QALY from the infliximab dose escalating strategy was similar to the clinically meaningful benefit seen in other decision-analysis models in CD.16, 25

However, the cost (ICER of $332 032/QALY) to achieve this benefit was considerable. Although no universally adopted standard for an ICER exists, an ICER of over $330 000/QALY was appreciably higher than typically accepted values reported in the literature.26, 27 Nevertheless, this ICER was similar to the over $350 000/QALY value identified by a Markov model that compared infliximab to mercaptopurine and metronidazole in treating perianal fistulizing CD;28 despite the high calculated ICER, infliximab is a key agent in the medical management for fistulizing CD.29 An additional Markov model demonstrated that in moderate-to-severe CD patients who failed to respond to standard medications, the ICER of maintenance therapy with infliximab compared to surgery was $762 245 (euros) per QALY.30 Ultimately, society and policy-makers will have to determine if this expenditure is acceptable or whether adalimumab should replace infliximab after a patient loses his response to 5 mg/kg dosing.

Sensitivity analysis was performed to confirm the robustness of the model, to identify threshold values when the adalimumab strategy yielded more QALYs than the dose escalating strategy, and to determine the threshold value where the ICER favoured the dose escalation strategy. The most important model parameters that influenced the QALYs were the initial response rate of dose escalation from 5 to 10 mg/kg and the remission rate at 1 year among initial responders to dose escalation. The initial response rate incorporated the probability of initially responding to dose escalation and the probability of discontinuing the drug because of an adverse event; both probabilities were based on the results of the crossover arm of 5 to 10 mg/kg in ACCENT 1.2, 15 Although the initial response rate for infliximab could be reasonably estimated from the literature, the 1-year remission rate following dose escalation was less clear. In the model, we utilized the 1-year remission (39%) and response (14%) rates reported in ACCENT 1 for the patients randomized to 10 mg/kg. Similarly, 59% of the patients who crossed-over from 5 to 10 mg/kg in ACCENT 1 completed the trial on medication; however, the proportion in remission vs. response was not reported and most patients were on the higher dose for 6 or less months and thus lower 1-year remission and response rates were probable.15 A recent single-centre study reported that 76% of patients who underwent dose intensification (dose escalation to 10 mg/kg, dose interval reduction or both) were still on infliximab at the end of the study period.24 Nonetheless, when we modelled a 1-year remission rate of 80%, in the sensitivity analysis, the ICER was still >$150 000.

The only model parameter estimates that had a meaningful impact on the ICER in the sensitivity analysis were adalimumab mortality rate and the cost of the drugs. Only when the mortality rate of adalimumab exceeded an unrealistic value of 27%, did the ICER drop below $80 000/QALY. Ultimately, the cost of infliximab and that of adalimumab were the most important parameters in the model. If the price of adalimumab was doubled, the ICER approached $100 000/QALY, and the infliximab strategy was dominant when the price was nearly tripled. The cost of infliximab was based on the average wholesale price of the drug to treat a 70 kg patient at a dose of 10 mg/kg and the additional charges for the infusion, but did not include indirect costs such as loss of work during an infusion. The ICER was near $80 000/QALY, when the estimated cost of infliximab was reduced by 1/3, and when the value was halved, the infliximab dose escalating strategy was dominant.

Average wholesale prices were used to estimate the cost of adalimumab and infliximab.19 However, the charged price of infliximab is typically not the actual price paid. Ollendorf and Lidsky31 estimated that US health plans on average paid 1/3 of the price charged for infliximab. The actual price paid for infliximab and adalimumab will differ based on the setting as different payers negotiate different prices. It was not possible to estimate a standardized cost for the drugs throughout the US and thus we chose to use the average wholesale price. We created a two-way sensitivity analysis that varied the cost of both drugs from $0 to the 2006 average wholesale price. This analysis allows payers to determine which strategy is cost-effective by plotting the actual cost they pay for each drug; alternatively, the graph could be used to negotiate a price that a payer would be willing to pay for either infliximab or adalimumab in patients who have failed to respond to a dose of 5 mg/kg.

