| Abstract | Uvod: Ekspresija PD-L1 predložena je kao prognostički biomarker kod kolorektalnoga
karcinoma, ali još nije uvedena u praksu zbog nedostataka standardizacije u procjeni
imunohistokemijske analize, metodi bodovanja, primjeni različitih vrsta imunohistokemijskih
testova, materijala i probira bolesnika (10, 11).
Nedavno su razvijena dva sustava bodovanja za procjenu ekspresije PD-L1; procjena
ekspresije pozitivnih tumorskih stanica (TPS) i kombinirani pozitivni rezultat (CPS).
Cilj: U ovoj studiji cilj je bio procijeniti ekspresiju PD-L1 primjenom dvaju sustava
bodovanja TPS-a i CPS-a, a rezultate smo usporedili s kliničko-patološkim karakteristikama,
MMR statusom te RAS i BRAF mutacijama.
Materijal i metode: Analizirali smo imunohistokemijsku ekspresiju PD-L1 u 102 uzorka
kolorektalnoga karcinoma u resekcijskim uzorcima tkiva. Kliničko-patološke karakteristike
korelirali smo na različitim graničnim vrijednostima < 1%, 1 – 5 %, 5 – 10 %, 10 – 50 % i ≥
50 % upotrebom TPS i CPS metode bodovanja i korelacijom s MMR statusom te RAS i
BRAF mutacijama.
Rezultati: Uspoređujući TPS s CPS-om, uočava se da je značajna razlika u raspodjeli
bolesnika prema pozitivnome nalazu s obzirom na skupine, tako da je TPS pozitivno 42
(41%) bolesnika, a CPS pozitivno jest njih 97 (95%), što je značajna razlika (McNemarBowkerov test, P < 0,001). Ekspresija PD-L1 ocijenjena je u svakome uzorku upotrebom TPS
i CPS metode bodovanja. U 60 (58,8 %) slučajeva bilo je negativno (< 1 %) upotrebom TPS
metode u usporedbi s 5 (4,9 %) upotrebom CPS metode. Upotrebom TPS-a pozitivnost veću
od 50 % imala su samo 2 (2 %) uzorka. Upotrebom CPS metode najveći postotak pozitivnosti
36 (35,3%) imali su uzorci s rezultatima od ≥ 5 do 10.
Iz ovoga proizlazi da je statistički značajni dio ispitanika bio negativan na PD-L1 procjenom
TPS-a (TPS < 1 %), dok je procjenom pomoću CPS metode bilo samo 5 % ispitanika manje
od 1 %, a najviše je bilo u rangu ≥ 5 – 10 %, što je statistički značajno u odnosu na druge postotke.
Značajne su razlike u TPS-u i CPS-u u skupini muškaraca (test marginalne homogenosti,
P < 0,001) i u skupini žena (McNemar-Bowkerov test, P < 0,001) (Tablica 5.14).
113
Ako se gleda raspodjela prema TPS-u, nema značajnijih razlika s obzirom na TNM
klasifikaciju, limfovaskularnu i perineuralnu invaziju, TIL i TAM te s obzirom na postojanje
metastaza. Bolesnici s adenokarcinomom (NOS) značajnije više imaju vrijednosti TPS ≥ 10 –
50 % ili > 50 % u odnosu na medularni tip (Fisherov egzaktni test, P = 0,02). Umjereno
diferencirani značajnije su više oni s TPS > 50 %, a slabo diferencirani s vrijednostima TPS ≥
10 – 50 % (Fisherov egzaktni test, P < 0,001). Bolesnici sa statusom limfnih čvorova 2
značajnije više jesu s vrijednostima TPS-a ≥ 10 – 50 % (Fisherov egzaktni test, P = 0,005), u
odnosu na druge statuse limfnih čvorova ( Tablica 5.31).
Ako se gleda raspodjela prema CPS-u, nema značajnijih razlika s obzirom na TNM
klasifikaciju, histološki tip, gradus, limfovaskularnu i perineuralnu invaziju, TAM te s
obzirom na postojanje metastaza. Bolesnici sa statusom limfnih čvorova 0 značajnije su više s
vrijednostima CPS-a > 50 % (Fisherov egzaktni test, P = 0,02) u odnosu na druge statuse
limfnih čvorova. Ako je puno limfocita koji infiltriraju tumor, (TIL) vrijednosti CPS-a
značajnije su više > 50 % u odnosu na to da malo infiltriraju ili da uopće ne infiltriraju tumor
(Fisherov egzaktni test, P = 0,01) ( Tablica 5.37).
Od ukupno 60 (59 %) bolesnika s TPS-om < 1 %, značajno je manje, njih 8 (42 %), s
nedostatkom dvaju ili više proteina, dok je značajno više bolesnika koji imaju nedostatak dva
i više proteina s TPS-om ≥ 1 – 5 %. Vrijednosti TPS-a ≥ 10 – 50 % značajnije više imaju
bolesnici s nedostatkom jednoga proteina (Fisherov egzaktni test, P = 0,02).st, P = 0,01) (
Tablica 5.37).
Nema značajne razlike u raspodjeli bolesnika prema postotku pozitivnih tumorskih stanica
(TPS) ili broja tumorskih i imunoloških stanica (CPS) cijeloga uzorka kod bolesnika koji
nemaju RAS ili BRAF mutaciju.
Što se tiče povezanosti ekspresije PD-L1 u morfološki različitim područjima tumora značajno
su zastupljeni niži postotci kod TPS-a na mjestu najdublje invazije, na mjestu prijelaza
tumor/tumorska stroma i na mjestu prijelaza uredne sluznice u tumorsko tkivo a zastupljenost
broja tumorskih i imunoloških stanica CPS-a na mjestu najdublje invazije značajno je viša.
