| 1. |
|
|
| 2. |
|
|
| 3. |
- Hatschek, T., et al.
(författare)
-
PREDIX HER2 trial : Event-free survival and pathologic complete response in clinical subgroups and stromal TILs levels
- 2020
-
Ingår i: Annals of Oncology. - : Elsevier. - 0923-7534 .- 1569-8041. ; 31:Suppl. 2, s. S49-S49
-
Tidskriftsartikel (övrigt vetenskapligt/konstnärligt)abstract
- Background: Neoadjuvant treatment with Trastuzumab-emtansine was associated with similar rates of pathological complete remission (pCR) as standard therapy withd ocetaxel, trastuzumab and pertuzumab in the PREDIX HER2 trial. Here, results of event-free survival (EFS), and pCR rates in key clinical-pathological subgroups and biomarkers including the abundance of stromal tumor infiltrating lymphocytes (TILs) are presented.Methods: PREDIX HER2 is a randomized, multicenter, open-label, phase 2 study involving 9 Swedish sites. Patients with HER2 positive breast cancer, verified by ISH, T>20 mm and/or verified lymph node metastases were randomized to six three-weekly courses of either docetaxel, trastuzumab SC and pertuzumab (group A), or trastuzumab emtansine (T-DM1, group B). Switch of treatment to the opposite arm was allowed in case of lack of response or severe toxicity. Radiological evaluation included 18F-FDG PET/CT. Patients in both groups received adjuvant chemotherapy with epirubicin and cyclophosphamide. TILs were evaluated using standard methodology, median 10%.Results: In total 197 pts. were evaluable, 99 in group A, and 98 in group B. pCR (ypT0/is ypN0) was achieved in 90 pts, 45.7%, with no significant difference between the two treatment groups. pCR rates were lower in the group of patients with hormone receptor (HR)epositive compared with HR-negative tumors but similar in both treatment groups. pCR rates did not differ between the two treatments in subgroups defined by age, menopausal status, tumor grade, T size, node status, HR-status, HER2 status and Ki67. Progressive disease was observed in 3 pts. (3%) during treatment with T-DM1, none in group A. After a median follow-up of 2.4 years 13 EFS events occurred, with no significant differences between the treatment groups. The presence of 10% TILs predicted pCR significantly (p¼0.009), similar in both treatment groups. We also found that a decrease of SUVmax by more than 80% was highly predictive of pCR. HRQoL was significantly better in pts. receiving T-DM1.Conclusions: Our data suggest that neoadjuvant T-DM1 may be as effective as standard neoadjuvant treatment in all clinical subgroups evaluated. Both TILs and PET/CT showed potential to predict pCR.Clinical trial identification: NCT02568839.
|
|
| 4. |
|
|
| 5. |
|
|
| 6. |
|
|
| 7. |
- Andersson, A, et al.
(författare)
-
Efficient Resource Allocation with Noisy Functions
- 2001
-
Ingår i: Algorithm Engineering: 5th international workshop; proceedings/WAE2001; Lecture notes in computer science. ; , s. 91-105
-
Konferensbidrag (refereegranskat)
|
|
| 8. |
|
|
| 9. |
- Andersson, A, et al.
(författare)
-
Resource allocation with wobbly functions
- 2002
-
Ingår i: Computational Optimization and Applications. - 0926-6003. ; 23:2, s. 171-200
-
Tidskriftsartikel (refereegranskat)abstract
- We consider resource allocation with separable objective functions defined over subranges of the integers. While it is well known that (the maximisation version of) this problem can be solved efficiently if the objective functions are concave, the general problem of resource allocation with functions that are not necessarily concave is difficult. In this article, we focus on a large class of problem instances, with objective functions that are close to a concave function or some other smooth function, but with small irregularities in their shape. It is described that these properties are important in many practical situations. The irregularities make it hard or impossible to use known, efficient resource allocation techniques. We show that, for this class of functions the optimal solution can be computed efficiently. We support our claims by experimental evidence. Our experiments show that our algorithm in hard and practically relevant cases runs up to 40-60 times faster than the standard method.
|
|
| 10. |
|
|
