I will take this from the beginning, assuming that I am seeing a patient not previously on thyroid replacement who has just had a total thyroidectomy. Your subject heading specifies athyreosis; however the principles and my practice are the same for a spontaneously hypothyroid patient with, say, Hashimoto's thyroiditis, except that the starting dose will be higher for the postsurgical patient, since they will obviously require full replacement—all of their thyroid needs met by exogenous replacement. As I say, most of what follows applies to both situations, although you are correct that there can be additional challenges in the athyreotic patient.
I'm going to take this from the beginning because—while your question specifies "not responding to monotherapy" (which I take to mean levothyroxine [LT4] monotherapy)—because I think "currently failing monotherapy" is different from saying "monotherapy is inappropriate." It depends on how effectively and skillfully the LT4-only treatment is being implemented. I am a firm believer based on over 30 years of doing this, that most doctors, including most endocrinologists, don't dose LT4 high enough. LT4-only Tx tends to get a bad rap on patient-oriented websites, not because it doesn't work in many, if not most, cases, but because it gets underdosed.
Why? There are a number of reasons:
—It is a common, but I believe wrong practice to monitor TSH only. I believe it is easy to miss underreplacement if you don't also look at a T4 level, usually a modern FreeT4 assay. FreeT3 is less often truly useful, although it can be very important as we will get to. So, to answer one of your questions, I usually get a TSH + FT4 + FT3, especially early on, until it is clear the additional information provided by the FT3 isn't worth the additional cost. But, I almost always get at least a TSH + FT4, even in people I've been following for years.
—And, regardless of what tests are ordered, I find it all too common that ANY level somewhere within the lab report's published reference range ("normal" range) is accepted as an indicator of adequate thyroid replacement. Often that ends up meaning dosage adjustments are stopped if the levels are JUST BARELY normal. That might be okay and it might not, and the determining factor is the patient's symptoms. If the patient feels good we're done, and if not, we're not. Although too often symptoms are ignored in the face of "normal" lab results. Endocrinologists are particularly guilty of this because we tend to be taught to treat the numbers not the patient, which flies in the face of my earlier internal medicine training to treat the patient not the numbers. The statistical disconnect is this: The lab's reference range represents what is STATISTICALLY normal for the population, NOT what is necessarily PHYSIOLOGICALLY ideal for an individual patient. That individual physiologically ideal range is always narrower than the population statistical range, usually, but not always lying within the population range. Sometimes an individual's ideal level will lie outside of the lab's reference. In fact, by statistical definition 5% of normal people will have an abnormal result on any given lab test (2.5% above, and 2.5% below).
—Relatedly, another reason LT4 gets underdosed is that low TSH's are almost universally interpreted to mean the patient is overreplaced, and same with high FT4's and FT3's. When in fact there is no question that some patient's need to be run with a low TSH or a high FT4 or FT3 (these are people with central hypothyroidism and thyroid hormone resistance, respectively).
—Lastly, it is not generally recognized that if a TSH does become suppressed by true overreplacement, or other reasons, it can be very slow to rise back to normal—leading to obsessive, repeated reductions in LT4 dose to get the TSH up. This is a good reason not to follow TSH's alone.
Okay, so I will start LT4 at some appropriate starting dose (I encourage Synthroid or one of the other branded products, but if cost is an issue, starting with generic LT4 is okay). It takes at least 5 weeks to see the full effect of any given adjustment, so it is almost never appropriate to check labs sooner than that. Generally speaking I will increase the LT4 dose in baby steps—one dosage at a time (125, to 137, to 150, to 175, etc., for example)—until the TSH is <2.0 and the FT4 is mid to high normal and the patient is no longer symptomatically hypothyroid.
If hypothyroid symptoms remain in spite of labs meeting those criteria, then I make further cautious upward adjustments, monitoring FT4, FT3, and TSH, and symptoms. If the TSH becomes low and the patient is not symptomatically HYPERthyroid, I often will not back off because of that—although I warn you most doctors will.
—if the patient is taking generic LT4 I will switch to synthroid or another name brand, and sometimes people feel better just from that.
If the above does not find a dose that resolves the hypothyroid symptoms—which is getting to the meat of your question—especially if the FT4-to-FT3 ratio is high, suggesting inadequate T4 to T3 conversion, then I will add brand-name Cytomel or generic liothyronine (LT3) 5 mcg once daily, taken first thing in the morning on an empty stomach. Depending on a number of factors I might or might not reduce the LT4 dose at the same time. If all that is added or adjusted is LT3, then labs can be checked in 3 weeks, but any adjustment in LT4 demands labs be delayed for 5 weeks or longer. When using combination therapy (LT4 + LT3) the labs I monitor are: TSH + FT4 + FT3 + SHBG, and I calculate a FT4-to-FT3 ratio.