Zaključak: Zaključno, pokazali smo da je CPS korisnija metoda procjene ekspresije PD-L1
nego TPS, pa bi kolorektalni karcinomi koji imaju viši postotak imunohistokemijske
ekspresije PD-L1 CPS metodom mogli imati poremećaj na putu imunološke kontrolne točke
114
PD-1/PD-L1 te bi kao takvi mogli predstavljati podskupinu koja može imati koristi od anti
PD-L1 terapije.
Također smo pokazali da su imunohistokemijska ekspresija PD-L1 te MMR i MSI status i
status TIL-ova neovisni prognostički prediktori kod bolesnika s CRC-om, pa bi se kao takvi
trebali uzeti u obzir u daljnjim istraživanjima. |
| Abstract (english) | Assessment of PD-L1 expresion in colorectal cancer by comparasion of scoring methods and
their significance in ralation to clinicopathological parameters
Introduction: PD-L1 expression has been proposed as a prognostic biomarker in colorectal
cancer, but it has not yet been put into practice due to the lack of standardization in the
assessment of immunohistochemical analysis, the scoring method, the application of different
types of immunohistochemical tests, materials and patient screening (10,11). Recently, two
scoring systems have been developed to assess PD-L1 expression; positive tumor cell
expression assessment (TPS) and combined positive score (CPS).
Objective: In this study, the objective was to evaluate PD-L1 expression using two scoring
systems, TPS and CPS The results were compared with clinical-pathological characteristics,
MMR status as well as RAS and BRAF mutations.
Results: Comparing TPS with CPS, it is observed that there is a significant difference in the
distribution of patients according to the positive findings with regard to the groups, so that
TPS is positive in 42 (41%) patients, and CPS is positive in 97 (95%), which is a significant
difference (McNemar-Bowker test, P < 0.001). PD-L1 expression was assessed in each
sample using the TPS and CPS scoring methods. In 60 (58.8%) cases it was negative (< 1%)
using the TPS method compared to 5 (4.9%) using the CPS method. Only 2 (2%) had a
positivity higher than 50% using the TPS of the sample. It follows from this that a statistically
significant part of the subjects was negative for PD-L1 by TPS assessment (TPS < 1%), while
only 5% of subjects were less than 1% by assessment using the CPS method, and most were
in the range ≥ 5 – 10%, which is statistically significant compared to other percentages.
Using the CPS method, the highest percentage of positivity (35.3%- 36 cases) had samples
with results from ≥ 5 to 10. There are significant differences in TPS and CPS in the group of
men (test of marginal homogeneity, P < 0.001) and in the group of women (McNemarBowker test, P < 0.001) (Table 5.14). If we look at the distribution according to TPS, there are
no significant differences with regard to TNM classification, lymphovascular and perineural
invasion, TIL and TAM, and with regard to the existence of metastases. The patients with
adenocarcinoma (NOS) significantly more often have TPS values ≥ 10 – 50% or > 50%
compared to the medullary type (Fisher's exact test, P = 0.02). The moderately differentiated
tumors significantly more often have TPS > 50%, and the poorly differentiated have values of
116
TPS ≥ 10 – 50% (Fisher's exact test, P < 0.001). Patients with lymph node status 2
significantly more often have TPS values ≥ 10 – 50 % (Fisher's exact test, P = 0.005),
compared to other lymph node statuses (Table 5.31). If we look at the distribution according
to CPS, there are no significant differences with regard to TNM classification, histological
type, grade, lymphovascular and perineural invasion, TAM and with regard to the existence of
metastases. There were significantly more patients with lymph node status 0 with CPS values
> 50% (Fisher's exact test, P = 0.02) compared to other lymph node statuses. If there are many
lymphocytes infiltrating the tumor, (TIL) CPS values are significantly more often > 50%
compared to if they infiltrate little or not at all (Fisher's exact test, P = 0.01) (Table 5.37). Out
of a total of 60 (59 %) patients with TPS < 1%, there are significantly fewer, 8 (42 %) of
them, with a deficiency of two or more proteins, while there are significantly more patients
with a deficiency of two or more proteins with TPS ≥ 1 – 5 %. TPS values ≥ 10 – 50% are
significantly more prevalent in patients with one protein deficiency (Fisher's exact test, P =
0.02).st, P = 0.01) (Table 5.37). There is no significant difference in the distribution of
patients according to the percentage of positive tumor cells (TPS) or the number of tumor and
immune cells (CPS) of the entire sample in patients who do not have RAS or BRAF mutation.
Regarding the association of PD-L1 expression in morphologically different areas of the
tumor, lower percentages are significantly represented in TPS at the site of the deepest
invasion, at the point of the tumor/tumor stroma transition and at the transition point of
normal mucosa into tumor tissue, and the representation of the number of tumor and immune
cells CPS at the site of the deepest invasion is significantly higher.
Conclusion: In conclusion, we have shown that CPS is a more useful method of assessing PDL1 expression than TPS. The colorectal cancers with a higher percentage of
immunohistochemical expression of PD-L1 (using the CPS method), could have a disorder in
the PD-1/PD-L1 immune checkpoint pathway and as such could represent a subgroup that
may benefit from anti PD-L1 therapy. We have also showed that co-histochemical expression
of PD-L1 and MMR and MSI status as well as the status of TILs are independent prognostic
predictors in patients with CRC, so they should be taken into account in further research. |
