Serum Triiodothyronine-to-thyroxine (T3/T4) Ratio Predicts Therapeutic Outcome to Biological Therapies in Elderly IBD Patients

Lorenzo Bertani; Domenico Tricò; Daniela Pugliese; Giuseppe Privitera; Giuseppe Linsalata; Federico Zanzi; Maria Gloria Mumolo; Brigida Barberio; Fabio Monzani; Santino Marchi; Luisa Guidi; Corrado Blandizzi; Alessandro Armuzzi; Francesco Costa


Aliment Pharmacol Ther. 2021;53(2):273-280. 

In This Article

Abstract and Introduction


Introduction: Rates of elderly patients with inflammatory bowel diseases (IBDs) are increasing, and biomarkers are needed to optimise their therapies. Serum triiodothyronine-to-thyroxine (T3/T4) ratio has been correlated with geriatric patient frailty.

Aim: To assess the suitability of T3/T4 ratio as a response marker to biologics in elderly patients with IBD.

Methods: Patients with IBD over 60 years old were enrolled, when starting biological therapy. Therapeutic outcome was assessed after 54 weeks of treatment as mucosal healing (Mayo endoscopic score < 2 for ulcerative colitis; ulcer disappearance for Crohn's disease) and clinical remission (Partial Mayo Score < 2 for ulcerative colitis; Harvey-Bradshaw Index < 5 for Crohn's disease). T3/T4 ratio was evaluated at baseline, and its association with therapeutic outcomes was tested by multivariable logistic regression and receiver operating characteristic (ROC).

Results: We enrolled 80 patients; 44 achieved clinical remission and 36 mucosal healing. Baseline T3/T4 ratio was higher in patients with mucosal healing, as compared with those without mucosal healing (P < 0.0001), regardless of the disease type or biological drug (OR 6.4 [2.9–14.3] for each T3/T4 unit increase, P < 0.0001). A cut point of 3.3 was identified as the optimal threshold of baseline T3/T4 ratio for predicting mucosal healing, providing 78% sensitivity and 89% specificity (area under the ROC curve 0.88 [0.79–0.94]; positive and negative likelihood ratios 6.8 [2.9–15.9] and 0.3 [0.1–0.5] respectively).

Conclusions: T3/T4 ratio seems a reliable tool for predicting therapeutic outcome of biological therapy in elderly patients with IBD. If validated, the assessment of this parameter before starting biological treatment might be suggested.


Ulcerative colitis and Crohn's disease are the main pathological conditions belonging to the class of inflammatory bowel diseases (IBDs). A dysregulation of immunologic response to enteric antigens is regarded as the pathogenic cornerstone of IBDs, and thus their therapeutic management is currently focused on immune-modulation.[1,2] In this respect, over the last two decades, the advent of biological therapy has modified significantly the natural history of IBD.[3] However, immune-modulation can be associated with the occurrence of serious adverse events, such as treatment-related cancers or serious systemic infections, particularly in elderly patients.[4,5]

Several biomarkers of therapeutic response have been proposed in an attempt of maximising the benefit-to-risk ratio of biological therapies in IBD. With regard for anti-tumour necrosis factor (TNF) drugs, recent evidence suggests that serum levels of oncostatin M[6] and whole blood expression of the TREM1 gene[7] could predict the therapeutic outcome. Conversely, the outcome of treatment with vedolizumab seems to be predictable through the detection of serum interleukins 6 and 8[8,9] as well as serum α4β7 concentration.[10] Unfortunately, however, the use of these biomarkers in clinical practice is currently hampered mainly by their poor availability in the large majority of hospitals.

In recent years, the ageing of population has led to a significant increase in the prevalence and incidence of elderly IBD patients.[5] Moreover, up to 20% of IBD patients are diagnosed after 60 years of age.[11] Owing to the higher risk of treatment-related complications,[12,13] the use of biological therapies in this particular population should be discussed carefully in terms of benefit-to-risk ratio. In this perspective, a reliable biomarker, able to predict the therapeutic effectiveness of biologics in the elderly, is lacking.

Age-dependent modifications of free triiodothyronine (fT3) serum levels have been described in home-dwelling healthy elderly.[14] fT3 levels are related with the activity of deiodinases, which convert free thyroxine (fT4) into fT3.[15] Of note, a relationship between low fT3 values, the so-called 'low T3 syndrome' or 'non-thyroidal illness syndrome', and worse outcomes has been documented in various clinical settings, regardless of patient age.[16] A slight decrease in serum fT3 values may represent a physiological, age-dependent modification of iodothyronine metabolism, even though markedly less than that observed during starvation or chronic diseases, when the excess of catabolism is counterbalanced by a substantial downregulation of deiodinase activity.[15] Indeed, several studies have shown an important relationship between low serum fT3 values and short-term survival in elderly patients under specific pathological conditions.[17–19] A recent study by Pasqualetti et al[20] pointed out that the serum triiodothyronine-to-thyroxine (T3/T4) ratio could be used as an independent biomarker of frailty in hospitalised geriatric patients, even in case of normal fT3 values. Since peripheral deiodinases are expressed mainly in lean muscles,[15] it is conceivable that patients with low muscle mass display a reduced expression of these enzymes, which can lead to lower fT3 levels, higher fT4 levels and, consequently, lower T3/T4 ratio.

Patients with IBD are often affected by reduction of lean muscle mass, a condition defined as 'sarcopenia', even when they are under disease remission.[21] However, sarcopenia is associated with a worst outcome of IBD, particularly in terms of surgical risk[22] and surgical comorbidities.[23] Notably, sarcopenia is more common in the elderly age, both in the general population[24] and among IBD patients.[23,25]

Based on the above background, the aim of the present study was to assess whether the assessment of T3/T4 ratio before starting a biological treatment could be useful as a biomarker for predicting the therapeutic outcome in elderly patients with IBD.