Treatment failure in hyperthyroid cats after radioiodine (I‐131) injection

Abstract Background There is limited published information on the outcome for cats where total thyroxine concentration (TT4) remains elevated after treatment with radioactive iodine (RAI). Objective To determine the frequency of, and predictors for, subsequent treatment failure in cats for which TT4 remains elevated at hospital discharge, and to report clinical outcomes for cats requiring repeat treatment. Animals One hundred twenty‐one cats with TT4 ≥40 nmol/L after treatment with RAI (out of an original, treated study sample of 959 cats). Methods Retrospective study. Data regarding signalment, weight, TT4 concentration (before RAI treatment, at discharge, and percentage change), day of sampling, and I‐131 dose were acquired. Logistic regression was performed to evaluate predictors of treatment failure. Results In the 87 cats for which classification was possible, 35 (40%) became euthyroid without further treatment. All TT4 variables and weight normalized RAI dose were independently predictive of subsequent treatment failure. In multivariate analysis, TT4 concentration at discharge (P < .001) and weight normalized RAI dose (P = .04) remained in the final model. All 28 cats with TT4 concentration ≥150 nmol/L at discharge ultimately failed treatment, compared with 13/40 (32.5%) and 11/19 (57.9%) cats with TT4 concentrations of 40‐100 nmol/L and 100‐150 nmol/L, respectively. Of the 52 cats that failed treatment, 14 were subsequently managed medically, 12 underwent thyroidectomy (4 with carcinoma), 14 had repeat RAI treatment which was successful in 12/14 (86%) cats, and 13 had no further treatment. Conclusions and Clinical Importance Cats with TT4 >150 nmol/L at discharge after RAI might be candidates for immediate repeat treatment.


| INTRODUCTION
Hyperthyroidism in cats is a commonly diagnosed endocrinopathy with an overall prevalence of 2.4% in England which increases to 8.7% in cats aged 10 years or older. 1 Radioactive iodine (RAI)] is generally considered the optimal treatment for hyperthyroidism in cats. 2 The goal of RAI treatment is to restore euthyroidism while avoiding the development of hypothyroidism. The response to RAI treatment is rapid with 84.7% (444/524) cats having total thyroxine concentration (TT4) within or below reference range by the time of hospital discharge and TT4 is within or below the reference range in 98.5% of cats 6 months later. 3 The development of iatrogenic hypothyroidism, after treatment with RAI, has been linked to the development of azotemia and reduced survival time. 4 Avoidance of hypothyroidism has led to increasing interest in treatment with lower doses of RAI but this has the potential to increase the risk of treatment failure, at least in humans. 5 Several studies have assessed the ideal dosing protocol for RAI. 6,7 A protocol for low-dose (74 MBq or 2 mCi) treatment of hyperthyroidism has a high success rate for resolution of hyperthyroidism overall, however, the power of the study to detect a decrease in effectiveness with the lower treatment dose was only 5.5%, and several cats in that study had "high-normal" thyroxine concentration (>3 μg/dL or >38 nmol/L) after treatment. 8 In many localities where short-hospital stays after RAI treatment are permitted, evaluation of the success of treatment is not performed until a month or more after discharge from the hospital. In the United Kingdom, regulation of radioisotope use in veterinary medicine is in general more stringent and cats are hospitalized for longer after RAI treatment. This means that initial evaluation is performed before the cat is discharged from the hospital, 2-3 weeks after injection of RAI.
The greatest reduction in TT4 concentration occurs in the first month after treatment with RAI, although TT4 concentration can continue to decline for up to 6 months. [9][10][11] If the TT4 concentration is high at the time of discharge it would be useful to know whether it is likely to continue to decrease resulting in the treatment being considered a success, or whether repeat treatment of the cat is necessary.
If repeat treatment is required it would also be useful to know the likelihood of treatment success with a second injection of RAI, or with other types of therapy (surgical or medical management).
The purpose of this study was to report outcomes for cats that had elevated TT4 concentration at the time of discharge from hospital after RAI treatment of hyperthyroidism and to assess for predictors of ultimate treatment failure in these cats which might assist in clinical decision making for these cats. Another aim of this study was to report outcomes for cats that were retreated with RAI after the failure of initial treatment with RAI.  Chi-squared test. Parametric Student's t test was used to compare normally distributed continuous data between 2 groups. Mann-Whitney U test was used to compare non-normally distributed continuous data.

| Statistical analysis
Factors that were individually associated with treatment outcome at P < .1 were then entered into binary logistic regression to build a multivariable model of treatment failure. The association between the TT4 concentration at discharge and failure of treatment was tested using receiver operating characteristic (ROC) curve analysis. The area under the receiver operating characteristic curve (AUROC) was calculated to evaluate the diagnostic performance of the TT4 concentration at discharge to predict failure of treatment. Youden's index was used to determine an ideal diagnostic cut-off with adjustment based on expected clinical utility. Time to development of euthyroidism was depicted graphically using Kaplan-Meier curves, with documentation of TT4 concentration <40 nmol/L defined as an event and cats censored if they died without documented euthyroidism, had repeat treatment with RAI, had thyroidectomy performed, or restarted medical management. If a cat was still alive but lost to follow-up, it was censored on the date that TT4 concentration was last measured. Normally distributed data are reported as mean ± SD (range) and non-normally distributed data are reported as median [25th, 75th percentile]. Statistical significance was defined as P < .05.

