Oral glucocorticoid pulse therapy for induction of treatment of canine pemphigus foliaceus – a comparative study

Introduction

Pemphigus foliaceus (PF) is the most common autoimmune skin disease of dogs, and, in most cases, it requires long-term (often life-long) treatment.1, 2 After more than three decades of experience of treating dogs with PF, glucocorticoids remain the most common therapeutic intervention.1-5 Traditionally, the induction dose of prednisone or prednisolone used for canine PF ranges from 2.2 to 6.6 mg/kg/day.6 An attempt to reduce this dose is made as soon as the disease undergoes remission, which can often take several months. Indeed, the average time to complete remission (CR) of skin lesions in dogs reported in the largest canine PF case series was 9 months (range 1–36 months). Surprisingly, the addition of azathioprine as adjuvant therapy did not significantly shorten this time.2 Unfortunately, as the treatment with glucocorticoids is often prolonged, numerous adverse effects usually develop, many of which are intolerable for owners and lead to requests for euthanasia.1, 2, 5 In two series, between 2 and 3% of dogs treated for PF were euthanized because of unacceptable treatment adverse effects.2, 5 Therefore, there is a need for alternative treatment protocols that would provide rapid remission and, potentially, reduce the need for prolonged courses of glucocorticoids at high dosages.

Pulse therapy with high-dose glucocorticoids for human pemphigus vulgaris (PV) was introduced more than three decades ago as a means to provide rapid disease control, which would then allow for earlier dose reduction of glucocorticoids and/or adjunctive immunosuppressive drugs.7-9 Originally, pulse therapy was defined as intermittent intravenous injections of very high doses of glucocorticoids (10–20 mg/kg of methylprednisolone or 2–5 mg/kg of dexamethasone) over a short period of time, with or without adjuvant drugs such as cyclophosphamide or azathioprine. While most open studies reported the exceptional benefits of pulse therapy, such as a high rate and a rapid onset of CR, as well as a low rate of adverse effects in patients with PV,9 a randomized double-blinded controlled trial did not identify any outcome benefits when compared to conventional PV therapy.10

The evidence of efficacy of glucocorticoid pulse therapy is limited in veterinary dermatology. An open study reported a dramatic and rapid improvement (time not specified) of skin lesions in one dog with PV and three with PF after a single pulse of intravenous methylprednisolone at 11 mg/kg for three consecutive days.11 Adverse effects associated with this pulse therapy were not observed in any of the four dogs.11 Although only reported in abstract form, a randomized, controlled trial of 20 dogs with PF did not suggest any additional benefits of a single intravenous pulse of methylprednisolone (10 mg/kg for three consecutive days) when compared to the outcomes achieved using conventional therapy.12

Our objectives were to compare the therapeutic outcomes of oral glucocorticoid pulses and traditional therapy during the first 3 months of disease management (i.e. induction phase). To reduce the relative imprecision associated with any retrospective study, only outcomes that were clearly identifiable were reported. These included, for each treatment group, the number of dogs achieving CR, the time to CR, the number of dogs requiring adjuvant immunosuppressive therapy and the number of dogs experiencing severe adverse drug reactions.

Materials and methods

Results

Study subjects

Details on the allocation of dogs with PF to the ‘traditional’ and the ‘pulse’ groups can be found in the flow chart presented as Figure 1. These groups were not significantly different in regards to age of the dogs at the time of disease onset, or the proportion of dogs with a generalized rather than facially restricted disease (data not shown).

The main outcome measures for each group can be found in Table 1. The proportion of dogs achieving CR during the first 12 weeks of PF treatment induction was significantly (P = 0.0063) higher in dogs from the ‘pulse’ group (61%) compared to those from the ‘traditional’ group (15%). The number needed to treat (NNT) for CR during pulse compared to traditional therapy (i.e. the inverse of the difference between percentages of CR, rounded upward) was calculated at three, a low number (i.e. three dogs would need to be treated with pulse therapy to get an additional CR compared to traditional protocols). The proportion of dogs reaching CR every week is depicted in a cumulative graph in Figure 2, and two examples of rapid improvement following the first oral glucocorticoid pulse are shown in Figure 3. The time to CR and the proportion of dogs needing adjuvant immunosuppression during the first 12 weeks of treatment induction were lower in dogs receiving glucocorticoid pulses, but they were not significantly different between groups (Table 1).