The conclusions of the analysis should be interpreted cautiously because the model is based, in part, on hypothetical assumptions and subjected to several limitations. The model was based on hypothetical 35-year-old CD patients who lost response to infliximab dosed at 5 mg/kg. The hypothetical patients may not represent all CD patients who fail infliximab therapy and the decision tree may not mimic the practice patterns of certain regions in the world. Additionally, a time horizon of 1 year was chosen for this study because data on efficacy for both adalimumab and infliximab beyond this point were limited. Additionally, in practice, as patients lose response to infliximab, the dose interval may be reduced from every 8 weeks. The model did not include an option for dose interval reduction because data on remission and response, short term and long term, do not exist. Furthermore, we assumed that the probability that a patient experienced an adverse event that led to discontinuation of a drug was modelled upfront instead of allowing for the possibility that the adverse events occurred equally over the 1-year period. Furthermore, for non-anti-TNFα therapy, the probabilities of events were estimated from a Markov model published before the widespread use of infliximab.17 Costs were also estimated from the model and inflated to 2006 dollars. Ideally, we would have used data that documented the natural history of infliximab failures; however, this information is not available. We attempted to overcome these limitations by designing a model that was representative of typical practice using probabilities from randomized-controlled trials wherever possible. Importantly, when we varied our estimates in our sensitivity analyses, the resulting QALYS and ICER did not materially alter from the base-case model. Finally, the long-term risk of lymphoma and other malignancies was not included in the model because we assumed that both infliximab and adalimumab would cause an equal number of cancers and thus the net difference would be negated.

In conclusion, among CD patients who lose response to 5 mg/kg of infliximab, increasing the dose to 10 mg/kg would lead to an excess of 0.03 QALY over a 1-year time frame compared to using adalimumab therapy alone. However, the difference in cost between these strategies resulted in an ICER of over $330 000/QALY. In sensitivity analysis, the most influential parameter estimates were the costs of adalimumab and infliximab, such that if the cost of dose escalation was halved, this strategy became dominant. The two-way sensitivity analysis graph of costs of infliximab and adalimumab allows payers to determine which strategy is cost-effective on the basis of actual cost they have negotiated with the manufacturers. Patients, doctors and healthcare policy-makers may use of this figure to evaluate the appropriate management for CD patients who fail standard dose infliximab therapy based on a cost-effectiveness approach.

Acknowledgements

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Gilaad Kaplan is a Canadian Institutes Health Research and Alberta Heritage Foundation for Medical Research Clinical Fellow. Declaration of personal interests: Gilaad Kaplan has served as an advisory board member for UCB Pharma and Abbot Laboratories. Bruce Sands: Abbot Laboratories – Research grants, honoraria for speaking, scientific advisory board. Centocor – Research grants, honoraria for speaking, scientific advisory board. UCB Pharma - Honoraria for speaking, scientific advisory board. Bristol-Myers Squibb – Research grants, scientific advisory board. Elan Pharmaceuticals - Research grants, honoraria for speaking, scientific advisory board. Prometheus Laboratories – Research grants, scientific advisory board. Otsuka America Pharmaceuticals Inc. - Research grants, honoraria for speaking, scientific advisory board. Joshua Korzenik: Consulting and lecture fees from Proctor & Gamble, Shire Pharmaceuticals, Isis Pharmaceuticals, Cytokine Pharmasciences, Berlex and Centocor and research support from Danisco. A US patent entitled “Stimulating Neutrophil Function to Treat Inflammatory Bowel Disease (6500418)” was issued December 31, 2002. Dr Korzenik is one of the inventors. The patent is owned by Washington University and licensed by Berlex Laboratories. Chin Hur: none to declare. Declaration of funding interests: None.