| RESULTS
Medical records of 959 cats that were treated with RAI from January 2007 to October 2018 were reviewed ( Figure 1). Data completeness for the variables to be assessed were as follows; breed 100%, age 100%, sex 100%, bodyweight 99.8%, RAI dose 99.5%, TT4 concentration before RAI treatment 99.9%, TT4 concentration at discharge 99.9%, percentage change in TT4 concentration 99.9%, stay duration 100%. Breed was recorded for each cat but because of the low number of purebred cats, breed was classified as purebred and nonpurebred. Sex and neuter status was recorded for each cat, but because of the small number of intact cats, sex was coded as male or female. Eight hundred thirty-seven (87.4%) cats had TT4 concentration < 40 nmol/L at the time of discharge and were considered to have been successfully treated and were not studied further. One hundred twenty-one cats had TT4 concentration at discharge ≥40 nmol/L (94.9 [59.3, 156.0] nmol/L).
There was no recorded TT4 concentration at discharge for 1 cat which was excluded from the logistic regression analysis, although this cat was documented to be euthyroid 1 month later. Of these 121 cats, 2 cats did not receive full dose of RAI due to misinjection. These cats were retreated with RAI and also excluded from analysis. Fifteen of the cats with TT4 concentration at discharge ≥40 nmol/L (70. 6 [48.8, 113.0] nmol/L) were described in follow-up visits to the referring vet as having clinical improvement with no biochemical documentation of euthyroidism. These cats were alive more than 1 year after treatment, had at least 2 revisits to their primary care practitioner, with no further medication for hyperthyroidism dispensed within the follow-up period, and were documented to have either a stable weight or weight gain and no reported clinical signs of hyperthyroidism. However, further TT4 concentration testing was not undertaken (in spite of our recommendation that this be performed) and these cats were excluded from statistical analysis. Seventeen cats were lost to follow-up. In total, 34 cats were excluded leaving 87 cats for further analysis. Fifty-two (60%) cats were classified as treatment failure and 35 (40%) were subsequently classified as treatment success. As a total of the study sample treated, with adequate follow-up, 52/925 (5.6%) cats were classified as treatment failure.
T A B L E 1 A scoring system was used to calculate the dose of radioactive iodine administered to each cat Note: A score was assigned to both TT4 concentration at the time of hospital admission and the clinical signs. These were then added together and on the basis of that figure the dose was given.
T A B L E 2 The scoring system used to calculate the dose of radioactive iodine administered to each cat was modified in 2015

| Thyroidectomy
Twelve cats had thyroidectomy performed after treatment with RAI.
Thyroid carcinoma was confirmed on histopathology in 5 of these cats

| Restarting medical management
Fourteen cats were medically managed (methimazole, thiamazole, carbimazole) and were reported to be well controlled, however, 1 cat was euthanized 6 months after restarting medical management due to difficulties medicating.

| Treatment failure with no further treatment
The remaining 13 cats that failed RAI treatment had no further treat-  Although the upper limit of the reference range for TT4 concentration in our laboratory is 65 nmol, TT4 concentration at discharge of ≥40 nmol/L was considered inappropriately high in this study sample.
The modal TT4 concentration measured 19 days after RAI treatment is reported as 10.0 nmol/L (IQR 10.0-13.6 nmol/L). 18 Furthermore, older cats with comorbidities are expected to have serum TT4 concentration lower than 40 nmol/L. 23 In this study sample, 121/959 (12.5%) cats had elevated TT4 concentration at the time of discharge from the hospital. Adequate follow-up was available for 87 of these cats and the overall, treatment failure rate of 5.6% (52/925) after a longer period of follow-up was similar to that reported previously. 3,8,9,24 Nine cats had an increase in TT4 concentration at the time of discharge from hospital after RAI treatment. This might be due to the fact that although it is recommended to discontinue medical treatment of hyperthyroidism for 2 weeks before treatment with RAI, some cats might have continued medical management due to clinician or owner concerns. Alternatively, their TT4 concentration might have been suppressed initially due to transient nonthyroidal illness or laboratory errors could have occurred.
In conclusion, cats with elevated TT4 concentration at the time of discharge from hospital after RAI treatment can later become euthyroid and the degree of elevation of TT4 concentration at discharge is a predictor of ultimate treatment failure, however, this prediction cannot be improved by also considering other parameters such as how high the TT4 concentration was before RAI treatment, or how long after treatment the sample was collected. Cats with mild elevation of TT4 concentration (40-100 nmol/L) might be good candidates for monitoring as they have a good chance of becoming euthyroid. In contrast, no cats with severe elevation of TT4 concentration (≥150 nmol/L) at the time of discharge from hospital after RAI treatment subsequently became euthyroid in this study and these cats are good candidates for retreatment with RAI.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Approved by the Royal Veterinary College clinical research ethical review board (URN: SR2019-0210).

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.