Table 1. The outcome measures for ‘traditional’ and ‘pulse’ treatment groups during the first 12 weeks of therapy

	‘Traditional’	‘Pulse’	P-value
Number of dogs	20	18
Number (%) of dogs achieving CR	3 (15%)	11 (61%)	0.0063
Time to CR; median (range) (weeks)	9.1 (2.0–11.9)	6.9 (1.7–11.9)	0.4353
Number (%) of dogs needing adjuvant therapy	12 (60%)	7 (39%)	0.3300
Number (%) of dogs experiencing severe ADEs	2 (10%)	2 (11%)	1.000

• CR, complete remission; ADE, adverse drug events.

The mean and/median number of pulses was two (range: 1–4). The maximal oral glucocorticoid dosage given to dogs from the ‘traditional’ group during treatment induction was significantly higher [median 3.2; 95% confidence interval (CI): 2.5–3.9 mg/kg] than that given between pulses to dogs from the second group (median 1.1; 95% CI: 1.1–1.7 mg/kg; P < 0.0001).

The percentage of dogs treated with topical glucocorticoids was not significantly different between ‘traditional’ and ‘pulse’ groups (40% versus 67%; P = 0.119). Similarly the percentage of dogs receiving niacinamide and tetracycline in addition to their glucocorticoids was not significantly different between ‘traditional’ and ‘pulse’ groups (20% versus 50%; P = 0.087). There was no significant difference in the prescription of antibiotics during the first 12 weeks of treatment induction in dogs from the ‘traditional’ (40%) and those from the ‘pulse’ group (44%) (P = 1.000).

The number of dogs with severe ADEs was identical in both groups (two dogs each). In the ‘traditional’ group, one dog died suddenly and one developed azathioprine-induced myelosuppression. In the ‘pulse’ group, one dog was euthanized at his owner's request due to the appearance of signs of iatrogenic Cushing's syndrome and one dog developed melena on its fourth pulse. In the ‘traditional’ group, there were no ADEs other that those expected to occur with glucocorticoid therapy (e.g. polyuria/polydipsia, polyphagia, weight gain). In the ‘pulse’ group, however, two dogs (11%) developed transient diarrhoea and two (11%) exhibited transient aggressive behaviour; these ADEs regressed spontaneously at the end of the pulses.

Discussion

The main goal of glucocorticoid pulse therapy in human pemphigus is to achieve a higher rate and speed of disease remission and, potentially, to establish a long-term CR. In this retrospective study, oral glucocorticoid pulse therapy resulted in a significantly higher proportion of dogs achieving CR during the first 3 months than in those using traditional protocols (pulse: 61%, traditional: 15%). Although a large case series of 91 PF-affected dogs reported CR in about 52% of dogs using conventional treatment approaches,2 a direct comparison with our CR data could not be made because the time frame of the reported CRs was not limited to the first 3 months of treatment, as in our study. Results of a randomized controlled trial evaluating the efficacy of a single pulse (three consecutive days) of intravenous methylprednisolone dosed at 10 mg/kg/day, showed no difference in response between groups (pulse dose 67% CR; traditional dose 75% CR). Again, the time frame of the reported CR was not limited to the first 3 months of treatment.12 Moreover, dogs in our study received repeated pulses, whereas dogs in the randomized control trial received only a single pulse at the beginning of their treatment.12

In the present study, the median time to CR in the pulse therapy group was 6.9 weeks. The mean time to CR reported in a randomized control trial was 8.2 weeks.12 In both studies, the time to CR was not significantly different between the pulse and the traditional groups. Furthermore, the median time to CR in both treatment groups in our study was markedly shorter than the median time to CR of 6.5 months previously reported.2 The reason for such differences remains unknown.