References

  1. Top of page
  2. Summary
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Present DH, Rutgeerts P, Targan S, et al. Infliximab for the treatment of fistulas in patients with Crohn’s disease. N Engl J Med 1999; 340: 1398405.
  • 2
    Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance infliximab for Crohn's disease: the ACCENT 1 randomized trial. Lancet 2002; 359: 15419.
  • 3
    Lichtenstein GR, Yan S, Bala M, Hanauer S. Remission in patients with Crohn’s disease is associated with improvement in employment and quality of life and a decrease in hospitalizations and surgeries. Am J Gastroenterol 2004; 99: 916.
    Direct Link:
  • 4
    Lichtenstein GR, Yan S, Bala M, Blank M, Sands BE. Infliximab maintenance treatment reduces hospitalizations, surgeries, and procedures in fistulizing Crohn’s disease. Gastroenterology 2005; 128: 8629.
  • 5
    Baert F, Noman M, Vermeire S, et al. Influence of immunogenicity on the long-term efficacy of infliximab in Crohn’s disease. N Engl J Med 2003; 348: 6018.
  • 6
    Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn’s disease: the CHARM Trial. Gastroenterology 2007; 132: 5265.
  • 7
    Barthel HR, Gille T, Halbsguth A, Kramer M. Successful treatment with adalimumab in infliximab-resistant Crohn’s disease. J Gastroenterol Hepatol 2005; 20: 14645.
  • 8
    Papadakis KA, Shaye OA, Vasiliauskas EA, et al. Safety and efficacy of adalimumab (D2E7) in Crohn’s disease patients with an attenuated response to infliximab. Am J Gastroenterol 2005; 100: 759.
    Direct Link:
  • 9
    Sandborn WJ, Hanauer S, Loftus EV Jr, et al. An open-label study of the human anti-TNF monoclonal antibody adalimumab in subjects with prior loss of response or intolerance to infliximab for Crohn’s disease. Am J Gastroenterol 2004; 99: 19849.
    Direct Link:
  • 10
    Stallmach A, Giese T, Schmidt C, Meuer SC, Zeuzem SS. Severe anaphylactic reaction to infliximab: successful treatment with adalimumab – report of a case. Eur J Gastroenterol Hepatol 2004; 16: 62730.
  • 11
    Youdim A, Vasiliauskas EA, Targan SR, et al. A pilot study of adalimumab in infliximab-allergic patients. Inflamm Bowel Dis 2004; 10: 3338.
  • 12
    Sandborn WJ, Rutgeerts P, Enns R, et al. Adalimumab rapidly induces clinical remission and response in patients with moderate to severe Crohn’s disease who had secondary failure to inflixmab therapy: results of the GAIN study. Am J Gastroenterol 2006; 101: S408 (A1147).
  • 13
    Best WR, Becktel JM, Singleton JW, Kern F Jr. Development of a Crohn’s disease activity index. National Cooperative Crohn’s Disease Study. Gastroenterology 1976; 70: 43944.
  • 14
    Sands BE, Steinhart AH, Lewis JD, Hanauer SB, Persson T, Sandborn WJ. Optimal Crohn's disease activity index (CDAI) response criteria is defined by decrease ≥100 points. Gastroenterology 2003; 124: A206.
  • 15
    Rutgeerts P, Feagan BG, Lichtenstein GR, et al. Comparison of scheduled and episodic treatment strategies of infliximab in Crohn's disease. Gastroenterology 2004; 126: 40213.
  • 16
    Siegel CA, Hur C, Korzenik JR, Gazelle GS, Sands BE Risks and benefits of infliximab for the treatment of Crohn’s disease. Clin Gastroenterol Hepatol 2006; 4: 101724; quiz 976.
  • 17
    Silverstein MD, Loftus EV, Sandborn WJ, et al. Clinical course and costs of care for Crohn’s disease: Markov model analysis of a population-based cohort. Gastroenterology 1999; 117: 4957.
  • 18
    Gregor J, McDonald W, Klar N, et al. An evaluation of utility measures in Crohn’s disease. Inflamm Bowel Dis 1997; 3: 26576.
  • 19
    2006 Drug Topics, Red Book. Montvale, NJ: Medical Economics, 2006.
  • 20
    AMA Website. Available at: http://www.ama-assn.org, accessed on 15 December 2006.
  • 21
    U.S. Department of Labour Bureau of Labor Statistics Data. Available at: http://www.bls.gov, accessed on 22 January 2007.
  • 22
    Winkelmayer WC, Weinstein MC, Mittleman MA, Glynn RJ, Pliskin JS. Health economic evaluations: the special case of end-stage renal disease treatment. Med Decis Making 2002; 22: 41730.
  • 23
    Corman S, Issa M, Beaulieu DB, et al. Escalation of inflixmab maintenance therapy in Crohn’s disease. Am J Gastroenterol 2006; 101: S470.
  • 24
    Regueiro M, Siemanowski B, Kip KE, Plevy S. Infliximab dose intensification in Crohn’s disease. Inflamm Bowel Dis 2007; 13: 10939.
  • 25
    Lewis JD, Schwartz JS, Lichtenstein GR. Azathioprine for maintenance of remission in Crohn’s disease: benefits outweigh the risk of lymphoma. Gastroenterology 2000; 118: 101824.
  • 26
    Azimi NA, Welch HG. The effectiveness of cost-effectiveness analysis in containing costs. J Gen Intern Med 1998; 13: 6649.
  • 27
    Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations CMAJ 1992; 146: 47381.
  • 28
    Arseneau KO, Cohn SM, Cominelli F, Connors AF Jr. Cost-utility of initial medical management for Crohn’s disease perianal fistulae. Gastroenterology 2001; 120: 164056.
  • 29
    Vader JP, Froehlich F, Juillerat P, et al. Appropriate treatment for Crohn’s disease: methodology and summary results of a multidisciplinary international expert panel approach – EPACT. Digestion 2006; 73: 23748.
  • 30
    Jaisson-Hot I, Flourie B, Descos L, Colin C. Management for severe Crohn’s disease: a lifetime cost-utility analysis. Int J Technol Assess Health Care 2004; 20: 2749.
  • 31
    Ollendorf DA, Lidsky L. Infliximab drug and infusion costs among patients with Crohn’s disease in a commercially-insured setting. Am J Ther 2006; 13: 5026.
  • 32
    Sandborn WJ, Hanauer SB, Rutgeerts PJ, et al. Adalimumab for maintenance treatment of Crohn’s disease: results of the CLASSIC II Trial. Gut 2007; 56: 12329.