What adjuvant chemotherapy regimen do you recommend for a patient with T1cN0 TNBC?

This is such a nuanced question that comes up with our fellows and affiliates a lot. I will give my take on this, although I want to acknowledge there are still pockets of uncertainty here!

I want to start by saying that in the purely adjuvant setting, we have the ABC trials (USOR 06-090, NSABP-B46I/USOR 07132, and NSABP-B49). These compared, in one way or another, anthracycline sparing TC x6 with an anthracycline containing regimen (TAC6 or AC-Tax). I think the B-49 trial is the closest to what we do in clinical practice. It is early, but there is a slight trend towards improved outcomes in the B-49 trial with anthracycline but this is with more advanced disease. In addition to the ABC trials, we do have a bit of a look with our colleagues in other countries with epirubicin where they are comparing anthracycline sparing regimens. It also seems that TC when compared to epirubicin regimens are at least close, if not equivalent, in early stage patients (Yu et al., PMID 34327363).

What is tough is deciding how much disease is "worth it" (anthracycline). I think this depends on a lot. What are the co-morbidities of the patient? You will face neuropathy versus heart failure. In a SEER analysis of older adults, stage III patients probably did a little better with ACT, but earlier stage did well with either (Mougalian et al., 2015 ASCO Annual Meeting I). Thus, for the early stage patient either seems reasonable and you have to pick which side effect you want to struggle with. For example, in my own experience giving TC x 6 to an older adult with diabetes is harder in some ways than ACT. The neuropathy that comes with six cycles of TC can be significant (the trials seem to only report grade 3-4 so the rates look low but they are not if you consider grade 1-2 as well).

On the points of Pembrolizumab and Olaparib, I would not offer either for a T1cN0 patient. Cutoffs in the OlympiA trial were T2 or node positive disease for patients with TN disease. We have also not had purely adjuvant immunotherapy results for TN disease where in one setting the tumor environment is intact versus removed. I am hopeful we will get SWOG 1418 data to help guide this decision. 

It is academic now as they have gone to surgery, but for the patient above, I would have offered neoadjuvant chemotherapy. In the neoadjuvant space, NCCN recommends all T1c should at least be offered ACT (Korde et al., PMID 33507815). I agree with this because it gives the opportunity to "test" biology and offer adjuvant therapy for incomplete responders (BRCA mutated or not). I do not give TCx6 in this space because then you are wondering what to do with incomplete responders (i.e. having to decide whether to give AC on the back end). I also do not offer neoadjuvant Pembrolizumab for T1c patients. My cutoff is 2cm or node positive in line with the current FDA approval. For me, the challenge in T1c is if I should add carboplatin to the ACT. I will usually add carboplatin for younger fit patients or known BRCA patients.

I disagree with Dr. @Dr. First Last. For many years, my 'go-to' regimen for stage I TNBC was TC (docetaxel and cyclophosphamide) x 4, and results of the ABC analysis didn't change that - the benefit from the addition of doxorubicin in node negative TNBC was so small that it is, in my opinion, outweighed by the risks of MDS/AML and cardiac toxicity, especially in patients with smaller cancers, and data doesn't support the need to administer 6 cycles over 4 for these lower risk patients. Now, however, I administer paclitaxel and carboplatin (PCb) days 1, 8 and 15 every 28 days x 4 cycles (the schedule can be amended to days 1 and 8 every 21 days x 6 cycles if a patient repeatedly has to delay treatment for low counts on day 15), which I find to be much better tolerated than TC, especially in older patients. This change was based on results from the PATTERN trial, where PCb bested FEC-Docetaxel in TNBC; it's not much of an extrapolation to assume that PCb is better than TC if it is better than FEC-D. I use a similar regimen in patients with clinical T1cN0 TNBC in whom neoadjuvant therapy is desired to shrink the breast primary before surgery, and don't typically add pembrolizumab, or give sequential therapy with AC as in KEYNOTE-522, in a stage I patient. 

My cut off for AC-Taxane in the adjuvant setting is typically stage IIA-III TNBC disease. Stage II-III disease is required for use of neoadjvuant KN522 and for adjuvant olaparib as well. The NCCN guidelines don't advocate for anthracycline over non-anthracycline regimens for T1cN0 disease although, the ASCO guidelines did mention this. Dr. @Dr. First Last cited multiple studies like ABC that showed an absolute 2.5% estimated difference with wide confidence intervals between TCx6 and TaxAC in node negative TNBC. The phase 3 PATTERN trial from China (Yu et al., PMID 32789480) compared adjuvant carbo/taxol x 6 to FEC-D in early stage TNBC (78% node negative) with 5 year DFS of 86.5% vs. 80.3% (p=.03). In this trial, the gBRCA patients appeared to do quite well with carbo/taxol. Other phase 3 trials like DBCG 07 READ in HER2- TOP2A normal patients didn't demonstrate an advantage of anthracycline/taxane vs non-anthracycline adjuvant therapy for all the patients including ER - disease and in those tumors less than 2cm. In general, for T1b-cN0 patients the small difference (if any) between anthracycline/taxane vs. non-anthracycline chemotherapy in long term outcomes should be weighed against the toxicities and discussed with the patient.

If the tumor is 1.5 cm or greater, I would treat with neoadjuvant chemotherapy consisting of weekly paclitaxel and carboplatin with pembro followed every 3 week AC with pembro followed by adjuvant pembro as in the KEYNOTE 522 trial (Schmid et al., PMID 32101663). If less than 1.5 cm, I would use DD AC followed by weekly paclitaxel. Yes, I agree with olaparib based on the OlympiA trial. However, in both KEYNOTE 522 and OlympiA trials, the population had either positive nodes or larger tumors or both. Higher risk than the patient described. Therefore, if prognosis is favorable, it is harder to demonstrate an absolute benefit that is clinically meaningful. So I would treat with KEYNOTE 522 and OlympiA trials (if BRCA positive) at tumor size of 1.5 cm or greater provided the patient's performance status and end organ function are okay. It may be overkill what I am suggesting. For tumors less than 1.5 cm, I would have trouble justifying the KEYNOTE 522 and olaparib treatment.