SHBG is sex-hormone binding globulin—which is simply a protein the liver produces more of, the more the thyroid-hormone receptors deep within the nuclei of the liver cells are stimulated. This makes SHBG a sometimes useful biomarker for TISSUE thyroid levels, as opposed to SERUM levels. Remember, what we really care about in the situation you are asking about is not how much thyroid hormone is in the blood—which is what we measure out of necessity—but rather how much the thyroid hormone receptor (THR) is being stimulated deep in the body's tissues. The SHBG reflects PERIPHERAL THR stimulation, and the TSH is a reflection of BRAIN and PITUITARY THR stimulation. One thing almost no doctor considers is that the amount of thyroid hormone that adequately stimulates those BRAIN THRs, thereby suppressing TSH, is less than what it takes to adequately stimulate the PERIPHERAL THRs all over the body outside the brain. This is how you can have a low TSH, suggesting overtreatment, while the patient still feels hypothyroid because the peripheral THRs want more thyroid hormone. (This represents relatively recent research that is not widely disseminated—but was discussed in a meeting I attended in Chicago a couple of years ago, conducted simultaneously and virtually with a meeting of endocrinologists in London).
Before finishing up, let me back up and talk about the general topic of LT4 + LT3 combination therapy. As you may know, T3 is the less abundant but more potent form of thyroid hormone—it is the more active form of the hormone. The thyroid gland directly secrets about 6% of the body's daily need for T3. The rest is produced all over the body (brain, liver, kidney, muscle) via T4 to T3 conversion. For this reason, most hypothyroid patients don't need to be given LT3. Their bodies take the LT4 and convert it to however much T3 is needed by any given organ at any given time. However, some patients do benefit from the combination if they convert less efficiently—sometime the LT4 replacement will suppress conversion—and a clue to that as I've said, can be a high FT4-to-FT3 ratio (although that is not a terribly commonly used parameter by most doctors), or simply the scenario that you present: failure to respond well to monotherapy. The athyreotic patient will of course lack that small percentage of T3 that the thyroid gland directly produces, and so they are at least theoretically more likely to benefit from the combination. Some not-particularly-well-informed pundits on the Internet will argue that "natural" thyroid products that contain both T4 and T3 are always better than LT4 alone, which is simply not true. In fact, LT4 monotherapy, with the body making its own T3, mimics normal human physiology more closely (where the thyroid makes a lot of T4 and very little T3) than giving a lot of LT3 does.
There are two ways to give LT4 + LT3 combo therapy. My preference is, as I presented above, adding once daily LT3, 5 mcg, to LT4. If needed a second or even third daily dose can be added, generally taken no later than 3 p.m. so as not to disrupt sleep. Some people do okay taking the 2 or 3 LT3 pills all together in the morning for convenience and better compliance, but usually spreading the doses out through the first half of the day is more effective. Unlike LT4, LT3 does not have a long enough serum half-life to last all day on just a single dose.
The other option some patients prefer—and this isn't my first choice, but I will not uncommonly switch to this route if the response to LT4 + LT3 is inadequate, or the patient really wants to try it—is a desiccated thyroid extract (DTE) product. There are a number of these out there but the best known—and most reliably available—is Armour Thyroid. These products get their thyroid hormone from pork thyroid glands harvested from slaughter houses (the "Armour" name literally came from the Armour meat packing company). Some argue that DTE products are better because they are "natural." Not true. Whether or not they work better for some individuals, the "natural" part has nothing to do with it. The main active ingredients are chemically identical to synthetic LT4 and LT3. That said, these animal thyroid glands probably contain some minor metabolites of thyroid hormone (T1, T2, sulfated TH, etc.) that would not contaminate the synthetic products, and it is at least theoretically possible some people do detect some benefit from them. It is very unclear however that these things have any significant biologic activity.
One problem with DTE monotherapy that is clear is that it contains more T3 relative to T4 than is appropriate for humans, resulting in a low T4-to-T3 ratio. The practical result of that is that in virtually all cases in my experience, if the FT3 is adjusted to normal, then the FT4 is low, and if the FT4 is adjusted to normal, the FT3 is too high. It may in fact be that a "stimulant" effect from that extra T3 results in the "cult following" DTEs have amongst some so-called experts, but that does not make the situation safe, or physiologic, or even "natural." If a patient really wants to be be on a DTE, or seems to truly benefit from it, what I will often do is add a small dose of LT4 to the DTE, and fine-tune the two drugs to achieve a normal FT4-to-FT3 ratio. The labs I monitor for either DTE monotherapy or combined LT4 + DTE are the same as I gave above for monitoring LT4 + LT3 combo therapy, because all the issues are the same.
I hope that helps. Perhaps more than you wanted to know, but I think one of the problems with run-of-the-mill hypothyroidism management these days is that most doctors, including many endocrinologists, think it is a lot simpler than it really is.