Pembrolizumab vs Atezolizumab: PD-1 vs PD-L1

They block the same immune brake at two different points. Pembrolizumab (Keytruda) is an anti-PD-1 antibody — it binds the PD-1 receptor on your T cells. Atezolizumab (Tecentriq) is an anti-PD-L1 antibody — it binds PD-L1, the protein the tumour and nearby immune cells display to press that receptor.

PD-1 sits on the T cell. PD-L1 sits on the tumour cell and on tumour-infiltrating immune cells. When the two meet, the T cell stands down. Blocking either side of that handshake keeps the T cell active, which is why both medicines are grouped as immune checkpoint inhibitors and why both can cause the same broad family of immune-related side effects.

There is one mechanistic difference worth knowing, because it explains why the two are not simply swapped for one another. PD-1 has a second partner, PD-L2. A PD-1 blocker such as pembrolizumab interrupts both the PD-L1 and the PD-L2 signal. A PD-L1 blocker such as atezolizumab leaves the PD-L2 route untouched, but it also prevents PD-L1 binding B7.1 (CD80), a separate brake. Whether that translates into any difference a patient would notice has not been settled by head-to-head trials — the two have rarely been compared directly against each other in the same population.

Class fact, not a claim about your case: both are given as day-care infusions and neither is a tablet, a course of injections you take at home, or a one-time procedure.

They overlap in lung cancer, but the two indication lists are not the same. Pembrolizumab carries the wider approved list internationally, including a tumour-agnostic MSI-High/dMMR approval. Atezolizumab's recognised roles are narrower and more combination-bound — extensive-stage small-cell lung cancer with chemotherapy, and hepatocellular carcinoma with bevacizumab.

The lists also move. Roche India's atezolizumab received DCGI approval in 2019 for first-line extensive-stage small-cell lung cancer, and the CDSCO expert committee subsequently reviewed two of its other Indian indications in 2022 — reported in the Indian pharmaceutical trade press at the time. Indications are approved, extended and sometimes withdrawn on a rolling basis in every market, and the Indian list is not a copy of the US or EU list.

Yes — and this is the difference patients are least often told about. Pembrolizumab eligibility in lung cancer is read from the Tumour Proportion Score (TPS), typically stained with the 22C3 clone. Atezolizumab's pivotal work used separate tumour cell (TC) and immune cell (IC) scores, typically stained with SP142. Different rule, different antibody, different threshold.

The numbers below are the thresholds used in the pivotal trials and reflected in the approved labels for first-line single-agent use in non-small-cell lung cancer. They are eligibility cut-offs, not a measure of how well anyone will do.

Not reliably. The same tumour block can read as high on one assay and low on another, because the antibody clones are not equally sensitive. Published comparisons have repeatedly found SP142 stains fewer tumour cells than 22C3, 28-8 or SP263 — so a TPS from a 22C3 report does not automatically translate into a TC or IC score.

This matters in India in a specific, practical way. Laboratories differ in which clone they stock and run, and a biopsy is often small. If the report in your file was run for one drug and the discussion has since moved to the other, the question to ask is simply: does this report answer the question for this drug, or does the block need re-staining? That is a two-minute conversation that can save a wasted decision.

Assay-comparison work of this kind is published in peer-reviewed pathology and immunotherapy literature and is reflected in IASLC pathology committee guidance. It is a recognised limitation, not a laboratory error.

The choice is rarely framed as pembrolizumab against atezolizumab. It follows from the cancer type and stage, which drug is approved for that exact indication, whether a combination partner is required, what the biomarker says on the relevant assay, and the treating hospital's protocol.

In most real situations only one of the two is even on the label for the diagnosis in front of the doctor. Extensive-stage small-cell lung cancer with chemotherapy points one way; a tumour-agnostic MSI-High result points the other. Where both are technically options, the deciding factors are usually the combination the regimen needs and the assay the report was run on — not a general ranking.

The same logic applies to combination questions inside a single molecule, which are a separate decision again: see Single-Agent vs Nivolumab Plus Ipilimumab for how adding a second checkpoint drug changes both the intent and the risk profile.

Broadly, no. Both release the same brake on the immune system, so both can trigger immune-related adverse events (irAEs) — inflammation of the gut, lungs, thyroid, liver, skin, joints or, less often, the heart or adrenal glands. Neither has a meaningfully safer reputation as a class matter.

What differs between two people on the same drug is far greater than what differs between the two drugs. Which organ is affected, when it appears, and how quickly it is recognised are the variables that actually shape a patient's experience. Both drugs require the same monitoring discipline: report a new symptom early, and never treat it at home as an ordinary infection or upset stomach.

For symptom-by-symptom guidance and when a symptom needs urgent assessment, use the side-effect and emergency-symptom pages on the main immunotherapy section rather than assuming a drug comparison covers that ground.

Both remain high-cost originator biologics in India, and as of August 2026 neither has an approved Indian biosimilar. That is the significant point: the price relief that arrived for nivolumab, whose Indian patent lapsed in May 2026 with a domestic biosimilar launched in January 2026, has not yet reached either of these two molecules.

What varies the total is the shape of the regimen rather than a headline vial price — the dose your weight and indication require, how many vials that dose consumes, the interval between cycles, whether a combination partner such as chemotherapy or bevacizumab is added, and how many cycles are planned. A subcutaneous form of atezolizumab approved by CDSCO and launched by Roche India shortens chair time considerably, which changes day-care cost, though not the cost of the molecule itself.

Any figure quoted anywhere for either drug is indicative, as of August 2026, and drawn from published manufacturer and trade-press reporting — not a CION rate card, and not a quotation for your treatment. For how domestically manufactured and imported immunotherapy actually differ on price, quality oversight and availability, see Indian-Made vs Imported Immunotherapy: What Actually Differs.

If your situation is stage 3 lung cancer rather than advanced disease, the PD-1 versus PD-L1 question takes a different shape entirely — durvalumab, another PD-L1 blocker, occupies that setting. Durvalumab vs Pembrolizumab in Stage 3 Lung Cancer works through it. For the wider map of checkpoint inhibitors, biosimilars and how the medicines in this class relate to one another, start at Immunotherapy Medicines Explained. And for the treatment itself — how it is given, what it involves, and what it does not do — see the main immunotherapy section.