Table 3:

Summary of Benefit for Immunotherapy and Chemotherapy Combinations (see also Supplemental Table 1 for biomarker-chemotherapy relationships)^a

Markerb	MSI-H	TMB-H	PD-L1 expression
ERCC1 (platinum)	MSI-H and ERCC1 negative are associated; therefore immunotherapy and platinum combinations are likely beneficial	TMB-H and ERCC1 negative are associated; therefore immunotherapy and platinum combination are likely beneficial	PD-L1 expression and ERCC1 positive were associated in 3 of 39 tumor types; hence combinations of platinum and immunotherapy are not beneficial In 1 of 39 tumor types, PD-L1 expression and ERCC1 negative were associated; hence combinations of platinum and immunotherapy are likely beneficial in this tumor type. In the other 35 tumor types there was no relationship between PDL1 expression and ERCC1.
MGMT (dacarbazine/temozolomide)	No significant associations between MGMT negative and MSI-H	No significant associations between MGMT negative and TMB-H	MGMT decreased was associated with PD-L1 expression. Hence the combination of dacarbazine or temozolomide with immunotherapy is likely of benefit
RRM1 (gemcitabine)	MSI-H was associated with RRM1 positivity. Hence gemcitabine and immunotherapy is unlikely to be of benefit.	TMB-H was associated with RRM1 positivity in 8 of 40 tumor types and hence the combination of immunotherapy and gemcitabine is unlikely to be of benefit in these tumor types. TMB-H was associated with RRM1 negativity in 1 of 40 tumor types and hence the combination of immunotherapy and gemcitabine is likely to be of benefit in this tumor types. In the other 31 tumor types there was no relationship between TMB-H and RRM1.	PD-L1 expression was associated with RRM1 positivity in 5 out of 40 tumor types; hence gemcitabine and immunotherapy is unlikely to be of benefit. In the other 35 tumor types there was no relationship between PD-L1 expression and RRM1.
TOP2A (doxorubicin/etoposide/epirubicin)	MSI-H was associated with TOP2A positivity in 4 of 40 tumor types; hence doxorubicin/etoposide/epirubicin and immunotherapy is likely to be of benefit. In the other 36 tumor types there was no relationship between MSI-H and TOP2A.	TMB-H and TOP2A positivity are associated; therefore immunotherapy and doxorubicin/etoposide/epirubicin combination are likely beneficial	PD-L1 expression was associated with TOP2A positivity in 10 of 40 tumor types; hence doxorubicin/etoposide/epirubicin and immunotherapy is likely to be of benefit. In the other 30 tumor types there was no relationship between PD-L1 expression and TOP2A.
TOPO1 (irinotecan/topotecan	MSI-H was associated with TOPO1 negativity. Hence irinotecan/topotecan and immunotherapy is unlikely to be of benefit.	TMB-H was associated with TOPO1 negativity. Hence irinotecan and immunotherapy is unlikely to be of benefit.	No significant associations between TOPO1 positive and PD-L1 expression
TS (fluorouracil/pemetrexed/capecitabine)	MSI-H was associated with TS positivity in 9 of 40 tumor types; hence fluorouracil/pemextrexed/capecitabine and immunotherapy is unlikely to be of benefit. In the other 31 tumor types there was no relationship between PDL1 expression and TS.	TMB-H expression and TS positive were associated in 9 of 40 tumor types; hence combinations of fluoruracil/pemetrexed/capecitabine and immunotherapy are not beneficial In 2 of 40 tumor types, TMB-H and TS positive were associated; hence combinations of fluorouacil/pemetrexed/capecitabine and immunotherapy are likely beneficial. In the other 29 tumor types there was no relationship between TMB-H expression and TS.	PD-L1 expression was associated with TS positivity in 15 of 40 tumor types; hence fluorouracil/pemextrexed/capecitabine and immunotherapy is unlikely to be of benefit. In the other 25 tumor types there was no relationship between PDL1 expression and TS.
TUBB3 (taxane)	MSI-H and TUBB3 negative are associated; therefore immunotherapy and doxorubicin/etoposide/epirubicin combination are likely beneficial	TMB-H and TUBB3 positive were associated in 2 of 40 tumor types; hence combinations of taxanes and immunotherapy are not beneficial In 2 of 40 tumor types, TMB-H and TUBB3 negative were associated; hence combinations of taxanes and immunotherapy are likely beneficial in this tumor type. In the other 36 tumor types there was no relationship between TMB-H and TUBB3.	PD-L1 expression was associated with TUBB3 positivity in 8 of 40 tumor types; hence taxanes and immunotherapy is unlikely to be of benefit. In the other 32 tumor types there was no relationship between PDL1 expression and TUBB3.

Bold “benefit” signifies the strongest evidence to for benefit of immunotherapy and chemotherapy combination

^aSummary from Mantel-Haenszel tests that pools data from the 40 histologies tested (Table 2). For patients where Mantel-Haenszel could not be used (i.e. different histologies showed different results), number of tumor types with significant results are listed—in that case, the Fisher exact test was used to calculate significant associations; See Supplemental Tables 5, 6, 7 for the exact tumor types in each case.

^bFor ERCC1, MGMT, RRM1, TS and TUBB3, it is low or negative expression of the marker that is associated with chemotherapy benefit; for TOP2A and TOPO1, it is positive expression that is associated with chemotherapy benefit. Therefore, as examples, in this Table, ERCC1 negativity (potential benefit from platinums) is associated with MSI-H (potential immunotherapy benefit); TOP2A positivity (potential benefit from doxorubicin, etoposide and epirubicin) is associated with MSI-H (potential benefit from immunotherapy) in 4 of 40 tumor types)

Abbreviations: ERCC1=excision repair complementation group 1, MGMT= O-6-methyl guanine DNA methyltransferase, MSI =microsatellite instability, PD-L1=programmed death-ligand 1, RRM1=ribonucleotide reductase regulatory subunit M1, TMB = tumor mutational burden; TOP2A=topoisomerase 2, TOPO1=topoisomerase 1, TS=thymidylate synthetase, TUBB3=tubulin beta 3