Adjuvant immunosuppressive drugs are commonly used for management of canine PF, even though their additional benefit has not been fully confirmed. Indeed, the analysis of treatment outcomes in the largest case series reported to date did not show a statistically significant difference between the numbers of dogs responding to glucocorticoid monotherapy and dogs treated with combination of glucocorticoids and azathioprine.2 In our study, adjuvant immunoregulatory therapy was allowed in dogs in which, after at least 4 weeks of oral glucocorticoid therapy (using either traditional or pulse therapy), new pustules or erosive lesions continued to appear. While the proportion of dogs needing adjuvant immunoregulation was lower in the ‘pulse’ (39%) compared to the ‘traditional’ (60%) group, the difference was not significantly different.

In addition to rapid establishment of remission, a goal of pulse therapy is to reduce the overall amount of glucocorticoids received. While the cumulative dose of glucocorticoids could not be calculated in our study due to its retrospective nature, the average maximal oral glucocorticoid dosage given to dogs from the ‘traditional’ group was significantly higher than that given between pulses to dogs from the ‘pulse’ group. In contrast, recent randomized control trials of glucocorticoid pulse therapy in human PV and canine PF failed to demonstrate any beneficial effect of pulse therapy on the average daily dose of glucocorticoids and/or the cumulative dose of glucocorticoids when compared to traditional therapy.10, 12 Nevertheless, authors of the human PV study reported a positive, but not statistically significant, trend with a lower mean prednisolone dose used to achieve CR in the ‘pulse’ group.10. Finally, we speculate that the lack of added benefit (shorter time to CR, lower total dose of glucocorticoids used) for pulse therapy reported previously for canine PF could be related to the single glucocorticoid pulse given at the beginning of the trial, which is in contrast to the often multiple pulses given to the dogs in our study.12 Similarly, reported protocols for glucocorticoid pulse therapy in humans with pemphigus involve repeated pulses until disease CR is reached.9, 10

Because of the contradictory results regarding the effect of antibiotic therapy on the outcome of canine PF,1, 2 the number of dogs receiving antibiotics concurrently with the immunosuppressive drugs was recorded for each group of dogs in our study. There was no significant difference between groups in the percentage of dogs having received antibiotics during the first 12 weeks of treatment.

Adverse drug effects related to the treatment of dogs with PF are the most common reason for euthanasia.1, 2 Because most ADEs are associated with the long-term use of medium to high dosages of glucocorticoids, the 3-day courses of pulse therapy in our study followed by low maintenance dosages could, potentially, result in fewer ADEs in the long term. In this study, severe ADEs were documented in an equal number of dogs from each group, which was similar to findings reported in the other two pulse studies.11, 12

In contrast to these previous two reports,11, 12 the route of administration of glucocorticoid pulses in our study was not intravenous but oral. A similar approach has been used in people.10 In dogs, the high oral absorption of prednisolone that very rapidly provides an adequate drug concentration (peak plasma concentration within about 1.5 h) fully justifies an oral route for pulse therapy in this species.13 When considering the dose equivalency between individual glucocorticoids and their route of administration reported in people (500 mg of methylprednisolone equals 625 mg of prednisone), it could be hypothesized that the oral dose of prednisone/prednisolone used in our study could have been slightly higher. Interestingly, several controlled comparative dose studies in people did not find significant differences in short-term outcomes induced by different doses of methylprednisolone (1000 mg versus 250 mg, 5 mg/kg versus 10 mg/kg, etc.),14, 15 suggesting that the small difference in our dose might have been clinically irrelevant. Finally, one study suggested that 11-β-hydroxydehydrogenases, an enzyme family responsible for the conversion of prednisone to its active form prednisolone, might get saturated with higher dosages of prednisone.16 This observation raises the question whether glucocorticoids such as prednisolone or dexamethasone might be more suitable for pulse therapy rather than prednisone itself.

The translatability of the results of this study is, to some extent, restricted by its limitations, such as the relatively small number of patients and its retrospective nature. Nonetheless, our results suggest several benefits that could be associated with oral glucocorticoid pulse therapy using the protocol reported herein, such as (i) a higher percentage of dogs achieving CR during the first three months, (ii) a lower average maximal oral glucocorticoid dosage given between pulses, and (iii) minimal ADEs.

Future studies of a prospective design, including more patients and the use of prednisolone rather than prednisone, could provide additional information on the benefit(s) of oral glucocorticoid pulse therapy for management of canine PF.